Enacting Perfection: Buddhajñānapāda’s Vision of a Tantric Buddhist World

Enacting Perfection: Buddhajñānapāda’s Vision of a Tantric Buddhist World

By Catherine Dalton

A dissertation submitted in partial satisfaction of requirements for the degree of Doctor of Philosophy in Buddhist Studies in the Graduate Division of the University of California, Berkeley Committee in charge:

Professor Jacob Dalton, Chair

Professor Alexander von Rospatt

Professor Robert Sharf

Professor Munis Faruqui

Professor Harunaga Isaacson

by Catherine Dalton

Abstract

Enacting Perfection:

Buddhajñānapāda’s Vision of a Tantric Buddhist World by Catherine Dalton

Doctor of Philosophy in Buddhist Studies

University of California, Berkeley

Professor Jacob Dalton, Chair

This dissertation focuses on the life, writings, and thought of the Indian Buddhist yogin and tantric exegete Buddhajñānapāda, remembered as the founder of the Jñānapāda School of tantric theory and practice. Through an in-depth study of his oeuvre, I attempt to excavate the late eighth-century world of tantric Buddhism as it emerges in his narrative, doctrinal, and ritual writings. I focus, in particular, on his most important composition, the Dvitīyakramatattvabhāvanā-mukhāgama in which Buddhajñānapāda uses autobiographical narrative and visionary revelation to frame assertions about the nature of reality and outline rituals that lead to its realization. I examine the key features of both the doctrinal positions articulated in his works, as well as the ritual systems through which a yogin was to embody and realize those doctrines, showing that Buddhajñānapāda was an important and innovative figure in the realms of both tantric Buddhist doctrine and ritual. The dissertation also includes a Tibetan edition and an annotated English translation of the Dvitīyakrama. i For Chökyi Nyima Rinpoche, whose kindness cannot be repaid. ii Acknowledgements To the Glorious One who has perfected bliss, who is endowed with the radiance of nondual profundity and luminosity; to his nature, which is peace, the blazing sixteenth part, the ultimate essence; and to the three supreme gurus who teach that, I constantly bow with my three activities of body, speech, and mind equally. -Buddhajñānapāda, Dvitīyakrama I would not have undertaken my doctoral studies in the first place had it not been for many years of encouragement from Chökyi Nyima Rinpoche to do so. His advice and support have remained instrumental throughout the process of researching and writing this dissertation, and indeed throughout many aspects of my life in the nearly twenty years that I have had the fortune to know him. The lightheartedness and joy of Tsikey Chokling Rinpoche has remained a constant source of inspiration for me through this whole process. I am also grateful for the advice and support of Dzongsar Khyentse Rinpoche and Dzigar Kongtrül Rinpoche, whose encouragement, along with that of Chökyi Nyima Rinpoche, helped me to move forward as I came up against what were, for me, some of the more difficult materials in this dissertation. I have done my best to adhere to their guidance, as well as to maintain the academic rigor that they encouraged me to uphold. My personal approach to the materials studied in this dissertation comes from a foundation in the Buddhist tradition that was built through years of study with these and other teachers. In addition, the lamas, khenpos, and monks of Ka-Nying Shedrub Ling Monastery and the nuns of Nagi Gompa Nunnery, both located in Kathmandu, Nepal, have, over a period of nearly twenty years now, taught me not only how to approach Buddhist texts and practices, but how to appreciate Buddhism as a living tradition, a lesson that has been of direct and great benefit in helping me come to understand Buddhajñānapāda’s 8th/9th-century world. My advisor at the University of California, Berkeley, Dr. Jacob Dalton, whom I met at the “Translating the Words of the Buddha” conference in India in 2009, has supported and guided my work with kindness and generosity throughout the entire process. It was in my coursework with Jake that I first came to know of Buddhajñānapāda, and where Jake’s passion and curiosity about the “intermediate period” of tantric Buddhism in which Buddhajñānapāda lived and wrote inspired my own. Jake’s support and guidance throughout the process of the dissertation research and writing has been essential in helping me bring this project to fruition. The other members of my dissertation committee have each contributed to this dissertation in essential ways. Dr. Alexander von Rospatt has been a kind and generous source of personal and academic inspiration and support since we first met in Kathmandu when I was applying to Berkeley. Meeting periodically with Alex throughout the years of my dissertation research and writing to discuss my work, whether in his office at Berkeley or a garden restaurant in Boudhanath, has invariably left my horizons widened and my understanding of these materials richer. Dr. Harunaga Isaacson was kind enough to join my committee just a few months before I submitted my dissertation, but his great generosity with comments, corrections, and guidance during this final period of my work has significantly improved, and even shaped, the final form and content of this dissertation. Haru’s kindness, and especially his unflagging enthusiasm for the material, provided much inspiration that enabled me to bring this work to completion. iii My coursework and exam on Chinese Buddhism with Dr. Robert Sharf broadened my knowledge of Buddhist traditions, and Bob’s guidance on the dissertation writing process was of great help to me in navigating my way through such a long and varied writing project. My exam on Indian History with Dr. Munis Faruqui, and his astute questions about my work at the outset of the research process helped me to frame my research within, and learn so much more about, the wider historical context of the Indian subcontinent. Munis also taught me much of value about the historian’s craft. I could not have completed my edition and translation of the Dvitīyakrama without the assistance of Khenchen Chodrak Tenphel, who patiently and insightfully answered my many questions on the text and its commentary nearly every day over a period of several months in Boudhanath, Nepal in the winter and spring of 2016. The series of teachings that Khenpo Rinpoche gave to the monks of Ka-Nying Shedrub Ling Monastery for several months that winter, and which I was generously permitted to attend, were also instrumental in opening Buddhajñānapāda’s writings to me. Khenpo Karma Gyurme of Ka- Nying Shedrub Ling Monastery kindly read through the entirety of the Tibetan translation of the Ātmasādhanāvatāra with me, clearing up confusing points in the text, as well as confirming my suspicion that it is a genuinely difficult work! Dr. Péter-Dániel Szántó very generously (and often unasked) shared with me quite a bit of his unpublished work related to Buddhajñānapāda. His draft Sanskrit editions of several texts were of tremendous help for me in writing the dissertation. A number of exciting Skype conversations with Péter as we worked together on a short encyclopedia article on Buddhajñānapāda were also very helpful for me as I moved forward with writing my dissertation. Likewise, Dr. Kimiaki Tanaka kindly shared with me an unpublished English translation of some of his work on Buddhajñānapāda, enabling me to read research that was otherwise inaccessible to me, as it was published in Japanese. Dr. Tanaka has also generously gifted me copies of several of his difficult-to-acquire publications on Buddhajñānapāda’s writings over the past few years. Dr. Christian Wedemeyer generously shared with me a written copy of a paper that he delivered in a conference on tantra at Berkeley in 2014 (an updated version of which is fortunately soon to be published in 2019), which turned out to be of great relevance to my own research. Hubert Decleer kindly shared with me several of his unpublished articles concerning Buddhajñānapāda’s life story. I am also grateful to Hubert for introducing me to Tibetan culture when I was an undergraduate student on the SIT Tibet Studies program that he directed in the spring of 2000, and for his continued friendship since then. Dr. Ryūta Kikuya shared with me several of his published but difficult to access articles (in Japanese) on topics related to Buddhajñānapāda, and has been kind and encouraging of my work. Dr. Adam Krug generously shared with me his newly submitted dissertation before it was otherwise available, and several other helpful articles. Dr. James Gentry offered helpful insights at several points in the research and writing process, as did the other members of Rangjung Yeshe Institute’s Faculty Research group in the spring of 2015—Dr. Gregory Sharkey SJ, Joanne Larson, Dr. Karin Meyers, and Dr. Philippe Turenne. Dr. Douglas Duckworth and Dr. Thomas Doctor kindly helped me with some queries on points of philosophy. Dr. Mattia Salvini gave some very helpful guidance on points of Sanskrit and kāmaśāstra. Dr. Takahiko Kameyama and Atsuki Nakagami generously helped me look through several articles written in Japanese. Ryan Damron patiently listened to quite a number of extensive and excited monologues (I iv occasionally let him get a word in) on my research over the years, and always offered constructive feedback once I finally stopped to breathe. Ryan has also been consistently enthusiastic regarding my requests to look at a number of short Sanskrit passages with me along the way. Somānanda Dharmanātha has been a generous source of all sorts of information, as well as a delightful sounding-board for many discoveries and ideas over the years of my research and writing. Our dear friends Zack Beer and Sara Rojo have given my husband Gerry and me a warm and welcome (and migrating!) second home during our many visits to the Bay Area while I have been working on this dissertation. Many, many more friends and family members, too numerous to name, have supported me with their friendship and love over the years of my doctoral work, and not complained too much when I gave overly long and detailed answers to their polite questions about my dissertation topic.

Dr. William Waldron was the first to introduce me to Buddhism in 1998 in a seminar

(on Nāgārjuna’s Mūlamadhyamakakārikās; what a place to start!) during my undergraduate studies at Middlebury. The inspiration from that and several other courses with Bill at Middlebury have carried me all the way through a PhD in Buddhist Studies, and I am grateful for his friendship throughout these years. Ten years of study, teaching, and translating at Kathmandu University’s Centre for Buddhist Studies at Rangjung Yeshe Institute prior to taking up my doctoral studies at Berkeley gave me a solid foundation in Buddhist Studies and Tibetan language, as well as leading to many lasting friendships in the unique and beautiful community there, and to making a real home in Kathmandu; I look forward to returning to the faculty at RYI next year. My MA thesis advisor at RYI, Dr. Andreas Doctor, helped extend my horizons beyond the world of Buddhist philosophy to the world of Buddhist ritual practice, a shift that has influenced the trajectory of all of my subsequent academic research. I am also grateful to the late Greg Whiteside, the principal at Rangjung Yeshe Institute for more than 20 years, for many years of friendship and for his always enthusiastic encouragement of my studies. I wish I could have celebrated the completion of these studies with Greg. I may be one among the last generation of Buddhist Studies graduate students working in Tibetan who had the tremendous pleasure to know the encyclopedic scholar and consummate gentleman, the late, and it is no exaggeration to say great, Gene Smith. I had the fortune to spend time with Gene on several occasions, always on his visits to Kathmandu where I was living at the time, but this was before I began my doctoral studies, so I was not able to pick his brain about the topic of my dissertation. However, I cannot even conceive of having written this dissertation without the aid of Gene’s BDRC (formerly TBRC), which has placed the entire Tibetan canon and more at my fingertips, and I have thought of Gene often and with gratitude throughout the process of my research and writing. My parents, Phil and Sara Dalton, have lovingly supported me in every single endeavor I have ever undertaken. They also provided a warm and cozy space with pleasant company for the writing of the early part of this dissertation in their beautiful mountain home. And last, but very, very far from least, my husband, Gerry Prindiville, has nurtured and supported me throughout the entire process of my graduate studies, especially during the dissertation phase, providing encouragement when I needed encouragement, care v when I needed care, space when I needed space, and love all the way through. It is difficult to imagine having made it through all of this without Gerry’s love and support. It has been a great personal pleasure for me to have had the time and energy to devote myself to the research and writing of this dissertation over the past several years. Generous fellowships from the Robert E. Ho Foundation, FLAS, UC Berkeley, and the Khyentse Foundation have provided the material assistance that enabled me to focus on research and writing, making the completion of this dissertation possible. Over the course of researching and writing this dissertation I have gratefully received advice, guidance, and clarifications from a number of eminent Buddhist teachers and learned scholars. The flaws and errors that inevitably remain are entirely my own. vi Table of Contents Dedication i Acknowledgements ii Table of Contents vi Introduction viii Part I A Study of Buddhajñānapāda’s Life, Works, and Thought Narrative Chapter One Meeting Mañjuśrī: Buddhajñānapāda’s Life and Works 3 Chapter Two Narrating Revelation: The Dvitīyakrama’s Unique Framing of Doctrine and Ritual 63 Doctrine Chapter Three Following the Tantric Path to the Reality of Nondual Wisdom: Buddhajñānapāda’s Doctrinal Positions 81 Chapter Four The Perfection Stage of the Perfection Stage: Buddhajñānapāda and the Great Perfection? 156 Ritual vii Chapter Five The Two Stages of Tantric Practice: Generating Self as Deity in Buddhajñānapāda’s Generation Stage Sādhanas 176 Chapter Six The Ultimate Path to Awakening: The Perfection Stage in Buddhajñānapāda’s Writings 209 Chapter Seven Revealing Reality: Tantric Initiation in Buddhajñānapāda’s and Vaidyapāda’s Writings 247 Conclusion Chapter Eight Buddhajñānapāda and Beyond: Buddhajñānapāda’s Thought Moving into the Later Tradition 285 Concluding Reflections 295 Part II Tibetan Edition and English Translation of Buddhajñānapāda’s Dvitīyakramatattvabhāvanā-mukhāgama Tibetan Edition of the Dvitīyakrama 300 Annotated English Translation of the Dvitīyakrama 355 Bibliography 446 viii Introduction You are the father and the mother of all beings! Protect me and others from great danger! Master, lord of beings, dispel suffering! Emptier of the three realms, greatest of the great, you protect beings! You are beginningless, unvoiced, lacking the upper part of the bindu, the revered, the letterless, producer of nectar, the empty bliss of great joy! In order to benefit beings, O you Great Protector, please bestow happiness— the happiness that is great bliss—upon all the buddhas. The path to awakening, not stained by faults, which pacifies all types of suffering, and quenches thirst, liberates from the waves of saṃsāra, and places one in happiness—please teach this path which is not fathomed [even] by those who are victorious over all things! -Buddhajñānapāda, Dvitīyakrama These are the words that Buddhajñānapāda (ca. 750-820), remembered as the founder of the Jñānapāda School of tantric exegesis and practice, records himself as having cried out during a vision of the awakened bodhisattva Mañjuśrī1 who laid out a tantric maṇḍala before Buddhajñānapāda’s very eyes. Mañjuśrī’s response to his supplication, also meticulously recorded by the Indian Buddhist yogin and tantric exegete, became the basis for Buddhajñānapāda’s system of tantric theory and practice focused on the Guhyasamāja-tantra. In his Dvitīyakramatattvabhāvanā-mukhāgama, The Oral Instructions on Training in the Suchness of the Second Stage,2 the text that records this visionary experience and its contents, Buddhajñānapāda articulates his vision of the tantric Buddhist path through a combination of autobiographical narrative, doctrinal reflection, and ritual liturgy. This dissertation focuses on the figure of Buddhajñānapāda and attempts to excavate the late eighth- and early ninth-century world of tantric Buddhism as it emerges in his narrative, doctrinal, and ritual writings. Among the most influential of the Buddhist tantras, the Guhyasamāja-tantra and the rituals it inspired revolutionized Buddhist practice in India in ways that are still coming to be fully appreciated. Buddhajñānapāda played an important role in shaping the doctrinal and practice traditions associated with it, and thus too tantric Buddhism more generally. Fortunately, quite a number of primary sources are available that allow access into Buddhajñānapāda’s life and thought. Eleven works that can be confidently attributed to him are preserved in Tibetan translation in the Tibetan canon, two of which survive in their original Sanskrit and three more of which survive partially in Sanskrit. It is, however, his Dvitīyakrama, quoted above, that serves here as the primary framework for my presentation of Buddhajñānapāda’s life, writings, and thought. This truly unique composition defies even the concept of literary genre, spanning as it does autobiographical narrative, visionary experience, doctrinal claims about the nature of reality, and detailed ritual instructions for a variety of tantric Buddhist practices. Consequently, every other piece of writing attributed to Buddhajñānapāda can be related in some way to the content of the Dvitīyakrama, and it serves as an exceptionally broad window through which we can see into his world—the world of tantric Buddhist north India at the close of the eighth century and the opening of the ninth. In the extraordinary framing narrative of his Dvitīyakrama, Buddhajñānapāda describes the years he spent traveling through India, studying and practicing with several of the important philosophical teachers and tantric gurus of his day, culminating in the vision of his most exalted 1 I discuss the identity of Mañjuśrī as he appears in Buddhajñānapāda’s works in Chapter Two. 2 I explain my departure from the title usually given for this work, the Dvikramatattvabhāvanā-mukhāgama, in Chapter Two, and also in note 3 of my translation of the Dvitīyakrama. ix guru, Mañjuśrī himself, who gave Buddhajñānapāda the instructions that form the core of his system of tantric theory and practice. In claiming such visionary inspiration for his writings and directly recording Mañjuśrī’s words ensconced within the narrative of his own personal autobiography, Buddhajñānapāda places the Dvitīyakrama at an unusual juncture between scripture and authored treatise. This unique position simultaneously gives the work the authority of scripture and yet links its contents, in a very personal way, with Buddhajñānapāda himself. His use of autobiography is unique within Indian Buddhist writings—apart from the Dvitīyakrama and a composition by one of Buddhajñānapāda’s disciples, Śākyamitra, which it clearly inspired, I know of no other autobiographical narratives in the whole of Indian Buddhist literature. The practice of Buddhist tantra, and indeed arguably of Buddhism as a whole, is about reenvisioning identity. Buddhajñānapāda’s writings participate in this process in a number of ways. His doctrinal reflections on the nature of both reality and personal identity, as well as his ritual and liturgical writings stem, our author himself tells us, from his personal experience of realization, a transformation effected through tantric Buddhist practice: “Due to abiding within the profound and luminous maṇḍala, I remain in unceasing nirvāṇa,”3 he boldly states in the Muktitilaka. Through his teachings and writings, he attempted to bring about such a state of realization in his disciples, as well. Thus Buddhajñānapāda’s works at once describe the necessary doctrinal knowledge and ritual means for the re-construal of identity that constitutes tantric Buddhist liberation, but they also—particularly his autobiographical narrative and statements—serve as his personal articulation of what it means to embody that awakening. The period in which Buddhajñānapāda was writing was one of great creativity and development within tantric Buddhism, as new kinds of yogic practices were emerging and being incorporated into the tantric yogin’s repertoire, and yet this period remains not well studied or understood by scholars of Indian tantric Buddhist history.4 Buddhajñānapāda’s writings provide us with an invaluable window into this important period. As we will see, much of what developed, both doctrinally and ritually, in this period continued to have a tremendous impact on later tantric Buddhist traditions, and further studies of works and authors from this period are essential to our coming to a better understanding of Indian Buddhist tantra. This dissertation contributes to an improved appreciation and understanding of the early development of the “mature” form of Indian tantric Buddhism that took place in this period. Previous Scholarship and Place of this Study in the Field The present study of the life, writings, and thought of Buddhajñānapāda will add to a number of recent studies on Indian tantric Buddhism. While the field of tantric Buddhist studies is still in its early stages, the past fifteen years have seen a significant upsurge in the number of scholars writing on Indian Buddhist tantra and the publication of translations and studies of Indic tantric Buddhist materials. Many of the book-length scholarly writings on Indian Buddhist tantra have taken the form of investigations into a single tantra or tantric system.5 More recently, 3 des na zab gsal dkyil ‘khor bas// nga ni rtag tu mya ngan ‘das// (Muktitilaka, D 47a.7-47b.1; P 57b.5-58a.1) The Dpe sdur ma edition (962) reads da, rather than nga in the beginning of the second line here but the Derge (47b.1), Peking (56b.8) and Vaidyapāda (Muktitilaka-vyākhyāna, D 49a.7) clearly read nga. Vaidyapāda’s commentary makes the point even more clear, adding “I and others....” (bdag sogs) (Muktitilaka-vyākhyāna, D 49a.7). 4 J. Dalton (2004), for example, has referred to the period from 750 to 850 CE as an “ill-defined” “intermediate period” in Indian tantric Buddhist history. 5 See Snellgrove (1959) on the Hevajra-tantra; Beyer (1973) on the practices of Tārā; Fremantle (1971), Wayman (1977), and Matsunaga (1978) on the Guhyasamāja-tantra; Skorupski (1983) on the Sarvadurgatipariśodhanatantra; Tsuda (1974) on the Saṃvarodaya Tantra, Dorje (1987) and Garson (2004) on the Guhyagarbha-tantra; Davidson (1982 and 1995) on the Mañjuśrīnāmasaṃgīti; Hodge (1995 and 2003) on the x others have focused on the translation or study of a single tantric commentary or practice manual.6 Only a few scholars have made a book-length study of a single topic within Indian tantric Buddhism7 or attempted to give a broader picture of tantric Buddhism as a whole.8 Several recent works have addressed tantric Buddhism from a topical perspective, examining questions of tantric Buddhist apologetics (Onians 2003) and hermeneutics (Campbell 2009; Kittay 2011; Wedemeyer 2013). Finally, Ronald Davidson (2002) has recently made some important initial steps towards outlining a social history of Buddhist tantra while Christian Wedemeyer (2006; 2013) has offered some helpful critiques of Davidson’s work, further refining our understanding of this topic.9 Prior to this dissertation, however, no systematic book-length scholarly study of a single Indian tantric Buddhist author has been undertaken.10 Moreover, despite his important role in the development of Buddhist tantra and the fact that a number of his writings are extant (most of them in Tibetan translation rather than their original Sanskrit) not a single one of Buddhajñānapāda’s eleven surviving works has been fully edited or translated into any modern language (though editions of short parts of two of his works have been published),11 and only four later Indian Jñānapāda School texts have received scholarly attention. The scholarly work on later Indian Jñānapāda School texts includes three editions of the Guhyasamājamaṇḍalavidhi, an important Jñanapāda School maṇḍala ritual composed by Buddhajñānapāda’s direct disciple Dīpaṃkarabhadra;12 two editions—one of Mahāvairocanābhisaṃbodhi-tantra; Wallis (2002) on the Mañjuśriyamūlakalpa; English (2002) on the tantric practices of Vajrayoginī; Kwon (2002) and Weinberger (2003) on the Sarvatathāgatatattvasaṃgraha; Wallace (2004 and 2010) on the Kālacakra-tantra; Gray (2007) on the Cakrasaṃvara-tantra; Szántó (2012) on the Caṭuṣpītha; Damron (2014) on the Mahāmāyā-tantra. 6 See Sferra (1990) on the Ṣaḍaṅgayoga; Skorupski (2002) on the Kriyāsaṃgraha; Onians (2003) on the Abhiṣekanirukti; Tomabechi (2006) on the Pañcakrama; Wedemeyer (2007) on the Caryāmelāpakapradīpa; Campbell (2009) on the Pradīpoddyotana; Wright (2010) on the Piṇḍīkṛta; Mori (2009) on the Vajrāvalī; Kittay (2011) on the Vajramālā; Klein-Schwind (2012) on the Daśatattvasaṃgraha; Isaacson and Sferra (2014) on the Sekanirdeśapañjikā; Tribe (2016) on the Nāmamantrārthāvalokinī; and Szántó and Mallinson (forthcoming) on the Amṛtasiddhi. 7 See Sakurai (1996 (in Japanese)) on tantric initiation and Tanaka (1996 (in Japanese)) on the maṇḍala. 8 See Snellgrove (1987), and Tribe (2002). Snellgrove’s several chapters on Buddhist Tantra are the most extensive attempt to present and overview of the topic. Tribe’s is one chapter in a larger book on Buddhist thought in India, but attempts to give a comprehensive introduction to and overview of Indian Buddhist tantra. 9 Péter Szántó will be soon concluding a five-year postdoctoral fellowship at Oxford with the purpose of pursuing further research in the social history of tantric Buddhism. 10 The hagiographies of the Indian tantric Buddhist masters Tilopa (Torricelli and Naga 1995) and Naropa (Guenther 1963) have been translated from Tibetan hagiographical sources, but this is a very different project from the systematic scholarly study of those figures based on Indic sources. 11 The surviving Sanskrit verses of Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana have, however, been edited in a number of publications by Tanaka and Kano (see note 13 below) and more still have been reconstructed, or partially reconstructed, in Péter Szántó’s draft edition of the Sāramañjarī, Samantabhadra’s commentary on Buddhajñānapāda’s sādhana. Sakurai (1996) has edited verses 85-125 of the Dvitīyakrama in an appendix to his book (in Japanese) on tantric initiation. 12 The three editions—none of which is accompanied by a study of the text—are, respectively, an edition prepared by Sabine Klein-Schwind, proof-read and revised by Harunaga Isaacson and circulated online (http://www.tantricstudies. uni-hamburg.de/e-texts/bauddha/GuSaMaVi.txt), one published in the journal Dhīḥ (2006), and, most recently, an edition edited by Bahulkar (2010). Szántó 2015 includes in a footnote a diplomatic edition of the final verses of the Sanskrit text as found in a newly discovered Cambridge manuscript of the text. These verses were missing (or reconstructed from the Tibetan) in the earlier editions which were based on the Göttingen manuscript, previously thought to be the only surviving recension, and which lacks the final folio of the text. Several scholars are currently working on this important composition. Ryūta Kikuya is in the process of bringing to publication his study (in Japanese) of the pūrvasevā section of the maṇḍalavidhi (Kikuya, personal communication, March 2014), and Harunaga Isaacson’s student, Daisy Cheung, is planning to conduct her doctoral dissertation research on the text (Isaacson, personal communication, March 2014). xi which also includes an introduction and a translation—of a fragmentary commentary of Samantabhadra’s Sāramañjarī, a commentary on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana;13 a study, including an edition and English translation, of Kṣitigarbha’s Daśatattavasaṃgraha, an eleventh-century Indic text pertaining to the Jñānapāda School;14 and a brief study and Sanskrit edition of the Mañjuvajramukhyākhyāna, a later Nepalese ritual manual based on the Jñānapāda School maṇḍala, but which also incorporates elements from Ārya School ritual texts.15 Buddhajñānapāda’s tradition and his works have been the subject of several articles (and received mention in several books) by Japanese scholars headed by Hakuyu Hadano, Kimiaki Tanaka, and Ryūta Kikuya, mainly in Japanese.16 Buddhajñānapāda has received brief mentioned in a number of Western-language publications,17 and more recently in two short Englishlanguage articles about his life and tradition,18 but prior to the present dissertation there have been no book-length studies of this important figure, his writings, or his tradition. In sum, the research that has thus far been done on Buddhajñānapāda and his writings has focused primarily on 1. establishing his dates and identifying his works and those of his main disciples, 2. publishing the fragmentary Sanskrit text of his important Caturaṅga/Samantabhadra-sādhana and identifying the few extant Sanskrit sources for his other writings or excerpts thereof, and 3. describing, in brief, some of the ritual structures found in his tantric works and in the works of other Indian authors of his tradition. No edition or comprehensive analysis of any single work by Buddhajñānapāda, let alone of his oeuvre as a whole, has yet been attempted, nor has any scholarly attention at all been paid to his doctrinal positions or his philosophical writings.19 Contents of this Dissertation 13 The editions of the Sāramañjarī are by Kimiaki Tanaka (2017), whose recent publication also includes an introduction and an English translation of the fragment of the commentary, and Péter-Dániel Szántó (unpublished), who has made an edition of a much longer recension of the same, still fragmentary, commentary, which he has very generously shared with me. The great advantage of the manuscript that Szántó’s edition is based upon is that it also contains many extensive quotations from other works, including Buddhajñānapāda’s Ātmasādhanāvatāra, which is otherwise not available to scholars in Sanskrit. 14 Klein-Schwind 2012. Klein-Schwind’s dissertation is important in its highlighting and unpacking of a number of ritual details related to the Jñānapāda tradition and is therefore an especially important resource for studying Buddhajñānapāda’s ritual writings. 15 Tanaka 2018. 16 The Japanese scholar Hakuyu Hadano (1950 and 1951) researched the dates and context of Buddhajñānapāda and outlined his writings. More recently Kimiaki Tanaka (1991, 1995, 1996, 2007, 2010, and 2017) has published some Sanskrit verses from Buddhajñānapāda’s Samantabhadra-sādhana and a brief analysis of the maṇḍala and some of the ritual structures from that sādhana. Munenobu Sakurai (2007 and 2009) has written two short articles on the ritual structures of the sbyor ba bdun and the pañcākarābhisaṃbodhi as handed down in the Jñānapāda School. Ryūta Kikuya (2000, 2000b, 2008, 2010, 2012, 2012b) has written a number of short articles on other ritual structures, including the ṣaḍaṅgayoga, the two stages of tantric practice (utpattikrama and utpannakrama), and the three bindu yoga in Buddhajñānapāda’s writings and his tradition. Kazuo Kano (2014) has recently published a Sanskrit edition of verses 19-55 of the Caturaṅga/Samantabhadra-sādhana. 17 The most substantial of these brief references are in Davidson 2002: 311-15, J. Dalton 2004: 13, Tomabechi 2008: 173-74, and Sanderson 2009: 93. 18 Szántó 2015 and C. Dalton and Szántó forthcoming. I also delivered a conference paper on the relationship of Buddhajñānapāda’s practice system with later tantric literature (C. Dalton 2014). 19 Tanaka (2018, 29) mentions a three-page 1985 article by Chizuko Yoshimizu that “argues the Jñānapāda school from the side of philosophy for the first time.” According to Tanaka’s bibliography, the article is titled “On the Yogācāra-Mādhyamika Theory in the Jñānapāda School,” but as the article is in Japanese I have been unable to consult it, and I am unsure whether it focuses on Buddhajñānapāda’s own writings or those of later authors in his tradition. xii This dissertation is a first attempt at beginning to fill this lacuna, and entails a two-fold approach to the material, both analytical and textual. Part I of the dissertation is a study, in which I introduce and give an overview of Buddhajñānapāda’s life and work, and of both the doctrinal positions and ritual systems set forth in his writings. Part II is a Tibetan edition and annotated English translation of his most important work, the Dvitīyakramatattvabhāvanā-mukhāgama. Although the Dvitīyakrama’s focus is—as indicated in its title—on the second stage of tantric practice, this work is so all-encompassing that each of the major topics discussed in the study of his thought is referenced at some point within the Dvitīyakrama, and the text therefore serves as a framework, or an anchor, around which my study of Buddhajñānapāda’s life and thought is arranged. Because so little scholarly research has been done on Buddhajñānapāda’s writings and his thought until now, what I have been able to include in this dissertation is really just an initial step towards coming to understand and appreciate the quite influential work of this unique individual. Part I: Study My study of Buddhajñānapāda’s life and thought is divided into three sections focusing on narrative, doctrine, and ritual. The section on narrative introduces Buddhajñanapāda’s life and writings and discusses the narrative structure of his Dvitīyakrama. Chapter One begins with a translation of Buddhajñanapāda’s own autobiographical narrative from the Dvitīyakrama, supplemented by the commentary on that narrative from his disciple and most prolific commentator, Vaidyapāda. The chapter gives a brief overview of the historical and doctrinal context for understanding Buddhajñānapāda’s life and writings—tantric Buddhism in 8th-century India—and then presents what we know about his life, including identifying some of his gurus and their possible influences on his thought, as well as his important disciples. I then address Buddhajñānapāda’s writings, assessing the question of the authorship of the works attributed to him, and provide a short summary of each of his extant compositions. Chapter Two looks at the Dvitīyakrama, specifically, examining the unique narrative structure of that work and its function in Buddhajñānapāda’s oeuvre. Here I take a look at the identity of Mañjuśrī as the “author” of most of the Dvitīyakrama’s content, and provide a brief overview of the narrative structure and contents of that text. The second major section of the study, on doctrine, takes a look at some of the doctrinal positions set forth in Buddhajñānapāda’s writings. I constrain myself in this section to examining the views set forth in his tantric works, looking at some of the possible influences—both Buddhist and non-Buddhist—on his thinking, and noting what I have found to be some of the most prominent or remarkable doctrinal features of his writings. Chapter Three first considers Buddhajñānapāda’s articulation of nondual wisdom as the nature of all phenomena and of the mind itself, as well as the very source of the phenomenal world. I then examine the structure of the higher tantric path, by means of which nondual wisdom is first “obtained” and then cultivated, as it is set forth in his writings, drawing attention to Buddhajñānapāda’s reformulation of Śākyamuni Buddha’s own awakening narrative to hew to what seem to have been the most important features of that path. I show how Buddhajñānapāda’s writings privilege the tantric path—especially that of the perfection stage—and its result, and finally examine the relationship between ritual and the rhetoric of non-action as found in his writings on the perfection stage. Chapter Four examines the question of the relationship between Buddhajñānapāda’s writings and the early literature of the Great Perfection, with which it shares some significant doctrinal similarities. xiii The third major section of the study focuses on ritual. Here I look at the ritual systems of the generation stage, the perfection stage, and finally tantric initiation according to Buddhajñānapāda’s writings. Parts of these chapters are more descriptive than analytical, primarily because the dearth of prior detailed studies of Buddhajñānapāda’s ritual systems made it necessary to first describe the practices in question before discussing them. Chapter Five begins by looking at the division of tantric practice into two stages, which Buddhajñānapāda was one of the early authors to make. I first examine the distinction between the two stages as we find it articulated in his writings, and then give an overview of the different generation stage sādhanas he composed, before focusing on the most important of these, his Caturaṅga/Samantabhadra-sādhana, and some of its unique features. Chapter Six, which focuses on the perfection stage practices in Buddhajñānapāda’s writings, first examines the way the term “the perfection stage,” and its synonym “the second stage,” are used in Buddhajñānapāda’s works, and then takes a look at the perfection stage practice of the three bindu yogas described in Buddhajñānapāda’s Dvitīyakrama and his Muktitilaka. In this chapter I also examine several points related to the sexual yogic practices that were an important part of the perfection stage in his system. Chapter Seven looks at the ritual sequences for the higher tantric initiations as they are described in both Buddhajñānapāda’s writings and those of his disciple Vaidyapāda, and shows that initiation in the early Jñānapāda School already included the major elements of tantric initiatory practice found in the mature form of Indian tantric Buddhism. In the Conclusion to Part I, I look at some of the ways in which Buddhjñānapāda’s thought moved into and influenced the later tantric tradition, and offer some concluding reflections about his life, work, and thought. Chapter Eight examines the relationship between Buddhajñānapāda’s writings and the Samājottara, showing the ways in which his thought, and in particular his ritual systems, have moved forward into the later tantric tradition. My concluding reflections bring together the major points of the preceding chapters, and offer some suggestions for further research. Part II: Tibetan Critical Edition and Annotated English Translation of the Dvitīyakrama Part II of the dissertation consists of a Tibetan critical edition and an English translation of the Dvitīyakrama. While this work was originally composed in Sanskrit (and most fortunately we find a scattered few parallel verses in later works that are still extant in Sanskrit), the Dvitīyakrama does not survive in its original language, and I have therefore made an edition and English translation based on the 11th-century Tibetan translation of the text by Kamalaguhya and Lha Yeshe Gyaltsen that is preserved in the Tibetan canon, using all five extant recensions of the Tibetan Tengyur. I was unable to find any extra-canonical witnesses of the work. I did, however, also consider the testimony of the citations of the Dvitīyakrama in Vaidyapāda’s commentary, the Sukusuma, and occasionally selected these over what was found in all available recensions of the root text. While making an edition of a translation rather than of the text in its original language is, of course, not ideal, a careful study of Buddhajñānapāda’s writings, including editing them—even those that are only available to us in Tibetan translation—is an important step in coming to understand his thought. Certainly, there are places where the Tibetan translation gives us an imperfect rendering of the original text, but on the whole the benefits of editing the Tibetan translations of Indic texts that only survive in Tibetan overwhelmingly outweigh the uncertainties and risks of occasional misunderstanding that are inevitable in studying such works without being able to consult the Indic originals. Understandably Buddhologists who are Sanskritists prefer to edit Sanskrit Buddhist texts, and Tibetologists tend to focus on works originally composed in Tibetan, but the fact remains that there are a xiv tremendous number of incredibly valuable Indian Buddhist texts—especially tantric ones— extant only in their Tibetan translations. While it is not the case that scholars completely ignore these works, scholars working in the West, at least, generally tend not to edit them.20 Editing the Tibetan translations of Indic works, though, entails approaching these texts very closely, thus allowing for a more precise study of their contents and, as in the case of the Dvitīyakrama, can mean opening up important aspects of the history of Indian tantric Buddhism that would otherwise remain obscure. For me, the process of editing the Dvitīyakrama was invaluable in coming to a better understanding of Buddhajñānapāda’s text. I hope that this edition of the Tibetan translation of the Dvitīyakrama will encourage and inspire scholars, despite the challenges of such an endeavor, to edit and carefully study more of the treasure trove of Indic Buddhist writings that survive only in Tibetan translation. The annotated English translation of the Dvitīyakrama that follows the edition in Part II was made to be able to stand alone. While this allows the interested reader to access the translation without reading my study of Buddhajñānapāda’s works, it also means that some of the notes therein repeat points that are made in the study of Buddhajñānapāda’s thought in Part I. There are also, however, many points taken up in the notes to the translation that are not examined in further detail in the study. Many of the annotations include short translations (with a basic edition of the translated passages, based on the Derge and Peking recensions of the work) from Vaidyapāda’s Sukusuma, a commentary on the Dvitīyakrama that I found indispensable in coming to understand Buddhajñānapāda’s incredible, but sometimes rather opaque text. VTao (talk) Buddhajñānapāda lived, practiced, and wrote during a period of great creativity and growth within the Indian tantric Buddhist tradition, but one that is nonetheless still in the process of emerging into the light of historical study and analysis. His writings provide an important window into this period, and it is my hope that both the study of his life and thought, and the edition and annotated translation of his Dvitīyakrama offered here will serve as a helpful contribution to shedding light on the historical development of tantric Buddhist traditions in India. It has been my great pleasure over the past years to venture a bit into Buddhajñānapāda’s world. I now invite you, the reader, to join me for some steps on that journey. 20 There are certainly exceptions; for example, Vogel 1965. More typically, however, Tibetan editions of Indic Buddhist texts are made by scholars who are also producing a Sanskrit edition of the same text; for example, Wedemeyer 2009 and Isaacson and Sferra 2014. 1 Part I: A Study of Buddhajñānapāda’s Life, Works, and Thought 2 Narrative 3 Chapter One Meeting Mañjuśrī: Buddhajñānapāda’s Life and Works The great bodhisattva Mañjuśrī looked upon me with a smiling face and said “Excellent!” three times. With this vajra song, like an echo, he taught to me the playful dance and the suchness of all phenomena. -Buddhajñānapāda, Dvitīyakrama A Life in Context 1. Buddhajñānapāda’s Autobiographical Account with Vaidyapāda’s Commentary1 In a town called Takṣaśilā,2 in the area of Khapir,3 in the land of Magadha, I pleased the guru Haribhadra,4 who had attained great fame. I received his instruction and studied many scriptures I investigated those and derived understanding.5 |3| Then, in order to inspire faith in beings the revered master speaks about the story of his own encounter with suchness with the verse beginning with, In a town called.... Our great guru first [went to] the region of Magadha, which is in the direction of Nālandā. The area is called Khapir. In that area there is a town called Takṣaśilā. In one part [of that town] there lived the very famous Haribhadra, a master who was respected by many noble beings and who had fatigued [himself] with [the study of] many different sections of the scriptures. Among these, he had received the Prajñāpāramitā and others. [[[Buddhajñānapāda]]] served him, received instruction on the Prajñāpāramitā, and studied many other scriptures [under his tutelage]. He investigated [all of] this using logic.6 1 What follows is the autobiographical narration, in verse, of his life extracted from Buddhajñānapāda’s Dvitīyakrama, which is translated in full in Part II of this dissertation. I have here interspersed Buddhajñānapāda’s verses with excerpts from his 9th-century Indian disciple Vaidyapāda’s prose commentary, the Sukusuma, on these specific verses. A portion of Vaidyapāda’s commentary on these verses has been translated in Davidson (2002, 311- 13). My reading of Vaidyapāda parts ways with Davidson’s translation in a number of places. Vaidyapāda’s commentary to the autobiographical sections of the Dvitīyakrama is the only Indian account of Buddhajñānapāda’s life that survives; all other extant accounts were composed by Tibetan authors several centuries or more after Buddhajñānapāda’s passing, and are all clearly based on Vaidyapāda (and, over time, each other). I address some of these later accounts in this chapter and in the notes to this translation. 2 I address all of the toponyms in Buddhajñānapāda’s account and Vaidyapāda’s commentary, along with their possible identifications, below in my discussion of his life, so I will not address them in the notes here. 3 kha pir] D C S V(P), kha bir P N V(D). As in the notes to my translation of the Dvitīyakrama in Part II of the dissertation I only record in the notes to the root verses here significant variants that affected my translation choices. For the full critical apparatus of the root verses see the edition of the Dvitīyakrama, also in Part II. 4 Haribhadra’s name is here and in Vaidyapāda’s Sukusuma given as Bzang po seng ge, rather than the more common Seng ge bzang po. There is little doubt, however, about his identity, as Vaidyapāda explains that Buddhajñānapāda studied Prajñāpāramitā with this guru, a well known Prajñāpāramitā scholar, and Buddhajñānapāda himself wrote Prajñāpāramitā works. Later Tibetan histories also corroborate that this guru is Haribhadra. 5 rig ‘byung. I remain slightly unsure about this line. Vaidyapāda’s commentary does not address this phrase, and concludes his comments on Buddhajñānapāda’s studies with a gloss of the term rnam dpyad, “I investigated.” 6 da ni ‘gro ba rnams dad par bya ba’i phyir/ rje btsun bdag nyid kyis de kho na nyid mnyes pa’i lo rgyus gsungs pa/ dbus kyi yul chen zhes pa la sogs pa’o// de la bdag cag gi bla ma chen pos dang por yul khams ni dbus zhes bya 4 At Śrī Nālandā, in response to the one of noble birth called *Guṇamitrā, With a [still] ignorant mind7 I composed some treatises joyfully. Thinking to benefit those who live there with those treatises, [While] I stayed there, I composed and taught. |4| At the great Buddhist Institute of Nālandā there lived a bhikṣunī called *Guṇamitrā, who was brahmin by birth and had stable faith. At her request [[[Buddhajñānapāda]]] composed several texts. [He writes that he composed these] with a [still] ignorant mind because, although his mind was engaged in the Prajñāpāramitā, he had not yet realized suchness exactly as it is. [The texts he wrote there] were a few [compositions] including a summary of the Prajñāpāramitā,8 which he wrote and taught joyfully. Those Prajñāpāramitā texts were intended to benefit [those who received them].9 Then I travelled to the land of Uḍḍiyāna, the source of all positive qualities, [Where there lives] someone known as Vilāsavajra10 From him I learned much and investigated, as well. And also, in that same place, I pleased a guru called Guṇeru11 |5| Then, two hundred and thirty yojanas to the north of Magadha is a place that is called The Source of All Qualities because most of those who have been blessed by the ḍākiṇīs come from there. It is [also] called Uḍḍiyāna. [[[Buddhajñānapāda]]] traveled to that place. The great ācārya Vilāsavajra, born in a part of Uḍḍiyāna called Ratnadvīpa, who had accomplished the mahāmudrā and was also known as Śrī Viśvarūpa,12 lived there. With him [[[Buddhajñānapāda]]] studied many Kriyā and Yoga tantras. He also put great effort into examining them. Also, in an[other] area of that very same sacred land [of Uḍḍiyāna] there lived the great yoginī who had ste/ na lendra’i phyogs so// yul gyi ming ni kha pir (pir] P, bir D) zhes bya‘o de ni grong khyer rdo ‘jog ces bya ba yod do// de yi phyogs gcig na ‘phags pa mang pos bkur ba’i sde pa gzhung sna tshogs kyis dub pa las/ shes rab kyi pha rol tu phyin pa la sogs pa thob pa/ bzang po seng ge zhes bya bar grags par rab tu thob pa zhig yod pa de mnyes par byas nas pha rol tu phyin pa’i lung thob nas de las gzhan pa’i gzhung yang mang du thos pas de nyid la ‘thad bsgrub kyis rnam par dpyad (dpyad] D, spyad P) cing gnas pa la (Sukusuma, D 89a.7-89b.2, P 107a.3-7). 7 blun blos. 8 The Prajñāpāramitā summary likely refers to the Sañcayagāthā-pañjikā (Tōh. 3798), a Prajñāpāramitā commentary, which does seem likely to have been a composition that Buddhajñānapāda wrote early in his career, and whose colophon mentions *Guṇamitrā as the petitioner (Sañcayagāthā-pañjikā, D 189a.5). 9śrī (śrī] D, śri P) na le ndrar rig pa’i chos sgwra chen po na gnas pa’i dge slong ma yon tan bshas gnyen zhes bya ba (ba] P,‘ D)/ bram ze’i rigs su skyes pa dad pa brten ba zhig yod pa de’i ngor/ blun blos zhes te pha rol tu phyin pa’i blos gnas pas de bzhin nyid ji lta ba bzhin du ma rtogs pa’i phyir ro// rab tu byed pa phyogs ‘ga’ rtsom byed pa’i zhes pa ni shes rab kyi pha rol tu phyin pa’i bsdus don la sogs pa nyung shas cig la blo ‘phrod (‘phrod] D, brod P) pas rtsom pa dang ston pa byas te/ pha rol tu phyin pa’i gzhung des phan gdags par bya ba’i phyir ro// (Sukusuma, D 89b.2-89b.4, P 107a.7-107b.1). 10 Jo sgeg rdo rje. The name Vilāsavajra is usually rendered into Tibetan as Sgeg pa’i rdo rje, but ‘jo sgeg is a synonym for sgeg pa, so the identification here seems rather certain, and is again corroborated by later Tibetan histories. 11 gu ne ru] S P V (D and P), gu ne nu D C N. 12 In the colophon to the Sanskrit text of Vilāsavajra’s commentary on the Mañjuśrīnāmasaṃgīti, it is stated that the author was also known as Śrī Viśvarūpa, and that he lived in Ratnadvīpa, exactly as Vaidyapāda here states (see Tribe 2016, 26). 5 encountered suchness called Guṇeru,13 who had received the instructions on the stages of the inconceivable.14 [[[Buddhajñānapāda]]] went to see [this master] and attended her.15 And received teachings from her.16 At the northern gate17 of that place I pleased a girl of sixteen years named Jātig Jālā,18 Mahālakṣmī. For eight months I took her instruction, and having received it, I achieved accomplishment. |6| 13 I believe the grammar of the passage clearly indicates that Guṇeru is herself the great yoginī, in which assessment it seems I am preceded by Gö Lotsāwa and Tāranātha (Deb ther sngon po, 447; Roerich 1976, 367; Bka’ babs bdun 103; Templeman 1983, 7; See note 15 for the Tibetan of the passage). Chögyal Phagpa, however, identifies Guṇeru as a yogin (rnal ‘byor pa) rather than a yoginī (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 611). 14 Bsam kyis mi khyab pa’i rim pa’i man ngag. It is unclear whether this is meant to refer to the title of a text or not. A text of precisely this title is extant in the Tengyur (*Acintyakramopadeśa, Bsam kyis mi khyab pa’i rim pa’i man ngag Tōh. 2228), where it is attributed to one *Kuddālīpāda (tog rtse zhabs). The same work survives in a second Tibetan translation, apparently of a slightly different recension of the Sanskrit text, in a compendium of Sakyapa works; within the Sakyapa tradition the work is understood to represent one among a series of eight subsidiary instructions connected to the Lamdre (lam ‘bras) root text (Davidson 2005, 194-95). The *Acintyakramopadeśa is also considered, in the Tibetan tradition, oneamong a set of six Indian mahāmudrā works called the Sixfold Corpus on the Essence (Snying po skor drug) (Krug 2018, 328-9). The Sanskrit of the work, under the title Acintyādvayakramopadeśa, survives and has been edited (Samdhong and Dwivedi). I have not had the opportunity to compare this against the Tibetan translations and am unaware of any such comparison having been reported in modern scholarship. (Thanks to Harunaga Isaacson for first drawing my attention to the existence of this Sanskrit edition.) Regarding its author and period of composition, in his History of Buddhism in India Tāranātha mentions a

• Kaudālika/Mahā-koṭali (tog rtse ba che ba), who Chimpa and Chattopadhyaya have taken to refer to the same

figure as Kuddālipāda, and who Tāranātha says lived during the reign of King Gopāla, the Pāla king who reigned prior to Devapāla and Dharmapāla, who ruled when Buddhajñānapāda composed his writings (Chimpa and Chattopadhyaya, 262). However, the work as it survives at present focuses on perfection stage practices and was understood, at least by the 15th-century Tibetan scholar Ngorchen, to be based on the Sampuṭa-tantra (though Davidson (2005, 196) notes that the connection is “only indirect” and Isaacson (personal communication) has also expressed some doubt as to the connection with the Sampuṭa) (See also Davidson 2005, 195-96; Stearns 2006, 135). It is questionable whether this text is early enough to be the referent here in Vaidyapāda’s commentary. Krug (2018, 341) identifies the work as focused on the generation and perfection stage yogas of the Yoginī tantras. Apart from its content, further clues to the period of the author may be found in a lineage list given in the work itself, culminating in the author’s own guru, who he styles Bhadrapāda (Krug 2018, 335-6). For now, whether or not Vaidyapāda is referencing this particular work must remain a question. It is possible, as well, that the text as written down was meant to preserve a tradition of oral instructions that had not yet been previously recorded, and that such a set of oral instructions (which of course would have been supplemented over time) could conceivably be Vaidyapāda’s intended referent here (Harunaga Isaacson, personal communication). 15 de nas yul dbus las byang du dpag tshad nyis brgya sum cu’i phyogs na yon tan kun ‘byung zhes te mkha’ ‘gro mas byin gyis rlob pa phal cher de las ‘byung ba’i phyir/ u rgyan gyi yul la bya’o// der bgrod nas de’i phyogs gcig nor bu’i gling du sku ‘khrungs pa’i phyag rgya chen po thob pa/ dpal sna tshogs gzugs zhes kyang grags pa/ slob dpon chen po ‘jo sgeg rdo rje zhes bya ba yod de/ de la bya ba dang rnal ‘byor gyi rgyud mang du thos nas/ ‘bad pas rnam par dpyad pa yang byas so// yang gnas de nyid kyi phyogs gcig na bsam kyis mi khyab pa’i rim pa’i man ngag thob pa/ rnal ‘byor ma chen mo de nyid brnyes pa gu ne ru zhes bya ba zhig gnas pa de’i drung du bgrod nas de mnyes par byas te/ (Sukusuma, D 89b.4-7, P 107b.1-107b.5). 16 While Buddhajñānapāda does not specify her gender, and the unusual name gives no indication of gender, either, I have followed Vaidyapāda’s identification of this guru as a great yoginī. 17 byang phyogs chab sgo. See C. Dalton and Szántó (forthcoming) for a differing reading of this term where the term chab sgo is interpreted as a proper noun. I prefer to read it here as simply “gate,” especially given Vaidyapāda’s reading, which includes some grammatical particles omitted for metrical reasons in the Dvitīyakrama itself. Vaidyapāda reads: u rgyan gyi gnas de yi byang phyogs kyi chab sgo na/ (Sukusuma, D D 89b.7; P 107b.6). 18 Dzā (dzā] D C V (D), dza S P N, ‘dza’ V (P)) thig dzā (dzā] sugg. em. based on V (D); dza D C S P N, dzva V (P)) lā (lā] sugg em; la D C S P N). 6 He studied many [[[Yoga]]]niruttara19 tantras [with this master] and, having received this guru’s instructions, including samayas and initiations, he immediately put them into practice. Then he had a dream in which a deity told him, “At the northern gate of the sacred land of Uḍḍiyāna there is a sixteen-year-old outcaste girl called Jātig Jālā. She is [actually] the high-born yoginī Mahālakṣmī. Go there and you will achieve your aims.” So [[[Buddhajñānapāda]]] went there and befriended her, and then served her for eight months. Realizing that [[[Buddhajñānapāda]]] had an interest in the mahāmudrā she bestowed upon him some subtle instructions, and simply by [receiving them Buddhajñānapāda] attained the accomplishment of Jambhala.20 21 Then I went to the village of Ko no dze22 in the area of Jālandhara23 And met Bālipāda,24 who had attained great renown. Having pleased him, I studied the scriptures and received many instructions. Then I went to “the place with sky trees” in the Koṅkana, to the south. |7| There is a village called Ko no dze in the area called Jālandhara. In one part [of that village] there lived one called Bālipāda whose understanding of the tantras that emphasize wisdom was like a river. [[[Buddhajñānapāda]]] went before that master, served him, studied scriptures and received instructions, and then arduously applied himself to training [in these]. Three hundred yojanas to the south of Madhya there is an area called the Koṅkana. There is a place there called The Place with Sky Trees. Why is it called that? Because it is a place where the trees appear to lack roots and [yet] coil and spread upward.25 19 Vaidyapāda here writes only Niruttara tantras (bla med rgyud) but given that he has in the earlier passage referred to Yoga tantras and in a later passage of the Sukusuma (D 108a.6-108b.1) he explicitly distinguishes between Yoga tantras (rnal ‘byor rgyud) and Yoganiruttara tantras (rnal ‘byor bla na med pa’i rgyud) (the latter of which he equates in that passage with Dākiṇī tantras (mkha’ ‘gro ma’i rgyud)), I believe it is likely that Yoganiruttara tantras is what is intended here. 20 I am not completely certain of the meaning of this passage. des kyang phyag rgya chen po la brod pa yod (sugg. em., yin D, P) par rtogs nas/ phra mo’i lung stsal ba tsam gyis dpal jaṃ bha (dzaṃ bha] D, ‘dza mbha P) la’i grub pa thob par gyur to// (Sukusuma, D 90a.2). Gö Lotsāwa rephrases the line from Vaidyapāda in a way that supports my reading (des kyang slob dpon phyag rgya chen po la brod par rtogs na) (Deb ther sngon po, 447). Later in the Dvitīyakrama Buddhajñānapāda himself mentions receiving provisions from Jambhala, and he is also credited with composing three Jambhala sādhanas. 21 bla na med pa’i rgyud mang du thos par byas nas ji skad du gsungs pa’i dam tshig dang dbang la sogs pas bla ma de’i lung thob nas gzod bsgom pa la zhugs so// lhas rmi lam du bstan pa u rgyan gyi gnas de yi byang phogs kyi chab sgo na/ gdol pa’i rigs dzā (dzā] D, ‘dza’ P) thig dzā (dzā] D, dzva P) la zhes bya ba la bu mo lo bcu drug lon pa zhig yod kyis/ de ni rigs las skyes pa’i rnal ‘byor ma la kṣmī chen mo zhes bya ba yin gyis der song dang khyod kyi dgongs pa ‘grub par ‘gyur ro zhes pa dang/ ‘phral du song nas de dang bshes su ‘thams nas/ zla ba brgyad kyi bar du mnyes par byas so// des kyang phyag rgya chen po la brod pa yod (yod] sugg. em., yin D, P) par rtogs nas/ phra mo’i lung stsal ba tsam gyis dpal dzaṃ bha (dzaṃ bha] D, ‘dza mbha P) la’i grub pa thob par gyur to// (Sukusuma, D 89b.7-90a.2, P 107b.5-8) 22 Ko no dze] D C S P N V(P), ka no dze V(D). 23 dzā lendha] D C, dzā lāndha S P N 24 bā li pā da] D C, ‘ba’ mo pa ta S P N. Vaidyapāda’s commentary has the name translated as byis pa chung ba’i zhabs which supports the reading from D and C (Sukusuma, D 90a.2). Szántó reconstructs the name as Bālikapāda and suggests that the name may even read Bālhikapāda as reflective of a master from the area of Balkh (Szántó 2015, 542; see also C. Dalton and Szántó, forthcoming). 25 de nas yul dzā lendha ra (dzā lendha ra] D, dza len tha ra P) zhes bya ba na grong khyer ka no dzer (ka no dzer] D, ko no dzer P) zhes bya ba yod de/ de’i phyogs gcig na shes rab gtso bor byed pa’i rgyud la bsam pa chu klung lta bur gyur pa byis pa chung ba’i zhabs zhes bya ba’i drung du bgrod nas/ de mnyes par byas nas de’i gzhung thos shing lung yang thob par byas te bsgom pa yang nan tan du byed do// de nas yul dbus nas lho phyogs su dpag tshad sum brgya yod pa na yul ko ngka na (ko ngka na] D, kong ka na P) zhes bya ba yod de/ de la nam mkha’i shing ldan zhes bya ste/ ci’i phyir zhe na/ rtsa ba med par shing rnams ‘khril (‘khril] P, la‘khris D) shing steng du bras (bras] P, bris D) pa lta bur gnas pa’o// (Sukusuma, D 90a.2-4, P 107b.8-108a.3). 7 [There] the lord of siddhas, renowned as Pālitapāda26 Was surrounded by his disciples who could perform miraculous feats. All of them regularly received requisites, clothing, food, and wealth. I bowed at the feet of this sublime guru for nine years. |8| In one area of that place there lived the lord of siddhas, named Pālitapāda, whose understanding of the tantras emphasizing method was like a great river, and who was held back [from achieving awakening] by only a single lifetime. He was surrounded by a retinue of disciples who could perform miracles. Who were they? They were the brahmin Catrara,27 the brahmin Guhyaparta,28 the kṣatriya Mañjuśrī, and the vaiśya Pūrṇabhadra, the śūdra Dīpaṃkara, the śūdra Karṇaputra, the prostitute Ālokī, and the prostitute Sādhuśīlā. All of their requisites [[[including]]] clothing and food were provided by the Goddess Vasudharā29 who provided them daily with ten māṣa of gold, half a string of pearls, and three hundred kārṣāpana.30 [[[Buddhajñānapāda]]] bowed before this sublime guru for nine years and [himself] became held back [from achieving awakening] by only a single lifetime.31 I listened to the great Samāja-tantra together with its commentaries for eighteen [months].32 [I said] “I have not realized it” and the great guru said the same. Thinking, “Until I realize this, 33 anything else is useless,” I affixed the volume around my neck and set off to the north. |9| With the verse, I listened to the great Samāja-tantra... the master’s intention is as follows. The Samāja-tantra here means [its] yogas. Its commentary means the butcher girl Vimalamutrī(!?)34 who was trained in [those] yogas. I listened...together...for eighteen means 26 bā li pā da] D C, ba li pa ta S P N. Vaidyapāda (Sukusuma, D 90a.4; P 108a.4) identifies the teacher as bsrung ba’i zhabs. Szántó has recently provided evidence from a Sanskrit manuscript of the Sāramañjarī, a commentary to another of Buddhajñānapāda’s works, that this teacher’s name was Pālitapāda (Szántó 2015, 542-50; see also C. Dalton and Szántó, forthcoming). 27 This is an unusual name. It could possibly be a corruption of Catura, which has the advantage of at least being attested as a name (Harunaga Isaacson, personal communication). 28 I concur here with Davidson that we might here prefer Guhyapatra. Harunaga Isaacson (personal communication) also suggests a perhaps slightly less likely possibility of Guhyāvarta, though neither of these appears to be attested as a personal name. 29 lha mo nor gyi rgyun zhes bya ba. Gö Lotsāwa (Deb ther sngon po, vol I, 448) normalizes the name to its more common lha mo nor rgyun ma. 30 Monier Williams indicates that a kārṣāpaṇa is a coin that weighs differently depending on the material it is made of. 31 yul de’i phyogs gcig na thabs (thabs] D, thab ma P) gtso bor byed pa’i rgyud la gongs pa chu klung chen po lta bur gyur pa skye ba gcig gis thogs pa grub pa’i dbang phyug bsrung ba’i zhabs zhes bya ba gnas te/ de yang rdzu ‘phrul dang ldan pa’i slob mas bskor ba’o// de dag kyang gang zhe na/ bram ze tsa tra ra zhes bya ba dang/ bram ze gu hya pa rta (hya pa rta] D, ha par ta P) zhes bya ba dang/ rgyal rigs ma ñju śrī zhes bya ba dang/ rje’u rigs pū rṇa bha dra (dra] D, tra P) zhes bya ba dang/ dmangs rigs (rigs] D, ris P) dī paṃ (paṃ] D, baṃ P) ka ra zhes bya ba dang/ dmangs rigs kha rṇa pu tra zhes bya ba dang/ smad ‘tshong ma ā lo ki (ki] D, gi P) zhes bya ba dang/ smad ‘tshong ma’i sa du shi la zhes bya ba ste/ de kun gyi yo byad dang gos zas ni lha mo nor gyi rgyun zhes bya ba des nyi ma re re zhing gser gyi ma sha bcu dang mu tig gi ha ra phyed dang kā rṣa pa ṇa (kā rṣa pa ṇa] D, ka rṣa pa na P) sum brgya sbyor ro// bla ma dam pa de’i drung du lo dgu’i bar (bar] D, par P) du ‘dud cing skye ba gcig gis thogs par byas so// (Sukusuma, D 90a.4-7, P 108a.3-7). 32 bar du mnyan] sugg. em. based on V (D and P), rab tu mnyan D C S P N. 33 ‘di] sugg. em. based on V (D and P), ‘dir D C S P N. 34 D bi ma la mu dri, P bi ma la mu tri. This is an unusual name, indeed. Her name is reported by Tāranātha as Mālamodi (mā la mo di) (Bka’ babs bdun, 104). Perhaps this should be Vimalamurtī, Vimalamudrā, or even

that Buddhajñānapāda practiced [with her] for eighteen months.35 “I have not realized it”

means that the waves of [Buddhajñānapāda’s] realization had not poured forth. When the great guru, as well, said “I have also not realized it,” he was somewhat discouraged. Thinking, “Until I realize this, anything else is useless,” he placed his seal on a volume of the Samājatantra36 and, tying this around his neck, he set off to the north.37

Behind Vajrāsana is the forest called Kuvaca Which is full of tigers and bears—a terrifying place. There I spent six months, and thus realized the suchness of phenomena. I met an emanated monk together with two gurus. |10|

[[[Buddhajñānapāda]]] went to a forest called Kuvaca, which is behind Vajrāsana. [His] intention was as follows: “I remain among sentient beings who turn their gaze away from the Essence of Enlightenment. There are many tigers and bears, and so forth, [which are the manifestations] of desire and the other [[[afflictions]]]; it is a truly terrifying place. Since I want to be free from that, I will remain [here] for six months invoking [the deity?]38 and practicing, by means of which I will realize the suchness of phenomena.” And how did he realize that? [This is explained] in the lines beginning with I met an emanated monk… This monk was an emanation of the Great Vajra Holder. His lower robe was open,39 he had made a turban out of his dharma robe and was plowing a field. And the two gurus were an ugly woman with a small child and a white female dog with markings [on her coat].40 When he met them, since he did not [yet] have waves Vimalamudrī? Or, taking Tāranātha’s reporting into account, Vimalāmodā or Vimalāmodinī? Thanks to Harunaga Isaacson for suggesting these possibilities on what this name may have originally been. 35 Gö Lotsāwa’s periphrasis of Vaidyapāda supports my translation of this phrase (gnas der rnal ‘byor ma rnams dang lhan cig pa’i spyod pa yang zla ba bco brgyad kyi bar du mdzad do//) (Deb ther sngon po, 448; Roerich 1976, 368-9). 36 This is one of the more enigmatic passages in Vaidyapāda’s text, and here I have not followed Gö Lotsāwa’s reading. Vaidyapāda’s commentary reads rang gi phyag rgya ‘dus pa’i glegs bam du byas nas. Gö Lotsāwa (Deb ther sngon po, 448; Roerich 1976, 369) has understood this to mean “he transformed his consort into the form of a volume of text” (rang gi phyag rgya ma glegs bam gyi gzugs su bsgyur te/), and Tāranātha (Bka’ babs bdun, 105) follows suit: “He had there a consort named Mālamodi whom he transformed into a volume of the Samāja[-tantra] and affixed to his neck...” (der mā la mo di zhes bya ba’i phyag rgya zhig yod pa ‘dus pa’i glegs bam du bsgyur te

mgul du btags nas/). (I believe that Templeman (1983, 72) has mistranslated this passage in Tāranātha.) Both readings of Vaidyapāda are grammatically possible, but I am somehow hesitant to translate following Gö’s and Tāranātha’s interpretation of the phrase, in part because a consort does not figure in any later part of the account. 37 ‘dus pa’i rgyud chen zhes pa la sogs pa ni bla ma’i dgongs pa ste/ de la ‘dus pa’i rgyud ni rnal ‘byor rnams so// de’i ‘grel pa ni rnal ‘byor bslabs pa sme (sme] D, dme P) sha can gyi bu mo’i bi ma la mu tri’o (tri’o] P, dri’o D)// de dang bcas par bco brgyad bar du mnyan pa ni zla ba bco brgyad kyi bar du bsgrub pa’o// bdag gis ma rtogs pa zhes pa ni rtogs pa’i dba’ rlabs ma ‘phros pa’o// bla ma chen pos kyang bdag gis kyang ma rtogs zhes gsungs pa dang/ thugs cung zad chad nas ‘di ma rtogs par gzhang ni don med do bsams nas/ rang gi phyag rgya ‘dus pa’i glegs bam (bam] D, baṁ P) du byas nas mgul du btags nas de las byang phyogs su bgrod de/ (Sukusuma, D 90a.7- 90b.2; P 108a.7-108b.2). 38 Gö Lotsāwa definitely takes this to mean invoking the deity, and he specifies that it is done by means of a wrathful ritual (lha drag tu skul ba’i cho ga la brtson pas) (Deb ther sngon po, Vol. I, 448-9; Roerich 1976, 369 has

omitted this detail in his translation.) 39 Here I have emended byi ba’i sham thabs can to bye ba’i sham thabs can following Gö Lotsāwa and Tāranātha

who both have this reading (Deb ther sngon po, Vol, I, 449; Bka’ babs bdun, 104). 40 This account is further embellished in Chögyal Phagpa’s 13th-century version of the encounter, which I have translated below, but already in Vaidyapāda’s telling, nearly everything that could be wrong with this “emanated monk” is already there: he is accompanied by a woman (monks are celibate!) who has a son (monks are celibate!!), is plowing a field (monks are prohibited from tilling the soil and other such farmwork!), and wearing his dharma robe on his head with his lower robe open (monks are to dress in a respectful and seemly fashion!).

of realization the guru [[[Buddhajñānapāda]]] felt no shame in front of them. Then, the monk knowing that [[[Buddhajñānapāda]]] was engaged in the supreme mantra conduct, in order to bring forth his vision [of true reality?],41 emanated the maṇḍala of Mañjuśrī.42 On the eighth day of the seventh month, during [the constellation] Puṣya At the time when Mṛgaśīrṣa and Hasta are fading,43 in the early morning, right at dawn, Towards the emanated maṇḍala-cakra of Mañjuśrī44

I made a fervent supplication to understand the meaning. |11| What was the date [when this happened]? On the eighth day during [the constellation] Puṣya, at the time when Mṛgaśīrṣa and Hasta are fading. What was the month? The seventh month. What was the time? In the early morning, right at dawn. [[[Buddhajñānapāda]]] was asked if he had faith in the emanated maṇḍala or the guru, and when he replied that he had faith in the maṇḍala the monk together with the gurus immediately [left and] entered a small house.45 Then 41 de la spyan ras kyis bca’ ba’i phyir. This line is also puzzling, and I am unsure of the translation. Gö Lotsāwa has paraphrased his understanding quite straightforwardly, “in order to benefit him...” (de la phan pa’i phyir) (Deb ther sngon po, Vol I, 449).

42 rdo rje gdan gyi rgyab na ku ba tsa zhes bya ba’i tshal yod de der phyin pa’o// de yi dgongs pa ni byang chub kyi snying po las kha phyir bltas pa’i sems can rnams kyi nang na bdag gnas te/ de na ‘dod chags la sogs pa’i stag dang dred la sogs pa mang zhing shin tu ‘jigs pa’i sa ste (ste] D, te P) / bdag de las thar par ‘dod pa’i phyir der zla ba drug bskul ba dang bcas pa’i bsgrub pas gnas pas (pas] P, pa’i D) chos rnams kyi de bzhin nyid rtogs so zhes so// ji ltar rtogs she na/ sprul pa’i dge slong zhes pa la sogs pa’o// de yang rdo rje ‘dzin pa chen pos sprul pa’i dge slong bye (bye] sugg. em. based on Deb ther sngon po and Bka’ babs bdun, byi D, P) ba’i sham thabs can chos gos las thod byas pa gcig zhing rmo zhing gnas pa dang/ bla ma gnyis te bu chung dang ldan pa’i bud med ngan pa (pa] D, ma P) dang khyi mo dkar ba mtshan ma can no// de rnams dang phrad pa las rtogs pa’i rlabs (rtogs pa’i rlabs] D, rtog rlabs P) mi mnga’ bas bla mas de rnams la ma khrel to// de nas dge slong gis sngags kyi spyod pa’i mchog la gnas par shes nas/ de la spyan ras kyis btsa’ ba’i phyir/ ‘jam dbyangs kyi dkyil ‘khor sprul lo// (Sukusuma, D 90b.2-5; P 108b.2-7). 43 Puṣya is the eighth lunar mansion in Indian astrology; Mṛgaśīrṣa is the fifth; Hasta is the thirteenth. 44 ‘jam dpal dbyangs kyi (kyi] S P N, kyis D C) dkyil ‘khor ‘khor lo (lo] S P N, lor D C) sprul pa la. 45 Gö Lotsāwa reports the account nearly verbatim from Vaidyapāda, but Roerich has understood it differently and translated it as follows: “(His teacher) asked him: “Do you have faith in the teacher or the maṇḍala?” and he replied: “I have faith in the maṇḍala.” (The maṇḍala then vanished), and he found himself and the teacher staying inside a small house.” (Deb ther sngon po, 449; Blue Annals, 369). This appears to be a misreading of the text on Roerich’s part, as neither Vaidyapāda nor Gö makes any indication that the maṇḍala vanished, nor indeed does Gö report that Buddhajñānapāda entered the small house. The account, in Gö Lotsāwa’s rendering simply states that “He replied that he had faith in the maṇḍala and then the monk together with the two gurus entered into a small house.” (dge slong bla ma gnyis dang bcas pa khang pa chung ngu zhig gi nang du zhugs par gyur to//) (Deb ther sngon po, 449). Because Gö did not earlier follow Vaidyapāda in clarifying that the “two gurus” referred to the woman and the dog, Roerich presumably had not seen the phrase “two gurus” before, and apparently took it to mean Buddhajñānapāda and Mañjuśrīmitra. However, in Vaidyapāda’s account, which Gö has in this section reproduced almost exactly, it was clear from the earlier reference that the two gurus are the woman and the dog and do not include Buddhajñānapāda. Gö goes on to explain that after Buddhajñānapāda made his supplication, the lord of the maṇḍala—and here the term used, dkyil ‘khor gyi gtso bo, more likely refers to a deity rather than a guru—gave him instructions. There is no indication in Gö’s account that the maṇḍala was somehow re-emanated, because he never indicates that it disappeared. In Gö’s account, just as in Vaidyapāda’s, the monk and the woman and dog simply responded to Buddhajñānapāda’s preference for the maṇḍala rather than the guru by leaving and going inside a house, and Buddhajñānapāda then received his instructions directly from Mañjuśrī, the main deity of the maṇḍala. The disappearance and reappearance of the maṇḍala is clearly articulated in the account by Chögyal Phagpa, who reports two versions of the story, the first in which Buddhajñānapāda says he wishes to receive initiation from the maṇḍala and the monk says, “Fine, receive it from the maṇḍala!” and leaves, and another version in which the maṇḍala vanishes after Buddhajñānapāda says he wishes to receive initiation from the deity, upon which Buddhajñānapāda supplicates the monk who then re-emanates the maṇḍala from his heart center at dawn (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 614). I have translated this full episode from Chögyal 10 the great guru [[[Buddhajñānapāda]]] made the following supplication to the maṇḍala of Mañjuśrī in order to [be able to] receive suchness.46 47 Then, the great bodhisattva Mañjuśrī Looked upon me with a smiling face and said, “Excellent” three times. With this vajra song, like an echo, he taught to me The playful dance and the suchness of all phenomena. |19| Then, as an introduction to Mañjuśrī’s speech the great guru said Then... Then means immediately after the supplication. He is called Mañjuśrī (“the gentle voiced one”) [because] he satisfies beings with his gentle and sweet voice, since he is the pure form of the great wisdom of all the buddhas. He is called a bodhisattva because he is integrated with awakening (bodhi), not because awakening is his goal. For that very reason he is called great, and is distinguished from the [[[bodhi]]]sattvas on the ten bhūmis. He looked upon me with a smiling face means he was quite delighted because of having realization of the ultimate state. [The fact that] he said “Excellent” three times indicates that he was pleased by [Buddhajñānapāda’s display of] various modes of conduct that accord with having obtained suchness, by his supplications made with speech that accords with that meaning, and by his having observed everything to be profound and genuinely luminous. The rest was already explained. Like an echo has the sense of being like an echo, which makes a sound but is not truly established. A song that is like a vajra is a vajra song, which is a pleasing song. With the words he taught [this] to me, the great guru makes others feel confident.48 [What follows, amounting to ninety percent of the text of the Dvitīyakrama, are Mañjuśrī’s instructions to Buddhajñānapāda, recorded in Mañjuśrī’s first-person speech, and concluding with a prediction and command given by Mañjuśrī, in which he addresses Buddhajñānapāda directly in the second person. With the conclusion of these teachings, and the dissolution of his vision, Buddhajñānapāda returns to his autobiographical account.] Phagpa’s account below. Amye Zhab gives both versions of the story from Chögyal Phagpa’s account (Gshin rje chos ‘byung, 48a.1-4) and Dudjom reports only the version of the account where the maṇḍala disappears (Dudjom 1991, 494-96). 46 tshes gang zhe na/ mgo dang lag gnyis yol dang tshed brgyad rgyal la bab ces (ces] D, zhes P) so// nam zla gang zhe na/ ston zla ra ba zhes so// dus gang zhe na/ tho rangs skya rengs shar dus su zhes’o// der sprul pa’i dkyil ‘khor dang bla ma la mos pa dris pa dang/ sprul pa’i dkyil ‘khor la mos par bka’ tsal pa dang/ dge slong bla ma dang bcas pa de nyid du khang pa chung du cig gi nang du zhugs so// de nas bla ma chen pos ‘jam pa’i dbyangs kyi dkyil ‘khor la de bzhin nyid blang bar bya pa’i phyir gsol ba ‘di skad du btab bo// (Sukusuma, D 90b.5-7; P 108b.7- 109a.2). 47 I have omitted here the seven verses of Buddhajñānapāda’s supplication to Mañjuśrī, as these verses (v 12-18) do not contain autobiographical content. See Part II for the full translation of the root text. 48 da ni ‘jam pa’i dbyangs kyi gsung la ‘jug pa’i tshig bla ma chen po’i zhal snga nas gsungs pa/ de nas zhes pa la sogs pa’o// de nas zhes pa ni gsol ba btab pa’i de ma thag pa’o// ‘jam dbyangs zhes pa ni ‘jam zhing mnyen pa’i dbyangs kyis ‘gro ba rnams tshim par byed pa ste/ sangs rgyas thams cad kyi shes rab chen po rnam par dag pa’i phyir ro// de nyid byang chub dang ‘dres pa’i phyir byang chub sems dpa’ ste/ byang chub la dmigs pa ni ma yin no// de nyid kyis ni chen po zhes te sa bcu’i sems dpa’ rnams dgar ba’o// ‘dzum pa’i bzhin bltas zhes pa ni shin tu rangs pa ste/ mthar thug pa’i gnas rtogs pa’i phyir ro// legs zhes lan gsum gsungs zhes pa ni/ de bzhin nyid thob pa dang rjes su mthun pa’i spyod pa ji snyed pa dang/ don gyi rjes su ‘brang ba’i gsung ji snyed pas gsol ba ‘debs pa dang/ thams cad zab mo dang yang dag par gsal bar dmigs pa la thugs rangs pa’o// gzhan ni bshad zin to/ /sgra brnyan lta bur zhes pa ni brag ca lta bu ste grags kyang ma grub ces pa’i don to// rdo rje lta bu dang ldan pa’i glu ni rdo rje glu (ni rdo rje glu] D, P om.) ste dga’ bar byed pa’i glu’o// de lta bus bdag la bstan zhes bla ma chen pos gzhan yid brtan par mdzad pa yin no// (Sukusuma, D 93a.1-5; sP 111b.1-7). 11 In this way with the vajra song like an echo, together with the playful dance And the [[[maṇḍala]]-]cakra, right then49 he sang and praised me. Then, right there, he disappeared like a cloud into the sky And the monk and two gurus also likewise disappeared. |374| Then, in order to conclude Mañjuśrī’s speech, the master spoke about the dissolution of the maṇḍala with the verse beginning, In this way... The playful dance and the rest have already been explained. And the [[[maṇḍala]]-]cakra refers to Akṣobhya and the others. Right then means at that very time. As for, He sang and praised me [the words of that song of praise] should be known from the Treasury of Verses.50 Right there means in that very place. Into the sky means into suchness.51 Disappeared like a cloud into the sky is said in order to indicate that, just like clouds and moisture arise from the sky and dissolve back into it, likewise the Bhagavan, as well, through the yoga of great compassion, appears out of suchness and dissolves back into it. This being the case, his ‘causal emanations’ should be known to [do] the same. Having understood that, in order to tell the story of how he carried out the benefit of fortunate [[[disciples]]] he said I... and the rest.52 Realized a little bit is said in order to abandon [the act of] holding back the teachings out of avarice from those who are suitable recipients, [since Buddhajñānapāda had, in fact] exhausted [the obscurations to realizing] the ultimate suchness of all phenomena together with their latent traces, and had, by means of the stages of mudrā as explained above, gained realization.53 54 In a place fifty krośas behind Vajrāsana I lived in the Parvata cave. In order to benefit beings I compiled this [text, the Dvitīyakrama], composed and taught all of the treatises, and so forth. Since excellent beings made extensive supplications, I was delighted [to do so]. |375| 49 de nyid. I am following Vaidyapāda in interpreting this as referring to the immediate moment (Sukusuma, D 134b.6). 50 This text is mentioned by Vaidyapāda earlier in the commentary as a composition of Buddhajñānapāda’s. To the best of my knowledge, it is unfortunately not extant. I address Vaidyapāda’s list of Buddhajñānapāda’s writings below. 51 nam mkha’i khams su zhes te de bzhin nyid [+ nyid sugg. em.; P and D om.] du’o//. I have emended the text very slightly here, adding nyid, where it seems to have been left out. This is because without the emendation the content of the sentence does not make much sense; it would simply read Into the sky means the same.” Also this emendation brings the meaning of the sentence in accord with what follows. 52 Unusually, this short section of Vaidyapāda’s commentary appears to be commenting on a line or lines of the root text that are not extant in our version of the Dvitīyakrama. 53 I remain unsure about the meaning of this last sentence and suspect that the text may be corrupt. 54 da ni ‘jam pa’i dbyangs kyi gsung bsdu ba’i phyir bla mas sprul pa’i dkyil ‘khor bsdu ba gsung pa/ de ltar zhes pa la sogs pa’o// rol pa’i gar zhes pa la sogs pa ni bshad zin to// ‘khor lor bcas pas zhes pa ni rtag pa la sogs pa’o// de nyid ces pa ni dus der ro// glu dbyangs kyis bdag la bstod pa ni tshigs su bcad pa’i mdzod las shes par bya’o// der zhes pa ni gnas de nyid du’o// nam mkha’i khams su zhes te de bzhin nyid [+ nyid sugg. em.; P and D om; see note 51] du’o// sprin rnams med pa lta bur thim (thim] sugg. em. based on Dvitīyakrama; shes D, P) par ‘gyur/ zhes pa ni ji ltar sprin rlan (D, P add las; I suggest omitting) nam mkha’ las byung zhing der zhi ba bzhin du/ bcom ldan ‘das ‘di yang thugs rjes chen po’i sbyor bas de bzhin nyid las (las] sugg. em.; la D, P) snang zhing yang der zhi bar bstan pa’i phyir ro// de bas na de’i rgyu’i sprul pa yang de bzhin du shes par bya’o// de ltar rang gi de shes nas skal ldan gyi don ji ltar byas pa’i lo rgyus gsungs ba bdag gi zhes pa la sogs pa’o// cung zad rtogs pa zhes pa ni/ dngos po thams cad kyi mtha’i pha rol du son pa’i de bzhin nyid du ni bag la nyal du bcas pa zad pa ste/ de phyag rgya’i rim pas gong nas gsungs pa ltar rtogs nas dpe mkhyud snod rung ‘ga’ la yang spang pa’i phyir ro// (Sukusuma, D 134b.5-135a.2; P 162a.5-162b.3 ) 12 Behind Vajrāsana means to its northeast. A krośa is fifty fathoms. Fifty of those is six yojanas plus two krośas. The Parvata cave is [also] called Ma ta hra ni tra, the Dharma Sprout, and is a place where great lords of practice of former times stayed. I lived there means it was [his] residence. For what reason? In order to benefit beings, which means those who stayed nearby. Since there were many who were suitable recipients, the master mentions that it was for their benefit that he compiled this [text], meaning the [Dvitīyakramatattvabhāvanā]-Mukhāgama. All of the treatises refers to those that were mentioned above.55 Composed means produced. Taught means explained. The words and so forth include bestowing samayas and other activities. The cause for doing this was that excellent being made extensive supplications, just like those above. I was delighted [to do so] means that [it was done] with confidence. And [thus] in this way he engaged in the composition of those [texts].56 Living there together, my retinue and I [received] necessities, Clothing, food, a treasury of jewels, and various vast offering substances for gaṇacakra. [From] the tenth-ground bodhisattva, the treasure guardian,57 great Jambhala Each day we regularly received seven hundred kārṣāpaṇa. |376| Living there means there in that cave. Together [with] my retinue refers to the disciples who followed him. Among them there were eighteen who acted as his regents, and among those there were four who attained nirvāṇa in this very life: Dīpaṃkarabhadra, *Praśāntamitra,

• Rāhulabhadra, and *Vajramahāsukha. They are like our primary scriptures as they [follow] the

stages [of practice] of the great master just as [he taught them].58 All of their necessities— clothing, food, a treasury of jewels like gold and so forth, a vast array of substances for making offerings to the Heart of Awakening,59 and the necessities for himself and his students to engage in gaṇacakra practice—were provided by the bodhisattva of the ten bhūmis, who is himself the lord of treasures and is therefore [called] the Treasure Guardian. He appears in the form of a yakṣa and is therefore called the yakṣa Jambhala. Each day he provided each of them with seven hundred karṣāpaṇa of cowries.60 55 This refers to a list of texts in the Dvitīyakrama that Mañjuśrī commanded Buddhajñānapāda to compose and the further elaboration of that list found in Vaidyapāda’s commentary. Many, but not all, of these texts can be identified and are extant. I discuss this list below. 56 rdo rje gdan gyi rgyab (rgyab] D, ‘gab’ P) ni byang shar gyi mtshams na’o// rgyang grags ni ‘dom lnga brgya’o// de lnga bcu ni dpag tshad drug dang rgyang grags gnyis so// parba (parba] D, spar ba P) ta’i phug ces pa ni ma ta hra ni tra (ma ta hra ni tra] D, ma ta hrin dra P) zhes te chos kyi myu gu zhes pa sngon gyi grub pa’i dbyang phyug chen po’i gnas so (so] D, P om.)/ de la brten te zhes pa ni gnas bcas pa’o// ci’i phyir sems can don bya’i phyir/ zhes pa ni de’i nye ‘khor rnams ni khyad par du snod du rung ba mang bas/ bla mas kyang de’i don du zhes so// ‘di bsdus zhes pa ni zhal gyi lung ngo// rab tu byed pa thams cad ces pa ni gong du smos pa rnams so// rtsom (rtsom] D, rtsam P) pa ni byed pa’o// ston pa ni bshad pa’o// sogs kyi sgras bsdus pa ni dam tshig sbyin pa la sogs pa’o// de’i rgyu yang dam pas gsol ba rgya chen po btab pas zhes gong ma ltar ro// bdag ni shin tu brod ces pa ni rang yid ches nas so// de rnams rtsom pa’i sbyor ba la zhugs pa’o// (Sukusuma, D 135a.2-5, P 162b.3-7). 57 srung] D C V (D), gsung S P N V (P). 58 de rnams bdag cag gi dang po’i gzhung ltar de bla ma chen po’i rim pa ji bzhin pa’i phyir ro//. I am unsure of the meaning of this line, which seems to be corrupt in some way. 59 Presumably here this term refers to Vajrāsana. 60 der gnas ni phug der ro// ‘khor bcas rnams zhes pa ni rang gi rjes su spyod pa’i slob ma rnams kyi nang na rgyal tshab kyis pa’i gang zag bco brgyad yod de de rnams kyi nang nas mthong ba’i chos la mya ngan las ‘da’ ba bzhi yod de/ mar me mdzas bzang po dang/ rab tu zhi ba’i bshes gnyen dang/ sgra gcan ‘dzin bzang po dang/ rdo rje bde ba chen po’o// de rnams bdag cag gi dang po’i gzhung ltar de bla ma chen po’i rim pa ji bzhin pa’i phyir ro// de thams cad kyi yo byad dang gos dang/ zas dang gser la sogs pa’i nor gyi mdzod dang/ byang chub kyi snying po la mchod pa’i yo byad rgya chen po dang/ rang dang slob ma’i tshogs kyi sna tshogs ‘khor lo’i bya ba rnams sbyor bar byed pa ni/ sa bcu’i byang chub sems dpa’ ste de nyid gter rnams kyi bdag po yin pas (yin pas] D, bas na P) 13 Then I traveled to meet the great guru Pālitapāda61 In order to please that guru, I compiled62 some short sādhanas And the guru and all the others there were pleased. I returned to the place I had come from and63 joyfully performed the benefit of some64 fortunate [[[Wikipedia:individuals|individuals]]]. |377| Then he tells the account of having been invited by his guru, who had come to know of his blessings with the verse beginning, Then... The statement I compiled some short sādhanas refers to those mentioned above. There means there in that place in the south [of India]. The guru was Pālitapāda. By the others, we should understand those who were gathered there, that is, those [[[dharma]]] relatives who were present. Were pleased means [[[pleased]]] by his dharma teachings and so forth. The place I had come from means the Parvata [[[cave]]]. I performed the benefit of some fortunate [[[Wikipedia:individuals|individuals]]] means those who hadn’t been included in his previous activity.65 Thus, in this way everyone, having come to know the detailed accounts [of my life], Should use all methods to please the sublime and sincere learned one, And listen to and contemplate his teachings, compositions, and so forth. |378| Thus, having generated faith in that way (i.e. by means of telling the story of his own encounter with suchness), he teaches about the training in nondual wisdom and its result with the verse beginning, Thus... Having come to know the detailed accounts means the detailed accounts about the great master: the taming of Nālandā, making offerings at Vajrāsana, the [account of] the consecration and the others.66 Through these accounts the faith of those who have fortune is mdzod srung ngo (srung ngo] D, gsungs so P) / gnod sbyin gyi cha lugs ‘chang bas na gnod sbyin gyi gnas so (gnod byin gyi gnas so] D, gnod gnas so P) // des nyin re ‘gron bu kā rṣā pa ṇa (kā rṣā pa ṇa] D, ka rṣa pa na P) bdun brgya re re la sbyor zhes so// (Sukusuma, D 135a.5-135b.1; P 162b.7-163a.4). 61 bā li pā da’i] D C , bha li pa trī S P N. Vaidyapāda’s commentary reads bsrung ba’i zhabs. I follow Szántó in giving his name as Pālitapāda, based on the presence of this guru’s name in an 11th-century Sanskrit manuscript of Samantabhadra’s Sāramañjarī (Szántó 2015, 542). Tāranātha has rendered it more or less correctly, as well, as Pā li ta pa da (Bka’ babs bdun, 104). In my edition of the Dvitīyakrama, however, I have left the rendering from the Derge and Cone Tengyurs—Bā li pā da—because to “correctly” phoneticize the teacher’s name would make the line unmetrical. 62 It is worth noting that Buddhajñānapāda uses the word “compile” (bsdus) rather than “compose” (rtsom). In an earlier verse, he also uses the term “compile” to describe the compilation of the Dvitīyakrama, but that is presumably because it is in fact Mañjuśrī’s teaching, which he is only compiling within the framework of his own narrative. In this case “compiling” rather than “composing” these sādhanas may hint at a process more revelatory than compositional, but more likely it is simply an acknowledgement that the sādhana was compiled, at least in part, from other sources, most prominently the Guhyasamāja-tantra itself. 63 nas] S P N, gnas D C. 64 ‘ga’] D C V (D and P), dga’ S P N 65 de nas byin rlabs shes pas bla mas spyan drangs pa’i lo rgyus gsungs pa/ de nas zhes pa la sogs pa’o// cung zad bsdus pa zhes pa ni gong du gsungs pa rnams so// de ru zhes pa ni lho phyogs kyi gnas der ro// bla ma ni bsrung ba’i zhabs so// sogs kyi sgras bsdus pa der rtogs so// (D + zhes) ‘tshal spun zlar gyur ba rnams so// mnyes par byas te zhes pa ni chos kyi gtam la sogs pa’o// sngon gnas zhes pa ni parba (parba] D, par pa P) ta’o// skal ldan don ‘ga’ byas zhes pa ni sngon las ma gtogs (ma gtogs] D, rtogs P) pa rnams kyi’o// (Sukusuma, D 135b.1-3; P 163a.4-6). 66 Vaidyapāda here refers to several accounts of Buddhajñānapāda’s life as if they are already well-known stories that will be understood by anyone reading his text. These same accounts are described in the later Tibetan histories in much more detail, though unfortunately only one such supportive detail is, to my knowledge, found in an extant Indian source, Atīśa’s *Bodhipathapradīpapañjikā, which I discuss below. Some of the Tibetan historians who 14 further increased. Then, the learned one who is learned in those scriptures that we uphold is the guru [[[Buddhajñānapāda]]] himself. Since he himself has overcome mental doubts he is sincere. [[[Beings]] should] please him, using all methods which were taught above. In order to familiarize themselves with these, they [should listen to and reflect upon] his teachings, which means his compositions, and so forth,—which [are called] compositions because they are very excellently composed—like the Samantabhadrī[-sādhana] and so forth. The and so forth includes the commentary on the tantra and other [texts]. Listen[ing] to these means also bring about attainment, since the stages of the grounds and paths come about through attainment. Contemplat[ing] them means repeatedly bringing about mental certainty through valid engagement [with them].67 Through relying upon that, remaining in isolated places and the rest, Training one’s mind in suchness, and genuinely realizing the way things are, [One can] attain awakening in this very life, or [even] in [just] six months, and so forth—who could refute this?! |379| Through relying upon that means relying upon those contemplations. In order to bring about suchness in a unique way one is meant to stay in isolated places, and so forth, as described above. Through training one’s mind in suchness means by means of the two stages, like the first [stage] and so forth. Through genuinely realizing the way things are means that through encountering signs of realization, realizing a little bit, genuinely realizing suchness, [and] by means of vratas and the like, one exhausts the remainders [of defilements] in this life, meaning during this very life. As for [even] in [just] six months, the text [also] states and so forth, which indicates an inferior [[[attainment]], i.e. longer time periods]. [Within the various time frames mentioned, one can] attain awakening, which is the realization of the ultimate state. Who could refute this achievement, enacted through such unique methods? Indeed, this being the way things are,68 [it] is difficult to refute, like a cascade of raindrops [falling] through the empty sky.69 provide the more detailed accounts of Buddhajñānapāda’s life, like Tāranātha, do list Indian sources that are no longer known to us. 67 da ni de lta bus dad pas byas te/ gnyis su med pa’i ye shes bsgom pa ‘bras bu dang bcas pa gsungs pa/ de bas zhes pa la sogs pa’o// gtam rgyud rgyas par shes byas nas/ zhes pa ni bla ma chen po’i gtam rgyud rgyas pa na landa ([landa] P, lendra D) ‘dul ba dang/ rdo rje gdan gyi mchod pa byas pa dang/ rab tu gnas pa byas pa la sogs pa’i lo rgyus kyis skal ba dang ldan pa cher dad par byas nas/ des kyang rang gi ‘dod pa’i gzhung la mkhas pa ni bla ma ste (ste] P, sta D)/ de nyid kyis blo’i som nyi bzlog pas na gzu po’o// (P +de) dgong du gsungs pa’i thabs kun gyis mnyes par byas te (te] P, ta D) zhes so// de la (la] sugg. em., las D, P)/ goms pa’i phyir na lung ste/ rab tu byed pa la sogs pa’o// rab tu byed pa ni shin tu legs par byad pa’i phyir na ste/ kun du bzang mo la sogs pa’o// sogs kyi sgras bsdus pa ni rgyud kyi rnam par bshad pa la sogs pa’o// de rnams nyan pa ni thob byed dang bcas pa sa lam gyi rim pa thob pa las byung bas so// bsam par byas zhes pa ni ‘thad sgrub kyis yang dang yang du blo nges par bya’o// (Sukusuma, D 135b.3-6; P 163a.6-163b.3). 68 de’i chos nyid 69 de la rab brtan zhes pa ni bsam pa de la rab tu brten te de nyid khyad par can du bya ba’i phyir dgon sogs rab tu brten (brten] D, bsten P) byas zhes te gong ma ltar ro// rang gi sems de nyid bsgoms pas/ zhes pa ni/ rim pa gnyis kyis zhes pa dang po la sogs pa ltar ro// ji bzhin rab tu rtogs par (par P] pa D) byas pa yis/ zhes pa ni rtogs pa’i rtags rnyed pa dang cung zad rtogs pa dang/ de nyid yang dag bar rtogs pas brtul zhugs la sogs pas lhag ma zad pas tshe ‘di nyid la zhes te/ mthong ba’i chos nyid la’o// zla ba drug gis zhes pa ni sogs pa zhes pa tha ma’i tshig tu’o// byang chub thob pa ni mthar thug pa’i gnas rtogs pa ste/ thabs khyad par can gyis byed pa ‘di ni su yis bzlog ces te de’i chos nyid dgag dka’ ba ste/ bar snang la char gyi rgyun ltar ro// (Sukusuma, D 135b.6-136a.2; P 163b.4- 7). 15 2. Tantric Buddhism in Late 8th-century India Playfully dancing the great dance, with your various arms twisting and holding tight, you open the eight soft lotus petals and insert the vajra, the cause of nondual bliss. The secret suchness, undefiled, becomes clear. The moon which is born from the vajra and petals is perfectly gathered: this is the supreme suchness of all phenomena, born from means and wisdom. Revered master, in order to benefit me, explain what is hidden! -Buddhajñānapāda, Dvitīyakrama This extraordinary autobiographical narrative of the yogin and tantric exegete Buddhajñānapāda’s travels throughout the Indian subcontinent studying with different gurus, receiving teachings directly from Mañjuśrī in a visionary encounter, and setting up a hermitage with his disciples in the latter part of his life provide us a rich picture of Indian tantric Buddhist practice in the late 8th and early 9th centuries. His account is further enriched by the additional details provided in Vaidyapāda’s 9th-century70 commentary. This was a period of immense creativity and development within tantric Buddhist traditions, and many doctrinal and especially ritual developments from precisely this period continue to frame the structure of tantric Buddhist practice up to the present day. While the cadence and timbre of Buddhajñānapāda’s own voice clearly emerges from his surviving writings—and not only from the autobiographical narratives therein—we can better appreciate his individuality when it is approached from within the context of the world in which he lived and wrote, so it is to this that we will first turn. The Political, Social, and Religious Climate of Early Medieval India It is at our peril that scholars of Indian Buddhist traditions have often focused too narowly on textual sources and developments only within the Buddhist world, and neglected to consider the wider social, political, and religious climate of the Indian subcontinent in which Buddhist doctrinal and ritual developments emerged. When we do thus widen our perspective, as is fortunately increasingly the case in recent scholarship, we have access to a much more holistic, and therefore deeper as well as broader, view of the traditions we seek to understand. Having let Buddhajñānapāda himself, and his disciple Vaidyapāda, speak their stories first, I would like to begin my own account of Buddhjñānapāda’s life and writings by widening the lens to take in the broader world of the Indian subcontinent into which his voice emerged and was first heard. The early medieval period in India (roughly the mid-6th to the early13th century) was a time of upheaval, change, and immense creativity. Following the fall of the “golden age” Gupta empire in the 6th century of the common era there was a period of significant political restructuring. This was described by earlier scholars using a rhetoric of decline and decentralization, in which the process was termed “feudalization,”71 while more recent scholarship describes the same period using more positive language as a process of incorporation, in which newly founded state polities were both incorporating new territories and expanding into territories that had previously not been touched by a state polity.72 The general political climate of the period was neither a centralized state nor fragmented regional kingdoms, but rather, “a series of diverse and uneven political orders which, while regionally based, sought 70 Vaidyapāda was likely a direct disciple or at furthest a grand-disciple of Buddhajñānapāda, thus placing him squarely in the 9th century. I address the dates of both in more detail below. 71 e.g. Kosambi 1956. 72 e.g. Chattopadhyaya 1994. 16 to relate themselves, in diverse ways, to ever more integrated political hierarchies which had as their ideal the notion of an imperial polity ruled over by a single supreme overlord, a king over kings.”73 While the political rhetoric championed the idea of digvijaya, “conquest over the directions” as frequently mentioned in praśasti, the eulogistic poems dedicated to leaders that became an important literary mark of the political culture of the day, in actual point of fact such “conquest” often did not involve direct rule of the conquored lands by the overlord.74 More commonly, the conquered areas continued to be ruled by their own, now “lesser” lords, who submitted to the “greater” lord, thus creating a complex system of social and political relationships. An important aspect of this system was the gifting of land by new or established rulers, as a way of showing favor to their constituents, and sometimes also to encourage the expansion of agriculture into uncultivated areas. These land grants were often given to religious institutions— initially to monasteries, or individual brahmin families or communities, and increasingly, with the development of what is often called “temple Hinduism” to temples. In fact, religious grants seem, in most areas, to have been far more frequent than grants to non-religious beneficiaries.75 This political situation of the early medieval period resulted in a number of important developments in the religious sphere. The first of these is a direct result of the expansion of state polities into new areas. This expansion involved the movement of peoples, which resulted in the meeting of more established and pan-regional religious traditions, such as the Vaiṣṇava, Śaiva, Buddhist, and Jain traditions, with more local forms of religion.76 This process sometimes occasioned the adoption of local deities into translocal traditions.77 In addition, this political 73 Ali 2004, 33. 74 Pollock emphasizes the fact that the political styles and the tradition of composing praśasti was common to rulers from diverse religious traditions—Buddhist, Vaiṣṇava, Śaiva, and Jain. He notes that, in fact, they “all wrote more or less similar poetry and engaged in identical political practices” (Pollock 2009, 572). 75 Thapar 2002, 451. The most common explanation of this land-granting practice, given in much of the historiography, is that this was a practice done for purposes of “legitimation.” The new rulers, because many of them were not from traditional ruling (kṣatriya) families, needed to justify their rule, and thus gave grants to brahmins who then wrote important “fictitious” genealogies (vaṃśa) legitimizing the rule of these families. While these developments which, described as “purāṇic,” may seem to apply only to religions that developed out of the brahmanical tradition, in fact the Buddhists and Jains both integrated themselves into these structures, as well, claiming descent from the so-called “Solar Dynasty” that is one of the two important lineages in the purāṇic genealogies (Samuel 2008, 68). What is more, the 8th-century Buddhist tantra, the Mañjuśrīmūlakalpa, also contains a predictive royal genealogy, very much in the purāṇic model (Sanderson 2009, 94). Sheldon Pollock’s critique of the rhetoric of legitimation found in earlier historiography—essentially arguing that such legitimation does not make sense in the pre-modern period when rulers could (and did) simply force their rule upon people and did not need to rely on documents like genealogies to convince the populous that their rule was legitimate—is well taken (Polluck 2009, 521). However, Pollock himself admits that if legitimation does anything at all it builds ruling class consensus, rather than that of the larger populous (Pollock 2009, 523). This observation, in fact, makes perfect sense of the popular practice of granting land and receiving genealogical confirmation of one’s right to rule. These genealogies, which begin to appear from the 6th century, just as the Gupta empire was falling, insist on birth into certain types of lineages as a requirement for being part of the ruling class. Thus there was indeed a need for legitimizing oneself as belonging to a certain type of family in order to engage in the elite political culture of the time. Because of the structure of the political order, with its enmeshed polities and the important and intricate relationships that involved the exchange of gifts and women (which Daud Ali has carefully described in his 2004 work), it was impossible for a ruler to exist as a completely independent polity—one could not rule in a social and political vacuum. Political relationships were crucial to the maintenance of power, and to engage in these relationships, it was necessary to hold claim to a certain type of birth. Thus while the rhetoric of legitimation does not, as Pollock suggested, make sense as legitimation on behalf of the larger populous, it does make perfect sense when understood as a requirement for participation in the elite political culture of the time. 76 Thapar 2002, 389. 77 See e.g. Granoff 2004. 17 reorganization also involved significant warfare, which displaced populations and may have had some influence on de-urbanization in certain areas. When people and communities are on the move, they meet with other groups, leading to the intermingling of beliefs and practices, and creating a perfect environment for religious and cultural creativity. We can see this in the mutual influence between Buddhist and Śaiva tantric traditions78 where we find extensive Buddhist textual borrowing from Śaiva tantras,79 as well as examples of traditional Buddhist iconography passing into the Śaiva tradition.80 Moreover, there is also documented evidence of Śaiva borrowing from earlier Buddhist tantras,81 and the Buddhist tantric use of transgression specifically as a method to cultivate nondual gnosis was later adopted by Śaiva authors.82 Certain techniques, such as the practice of utkrānti, in which the consciousness of a yogin is ejected from his body, and sometimes transferred into the body of another individual (or more frequently a corpse), are known in Śaiva, Buddhist, and Jain texts, indicating a culture in which yogic techniques were shared.83 Many pilgrimage sites were also shared commonly among multiple traditions. Another important way in which the political environment affected the religious trends in early medieval India involves what has been described as the parallel developments of the “apotheosis of the king” and the “feudalization of the gods.”84 The early medieval period is thus characterized by the emergence of the idea of divine kingship. This process also involved the shift in ritual practices connected with kingship in which earlier Vedic rituals of royal consecration were replaced by purāṇic (Vaiṣṇava, Śaiva), and Buddhist versions.85 The purāṇic legends of the time describe the gods in ways that reflect the political culture of the time—they marry, live in fortresses, and so forth, like kings and queens.86 Unsurprisingly, rulers were enthusiastic patrons of these religious developments. The specific medieval Buddhist response to the political developments of the time has been studied by Ronald Davidson in his important work Indian Esoteric Buddhism, in which he asserts that the metaphor of kingship is the defining metaphor of Buddhist tantric systems. Davidson’s analysis of the role of the political environment in informing the ritual world of the tantric maṇḍala remains an astute and important observation. In fact, Daud Ali’s work on early medieval court culture provides a number of examples, easily visible to the scholar of Buddhist tantra, which further corroborate Davidson’s thesis.87 Indeed it does appear that one factor in the development of these particular forms of Buddhist practice was the current political climate; they appear to constitute a method for securing a place, as well as patronage, for Buddhism in the 78 The relationships between these two traditions has been studied by Sanderson and Davidson, who take different, if not exactly opposing, perspectives (See Sanderson 2001 and 2009 and Davidson 2002). 79 See Sanderson 2001 and 2009 and Hatley 2016. 80 Davidson 2002, 86. 81 Hatley 2016. 82 Wedemeyer 2013, 166-67. 83 See Smith (2006, 289) for an excellent description of this process in a Jain text, and for a description of the practice as allegedly performed by the 8th-century yogin Śaṅkara (Smith 2006, 294). The earliest description of the process of the ejection of consciousness that I am familiar with in a Buddhist text is in Buddhajñānapāda’s Dvitīyakrama, which has strong echos in the Catuṣpīṭha-tantra (On which see Szántó 2012a, esp. pp. 455-62). I address the topic of utkrānti in Buddhajñanapāda’s writings in Chapter Six. 84 Davidson 2002, 71. 85 Davidson 2002, 127-31. 86 ibid., 71. 87 For example, the king possesses a “seven-walled palace” (Ali 2004, 42); he has messengers (Skt. dūtaka, =Tib. pho nya) who carry out his business and doorkeepers (dvārapāla) guarding each door of the palace (ibid., 45); he bestows favor on a supplicant if he is pleased or satisfied (ibid., 106); sits on a lion throne and is fanned by whisk bearers (ibid., 112). 18 changing political and social environment. However, it is also important to take seriously the soteriological aims that these writings themselves explicitly claim to pursue. Another arena in which the social and political climate influenced religious developments, and one that is just beginning to be studied, is the realm of courtly culture. As Ali has shown with his groundbreaking work, the early medieval court “formed a key context for the production of knowledges that have more commonly been attributed to a generalised ‘society’ in ancient India.”88 Many of these developments have been an important influence in the religious sphere, as well, and Ali has documented significant contact between the courtly and religious worlds. It seems that a substantial number of men of the court came from monasteries, hermitages or brahmin households that were supported by the king and many prominent courtly gentlemen became hermits or monks when their masters died, or when they themselves entered old age.89 This certainly indicates a climate in which ideas could move freely between those two domains. Indeed Vatsyāyana, in his well-known Kāmasūtra, which typifies the pursuit of erotic and aesthetic pleasure central to the courtly life of the day, suggests that a young woman should learn erotic skills discreetly from an older sister, or from a nun (presumably one who had an earlier adult life as a laywoman!).90 Ali also shows that aspects of court protocol “intersected with codes of conduct from a wider domain, particularly a religious one.”91 These religious developments in turn influenced the political culture, because, in large part, religious masters and institutions were successful in their aims to secure patronage. That is to say, the kings of the time supported these traditions, and incorporated their rituals into rituals of state, effectively replacing the earlier Vedic model.92 Kings spent tremendous amounts of wealth supporting religious institutions and religious specialists. Royal preceptors often became wealthy in their own right, allowing them to support the development of their own traditions, to build temples, or support monasteries. Indeed the enormous Buddhist monasteries of the early medieval period such as Nālandā and Vikramaśīla, and the breathtaking Śaiva, Vaiṣṇava, and Śākta temples constructed during this period—the incredible outpouring of religious art, architecture, and literature—was primarily possible due to royal patronage. We will see resonances of many of these broader developments as we look more closely at Buddhjñānapāda’s life, writings, and thought: clear evidence of his movement throughout the subcontinent; engagement with the large state-supported Buddhist institutions of his time and 88 Ali 2004, 25. 89 ibid., 49. 90 ibid., 218. 91 ibid., 103. The influence of courtly culture (rather than politics, which Davidson has examined) on specifically Buddhist literature and ideas is a tantalizing but little explored area. We can easily see the influence of the sumptuary culture of the court on the Mahāyāna sūtras, with their imagery of worlds of jewels and gems, and the posture of royal ease adopted by the bodhisattva imagery of the period (Ali 2002, 159). This culture continues to appear in later śāstras, as in the elaborate bathing and dressing pavilions described by the 7th-century Buddhist author Śāntideva in his famed Bodhicāryāvatāra. The literary theory of rasa, developed in the nātyaśāstra literature and very important to the courtly aesthetic, was brought into the Buddhist tantras in the 8th-century Sarvabuddhasamāyoga-tantra (Smith 2006, 333). Moreover, with regards to erotology—another important cultural development that was made and refined in the courtly context— as we will see, Buddhajñānapāda, in his Dvitīyakrama, uses the classic four-fold female erotic typology from kāmaśāstra in his description of tantric Buddhist consorts (Dvitīyakrama, verses 50-67), in what is a very early instance of this classification system in Indian literature, preceding its appearance in extant kāmaśāstra literature by several hundred years (I discuss this further in Chapter Six). The same four-fold typology is also found in the later Saṃvarodaya-tantra (See Tsuda 1994, 155). Ali (2011, esp. pp 54-55) has explored some of the ways in which Buddhist tantric literature and practice seems to have influenced kāmaśāstra. 92 This is not to say that the earlier Vedic rituals were completely left behind. Indeed, many aspects of these much older rituals were incorporated into newer purāṇic, Buddhist, and Jain rituals. See, for example, Marko Geslani’s work on the incorporation of śānti rites into post-Vedic ritual contexts (Geslani 2011 and 2012). 19 with the political elite, as well as with other types of patrons and systems of patronage; evidence of his participation in a doctrinally and ritually eclectic millieu; and the incorporation of aspects of courtly culture into religious doctrine and ritual. Let us now narrow our focus one notch to survey the specifically Buddhist doctrinal and ritual context in which Buddhajñānapāda lived and wrote. 8th-Century Indian Buddhism While his writings indicate, through their many references to and much terminology from non-Buddhist traditions, that Buddhajñānapāda was living in a religiously eclectic environment, they leave no doubt about his self-identification as a Buddhist practitioner. Within the Buddhist tradition, he also clearly identifies as a practitioner of Buddhist tantra, and indeed appears to have been participating in, and likely even contributing to, the cutting edge of Buddhist tantric ritual technologies of his time. His life and work, therefore, must also be understood within the framework of the Buddhist tradition as it existed in India in the 8th century. The rise of Mahāyāna Buddhism around the turn of the common era brought with it a revelatory and visionary turn in the Buddhist tradition,93 expanding the scope of Buddhist worlds, and bringing newly expressed philosophical orientations to the fore. The Prajñāpāramitā-sūtras and Nāgārjuna’s Madhyamaka philosophy articulated a vision of reality whose emptiness of inherent nature allowed for infinite possibility, while the slightly later sūtras on buddha nature and those incorporatingYogācāra thought, along with their accompanying commentarial literature, attended more closely to the identity and nature of the mind that had access to such a reality. It was within the context of these doctrinal systems that Buddhist tantra began gradually to develop in the 7th century of the common era. However, while certainly it was in terms of such Buddhist doctrinal systems that the tantras were interpreted, the texts themselves appear to have emerged out of a more practical, that is to say a ritual, context.94 By the 8th century when Buddhajñānapāda lived and wrote, Buddhist tantric traditions had developed to the point that contemporary authors were beginning to classify the tantras into different categories.95 Buddhaguhya, a contemporary of Buddhajñānapāda’s, divides the tantras into Kriyā tantras, which involve more outward practice, and Yoga tantras, which involve more inner yogic practices.96 He also mentions what has sometimes been interpreted as a separate category, Ubhaya, or “both” (also sometimes called “Upa” or “Caryā”), which involve a mixture of external and internal practice.97 The Yoga tantras, the most prominent early example of which is the 7th-century Sarvatathāgatatattvasaṃgraha-tantra, showed a significant turn towards concern with 93 See, e.g. Harrison 1978 and 2003. 94 See J. Dalton 2016; Dharmachakra Translation Committee http://read.84000.co/translation/toh498.html, i.21; and Shinohara 2014. 95 See J. Dalton 2005, 121-31 for a summary of 8th-century Indian authors Buddhaguhya’s and Vilāsavajra’s tantric doxographies. These systems were almost certainly known to Buddhajñānapāda as Vilāsavajra is named in the Dvitīyakrama as Buddhajñānapāda’s own guru, and Buddhaguhya is sometimes mentioned, at least in the later Tibetan histories, as his disciple, though modern scholars have questioned this claim. Buddhajñānapāda’s own Dvitīyakrama contains a doxography, though not a tantric one (Dvitīyakrama, 126-43). He simply places tantra as a whole above all non-Buddhist and Buddhist philosophical systems, which is rather unique given that, as Dalton points out in the article just referenced, the Indian systems of classifying different systems of tantra rely primarily on ritual, rather than doctrinal, distinctions (J. Dalton 2005, 119-20). See also note 100 below for the wide variety of doxographical categories found in Vaidyapāda’s 9th-century Sukusuma. 96 J. Dalton 2005, 123-4. 97 ibid., 123-5. 20 soteriological goals,98 set forth the “mature” five-family maṇḍala system, and demonstrated full self-awareness of being a unique system of Buddhist practice.99 However, in the 8th century Buddhist tantric systems began to undergo a futher shift, with the development of what came to be called the Mahāyoga tantras. These texts, the most prominent of which is the Guhyasamājatantra, bring antinomian elements of sex and violence that were marginal in the Yoga tantras into the fore.100 The Mahāyoga tantras are furthermore characterized by the quite literal shift in the five-family maṇḍala arrangement from the centrality of Vairocana, the main buddha of the socalled buddha family, to that of Akṣobhya, of the vajra family. The 8th century also saw the emergence of the important Sarvabuddhasamāyoga-tantra, sometimes classified as a Yoga or Mahāyoga tantra, and other times as pertaining to the other newly emerging category of the Yoganiruttara or Yoginī tantras, which came into their full flourit in the 9th and 10th centuries with the Cakrasaṃvara and Hevajra-tantras, among others. The 8th and early 9th centuries were also a particularly important period in terms of the unfolding of the ritual structures and frameworks within which tantric Buddhist practice took place. That is, the system of tantric initiations was developing precisely in this period from the five-fold series of initiations that characterized the earlier Yoga tantras,101 to the addition of the later sexual initiations: the guhyābhiṣeka, then the prajñājñānābhiṣeka, and finally the so-called “fourth” initiation.102 The addition of these higher initiations corresponded with, and was likely necessitated by, the appearance of new modes of practice. The 8th century saw the use and development within the Buddhist tradition of sexual yogas, and the further development of practices that involved the manipulation of subtle energies within the body, which were often performed within the context of sexual practice. The addition of these new techniques into the tantric practitioner’s repertoire resulted in the division of tantric practice into the now ubiquitous categories of the generation (utpattikrama, bskyed rim) and perfection stages (niṣpannakrama, utpannakrama, rdzogs rim).103 Though of course these newly developed categories were in flux, the generation stage can be loosely characterized by the visualization of oneself as a deity and the worship thereof, and the perfection stage by the manipulation of internal energies while in 98 Tribe 2000, 209. 99 See Weinberger 2003 (esp. pp. 185-89) which draws attention to the rewriting of the Buddha Śākyamuni’s awakening story as a narrative of tantric practice in the Sarvatathāgatatattva-saṃgraha. Weinberger argues that this represents tantra’s “coming out party” or its “declaration of independence” as something distinct from earlier Buddhist traditions (Weinberger 2003, 189). 100 While the term Mahāyoga tantra was certainly used to describe the Guhyasamāja-tantra and other tantras pertaining to this class, they also continued to be referred to by some authors as Yoga tantras. The tantric doxographical categories of the time were indeed so variable that even within a single commentary by a single author, such as Vaidyapāda’s 9th-century commentary on the Dvitīyakrama, tantras are distinguished in multiple ways including those tantras “that emphasize wisdom” and those “that emphasize method” (Sukusuma, D 90a.2-4); Yoga and Mahāyoga tantras (Sukusuma, D 107a.6-7); Yoga and Yoganiruttara a.k.a. Dākiṇī tantras (Sukusuma, 108a.6-108b.1); and Krīya, Caryā, and Yoga tantras (Sukusuma, D 112a.4-5). 101 These are the water, crown, vajra, bell, and name initiations. See Isaacson 2010b, 264. When the later initiations were added, these five were re-classified as the “first” or vase initiation (kalaśābhiṣeka). I discuss initiation in Buddhajñānapāda’s writings in Chapter Seven 102 Isaacson (2010b) gives a summary of the gradual development of these initiatory systems. I address this topic in more detail in the context of Buddhajñanapāda’s writings on initiation, along with those of several of his direct disciples, in Chapter Seven 103 The well-known scriptural locus classicus of the two stages of tantric practice is the Samājottara, though as I have pointed out in an earlier conference paper and will discuss further in Chapter Eight, this important verse in fact seems to have originated in Buddhajñānapāda’s writings (C. Dalton 2013). The terms seem to have been in somewhat general use, though, by the middle of the 8th century, as they are found in other texts from the time, such as Padmasambhava’s Man ngag lta ba’i phreng ba (one of the few texts that scholars accept as having been composed by the historical Padmasambhava). 21 that form, often performed while in sexual union with a consort. As we will see, Buddhajñānapāda’s writings are important for refining our understanding of both of the development of initiatory rituals and the two stages of tantric practice. While these emerging systems certainly became popular modes of Buddhist practice that survive to this day as a central component of particularly the Tibetan and Newar Buddhist traditions, it is also clear that Buddhist tantra, like the Mahāyana before it, was not universally accepted among Buddhist communities. The tantras themselves and their exegetical commentaries, including Buddhajñānapāda’s own writings, include a number of features and strategies that appear to be aimed at legitimizing these newly emerging and unsurprisingly controversial practice systems. There are, moreover, records of discord between those who accepted the new traditions and those who did not, including reports of Buddhist monks publicly burning tantric Buddhist scriptures and destroying tantric Buddhist images at Vajrāsana in the late 8th/ early 9th century.104 From his writings we can see that Buddhajñānapāda, like his contemporary tantric exegetes, was well versed in a great deal of the Buddhist literature, doctrine, and practice systems that preceded him. This includes both exoteric Mahāyana sūtras and philosophical systems as well as earlier tantric traditions like those of the Mahāvairocana-tantra, the Sarvatathāgatatattvasaṃgraha-tantra and the Sarvabuddhasamāyoga-tantra, citations from the latter two of which appear in his writings. But it is the Guhyasamāja-tantra that is most central to his oeuvre, and especially to the ritual systems he set forth. Indeed, Buddhajñānapāda became known as the founder of the eponymous Jñānapāda School of Guhyasamāja practice, one of several lineages of practice associated with this tantra that first flourished in India and were later brought to Tibet. Thus, in order to further contextualize Buddhajñānapāda’s life and writings, we must again narrow our lens even more, to look at the emergence of and practice systems associated with the Guhyasamāja-tantra. The Guhyasamāja-tantra One indication of the Guhyasamāja-tantra’s importance is its consistent inclusion in every known version of an otherwise variable list of eighteen quasi-canonical tantric compositions, that circulated from India into both China and Tibet during the 8th century.105 The so-called Vajroṣṇīṣa (erroneously rendered as Vajraśekhara in earlier scholarship)106 scriptures were transmitted to China by Amoghavajra and Vajrabodhi, and are described in a short summary text by Amoghavajra as eighteen “assemblies,” of which the fifteenth has been identified as the Guhyasamāja.107 The idea of such a group of eighteen tantras, or tantric cycles, was also passed on to Tibet, where they were known there as the Māyājāla tantras and ascribed to the class of Mahāyoga tantra. However, Orna Almogi has argued, on the basis of the great variety in terms of both content and organization of thee lists preserved in Tibetan literature, both historical and doctrinal, that an actual list of the specific tantras that the group of eighteen 104 Chimpa and Chattopadhyaya 2010, 279; Maclean 1989, 12; and Flood 2009, 34. Incidentally, Flood interprets this as an act of hylotheism, which it clearly was not. The monks would not have been concerned that the tantric Buddhists were “confusing a transcendental god with matter” as Flood suggests (they certainly would have made and revered images of the Buddha, too), but that they were worshipping deities and engaging in modes of practice that the monks deemed non-Buddhist. Indeed, Tāranātha reports the monks to have said that the texts were “composed by Māra” (Chimpa and Chattopadhyaya 2010, 279). 105 It is, in fact, one of only three texts that appear in all known lists of this group; the other two are the Śrīparamādya-tantra and the Sarvabuddhasamāyoga-tantra. 106 On this issue see Geibel 1995, 109 and Davidson 2011, 24. 107 See Eastman 1981 and Geibel 1995. 22 contained was likely not.108 Nonetheless, all of the different Tibetan lists studied by Almogi contain the Guhyasamāja-tantra.109 An Indian version of this list is found in a commentary on the Āryaprajñāpāramitānayaśatapañcāśataka (Tōh. 2647) by the late 8th-century Indian scholar Jñānamitra who mentions “eighteen sections” headed by the Sarvabuddhasamayoga-tantra, but also including the Guhyasamāja.110 The Guhyasamāja-tantra survives in a number of Sanskrit manuscripts, at least three Tibetan translations, and one Chinese translation, again attesting to its importance.111 Yukei Matsunaga’s studies of the historical development of the tantra (1964, 1977a, 1977b, 1978) still serve as the primary basis for research in the field. Like most Buddhist tantras, the Guhyasamāja is an accretive text: the first twelve chapters of the tantra comprise an earlier level of textual composition, chapters thirteen through seventeen constitute an additional level,112 and the socalled eighteenth chapter, the Samājottara, first circulated in India as an independent text before being appended to the root tantra.113 Based on the presence of the summary of the “Guhyasamāja-yoga” among the eighteen “assemblies” noted in the account translated into Chinese (or perhaps composed)114 by Amoghavajra, who travelled in India between 744 and 746, it is clear that some form of the Guhyasamāja-tantra was in circulation in the first half of the 8th century.115 However, the Guhyasamāja as described in Amoghavajra’s “eighteen assemblies” only covers the basic maṇḍala structure and neglects the more antinomian elements of the 108 Almogi 2014, 51. 109Almogi 2014. On versions of the groups of eighteen tantras see also Eastman 1981 and J. Dalton 2005, 126 n32. 110 Geibel 1994, 114. 111 The extant Tibetan translations are the canonical translation in the Kangyur (including the so-called eighteenth chapter of the tantra, the Samājottara); the translation preserved in the Collected Nyingma Tantras (Rnying ma rgyud ‘bum v 12, 89a-157a), which also includes the Samājottara; and a manuscript of the translation of the root tantra alone, without the Samājottara, from Dunhuang. Kenneth Eastman (1979) has studied these different recensions of the Tibetan translations, concluding that the Dunhuang translation is the basis for the other two. The Sanskrit of the tantra was first edited by Bhattacharya (1931) on the basis of four Sanskrit manuscripts, and subsequently by Bagchi (1965). The first Western-language translation was made by Fremantle, whose doctoral dissertation (1971) included a Sanskrit edition and a romanized transcription of the Tibetan translation of the root tantra from the Peking Kangyur, as well as an English translation of the root tantra. Her work does not address the Samājottara. More recently Matsunaga (1978) has made a more comprehensive Sanskrit edition, including the Samājottara. His edition takes into account not only the Sanskrit and Tibetan, but also the Chinese translation of the tantra. 112 Indeed, even a quick glance at the composition of the Guhyasamāja-tantra bears witness to the fact that the chapters, starting from chapter thirteen onwards suddenly become much longer than the first twelve. 113 Matsunaga 1977b. In fact, the Samājottara is still preserved as an independent text in the Derge edition of the Tibetan Kangyur, where it is entitled the Rgyud phyi ma (Tōh. 443). According to Matsunaga the Samājottara also underwent stages of development. He notes that, “...the Uttaratantra [=Samājottara] text which is quoted in Viśvāmitra’s commentary and which remains as an old Tibetan translation differs with the present text. Accordingly, it is likely that a small process of development occurred before the present form of the Uttaratantra was completed” (Matsunaga 1977, 117). Matsunaga’s observations with respect to the Samājottara as preserved in the Nyingma Canon may be accurate. However, with respect to Viśvamitra’s commentary, Dpal gsang ba 'dus pa'i rgyud kyi man ngag gi rgya mtsho thigs pa (Tōh. 1844), my reading of this text has led me to the conclusion (which I hold with considerable certainty) that it is, in fact, not a translation from the Sanskrit at all, but rather an indigenous Tibetan composition. In addition to lacking both a Sanskrit title at the beginning and a translator’s colophon at the end (which would not in and of itself preclude its being an Indic text), the commentary, which deals only with the Samājottara and not with the root tantra, is nearly twice the length of most Indic commentaries on the tantra and shows a number of linguistic features that I believe could only have arisen in an indigenous Tibetan composition commenting on a Tibetan translation of the Samājottara, rather than on the Sanskrit text. 114 Even traditional Japanese Shingon sources consider this text to be a composition by Amoghavajra rather than a translation (Geibel 1995, 108). Nonetheless, he is understood to be summarizing Indian sources with which he was familiar. 115 Matsunaga 1977b, 112. See also Geibel 1995. 23 present day form of the tantra, sharing more features with earlier Yoga tantras such as the Sarvatathāgatatattva-saṃgraha.116 The text transmitted to Tibet in the period of the early translations (prior to the collapse of the Tibetan empire in 843), however, was the full root tantra.117 It is thus likely that while the Guhyasamāja-tantra originated in the early part of the 8th century, the text as we know it today was developed in the later part of that century. Buddhajñānapāda’s life falls directly towards the end of the period in which we surmise that the root tantra took its final form and given that the Samājottara first circulated somewhat later than the root tantra, his relationship to both of these texts is an important question that I will address below. The format of the tantras, with their compilatory nature, diverse topics, and swift jumps from narrative, to doctrinal content, to ritual, and back makes it difficult to summarize their contents.118 I will, however, just briefly outline some of the contents of the Guhyasamāja-tantra to give a sense of the scripture that Buddhajñānapāda references in many of his works. The tantra begins rather dramatically with the statement that the Bhagavan was abiding in the bhaga (vagina) of the Vajra Consort, which was so novel that just the tantra’s opening section was the subject of an entire commentary in its own right.119 The first chapter continues with the emergence of the Guhyasamāja maṇḍala which is generated by the speaker of the tantra, alternately called the Bhagavan, Mahāvairocana, and Bodhicittavajra. The deities produced are the five buddhas, beginning with Akṣobhya, who takes the central place,120 the four buddha consorts, and the four wrathful gatekeepers, each of whom is generated by means of his or her own mantra.121 The tantra alternately discusses doctrinal questions, like the nature of awakening (usually in the narrative frame of a discussion between the main promulgator of the tantra and his 116 Matsunaga 1977b, 112. 117 Also of note for the late 8th-century dating of the root tantra is the fact that the translation of the Guhyasamājatantra (but not the Samājottara) preserved in the Rnying ma rgyud ‘bum was, according to its colophon, translated by the early-period translators Vimalamitra and Kawa Paltsek (Ka ba dpal brtsegs), though Eastman notes that it was extensively altered after the thirteenth century (Eastman 1979, 3). The Rnying ma rgyud ‘bum translation of the Samājottara states, confusingly, that it was translated by the translators Buddhaguhya, who lived in the 8th century, and Drogmi Palgyi Yeshe (‘Brog mi dpal gyi ye shes) who lived in the 11th—while the translation of the Samājottara preserved in the Derge Kangyur states that it was translated by Śraddhākaravarman and Rinchen Zangpo (Rin chen bzang po) (958-1055), of the later translation period. It thus seems likely that the Samājottara was not translated until the later period. The Blue Annals also confirms the early-period translation of the Guhyasamāja-tantra, but attributes the translation to the translator Che Tashi (Lce bkra shis), and makes no mention of the translation of the Samājottara (Roerich 1976, 359). Additionally, Campbell notes that a translation of Vajrahāsa’s commentary on the root tantra is preserved in the Ldan dkar ma catalogue, thus dating the root tantra definitively before the early 9th century (Campbell 2009, 46). 118 Indeed most scholarship even on specific tantras makes no attempt to do so. The format of the Introductions to the 84000 Project translations of the Tibetan canon, which require the translator to provide at least some summary of the text she has translated, are therefore a welcome addition to the scholarship on Buddhist canonical literature (See http://read.84000.co/). 119 This is Vilāsavajra’s Śrīguhyasamājatantranidāṇagurūpadeśabhāsya (Tōh. 1910). The introductory narrative of the Guhyasamāja-tantra of course follows the traditional sūtric narrative stating “Thus I have heard at one time, the Bhagavan was residing at....” Usually the location of his residence in the sūtras is a location in India, such as Rajgir in the instance of many of the Prajñāpāramitā-sūtras. The Guhyasamāja-tantra thus shakes up this traditional narrative structure and brings tantric sexual imagery immediately to the fore by locating the Bhagavan’s residence not in an identifiable geographical location in India, but rather in his consort’s bhaga. 120 Other sections of the tantra, however, clearly prioritize Vairocana (see eg. Chapter 3). In some places we can see signs of the earlier three-family system, with just Vairocana, Akṣobhya, and Amitayus (see e.g. Chapter 9), as well as of a version of the five-family system where Vairocana is primary (see e.g. Chapter 12). 121 This basic thirteen-deity maṇḍala is not represented in either the sādhana literature of the Jñānapāda School or the Ārya School of the Guhyasamāja-tantra, which both have more elaborate maṇḍala structures. I have not, in fact, seen any Guhyasamāja sādhana with this simple thirteen-deity maṇḍala that directly accords with the tantra. 24 retinue of tathāgatas and bodhisattvas), and sets forth short sādhana-like practices including rites for the visualization of the buddhas, akin to those found in generation stage practice manuals, and yogic techniques like the sūkṣma-yoga that we find in perfection stage manuals. However, no full-fledged sādhana for either stage of practice appears in the tantra.122 Additionally, there are maṇḍala rituals, descriptions of sexual yogas, and practices including the injunction to consume the sexual fluids produced from union; injunctions to engage in antinomian behaviors like the consumption of traditionally impure substances; and injunctions to engage in behaviors associated with the three poisons of passion, aggression, and delusion, normally strictly proscribed in Buddhist practice.123 The tantra also describes the practice of mantra recitation, as well as a number of wrathful rituals.124 Like many tantras, the Guhyasamāja-tantra also contains “recipes” and instructions for the ritual production of certain substances that are to be consumed in order to gain (primarily) worldly accomplishments. There is also mention, but not a clear description, of initiation, as well as references to four stages of practice—sevā, upasādhana, sādhana, and mahāsādhana—that have been variously interpreted and used by different authors, including Buddhajñānapāda, to structure tantric sādhana. Given the almost chaotic nature of the Buddhist tantras, many of which are eclectic just like the Guhyasamāja-tantra, commentaries providing interpretive frameworks and liturgical manuals providing practical details were necessary to give practitioners avenues for engagement in these complex systems. These were always supplemented by oral instructions, as well, passed down through a lineage.125 According to the later Tibetan tradition, there were two main schools of Guhyasamāja exegesis and practice in India: the Jñānapāda School, eponymously named for Buddhajñānapāda, and the Ārya School, presumably named after its main exponent (Ārya) Nāgārjuna. These two were, however, not the only Guhyasamāja traditions practiced in India. A Guhyasamāja sādhana by the Indian tantric exegete Candrakīrti, extant in its original Sanskrit, refers to four schools of Guhyasamāja practice distinguished by the number of deities in their maṇḍalas; however, only the two mentioned above seem to have been passed on into the Tibetan tradition and we have little evidence of these other systems.126 Among these two major 122 The tantra does, however, contain many short ritual sequences describing practices that seem to pertain to what later became classified as both of the two stages of practice. The nature of the relationship between sādhanas and other such authored ritual manuals and the tantras to which they are connected remains a topic that merits further study. 123 These more antinomian elements of the Buddhist tantras have been interpreted variously by both traditional and modern scholars alike. Wedemeyer (2013) provides a helpful analysis of the various ways in which such practices have been, and ought to be read. 124 Wrathful rituals gain prominence from chapters thirteen onwards. 125 Indeed, it is said that the tantras are actually intentionally structured in a chaotic and confusing way specifically so that their content cannot be approached or practiced without the assistance of not only commentaries, but also oral instructions on the practices received directly from a guru. 126 Tomabechi 2008, 171n1. His reference is to Candrakīrti’s Vajrasattvaniṣpādanasūtra. See Hong and Tomabechi 2009, 35 for the Tibetan edition. The Guhyasamāja-based sādhanas found at Dunhuang, however, are exceptional in displaying no evidence of a distinction between the Jñānapāda and Ārya schools, and thus perhaps represent a stage of Guhyasamāja practice before this distinction developed. See J. Dalton and Van Schaik 2006 for a catalogue of the Tibetan tantric manuscripts from Dunhuang. I have, however, noticed what I find to be a curious distinction in ritual structure between the Indian Guhyasamāja sādhanas (and in fact almost all other tantric sādhanas) preserved in the Tibetan Tengyur and those preserved at Dunhuang, which I believe suggests that the Dunhuang Guhyasamājarelated manuscripts somehow relate to a different “strand” of liturgical works than those preserved in the Indic sources. That is, throughout Dalton and van Schaik’s catalogue there are references to the three samādhis “of Mahāyoga,” by which Dalton and van Schaik mean the samādhi of suchness (de bzhin nyid kyi ting nge ‘dzin), the all-illuminating samādhi (kun tu snang ba’i ting nge ‘dzin) and the causal samādhi (rgyu’i ting nge ‘dzin). Dalton (2004, 9) takes these same three samādhis to be characteristic of the generation stage practice of the Mahāyoga tantras on the whole. Indeed, in the Tibetan Mahāyoga practices of the Nyingma tradition up until the present day 25 Guhyasamāja practice traditions, Buddhajñānapāda’s is the earlier one. The Ārya School’s later date is determined both by its reliance not on the Guhyasaṃāja-tantra itself for its maṇḍala and ritual system, but on the later Guhyasamāja explanatory tantras—the Vajramālā, the Caturdevīparipṛcchā, the Saṃdhivyākaraṇa, and the Vajrajñānasamuccaya—and also by its more developed exegetical tradition.127 these three are exactly the three samādhis that pertain to the generation of the deity in generation stage practice, and indeed these are the three samādhis as they are represented in sādhanas and other Mahāyoga works at Dunhuang, including those that pertain to the Guhyasamāja-tantra. But when we come to Indic liturgical texts, we find a different story. A search of the Tibetan Kangyur and Tengyur shows only four texts—two tantras and two commentaries—that set forth the particular system of three samādhis common at Dunhuang and in the later Tibetan Nyingma tradition. One of these four, though, is the Dgongs pa ‘dus pa’i mdo, which Dalton (2002) has shown to be a Tibetan composition rather than an Indian text. The other three are the Śrīherukakaruṇākrīḍita-tantra, translated by Śrīkirti, which is one of the Rnying rgyud; Mañjuśrīmitra’s Bodhicittabhāvanādvādaśārthanirdeśa (Tōh. 2578), which has no translator’s colophon; and Vajravarman’s Vajravidāraṇādhāraṇīvṛtti, translated by Śraddhākaravarman and Rinchen Zangpo. (The Śrīherukakaruṇākrīḍita-tantra and Mañjuśrīmitra’s text both list only the second and third samādhis—the kun tu snang ba’i ting nge ‘dzin and the rgyu’i ting nge ‘dzin by name— but both also describe a samādhi prior to these two. The description of that first samādhi in both of these texts can, however, be correlated to the de bzhin nyid kyi ting nge ‘dzin.) Thus we find this particular set of three “Mahāyoga” samādhis in only three Indic texts surviving in the Tibetan canon, one of which is a from the Rnying rgyud section of the Kangyur, and another of which lacks a translator’s colophon. What we do find frequently in the Kangyur and Tengyur, as well as in the surviving Sanskrit manuscripts, like those of the Sāramañjarī, is a different set of three samādhis, those already well known from the Yoga tantras. These three samādhis continue to be used throughout the liturgical literature preserved in Sanskrit and in Tibetan translation of the so-called Mahāyoga tantras (including every single reference in the Kangyur and Tengyur to the three samādhis within the Guhyasamāja corpus), all the way up through works on the later Yoginī tantras like the Cakrasaṃvara, Hevajra, Sampuṭa-tantras. These other three samādhis, which are used in the Jñānapāda School and the Ārya School alike to structure their generation stage practices are the preliminary yoga samādhi (ādiyoga-samādhi), the supremely victorious maṇḍala samādhi (maṇḍalarājāgrī-samādhi), and the supremely victorious action samādhi (karmarājāgrī-samādhi). Tanaka (1996, 259), who works primarily with Indic texts, has observed that, “The stage of generation in late Tantric Buddhism is divided into three further stages, called ādiyoga nāma samādhi, maṇḍalarājāgrī nāma samādhi, and karmarājāgrī nāma samādhi...” We can compare this to the observation of van Schaik, who works primarily with Dunhuang texts, that “Meditation in Mahāyoga sādhanas tends to proceed along the structure of the three concentrations (ting nge ‘dzin, Skt. samādhi)...These three are: (i) The concentration on suchness (de bzhin nyid), (ii) the concentration on total illumination (kun tu snang ba), and (iii) the concentration on the cause (rgyu)” (van Schaik 2012, 13). The fact that there is such a clear difference in the set of three samādhis that are consistently used in Tibetan Dunhuang tantric texts and those that are consistently used in the Indic tantric texts suggests to me that these two groups of texts pertain to, for lack of a better term, different ritual “strands.” Certainly, more research is necessary to determine more about the relationship between the Guhyasamāja-related texts from Dunhuang and those preserved in the Tibetan canon. Christian Wedemeyer is currently working on some of the Dunhuang Guhyasamāja sādhanas; his work will be a very welcome addition to the very little research that has thus far been done on this topic. 127 See Matsunaga 1977, 115-16. Despite the fact that it is historically later, the Ārya School has received by far the most scholarly attention, both within the Tibetan scholastic tradition and the modern academy. There are several reasons for this. First, it was the Ārya School’s practice tradition of Guhyasamāja that became popular in Tibet, especially in the Tibetan Gelugpa school, and consequently a number of scholarly commentaries and sādhanas have been composed following that tradition up to the present day. The presence of a living Tibetan tradition and indigenous Tibetan commentarial literature on the Ārya School has made its Indian source texts (whether in the original Sanskrit or in Tibetan translation) much more approachable for modern scholars. Additionally, the Ārya School has fascinated modern scholars because of the curious names of its major exponents—Nāgārjuna, Āryadeva, and Candrakīrti—which are the very names of earlier Indian philosophers associated with the Madhyamaka philosophical tradition founded by the “original” Nāgārjuna, i.e. the non-tantric author of the Mūlamadhyamikakārikās. See Wedemeyer 2007, 7-43 for a detailed treatment of this issue. Moreover, a larger number of Indian texts pertaining to this tradition survive in their original Sanskrit. Following Alex Wayman’s (1977) pioneering inquiry into the Ārya School of Guhyasamāja exegesis and practice—an informative if somewhat rambling monograph—more recent scholarship (Tanaka 1999-2004c; Tomabechi 2006; Wedemeyer 2007; Campbell 2009; Hong and Tomabechi 2009; Wright 2010; Bentor 2010; Kittay 2011; and Columbia University team, forthcoming) has devoted some further attention to the study and translation Ārya School texts. With respect 26 The two systems can be generally distinguished by their distinct generation stage maṇḍalas, their perfection stage practices, and the number of initiations in their respective systems. The Jñānapāda School maṇḍala has nineteen deities and centers on Mañjuvajra, as laid out in Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, whereas the Ārya School has a thirty-two deity maṇḍala centering on Akṣobhyavajra as laid out in Nāgārjuna’s Piṇḍīkṛtasādhana. Regarding the perfection stage practices, Buddhajñānapāda’s system (which we must again note has been far, far less studied in the modern academy) is less formally structured than that of the Ārya School but is characterized by a system of three bindu yogas that include the practice of vajrajāpa and several (but not all) of the yogas from the classical ṣaḍanga or sixbranch yogas. These practices are detailed both in his Dvitīyakrama and his Muktitilaka. The Ārya School’s perfection stage system follows the structure of the five stages (pañcakrama) according to Nāgārjuna’s Pañcakrama and Candrakīrti’s Caryāmelāpakapradīpa.128 The early texts of the Jñānapāda School preserve a system of just three initiations, concluding with the prajñājñāna initiation (as it seems the fourth initiation had not yet fully developed in Buddhajñānapāda’s time), whereas the Ārya School initiatory rituals include the fully developed system of four initiations.129 Given the paucity of modern scholarship on the Jñānapāda School our understanding of the relationship between the two traditions continues to develop, and the present study hopes to make some steps toward that understanding. 3. Buddhajñānapāda’s Life Everyone having come, in this way, to know the detailed accounts of my life, should use all methods to please the sublime and sincere learned one, and listen to and contemplate his teachings and compositions. Through relying upon this, remaining in isolated places, training one’s mind in suchness, and genuinely realizing the way things are, it is possible to attain awakening in this very life, or even in just six months— who could refute this!? -Buddhajñānapāda, Dvitīyakrama Name to the study of Ārya School texts Wright (2010) and Tomabechi (2006) have made editions, translations, and studies of, respectively, Nāgārjuna’s Piṇḍīkṛta, the central Ārya School generation stage sādhana, and his Pañcakrama, a perfection stage manual. Their studies of these systems are supplemented by Wedemeyer’s (2007) Tibetan and Sanskrit edition, translation, and study of Āryadeva’s Caryāmelāpakapradīpa, which further elaborates Nāgārjuna’s perfection stage system, and Tomabechi’s (2006) Tibetan edition of the same text. In a series of articles Tanaka has presented an edition of Nāgabodhi/Nāgabuddhi’s Śrī-guhyasamājamaṇḍalopāyikā-viṃśati-vidhi, a maṇḍala ritual according to the Ārya School’s Guhyasamāja maṇḍala (See Tanaka 1999-2004c). Hong and Tomabechi (2009) present an edition of and brief introduction to Candrakīrti’s Ārya School Vajrasattvasādhana. Campbell (2009), Bentor (2010), Kittay (2011), and the team from Columbia (forthcoming) all deal primarily with the hermeneutical approach of the Ārya School, rather than its ritual systems. Campbell’s work considers the hermeneutics of Candrakīrti’s Pradīpoddyotana, an Ārya School commentary on the Guhyasamāja root tantra, an edition and translation of which are currently being prepared by a team at Columbia University. Bentor examines the Ārya School interpretations of a single verse from the root tantra, while Kittay looks at the question of hermeneutics with respect to the Vajramālā, an explanatory tantra of the Guhyasamāja-tantra that sets out the maṇḍala and ritual system of the Ārya School. Kittay’s study includes a complete English translation of the Vajramālā from the extant Tibetan translation of the text (the Sanskrit is no longer extant) but does not include a Tibetan edition. 128 Although the Tibetan tradition asserts the Caryāmelāpakapradīpa to be a “meaning commentary” (don ‘grel) on Nāgārjuna’s Pañcakrama, Tomabechi’s analysis of the relationship between these texts has led him to conclude that the Caryāmelāpakapradīpa is actually earlier than the Pañcakrama (Tomabechi 2006, 36-38). 129 Vaidyapāda’s Yogasapta, which I examine in Chapter Seven, shows that “the fourth,” though not yet identified as a separate initiation was indeed part of early Jñānapāda School practice. 27 Even before we address the details of his life and work, a few words ought to be said about Buddhajñānapāda’s name, which is reported with quite some variety in the works available to us which include Buddhajñānapāda’s own writings, the colophons of the Tibetan translations of his works, the writings of his direct disciples and other Indian authors (in Tibetan translation and in Sanskrit), and indigenous Tibetan texts.130 One feature of some, but not all, of Buddhajñānapāda’s writings, is the inclusion of his own name, cleverly inscribed within the dedicatory verses. In all five of the works in which he does this, he uses the words buddhajñāna (sangs rgyas ye shes), making it likely that this was the name he used for himself.131 Pāda is added as an honorific suffix to the names of quite a number of Indian masters, including many in Buddhajñānapāda’s lineage. Although Buddhajñānapāda (sangs rgyas ye shes zhabs), 132 and Buddhajñāna (Sangs rgyas ye shes),133 are the most common names found in the colophons to his works, we find other versions of his name in the Tibetan colophons, as well: Buddhaśrījñāna (Sangs rgyas dpal ye shes),134 and Buddhaśrījñānapāda (Sangs rgyas dpal ye shes zhabs/ Sangs rgyas dpal kyi ye shes zhabs).135 In Indic works by his direct disciples and later writers we find a similar variety: Buddhajñanapāda, Buddhajñāna, Jñānapāda (ye shes zhabs), Śrījñānapāda (Dpal ldan ye shes zhabs), and Buddhaśrījñānapāda. Tibetan authors similarly run the gamut, naming him Buddhajñānapāda, Buddhajñāna, Buddhaśrījñāna, and Jñānapāda. Adding śrī to various places in his name was presumably done out of respect, and shortening names is also common practice. We can certainly not assume that all of the works in the canon ascribed to authors under these names were written by one and the same Buddhajñānapāda (and I will address the attribution of each of the extant works under the whole variety of these names below), but it is clear that a wide variety of names was used in various contexts to refer to the particular individual whom I here refer to as Buddhajñānapāda. Not only does this seem to be the most commonly used of his various names (though the shortened version, Jñānapāda may also be in the running for common usage), but since it seems he called himself Buddhajñāna, I have settled on Buddhajñānapāda. Dates and Early Life as a Student We have already heard the story of Buddhajñānapāda’s life as he himself tells it, along with the details that Vaidyapāda adds to his guru’s autobiographical narrative. These early Indian accounts were supplemented by later Tibetan histories whose reports of Buddhajñānapāda’s life, albeit written hundreds of years after his death,136 contain additional details not found in the 130 In the secondary literature he is usually referred to as Jñānapāda, Buddhaśrījñāna, Buddhaśrījñānapāda, or Buddhajñānapāda. 131 The use of his name in the dedicatory verses is found in his Caturaṅga/Samantabhadra-sādhana, Muktitilaka, Ātmasādhanāvatāra, *Gativyūha, and Sañcayagāthā-pañjikā. 132 In the colophons of his Guhyajambhalasādhana (D and P), *Gativyūha (D), and Mahāyānalakṣaṇasamuccaya (D and P). It is also found in the colophon of Śākyamitra’s Mukhāgama, which is often attributed to Buddhajñānapāda. 133 In the colophons of his Caturaṅga/Samantabhadra-sādhana (D and P), Ātmasādhanāvatāra (D and P), Śrīherukasādhana (D and P), Bhaṭṭārakāryajambhalajalendra-sādhana (D and P), and *Gativyūha (P). 134 In transliteration as Bud dha śrī jñā na in his Sañcayagāthā-pañjikā (D), and probably erroneously in transliteration as Bud dha śrī ka jñā na in the Sañcayagāthā-pañjikā (P). 135 In the colophons of the Dvitīyakrama (D and P), the Muktitilaka (D and P), and the Caturaṅgasādhana (D and P). 136 The Tibetan histories I have relied upon in this study are Chögyal Phagpa’s 13th-century Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa; Gö Lotsāwa Zhonnu Pel’s 15th-century Deb ther sngon po (see also Roerich 1976); Tāranātha’s 17th-century Rgya gar chos ‘byung (See also Chimpa and Chattopadhyaya 1970) and Bka’ babs bdun (See also Templeman 1983); Amye Zhab’s 17th-century Gsang ‘dus chos ‘byung and his Gshin rje chos ‘byung; and Dudjom Jigdral Yeshe Dorje 20th-century History of the Nyingma School. This is certainly not an exhaustive list of the Tibetan accounts of Buddhajñānapāda’s life. I have not been able to include here all of the 28 Indian sources and appear to thus be based both on the Indian sources (i.e. Buddhajñānapāda, Vaidyapāda, and other Sanskrit sources no longer available to us), as well as possibly on oral history.137 Vaidyapāda seems to assume knowledge of such oral histories already in his commentary on the Dvitīyakrama when he suggests that the “detailed accounts” that Buddhajñānapāda mentions include “the taming of Nālandā, making offerings at Vajrāsana, [the account] of the consecration and others.”138 Since Vaidyapāda himself offers no further details, it seems he assumes his reader will know of the events he references simply by giving these names. As we shall see, although some of the Tibetan histories elaborate on and interpret aspects of Buddhajñānapāda’s life story in ways that seem to serve an apologetic or polemical function, we have no outright reason to doubt some of the details that are added in others of them, and they certainly add richer texture to what we know of Buddhajñānapāda’s life. We are unusually fortunate in the realm of medieval Indian history to be able to give fairly precise dates for Buddhajñānapāda’s life. Of the six human teachers he mentions in the Dvitīyakrama (setting aside for the moment his most important guru, Mañjuśrī himself), three are historically identifiable figures—Haribhadra, Vilāsavajra, and Pālitapāda—all of whom can be dated to the 8th century. In the colophon of his Abhisamayālaṃkārālokā Haribhadra states that it was completed at the Trikaṭuka Monastery under the reign of the Pāla king Dharmapāla (r. ca. 775-812),139 and Tāranātha notes that Dharmapāla was a patron of both Haribhadra and Buddhajñānapāda, describing both masters as Dharmapāla’s gurus.140 Vilāsavajra likewise lived in the latter part of the 8th century.141 Moreover one of Buddhajñānapāda’s own works, completed in the early part of his life, appears in the lDan kar ma catalogue which was completed in 824.142 On this basis we can reliably place Buddhajñānapāda’s life in the latter part of the 8th and early part of the 9th century. In a recent article Péter Szántó has drawn attention to an account in Atīśa’s

• Bodhipathapradīpapañjikā in which King Devapāla, Dharmapāla’s heir, is said to have made

an offering of his kingdom, his queen, and himself, to Buddhajñānapāda, which he later ransomed back with their weight in gold.143 This would mean that Buddhajñānapāda’s activity details included in these histories that differ from or supplement Buddhajñānapāda’s and Vaidyapāda’s accounts— they are divergent enough that such an endeavor would require a separate study, and my main concern here is with Buddhajñānapāda’s life and thought as we find it presented in the Indic tradition, rather than the later Tibetan interpretations of his life and tradition. However, I have included some accounts from the Tibetan sources, as the Tibtean historians do significantly supplement our knowledge of Buddhajñānapāda’s life, particulary several accounts from his later life that are only briefly mentioned in Vaidyapāda. 137 With regard to other no longer extant or accessible Indic sources, Tāranātha, whose accounts of Buddhajñānapāda’s life include quite a number of details not found in the Indian texts known to us, explains that his work was based on several Sanskrit sources that now appear to be lost, including one of the Magadhan scholar Sa dbang bzang po, whose work covers the history up to the reign of Rāmapāla (r. ca. 1072-1126) (Sanderson 2009, 89). The later Tibetan accounts, moreover, unsurprisingly draw from the earlier Tibetan accounts, as well. 138 na landa ([landa] P, lendra D) ‘dul ba dang/ rdo rje gdan gyi mchod pa byas pa dang/ rab tu gnas pa byas pa la sogs pa’i lo rgyus (Sukusuma, D 135b.4; P 163a.8). Elsewhere in his commentary Vaidyapāda references and cites from a number of textual sources, so it seems that these references to stories of Buddhajñānapāda’s life, which are mentioned by topic rather than by referening a particular text, do probably refer to oral accounts. 139 Chimpa and Chattopadhyaya 1970, 266n4; Ruegg 1981, 101n320; and Sanderson 2009, 93-4. 140 Rgya gar chos ‘byung, 262; Chimpa and Chattopadhyaya 1970, 274. The Pālas were, to quote Sanderson, a “robustly” Buddhist East Indian dynasty who used the Buddhist dharmacakra as a royal emblem, whose inscriptions began with praise to the Buddha, and a number of whose rulers are described in the manuscript and inscriptional record with the epithet paramasaugataḥ (Sandserson 2009, 87). 141 Tribe 1994, 9-23. 142 Tomabechi (2008, 175) cites Hadano on this point. This may refer to his Sañcayagāthā-pañjikā, the translation of which is indeed attributed to the Imperial era translators Kawa Paltsek and Vidyākarasiṃha. 143 Szántó 2015, 539. 29 stretched perhaps further into the 9th century than previously thought, that is at least until some time after 812 CE when Devapāla ascended the throne. However, an identical account of the king offering himself and his queen to Buddhajñānapāda and later ransoming themselves back by paying their weight in gold is also recorded in Tāranātha’s Seven Transmissions (Bka’ babs bdun) and a different version of the story in which the king builds a temple in gratitude to Buddhajñānapāda is found in Chögyal Phagpa’s Biography and Lineage History of Jñānapāda’s Guhyasamāja. In both of those sources, though, the king in question is identified as Dharmapāla, rather than Devapāla.144 While these accounts at least place in question Atīśa’s version of the story involving King Devapāla, Tāranātha seems to have either received mistaken information or misunderstood his sources on the Pāla succesion, as he incorrectly places Dharmapāla after Devapāla rather than before him and reverses some (but not all!) information about their respective reigns and activities.145 Moreover, Atīśa was both geographically and historically much closer to the events than either Chögyal Phagpa or Tāranātha, which perhaps makes his report more likely to be accurate. Another factor to consider is that Buddhajñānapāda is asserted by Tāranātha to have been the first tantric ācārya at Vikramaśīla, and it seems likely that it was Devapāla, not Dharmapāla (as Tāranātha states) who founded Vikramaśīla Monastery. Devapāla’s having constructed Vikramaśīla is mentioned in the colophon of Anupamavajra’s Ādikarmapradīpa,146 and this is further substantiated by the colophon of Atīśa’s Ratnakaraṇḍodghāta, where he states that he composed the work “At the great temple called Vikramaśīla, the commitment of Devapāla,” presumably indicating that the construction of Vikramaśīla was the fulfilment of one of King Devapāla’s tantric commitments.147 If that is the case, and if Buddhajñānapāda was indeed Vikramaśīla’s first tantric ācārya, then he did very likely live into Devapāla’s reign. The late 8thcentury dating of his gurus Haribhadra and Vilāsavajra would also suggest the likelihood of Buddhajñānapāda’s having lived at the end of the 8th century, and well into the 9th.148 But if Tāranātha’s account of his life, in which Buddhajñānapāda is said to have lived for 80 years, is correct, then his life could easily have spanned a good part of both of those centuries. Buddhajñānapāda’s own account of his life tells us nothing of his birth or childhood but begins only with his studies. It appears not to be until an account of his life written by Chögyal Phagpa in the 13th century, that we learn more about his early life, though the information in our sources here varies significantly. Chögyal Phagpa tells us that Buddhajñānapāda was born in a place called Sindhura in southeast India to a king called Gyaparuprabhava (!?),149 whereas Tāranātha, in the 17th century, reports that according to some sources (to which we unfortunately no longer have access) Buddhajñānapāda was a brahmin who was ordained at Nālandā into the Mahāsāṃghika school, while according to others he was a kṣatriya “reader” (scribe?).150 As 144 (Bka’ babs bdun, 106; Templeman 1983, 74; Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 617 145 See Sanderson 2009, 90-1. Tāranātha does, though, correctly assert Dharmapāla to roughly be a contemporary of King Trisong Deutsen (r. ca. 755-797) (Rgya gar chos ‘byung, 264; Chimpa and Chattopadhyaya 1970, 276). 146 Sanderson 2009, 91. 147 de wa pā la’i thugs dam bi kra ma/ shī la zhes bya’i gtsug lag khang chen du (Ratnakaraṇḍodghāta, D 116b.4). 148 See Sparham 1989, 3 on Haribhadra’s dates and Tribe 1994, 9-23 on Vilāsavajra’s. 149 Rgyal po gya pa ru pra bha wa. (Gsang ‘dus ye shes zhabs kyi rnam thar dang brgyud pa’i rim pa, 610). Amye Zhab’s 17th century Gshin rje chos ‘byung repeats this information (Gshin rje chos ‘byung, 304a.4). His account begins as a slightly shortened, but otherwise word-for-word, copy of Chögyal Phagpa’s account though there are places where Amye Zhab follows Vaidyapāda more or less word-for-word rather than Chögyal Phagpa. When he comes to the account of Buddhajñānapāda’s meeting with the monk in the Kuvaca forest, however, Amye Zhab’s account diverges from both. 150 Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 610; Bka’ babs bdun, 103; Templeman 1983, 71. On Tāranātha’s sources, now lost to us, see note 137. Tāranātha’s statement about Buddhajñānapāda’s ordination 30 Buddhajñānapāda himself tells us, he spent a number of years in the earlier part of his life traveling quite extensively throughout the subcontinent studying and practicing under the guidance of different gurus with whom he remained for varying amounts of time—he mentions staying with one guru for just eight months and another for nine years. The teachers Buddhajñānapāda mentions in his account are Haribhadra, with whom he studied in the town of Takṣaśilā, in the area of Khapir, in Magadha;151 Vilāsavajra, Guṇeru and Jātig Jālā, with whom he studied in Uḍḍiyāna;152 Bālipāda, who lived in Ko no dze in the area of Jālandhara;153 and Pālitapāda, who stayed at “the place with sky trees” in the Koṅkan, most probably at modern-day in the Mahāsāṃghika school is the only one I have seen that identifies him as having taken monastic ordination, though his early career studying with Haribhadra and teaching at Nālandā are suggestive of his having monastic status, at least in the early part of his life. As for Chögyal Phagpa’s account of Buddhajñānapāda’s life, he closely follows Vaidyapāda’s narrative from the account of Buddhajñānapāda’s studies with Haribhadra onward, but also includes episodes not reported in Vaidyapāda. His source for the earlier information on Buddhajñānapāda’s birth and parentage is unclear, but its style of delivery (in addition to his being the son of a king, he also notes that from a young age Buddhajñānapāda was handsome and intelligent, etc.) is classically hagiographical. Phagpa also implausably states that from Pālitapāda (who he styles Balipata) Buddhajñānapāda received initiation into the thirtytwo deity Guhyasamāja maṇḍala (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 612). (Tāranātha also states that in the Koṅkan Pālitapāda heard teachings directly from Candrakīrti and received the text of the Pradīpodyotana, so there do appear to be some accounts of Pālitapāda’s having been an Ārya School practitioner (Rgya gar chos ‘byung, 260; I disagree with the translation of this passage in Chimpa and Chattopadhyaya 1970, 273).) Phagpa’s account also contains the first version that I have seen of the expanded (from Vaidyapāda’s narrative) story of Buddhajñānapāda’s meeting with Mañjuśrī, in which he travels toward Wutaishan in hopes of meeting Mañjuvajra. I have translated this episode in full below. Tāranātha, Amye Zhab, and Dudjom all report abbreviated forms of this same account, which seem to be based on Phagpa’s, or else on the same sources he was using, though Phagpa’s hagiography contains several accounts not found in either of the others. See Templeman 1983, 72-3 and Dudjom 1991, 495-6 for English translations of Tāranātha’s and Dudjom’s versions of this story, respectively. (Templeman’s translation, in particular, suffers from a number of errors, some of which are understandable given that he was apparently not familiar with Vaidyapāda’s text, from which Tāranātha derives much of his material). While Phagpa’s account of Buddhajñānapāda’s life in some places seems implausible, and in this particular episode is so detailed as to seem suspicious (it even includes dialogue between Buddhajñānapāda, the monk, and the woman!), he reports multiple versions of certain portions of the account (some of which are included in Tāranātha’s, Amye Zhab’s, and Dudjom’s later tellings, and others of which are not), suggesting he was relying on multiple sources. At one point Phagpa details an account of Buddhajñānapāda’s visit to his guru Pālitapāda’s residence which is reported in Samantabhadra’s Sāramañjarī. It is possible, then, that other of these more detailed accounts of certain episodes of Buddhajñānapāda’s life were present in Indian sources we no longer have access to, or are at present buried in extant texts, like the account of the visit to Pālitapāda in the Sāramañjarī, and are simply waiting to be unearthed. 151 The toponyms mentioned in the Dvitīyakrama and the Sukusuma are difficult to ascertain with certainty. The place names mentioned here in the context of Buddhajñānapāda’s studies with Haribhadra are Dbu kyi yul chen, Kha pir (pir] P, bir D), and Rdo ‘jog. Rdo ‘jog is a common translation of Takṣaśilā (see C. Dalton and Szántó, forthcoming). However, the region in which the town is said to be located, Kha pir, calls into question this being the commonly-known city of Takṣaśilā located in modern-day Pakistan. On one hand, Kha pir may be a corrupted rendering of Kaspir, i.e. Kaśmir (see C. Dalton and Szántó, forthcoming). However, since this particular Khapir is specified as being in Magadha, such an identification is only possible if Magadha is understood to mean the Indian subcontinent more broadly, rather than the region of Magadha, which is not near Kaśmir. But Vaidyapāda describes Magadha as “in the area of Nālandā,” which again renders the identification of Khapir as Kaśmir difficult (Sukusuma, D 89a.7). 152 We may assume Uḍḍiyāna to be the region in the northwest of the subcontinent, often identifed as the Swat Valley in modern-day Pakistan. 153 Again here we have some difficulty identifying these locations. At first glance Ko no dze, or in the Derge edition of Vaidyapāda’s commentary Ka no dze (P reads ko no dze), does seem to be a transliteration of Kannauj, and Davidson (2002, 312) has rendered it as such. However, Szántó (2015, 542-3) places some doubt on this identification, since modern-day Kannauj is not near the modern-day city of Jalandhar, and C. Dalton and Szántó (forthcoming) note that at the time Kannauj was referred to as Kanyākubja, making the identification even less likely. 31 Kadri.154 His most important guru was Mañjuśrī himself, who appeared to Buddhajñānapāda as an “emanated monk” who then emanated Mañjuśrī and his maṇḍala for Buddhajñānapāda in the Kuvaca forest behind Vajrāsana.155 While the exact locations of these encounters are difficult to ascertain with certainty based on the toponyms given in his account, what is clear is that Buddhajñānapāda’s travels took him for thousands of kilometers across a wide swath of the subcontinent from its north-central area to the far northeast, to the southeast, and finally to the northeast. Buddhajñānapāda’s Gurus and Their Possible Influence on his Thought Haribhadra As for his teachers and his studies, among those figures we are able to identify and about whom we know more from other sources, Haribhadra, the only non-tantric guru that Buddhajñānapāda mentions studying under,156 is a well-known late 8th-century scholar of Prajñāpāramitā literature whose Abhisamayalaṃkārālokā remains important in many Tibetan monastic curricula even up to the modern day. Despite his own preference for writing on tantric topics, Buddhajñānapāda, who is considered Haribhadra’s principle disciple, was clearly influenced by this master. One of Buddhajñānapāda’s two non-tantric works, the Sañcayagāthāpañjikā, composed while he was staying at Nālandā early in his life and which is by far the longest of all his extant writings, is an extensive Prajñāpāramitā commentary. In this work he follows in Haribhadra’s footsteps in synthesizing Madhyamaka philosophy with the Abhisamayālaṃkāra.157 Vilāsavajra Vilāsavajra, the first tantric guru that Buddhajñānapāda mentions, is likewise a wellknown late 8th-century author158 who wrote treatises on a number of tantras including the Mañjuśrīnāmasaṃgīti, the Guhyagarbha-tantra, and the introductory section (nidāna) of the Guhyasamāja-tantra.159 If indeed the Vilāsavajra who authored the Nāmamantrārthāvalokinī 154 nam mkha’i shing ldan. Vaidyapāda’s etymology for the toponym reads: de la nam mkha’i shing ldan zhes bya ba ste ci phyir zhe na/ rtsa ba med par shing rnam ‘khril (P] ‘khril; D la ‘khris) shing steng du bras (P] bras; D bris) pa lta bur gnas pa’o// (D 90a.4). According to Khenchen Chodrak Tenphel the term “sky tree” (nam mkha’i shing) means mangrove (or perhaps banyan?) (personal communication, March, 2016). The toponym was earlier identified by Davidson as Kāṇherī, based upon his reading of Vaidyapāda’s gloss of the name (2002, 313). I am more convinced, however, by Szántó’s more recent work, which suggests that the place mentioned may rather be Kadri, which is in the Koṅkan near Mangalore, rather than Kānherī (i.e. Kṛṣṇāgiri) which is not usually understood to be part of the Koṅkan. See Szántó (2015, 550-52) for the full details of his assessment, which I also discuss in more detail below in conjunction with Buddhajñānapāda’s guru Pālitapāda. 155 Vajrāsana, we can safely say, is at Bodhgaya in modern-day Bihar. 156 Haribhadra wrote four Prajñāpāramitā texts, but no tantric work attributed to him survives either in Sanskrit or in Tibetan translation in the Tengyur. He was, however, certainly familiar with tantric systems, as he advocates in one of his works that practitioners should visualize their meditational deity in the form of Vajradhara (Sparham 1989, 3). Haribhadra was himself the disciple of the unknown Vairocanabhadra and of the well-known Śāntarakṣita, under whose tutelage he was a co-disciple of Kamalaśīla (ibid., 2-3). 157 Ruegg 1981, 101-2. However, it appears that Buddhajñānapāda did not follow his guru Haribhadra’s interpretation of there being four kāyas instead of the three taught in the Abhisamayālaṃkāra (Makransky 1997, 6). The Sañcayagāthā-pañjikā is the only one of Buddhajñānapāda’s extant works that I did not read in full as part of this study. This work, and particularly its relationship to Haribhadra’s Prajñāpāramitā writings, is a topic that certainly deserves further attention. 158 On Vilāsavajra’s dates see Tribe 1994, 9-23. 159 Vilāsavajra’s Mañjuśrīnāmasaṃgīti commentary is the Nāmamantrārthāvalokinī, his Guhyagarbha commentary is the Spar khab, and his Guhyasamāja commentary the Śrīguhyasamājatantranidāṇagurūpadeśabhāṣya. In his 32 commentary on the Mañjuśrīnāmasaṃgīti was Buddhajñānapāda’s guru, we find in that text the unusual instance of a guru citing his disciple’s work.160 However, this indeed appears to be the case here; the work cited by Vilāsavajra, the Mahāyānalakṣaṇasamuccaya, was composed by Buddhajñānapāda early in his career, most probably prior to his discipleship under Vilāsavajra.161 Moreover, we can say with certainty that Buddhajñānapāda knew Vilāsavajra’s writing on the Mañjuśrīnāmasaṃgīti, since in his Ātmasādhanāvatāra Buddhajñānapāda reproduces (without telling us he is doing so) a lengthy section of Chapter Five of Vilāsavajra’s Nāmamantrārthāvalokinī concerning the correspondences between the deities from the Mañjuśrīnāmasāṃgīti maṇḍala and a number of Mahāyāna categories.162 We can discern a number of possible ways in which his studies with Vilāsavajra may have influenced Buddhajñānapāda’s thought. Vilāsavajra’s Nāmamantrārthāvalokinī is written with an emphasis on Vijñānavādin perspectives combined with an acknowledgement of Madhyamaka systems163 that is likewise prominent in Buddhajñānapāda’s writings, and the central figure of Mañjuśrījñānasattva in the sādhana found in the fourth chapter of that work, which we now know that Buddhajñānapāda was familiar with, may perhaps have influenced Buddhajñānapāda’s own use of Mañjuvajra as the central figure of the generation stage maṇḍala in his Samantabhadra-sādhana.164 In any case, it seems likely that the identification of Mañjuśrī as an Ādibuddha figure in the Mañjuśrīnāmasaṃgiti and the fact that this practice system was taught by his guru may have been a factor in Buddhajñānapāda’s important personal connection with Mañjuśrī. And yet while Vilāsavajra also composed a Guhyasamāja-tantra commentary, and the Guhyasamāja was clearly the most important tantra for Buddhajñānapāda, there is no mention of Buddhajñānapāda’s having studied the Guhyasamāja with him.165 The most intriguing possibility of Vilāsavajra’s thought influencing Buddhajñānapāda’s concerns the former’s relationship with the Guhyagarbha-tantra. While Buddhajñānapāda makes no mention at all of the Guhyagarbha in his oeuvre, he is mentioned in some later Tibetan histories and polemical treatises as an Indian author who uses the term “great perfection” (rdzogs chen) in his works. The term indeed appears in two instances in Buddhajñānapāda’s writings in Tibetan translation (unfortunately the Sanskrit of neither passage is extant), but the context of its usage in these instances is less suggestive of a connection with great perfection traditions than are a number of other strains of his thought, including references to the immediate and direct study of the Nāmamantrārthāvalokinī Tribe notes that while the ascription of the Guhyasamāja commentary and the Guhyagarbha commentary cannot be definitively made to the author of the Nāmamantrārthāvalokinī, there are also no substantial reasons for doubting the ascription (Tribe 1994, 14). J. Dalton likewise asserts the same with regard to the attribution of authorship of the Spar khab to Vilāsavajra (J. Dalton 2005, 125n28). 160 This was first noted by Tribe (1994, 16). 161 Szántó 2015, 541 and C. Dalton and Szántó forthcoming. 162 This passage in the Ātmasādhanāvatāra reproduces not a continuous segment of Chapter Five of the Nāmamantrārthāvalokinī, but rather a number of shorter passages from that chapter that Buddhajñānapāda has strung together to create several pages of the Ātmasādhanāvatāra. The presence of this passage is especially significant since the Nāmamantrārthāvalokinī is extant in Sanskrit, so we now have access to yet another passage of the Ātmasādhanāvatāra in its original Sanskrit. I address this further below in my discussion of the Ātmasādhanāvatāra. 163 Tribe 2016, 11. 164 Tribe (1994, 8n20) suspects as much. Obviously Buddhajñānapāda’s vision of Mañjuśrī would have been a more important factor in this decision. 165 Vaidyapāda mentions only that Buddhajñānapāda studied “many Kriyā and Yoga tantras” with Vilāsavajra (Sukusuma, D 89b.5). As mentioned above, Vaidyapāda classifies the tantras in a number of different ways in the Sukusuma (see note 100), and it is not clear whether he would have included the Guhyasamāja-tantra under the category of Yoga tantras or held it to be exclusively a Mahāyoga tantra. In any case the Guhyasamāja is not explicitly mentioned as a topic of study with Vilāsavajra as it is in the case of Buddhajñānapāda’s later guru Pālitapāda. 33 pointing out of reality by a guru to a disciple, an emphasis on the immediacy of awakening, and overtly anti-ritual rhetoric in the midst of a deeply ritual practice system.166 Pālitapāda The guru with whom Buddhajñānapāda studied the Guhyasamāja-tantra and whom he reports attending for nine years and then returning to visit again later in his life, Pālitapāda, has only recently emerged as a figure about whom more is known. Buddhajñānapāda describes studying with this guru in the Koṅkan in a location called nam mkha’ shing ldan, literally “the place with sky trees,” and reports that he was surrounded by disciples who had miraculous abilities and that the whole entourage regularly recieved extensive support for their livelihood and practice. The unusual toponym mentioned for the location in the Koṅkan where Pālitapāda resided has been identified by Davidson as Kāṇherī and more recently by Szántó as Kadri, near present-day Mangalore.167 Szántó’s assessment has the advantage of being the location of a Śaiva temple whose deity is called Mañjunātha, and an ancient inscription on the site identifies the place as a vihāra.168 These details suggest that the temple was originally a Buddhist site devoted to Mañjuśrī, which of course fits well with Buddhajñānapāda’s system of Guhyasamāja whose central deity is Mañjuvajra, and suggests the possibility that yet another of Buddhajñānapāda’s gurus, in addition to Vilāsavajra, may have taught tantric practices centered on Mañjuśrī. Such a possibility is furthered by Szántó’s report of a Sanskrit initiation manual called the Parikramapadopāyikā composed by a certain Śrīkīrtipāda who identifies his own guru as Pālitapāda, and mentions that guru as having taught maṇḍala rituals, presumably an initiation manual, in whose spirit Śrīkīrtipāda composed his own work.169 Szántó further reports that the anonymous Mañjuvajrodaya (Tōh. 2590) shares significant parallels with Śrīkīrti’s manual, leaving the alluring possibility that the Mañjuvajrodaya might possibly be Pālitapāda’s work, on which his disciple Śrīkīrti based his own initiation manual.170 Though Szántó does not report this, the Mañjuvajrodaya appears to be a maṇḍala rite for the Mañjuśrīnāmasaṃgīti.171 The fact that its title, and a number of instances in the text, refer to Mañjuvajra places the Mañjuvajrodaya at a very interesting crossroads between the Mañjuśrīnāmasaṃgīti tradition and Buddhajñānapāda’s tradition of the Guhyasamāja. None of our sources report Pālitapāda to have taught the Mañjuśrīnāmasaṃgīti, but that certainly does not preclude his having done so. That the Pālitapāda mentioned by Śrīkīrti and Buddhajñānapāda’s guru by that name are one and the same teacher is especially likely, given that the Indian author Samantabhadra states that his Sāramañjarī, a commentary on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, was composed at the command of one Kīrtipāda.172 Presumably the Kīrtipāda (a.k.a. Śrīkīrti; adding pāda as a marker of respect to the names of members of Buddhajñānapāda’s lineage was 166 I discuss these issues in more detail in Chapter Three. 167 Davidson 2002, 313; Szántó 215, 550-52. 168 Szántó 2015, 551-2. 169 ibid., 552-3. 170 ibid.,553. 171 I have not had the opportunity to study the text in detail, but its location in the Tengyur among quite a few other works devoted to the Mañjuśrīnāmasamgīti (and none devoted to the Guhyasamāja) suggests as much. In addition, the Tohoku Catalogue notes to Buton’s Mtshan brjod kyi dkyil ‘khor gyi bkod pa specify that Buton’s text, a commentary on the Mañjuśrīnāmasaṃgīti from the Yoga tantra perspective and including a presentation of its maṇḍala structure, makes reference to a number of works, including the Mañjuvajrodaya (databases.aibs.columbia.edu). 172 Sāramañjarī, D 1a.3; Szántó 2015, 554. And, yes, it is confusing that a person named Samantabhadra composed a commentary on a sādhana called the Samantabhadra, but he did. 34 common173) who commanded Samantabhadra to compose a commentary on Buddhajñānapāda’s sādhana was a figure in his community who was senior to Samantabhadra, and a co-disciple of Buddhajñānapāda’s under Pālitapāda would fit that role well. Moreover, Szántó reports that Śrīkīrtipāda’s Parikramapadopāyikā is devoted in particular to the choreographical details of the maṇḍala ritual, and Buddhajñānapāda composed an entire text, the *Gativyūha, devoted to such choreographical details, suggesting that this topic may have been a speciality of their common guru. Buddhajñānapāda reports leaving Pālitapāda’s company after nine years when his guru admitted that he himself did not have full realization of the Guhyasamāja-tantra. He then traveled back to North India where he eventually had a vision of Mañjuśrī in the Kuvaca forest behind Bodhgaya, in which the instructions that formed the basis for Buddhajñānapāda’s unique system of Guhyasamāja practice were revealed to him. Nonetheless, Pālitapāda must have had a great influence on his disciple, as Buddhajñanapāda describes returning to visit Pālitapāda in the Koṅkan later in his life, on which occasion he composed some sādhanas to please his guru. This visit took place subsequent to his vision of Mañjuśrī, but presumably prior to his composition of the Dvitīyakrama, given that the visit is mentioned in that text. Vaidyapāda reports that it was the Samantabhadra-sādhana that Buddhajñānapāda composed at Pālitapāda’s request, though Mañjuśrī himself had already commanded its composition in Buddhajñānapāda’s vision.174 Other Gurus The other three human gurus Buddhajñānapāda mentions studying with—Guṇeru, and Jātig Jvālā (the latter also seems to have been his consort), and Bālipāda—are at this point only known to us from their names and the details given in Buddhajñānapāda’s own account. It is worth noting, however, that two of these (unsurprisingly) historically unidentifiable gurus—that is, two among the six human teachers that Buddhajñānapāda names—Guṇeru and Jātig Jvālā, were women. Vaidyapāda also mentions two female co-disciples of Buddhajñānapāda’s under his guru Pālitapāda, the prostitutes Ālokī and Sādhuśīlā; a tantric consort trained in the Guhyasamāja yogas with whom he practiced also under Pālitapāda’s guidance, the butcher girl Vimalamutrī; and one female student of Buddhajñānapāda’s, the nun Guṇamitrā who requested him to compose a text at Nālandā early in his career. Thus we find women occupying the full variety of tantric roles in Buddhajñānapāda’s life—as gurus, consorts, co-disciples, and disciples. Mañjuśrī[-mitra!?] The most important of Buddhajñānapāda’s gurus, whose direct first-person speech makes up ninety percent of the Dvitīyakrama and whose instructions form the basis for Buddhajñānapāda’s system of Guhyasamāja practice, is none other than Mañjuśrī himself who appeared to Buddhajñānapāda in a vision—or so it seems in the Dvitīyakrama and in Vaidyapāda’s Sukusuma. In both of their accounts, the “emanated monk” (in Vaidyapāda’s rendering he is further identified as an emanation of the “Great Vajra Holder”175) who revealed Mañjuśrī and his maṇḍala to Buddhajñānapāda’s is eclipsed by the figure of Mañjuśrī himself as the direct source of Buddhajñānapāda’s revelation. In some (but, tellingly, not all) of the later Tibetan histories, though, the identity of this monk becomes much more central, and it is he who 173 There are several instances of this in addition, of course, to Buddhajñānapāda’s own name. There is his disciple Vaidyapāda, as well as repeated references to Dīpaṃkarabhadra as “Bhadrapāda” in Samantabhadra’s Sāramañjarī. 174 Sukusuma, D135b.1; D 133b.6; Samantabhadrī-ṭīkā, D 131b.6-7. See also Szántó 2015, 547. 175 It is not impossible that this could be translating Vajradhara, as Davidson (2002, 313) has rendered the appellation in English, but the term that the Tibetan translators used, rdo rje ‘dzin pa, is not the common one for Vajradhara, which is rdo rje ‘chang. 35 is celebrated as Buddhajñānapāda’s most important teacher, rather than Mañjuśrī. The story of their meeting is, in several of the Tibetan accounts—starting with Chögyal Phagpa’s and including Tāranātha’s, Amye Zhab’s, and Dudjom’s—significantly expanded upon from Vaidyapāda’s account, on which it is clearly based. Let us first witness this encounter in Chögyal Phagpa’s delightful telling, the earliest expanded version I am aware of, before we address the question of the monk’s identity.176 The episode is preceded by his teacher Pālitapāda being unable to cut through Buddhajñānapāda’s doubts about the meaning of the Guhyasamāja-tantra. [[[Buddhajñānapāda]]] went to Vajrāsana and made supplications to attain great awakening. Then he heard a voice in the sky that made a prophecy: “Son of good family, you must search for Mañjuvajra and through his blessings you will be freed from all of your doubts.” He asked the Ācārya Pālitapāda177 for permission to go to Wutaishan in China, since that is where Mañjuvajra resides, and permission was granted. Setting off from Vajrāsana he headed north, and in the forest called Kuvaca178 he saw that there was a woman and a female dog in front of a hut. Nearby a monk who had made his dharma robe into a turban was plowing a field. [[[Buddhajñānapāda]]] thought to himself, “Alas! There is a monk with a woman plowing a field! The Buddha’s teaching is certainly in decline!” and his heart was heavy. However, it was almost noon, so he was thinking to go there to request alms when the monk said, “Go get the ācārya some food for his alms.” The ācārya [[[Buddhajñānapāda]]] sat down and the monk commanded the woman, “Serve this monk his lunch.” The woman had taken a fishing net to a stream, caught a fish, and cooked it. Then she had placed [it upon] a leaf in front of the female dog. [The monk] having commanded “Serve [his] alms,” the dog vomited [onto the leaf] and [the woman] brought that [vomit] along with the fish to the ācārya. This [meal], which was meant specifically for him, the ācārya [[[Buddhajñānapāda]]] regarded as flesh and filth, and he did not eat it. Another version of the story says that the woman killed many birds and cooked their flesh and brought it [to Buddhajñānapāda] who did not eat it. The woman then snapped her fingers and the cooked [birds] flew away, upon which the ācārya began to have some doubts [about his previous judgments regarding his lunch companions]. Then the monk said, “Alas [he is a] so-called worldly being; give him some ordinary food.” [The woman] then brought some cooked rice and yogurt, which the ācārya did eat. He then thought to depart, but the monk said, “If you leave from here now you won’t find a place to stay tonight; go tomorrow.” So [[[Buddhajñānapāda]]] stayed, but the monk went off somewhere else. The ācārya was reciting the Guhyasamāja aloud, and whenever he came to a place [in the text] that he did not understand, the woman showed a displeased and sad 176 Other versions of this account are found in Tāranātha’s Bka’ babs bdun (104-6), Amye Zhab’s Gshin rje chos ‘byung (46a.3-48a.4) and Dudjom’s History of the Nyingma School. See Templeman (1983, 72-3) for an English translation of Tāranātha’s version of the same encounter, which unfortunately has a number of translation errors. See Dudjom (1990, 494-6) for an English translation of Dudjom Rinpoche’s version. Amye Zhab’s version of the account, in his Gshin rje chos ‘byung, is is reported to be translated in an unpublished article by Hubert Decleer (Decleer, upublished), and is also discussed in Decleer 1998. Decleer’s 1998 article notes the strong parallels between the account in Tāranātha and Amye Zhab’s works and the narrative about Dharmaśrīmitra in the ca. 15th Century Svayambhūpurāṇa (on the Svayambhūpurāṇa’s dates see von Rospatt 2015, 827). 177 Chögyal renders the name Pa li pa ta, but I have corrected it here to what we know to be the name of Buddhajñānapāda’s Guhyasamāja teacher from the Koṅkan. 178 Chögyal renders the name of the forest as ku pa, but I have corrected to the form given in the Dvitīyakrama. 36 expression. The ācārya thought, “This woman certainly is clairvoyant and knows the minds of others! She will be able to cut through my doubts!” He prostrated to her and supplicated [for instruction] but the woman said, “I don’t know. But that monk from before, because he is my husband, is extremely learned. You should ask him.” “Where did he go?” asked Buddhajñānapāda, and [the woman] replied, “[He went] to buy beer.” “When is he coming back?” he asked, and [she] replied “He’ll be back in the late afternoon.” So [[[Buddhajñānapāda]]] waited. In the late afternoon [the monk] came back staggering179 drunk. Seeing that, [[[Buddhajñānapāda]] felt] a lack of faith, but he swallowed his pride, prostrated and made the request, “Please teach me the Guhyasamāja.” “Ask for the initiation,” [the monk] commanded. “I’ve already gotten the initiation,” [[[Buddhajñānapāda]]] replied, upon which [the monk] said, “You must receive the initiation directly from me.” The ācārya searched for the ritual articles and requested initiation. According to a different account he gave [all] the money180 he had to the woman and she emanated the [necessary] ritual articles, so he obtained them [that way]. Then, at midnight on the eighth day of the seventh month [the monk] emanated a celestial palace and at its center was the nineteen-deity maṇḍala of Mañjuvajra, which he emanated and showed directly [to Buddhajñānapāda]. The monk, however, remained in front of the maṇḍala in the very same form in which he had appeared before and asked the ācārya, “Will you receive the initiation from me or from the maṇḍala?” The ācārya, even though he knew that the maṇḍala had been emanated by the monk himself, was inspired181 by the form of the deity and replied, “I’ll receive it from the maṇḍala.” “Fine, take it, then,” said [the monk], and [[[Buddhajñānapāda]]] received the full initiation from the maṇḍala. According to a different account when [[[Buddhajñānapāda]]] said “I’ll receive it from the maṇḍala,” the maṇḍala disappeared and [the monk said] “I emanated the maṇḍala,” upon which Buddhajñānapāda understood the maṇḍala to be the monk’s emanation and prostrated [to him] saying, “You are the father and the mother of all beings...”182 In this way he praised him, asked his forgiveness,183 and supplicated him. At dawn, from [the monk’s] heart-center the maṇḍala was [re-]emanated, and with a smile he said “Excellent!” and bestowed the initiation on him. Then [the monk] began [to explain]184 the condensed meaning of the Guhyasamāja, the [Dvitīyakramatattvabhāvanā]-Mukhāgama and other [instructions], and [thus] brought [[[Buddhajñānapāda]]] to realize all of the key points of the tantra. The ācārya was satisfied and delighted, so he said, “I want to make an offering; what will you accept?” “I don’t want anything,” the monk replied. “Take anything at all!” [[[Buddhajñānapāda]]] beseeched, so [the monk] said, “Alright, make the offering that you will prostrate every time you see [me].” The ācārya made that promise right on the spot as his offering. Then the monk said, “Because of [your] conduct with respect to food/ And holding a slight delusion with respect to me/ You will not, in this very life, gain accomplishment/ With those embodied aggregates/ But your mind will take on the vajra 179 ‘khyor] sugg. em. based on Gshin rje chos ‘byung which follows Chögyal Phagpa’s account word for word here; Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa reads ‘khyol. 180 Lit. kārṣāpāṇa, a unit of currency mentioned in Buddhajñānapāda’s autobiography in the context of the offerings received by Buddhajñānapāda and his retinue from the wealth deity Jambhala when they were residing in the Parvata cave later in his life. 181 mos 182 This is the first line of Dvitīyakrama, verse 12. 183 bzod par gsol 184 brtsams. This could also mean that he “composed” the texts mentioned. 37 body/ And you will be liberated in the bardo.”185 Then he said, “Although you practice you will not attain buddhahood in this lifetime, but benefit others and you will be liberated in the bardo.” And then he disappeared.186 All accounts of Buddhajñānapāda’s life up to the 15th century with which I am familiar— that is, Buddhajñānapāda’s own, Vaidyapāda’s, and Chögyal Phagpa’s—refer to the figure that he met in the Kuvaca forest outside of Bodhgaya simply as a “monk,” or an “emanated monk.” It is Gö Lotsāwa in the 15th century who gives him a name: the master Mañjuśrīmitra.187 I am unsure of Gö Lotsāwa’s source for this information, but I can (and will, below) speculate on his possible motivations for including it. This identification of the monk who was Buddhajñānapāda’s teacher as Mañjuśrīmitra persists in some of the later Tibetan biographies, but not all. Tāranātha who, as mentioned above reports having access to several Indic records that no longer survive, does not identify the monk as Mañjuśrīmitra, referring to him only with the unusual term “householder monk” (khyim btsun), presumably because in Vaidyapāda’s Sukusuma, on which Tāranātha explicitly states he relied,188 he is both identified as a monk and as accompanied by a woman.189 The Sakyapa scholar Amye Zhab, who like Tāranātha was writing in the 17th century, provides a somewhat conflicted account. In his Gsang ‘dus chos ‘byung he reports on the one hand that the Jñānapāda tradition “was bestowed upon Jñānapāda by the Ācārya Mañjuśrīmitra who was indivisible from Mañjuśrī,” 190 but when reporting the encounter with the monk in the forest he makes the strange statement that Buddhajñānapāda “directly saw the face of the emanated monk together with two gurus—Vilāsavajra and Mañjuvajra.”191 Amye Zhab then goes on to report that it was directly from the mouth of Mañjuvajra that Buddhajñānapāda received the teachings recorded in the Dvitīyakrama.192 In his Gshin rje chos ‘byung, which contains an even more extensive account of Buddhajñānapāda’s life, Amye Zhab provides an equally confusing report that in the Kuvaca forest Buddhajñānapāda “had a vision of the maṇḍala of Mañjuśrī together with *Vilāsalīla (?! ‘jol sgeg rol pa)193 185 This is a rephrasing of Mañjuśrī’s comments to Buddhajñānapāda in verse 366 of the Dvitīyakrama, with an extra half verse about the bardo that is not present in the Dvitīyakrama added on here. 186 Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 612-15. The edition of Chögyal Phagpa’s text that I have translated from here includes readings from a second manuscript that occasionally adds short phrases with additional details. I have translated these, as well, wherever they are present. 187 Deb ther sngon po, 449; Roerich 1976, 369. 188 He notes in an earlier portion of the biography that Vaidyapāda’s commentary on the Dvitīyakrama identifies Guṇamitrā as a bhikṣunī (Bka’ babs bdun, 103). 189 Tāranātha, in his Bka’ babs bdun (104-6), follows Chögyal Phagpa, or perhaps a common Indic source, in providing a version of the much more detailed account of Buddhajñānapāda’s meeting with this monk, translated above. 190 ye shes zhabs lugs ‘phags yul du dar tshul ni/ mgon po ‘jam pa’i dbyangs dang dbyer med pa’i/ slob dpon ‘jam dpal bshes gnyen gyis ye shes zhabs la gnang ba yin te/ (Gsang ‘dus chos ‘byung, 56b.3-4). 191 sprul pa’i dge slong bla ma gnyis ‘jol sgeg rdo rje ‘jam rdo dang bcas pa’i zhal mgon sum du gzigs te/. (Gsang ‘dus chos ‘byung, 56b.4.) Earlier in his text Amye Zhab uses the same uncommon moniker for Vilāsavajra. 192 ibid., 56b.4. Amye Zhab’s rather odd account of Buddhajñānapāda’s life is further complicated by his assertion that Buddhajñānapāda received teachings on the Guhyasamāja-tantra from some twenty-five masters, including every guru Amye Zhab mentions in the context of his life story apart from Haribhadra—that is Vilāsavajra, Guṇeru (who he styles Gu ni ni), Bālipada (whose name he inexplicably renders Ba mo la tsam pa ta), and Pālitapāda (who he calls Baliṃta Ācārya) (Gsang ‘dus chos ‘byung, 56b.6-57a-3). 193 In his Gsang ‘dus chos ‘byung Amye Zhab uses the unusual ‘jol sgeg rdo rje for Vilāsavajra. I am assuming ‘jol sgeg rol pa here is an error for ‘jol sgeg rdo rje. 38 together with two emanated nuns (!) in front of an ācārya’s small hut.”194 In several short references to Buddhajñānapāda Jamgön Kongtrül, closely associated with the nonsectarian (ris med) movement in 18th-century Tibet identifies Buddhajñānapāda’s teacher in two different ways, in one instance as Mañjuśrīmitra, and two others as Mañjuśrī himself.195 Dudjom Jigdral Yeshe Dorje, the great 20th-century Nyingma scholar, follows Gö Lotsāwa in very clearly identifying the monk who is Buddhajñānapāda’s guru as Mañjuśrīmitra.196 Indeed, Dudjom’s biography of Buddhajñānapāda occurs in his encyclopedic History of the Nyingma School in the context of a list of lineage biographies of Great Perfection masters, where Buddhajñānapāda’s primary role appears specifically to be a disciple of Mañjuśrīmitra in the Great Perfection lineage, given that no attention is given to Buddhajñānapāda’s own disciples in the subsequent lineage history.197 This brief, and certainly incomplete, survey of Tibetan historians’ accounts of Buddhajñānapāda’s life suggests that (with the exception of the Sakya scholar Amye Zhab, whose accounts of this particular episode of Buddhajñānapāda’s life are anyway somewhat perplexing) it is primarily authors who are connected with traditions of the Great Perfection who identify the monk that Buddhajñānapāda met in the Kuvaca forest as Mañjuśrīmitra. Historians who are not so connected do not make this identification, and Kongtrül—who had a strong relationship to both the Nyingma and Sarma traditions—appears to have asserted both positions. This propensity of Tibetan scholars connected to the Great Perfection traditions to identify Buddhajñānapāda’s teacher as Mañjuśrīmitra is almost certainly connected to the importance for these authors of further connecting Buddhajñānapāda—two of whose compositions contain in their Tibetan translations the word “great perfection” (rdzogs chen)—to the Great Perfection tradition. Given that the Indian origins of this tradition have been questioned by scholars from other Tibetan traditions and that the word “great perfection” is not found in many other Indic sources, placing Buddhajñānapāda in the Great Perfection lineage would indeed have been compelling. While we do not have the Sanskrit for either of the two passages where the term “great perfection” appears in the Tibetan translations of Buddhajñānapāda’s writings,198 on my 194 ‘jam pa’i dbyangs ‘jol sgeg rol pa dang bcas pa gtso bor gyur pa’i dkyil ‘khor dang/ sprul pa’i dge slong ma gnyis dang bcas pa slob dpon gyi khan gpa chung du zhing tu snang bar gyur to// (Gshin rje chos ‘byung, 47a.2-3) In this account of Buddhajñānapāda’s life Amye Zhab relies heavily on both Vaidyapāda and Chögyal Phagpa, both of whose accounts he follows word-for-word at different parts of his narrative. However, with his unusual version of the account of the encounter in the forest behind Vajrāsana, he departs from both. 195 In his great Treasury of Knowledge, in the volume on Buddhism’s Journey to Tibet (2010, 229), Jamgön Kongtrül states that Buddhajñānapāda received Guhyasamāja instructions from Mañjuśrīmitra, while in his Torch of Certainty, Kongtrül states that “Buddhajñānapāda’s faithless perception caused him to see Mañjuśrī as a married monk with children,” thus appearing to identify the master as none other than Mañjuśrī himself (Kongtrül 1994, 130), and in the Treasury of Knowledge in the volume on The Elements of Tantric Practice (2009, 145) he states directly that “The Guhyasamaja completion phase in the tradition of Jnanapada (Buddhashrijnana) [stems from] what are called the Oral Teachings of Manjushri. These Teachings were directly transmitted by Arya [Mañjuśhri] himself to the master Buddhashrijnana.” 196 Dudjom 1990, 494-6. 197 The next biography given is that of Śrī Siṅgha, who is also said to be Mañjuśrīmitra’s disciple, and who Dudjom reports may have even been one and the same individual as Buddhajñānapāda (!) (Dudjom 1990, 496). The only mention of Buddhajñānapāda’s disciples is their own role as Great Perfection practitioners: “...it is implicit that the host of his followers and disciples belonged to the lineage of the Great Perfection” (Dudjom 1990, 496). 198 What is more, in the Caturaṅga/Samantabhadra-sādhana which was translated twice into Tibetan under two different titles, the word rdzog chen is found in only one of the two translations of the text, the one by Smṛtijñānakīrti. (But it also appears in one of the commentaries on the Samantabhadra, by Vaidyapāda, which was translated by a different set of translators, Kamalaguhya and Ngadak Yeshe Gyaltsen (Snga bdag ye shes rgyal mtshan)). The other translation of that same passage in the sādhana, by Śraddhākaravarman and Rinchen Zangpo (Rin chen bzang po), instead uses the word bsam yas. I address this point in further detail in Chapter Four. 39 reading, both of these instances seem to use the term in a sort of general way to refer to the result of practice, which is certainly not antithetical to, but also not precisely the way it was being used in the early (or proto-) Great Perfection literature of the late 8th and early 9th centuries (some of which is Indic, but all of which survives only in Tibetan). Vaidyapāda’s commentaries on those passages, however, do interpret the term in a way that brings it closer to—and in one case precisely in line with—contemporary 8th/early 9th-century usage of the term “great perfection.” I examine this point in more detail in Chapter Four, where I conclude that it is nonetheless unlikely that Buddhajñānapāda used a Sanskrit word with the semantic content of “great perfection” at any point in his writings. But whatever we make of the infrequent use of that particular term in the Tibetan translations of his oeuvre, we would certainly be too quick to write off as the mere apologetics of later Tibetan authors what does appear to be a genuine doctrinal affinity between Buddhajñānapāda’s writings and several early or proto- Great Perfection works from around the same period. I return to this connection below in Chapter Four. Later Life as a Teacher and Author Following his visionary encounter with Mañjuśrī, Buddhajñānapāda reports setting up residence with his students in the Parvata cave not far from Bodhgaya. The details that Vaidyapāda gives regarding the location of this residence, which he further specifies as the practice place of “great practitioners of former times,” enable us to identify it as being in the region of the Rajgir hills.199 There, Buddhajñānapāda himself tells us, he and his disciples received daily donations from the wealth deity Jambhala, and it was there that Buddhajñānapāda compiled the instructions he had received from Mañjuśrī in the Dvitīyakrama and composed other works. He likewise reports traveling back to the Koṅkan to visit his guru Pālitapāda who requested Buddhajñānapāda to compose a sādhana, which Vaidyapāda identifies as the Samantabhadra.200 As noted before, Buddhajñānapāda further mentions some “detailed accounts” of his life, which Vaidyapāda specifies to be “the taming at Nālandā, making offerings at Varjāsana, the consecration, and the others,” but gives no further information. It is only in the Tibetan histories, starting with Chögyal Phagpa, that we begin to find descriptions of these and other events from his later life. Some of the most detailed accounts of Buddhajñānapāda’s later life are found in the writings of Tāranātha. His Seven Transmissions contains a biography which references Vaidyapāda’s Sukusuma directly and elaborates on each of the three episodes mentioned by Vaidayapāda by name. Tāranātha’s account of the offerings at Vajrāsana201 is as follows: Once at a time when the ācārya [[[Buddhajñānapāda]]] had built a straw hut near Vajrāsana and was staying there, King Dharmapāla came to Vajrāsana to make offerings, and all the other Buddhist ācāryas also came to offer. Seeing that this ācārya [[[Buddhajñānapāda]]] had not made offerings, [the king] thought he ought to punish him. He entered into the ācārya’s hut, but the ācārya was not there, and [instead] he saw an image of Mañjuśrī. He came outside and asked the [[[master’s]]] retinue [where he was], and they replied that he was right in [that hut]. [The king] again entered [the hut] and [the image of Mañjuśrī] 199 See C. Dalton and Szántó, forthcoming. 200 Samantabhadra and later Chögyal Phagpa report that on that journey Buddhajñānapāda initially followed proper decorum and refused to teach in the presence of his guru and agreed only once Pālitapāda had himself given his assent (Sāramañjarī, D 2b.6-3a.2; Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 615). This episode is also discussed in Szántó (2015, 548-9). 201 An earlier and slightly variant version of this account is found in Chögyal Phagpa (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 617). 40 appeared as the ācārya. When [the king] asked “Why did you not make offerings?” [[[Buddhajñānapāda]]] replied, “I made them from here.” [The king] asked “How did you offer [from here]?” upon which the ācārya entered into equipoise and all of the images from Vajrāsana appeared directly in front of the ācārya like invited guests. [The king] then saw the ācārya make vast offerings to them. At that point the king felt confidence [in him] and requested initiation. Since he didn’t have anything else to offer as the initiation fee, he offered himself and his wife as [the master’s] servants. The next day, from the palace, gold equal in weight to both of their bodies [was brought] and offered as ransom.202 This account, elsewhere recounted as involving king Devapāla rather than Dharmapāla, as discussed above, is important in linking Buddhajñānapāda with the Pāla royalty, suggesting he may indeed have been a royal guru to one (or more) of the Pāla kings. In his History of Buddhism in India (Rgya gar chos ‘byung) Tāranātha further describes a close relationship between Buddhajñānapāda and King Dharmapāla in which the master gave predictions and advice to the king, advising him to have a great homa performed regularly to ensure the longevity of his dynasty as well as of the Buddhadharma. King Dharmapāla is said to have taken this advice and maintained the regular homa ceremonies at great expense.203As described above, royal patronage of monasteries and religious masters was common in the medieval period, and the wealth and renown such patronage must have brought would have helped Buddhajñānapāda spread his teachings more widely. Such a position in connection to the king might also partly explain the great respect Buddhajñānapāda is said to have been shown by his own gurus, Vilāsavajra and Pālitapāda, who respectively cite and are said to have requested his writings. Atīśa’s brief reference to this episode (which he links to king Devapāla, rather than Dharmapāla) is important, as it is the only Indic account I am familar with that links Buddhajñānapāda to the Pāla kings.204 Tāranātha goes on to recount “the consecration,” mentioned by Vaidyapāda, and specifies that it was the consecration of the great Vikramaśīla monastery: The four temples at Vikramalaśīla,205 Odyantapūri, Śrī Nālandā, and Somapuri were cut off from one another by days of travel. Vikramalaśīla had been newly constructed, Somapuri had undergone reconstruction, and the other two [[[monasteries]]] had temples that had been newly built, so the king requested consecration for these many [[[temples]]]. The ācārya [[[Buddhajñānapāda]]] emanated four bodies simultaneously and performed the consecration of all four at once.”206 Tāranātha, in his History of Buddhism in India identifies Buddhajñānapāda as not only having consecrated Vikramaśīla but also having served as its first tantric ācārya, though to my knowledge this is not reported in any of the extant earlier histories.207 Buddhajñānapāda’s Guhyasamāja tradition was passed down by a number of masters who were likewise connected 202 Bka’ babs bdun, 106-7. My translation here differs only in minor points from Templeman’s (1983, 74). 203 Rgya gar chos ‘byung, 266-67; Chimpa and Chattopadhyaya 1970, 278-79; See also C. Dalton and Szántó forthcoming. 204 Bodhipathapradīpapañjikā, D 288b.7-289a.1; See also Szántó 2015, 539. 205 As in much of the literature in Tibetan, Vikramaśīla is referred to by Tāranātha as Vikramalaśīla. 206 Bka’ babs bdun, 107. Again my translation differs only in minor points from Templeman’s (1983, 75). Tāranātha continues with a further account of a non-Buddhist yoginī who attempts to derail the consecration at Vikramaśīla, but whose attempts are magically foiled by Buddhajñānapāda. For this account see Templeman 1983, 75. 207 Rgya gar chos ‘byung, 7, 265-66; Chimpa and Chattopadhyaya 1970, 18, 278. 41 with Vikramaśīla, including Dīpaṃkarabhadra, Vaidyapāda, Ratnākaraśānti and Abhayākaragupta. I must admit that I find it somewhat discouraging that neither Buddhajñānapāda nor Vaidyapāda makes any direct statement about Buddhajñānapāda’s having had a connection with the Pāla royalty or having acted as the vajrācārya of Vikramaśīla, both of which seem accomplishments worthy of mention in an auto-/biography. Perhaps if these events did occur they happened after the composition of the Dvitīyakrama and were thus not described there by Buddhajñānapāda, and Vaidyapāda felt compelled to constrain his elaborations in the Sukusuma to events Buddhajñānapāda had mentioned directly. The fact that at least one Indic source, Atīśa’s *Bodhipathapradīpapañjikā discussed above, recounts an encounter between Buddhajñānapāda and the Pāla king Devapāla is, however, supportive of the later accounts in the Tibetan histories. The third episode mentioned by Vaidyapāda, the “taming at Nālandā” is elaborated upon in Tāranātha’s Seven Transmissions, and the same events are likewise mentioned in his History. The episode centers around the criticism of Vajrayāna practices by śrāvaka monks. The great ācārya was acting as the head of both Nālandā and Vikramaśīla. At that time some śrāvaka Sendhapas208 who resided at Udyantapuri and some monks who were distracted by conceptuality spoke negatively about him. One time when the ācārya was residing at Nālandā those monks repeatedly criticized him, saying “Buddhajñāna lacks discipline and it is therefore not suitable for him to be a preceptor who presides over the monastic saṇgha.” They also criticized the Vajrayāna, and it is said that [also at that time] many Singhalese Sendhapas at Vajrāsana destroyed a silver image of Heruka. The king [subsequently] killed many Singhalese from Vajrāsana and as he was beginning to impose punishment on some other Sendhapas the ācārya, out of great compassion, successfully protected them from being harmed by the king. In order to overcome their lack of faith he miraculously traveled under the earth, and on several occasions many nonhuman beings made offerings to the ācārya: he displayed quite a variety of miracles. Moreover, he composed many treatises on the supreme conduct of [secret] mantra, establishing it as not being in contradiction to the śrāvaka-piṭaka.209 The reference in this episode to Buddhajñānapāda’s lack of discipline presumably refers to his having a consort, which Buddhajñānapāda and Vaidyapāda both mention being the case at several points earlier in his life. As regards the general antipathy toward the Vajrayāna and Buddhajñānapāda’s efforts to overcome this, we certainly see in his writings an effort to present the tantric teachings within the context of the general Mahāyāna and to validate Vajrayāna doctrine and practices, though in his extant work there is nothing that particularly addresses the śrāvaka teachings.210 In his History Tāranātha further mentions that Buddhajñānapāda principally taught five among the “inner tantras:” the Guhyasamāja-tantra, the Māyājāla-tantra, the Sarvabuddhasamāyoga-tantra, the Guhyendratilaka-tantra, and the Yamāri-tantra,211 with 208 Sendha pa. The referent of this term is unclear, but from the way that Tāranātha uses it, it appears to refer to a particular group or sect of śrāvaka monks, primarily from Ceylon. See Templeman 1983, 143n144 for a further discussion of their possible identity. 209 Bka’ babs bdun, 108. Again, my translation differs only in minor points from Templeman’s (1983, 75-6). 210 The possible exception to this are his and Vaidyapāda’s use of the unusual terms thal byung blo can, thal byung zab mo and thal byung gyi stong pa nyid, all of which seem to refer to a śrāvaka-like fixation on the quiescent aspect of meditative equipoise, a position that Buddhajñānapāda rejects. See note 148 in Chapter Three. 211 The latter is rendered ‘jam dpal ‘kros pa, but presumably refers to the Krṣṇayamāri-tantra. Rgya gar chos ‘byung, 267; Chimpa and Chattopadhyaya 1970, 279. 42 the most emphasis on the Guhyasamāja. Though his surviving tantric writings are primarily Guhyasamāja-based, other Tibetan scholars including Amye Zhab include him as an important figure in the Kṛṣṇayamāri lineage.212 I will discuss Buddhajñānapāda’s surviving oeuvre in more detail below, but let us first turn our attention to his principal disciples. Disciples The earliest reference we have to Buddhajñānapāda’s disciples is in Vaidyapāda’s Sukusuma, which states that Buddhajñānapāda had eighteen disciples who functioned as his regents, among whom there were four who attained nirvāṇa in this lifetime: Dīpaṃkarabhadra (Mar me mdzad bzang po), Praśāntamitra (Rab tu zhi ba’i bshes gnyen), *Rahulabhadra (sgra gcan ‘dzin bzang po), and *Vajramahāsukha (Rdo rje bde ba chen po).213 These same four principal disciples are mentioned in Śrīphalavajra’s commentary on the Caturaṅga/Samantabhadra-sādhana.214 Dīpaṃkarabhara Of the four, we know the most about Dīpaṃkarabhadra, who composed the well-known and influential Guhyasamājamaṇḍalavidhi, an initiatory ritual according to the Jñānapāda School that continued to influence later ritual manuals such as Abhayākaragupta’s Vajrāvalī and 212 Gshin rje chos ‘byung, 16a.4. As regards the Yamāri tantra connection, in one lineage description Amye Zhab explains that Buddhajñānapāda’s Kṛṣṇayamāri lineage originated with Mañjuśrī and was passed from Buddhajñānapāda to the siddha Śrīdhara (Dpal ‘dzin), who passed it on to Nāropa (Gshin rje chos ‘byung, 16a.4). Later in the same work, at the conclusion of a biography of Buddhajñānapāda Amye Zhab writes that Buddhajñānapāda received from (emending la to las to accord with the verb gsan) Mañjuśrī teachings on the cycles of the Guhyasamāja as well as Vajrabhairava. He continues, explaining that Buddhajñānapāda passed the Vajrabhairava cycle on to Dīpaṃkarabhadra, who gave it to the siddha Śrīdhara, who composed many Yamārirelated works, which he then passed on to Nāropa (Gshin rje chos ‘byung, 48a.4-5). (Dīpaṃkarabhadra is credited with at least one Yamāri-tantra related text in the Tibetan canon, a work on the protection circle according to that tradition (Tōh. 1928), and many Yamāri-related compositions attributed to Śrīdhara survive in the canon.) Jamgön Kongtrül also briefly mentions Buddhajñānapāda’s view on the perfection stage yogas of the Yamāri-tantra as being consistent with that of Saraha (Kongtrül 200, 149). Buddhajñānapāda’s connection with the Yamāri-tantra tradition has been almost entirely ignored in the secondary literature (at least in English; there may exist references in Japanese of which I am unaware), with the exception of an unpublished article by Hubert Decleer that notes Buddhajñānapāda’s importance for the Sakyapas of the Ngor tradition as being the source of both their Guhyasamāja lineage and their Kṛṣṇayamāri lineage. In that article Decleer identifies an image from a Sakyapa thanka as depicting scenes from Buddhajñānapāda’s life (Decleer, unpublished). I am grateful to Hubert Decleer for sharing this unpublished work with me. Three Yamāri-tantra related compositions (Tōh. 2084, 2085, and 2086) are attributed to Buddhajñānapāda in the Tibetan canon, but I believe it is not likely that these are by the same Buddhajñānapāda who composed the many Guhyasamāja works that we know. I assess the attribution of these three texts to Buddhajñānapāda below in the section on his writings. To my knowledge there is not any other mention of Buddhajñānapāda’s having taught the other three tantras mentioned by Tāranātha—the Māyājāla, Sarvabuddhasamāyoga, or Guhyendratilaka—but he certainly knew the Sarvabuddhasamāyoga as several lines of his Dvitīyakrama (v. 50; see also notes 123 and 276 of my Dvitīyakrama translation) are a paraphrasis of verses in that tantra, and his disciple Praśāntamitra probably wrote a commentary on it, as well as one on the Māyājāla-tantra (which I discuss below). And Buddhajñānapāda’s guru Vilāsavajra taught parts of the Mañjuśrīnāmasaṃgīti in accordance with the Māyājāla (See Tribe 1994, 24). 213 Sukusuma, 135a.5-6. 214 Samantabhadrasādhana-vṛtti, D141a.6. The order of the list of the four main disciples is different here from Vaidyapāda’s Sukusuma, perhaps indicating that one list is not merely derivative of the other. Śrīphalavajra here also mentions three, but then names only two, of Buddhajñānapāda’s co-disciples: *Dharmākara (Chos kyi ‘byung gnas) of the Koṅkan and *Uṣṇīṣavajra (Gtsug tor rdo rje) of Mount Hasara (ri bo ha sa ra) (ibid., D 141a.5-6). The list of four principal disciples is repeated in several of the Tibetan histories, as well, and Tāranātha cites Śrīphalavajra’s statement about Buddhajñānapāda’s co-disciples (Bka’ babs bdun, 109; Templeman 1983. 76). 43 Jagaddarpaṇa’s Kriyāsamuccaya.215 A very substantial portion of the Guhyasamājamaṇḍalavidhi is a direct paraphrasis of Buddhajñanapāda’s Caturaṅga/Samantabhadra-sādhana and a good ninety percent of the Caturaṅga/Samantabhadra’s verses appear rephrased therein. Tāranātha reports that Dīpaṃkarabhadra succeded his guru Buddhajñānapāda as the second tantric ācārya at Vikramaśīla,216 and that he was said to have achieved even higher realization than his master.217 Tāranātha also identifies Vaidyapāda as Dīpaṃkarabhadra’s disciple.218 Praśāntamitra Buddhajñānapāda’s disciple Praśāntamitra is likely the author of a commentary on the Sarvabuddhasamāyoga-tantra, one on the Māyājāla-tantra, and a third on the Vajramaṇḍālaṃkāra.219 Tāranātha reports that Praśāntamitra received initiation from Buddhajñānapāda and had a meditative vision of Yamāri,220 and places him as the first disciple after Buddhajñānapāda in Jñānapāda’s lineage of the “word” instruction.221 He also identifies Vaidyapāda as a disciple of Praśāntamitra’s.222 About the other two among the four principal disciples mentioned in Vaidyapāda’s and Śrīphalavajra’s accounts, *Rahulabhadra and

• Vajramahāsukha, we unfortunately can say nothing more.

Buddhaguhya and Buddhaśānti Although I am not aware of any such reference in an extant Indic text, Gö Lotsāwa and Tāranātha mention Buddhaguhya and Buddhaśānti as disciples of Buddhajñānapāda’s, with Tāranātha specifying that this discipleship took place early in Buddhajñānapāda’s life.223 Buddhaguhya is a well known early 8th-century commentator on a number of Kriyā and Yoga tantras, including an important commentary on the Mahāvairocanābhisambodhi. A number of works on the Guhyagarbha-tantra are also attributed to him, though it is unclear whether it is the same author who wrote the Kriyā and Yoga tantra works and those on the Guhyagarbha.224 215 Szántó 2015, 554. This work survives in Sanskrit and has been edited. It was first circulated in an e-text (Klein- Schwind and Isaacson), edited in Dhīḥ (2006), and later by Bahulkar (2010). Szántó 2015, 556n34 gives a diplomatic transcript of verses that were missing the Dhīḥ edition. Daisy Cheung at Hamburg University is working on this text for her dissertation. Quite a number of other works in the Tengyur are ascribed to Dīpaṃkarabhadra (Mar me mdzad bzang po), including a short sādhana for the protection circle in the Yamāri-tantra tradition (Tōh. 1928), but I have not had the opportunity to look at these and assess their content or authorship. 216 Rgya gar chos ‘byung, 7. 217 Bka’ babs bdun, 110. Tāranātha gives a short biographical sketch of Dīpaṃkarabhadra (Bka’ babs bdun, 109- 112; Templeman 1983, 76-8). 218 Bka’ babs bdun, 112; Templeman 1983, 78. 219 Szántó 2015, 547n22. The commentaries are Tōh. 1663, 2514, and 2515, respectively. I agree with Szántó here on the likelihood that these commentaries are indeed composed by the Praśāntamitra who was Buddhajñānapāda’s disciple, especially given his note that this author includes a line in the dedicatory verses at the end of each text identifying himself as the author, a stylistic feature of many of Buddhajñānapāda’s own writings and those of his other disciples. 220 Rgya gar chos ‘byung, 267-68; Chimpa and Chattopadhyaya 1970, 279-80. 221 tshig gi brgyud pa’i brgyud pa’i bka’ babs. Bka’ babs bdun, 117; Templeman 1983, 66. It is unclear from Tāranātha’s description precisely what this “word lineage” is, but it is perhaps worth noting that the fourth initiation is often referred to as the “precious word initiation” and Vaidyapāda’s Yogasapta shows evidence of the early development of this initiation in Buddhajñānapāda’s tradition. 222 ibid. 223 Deb ther sngon po, Vol I, 451; Roerich 1976, 372; Rgya gar chos ‘byung, 269; Chimpa and Chattopadhyaya 1970, 280. Dudjom (1990, 465-6) repeats Tāranātha’s assertion. 224 Hodge 2003, 23 and 1995, 69. Scholarly opinion on whether the Yoga tantra and Mahāyoga tantra author were one and the same or not appears to be divided between those who assert “maybe no” and those who assert “maybe yes,” but there seems to be no certainty on the topic. Hodge (1995, 69) notes that Buddhaguhya does not mention the Guhyagarbha in any of his other tantric works and that the two corpi seem to him different on stylistic grounds. 44 Buddhaguhya is reported in a number of Nyingma sources to have been a student of Vilāsavajra’s, from whom he received teachings on the Māyājāla-tantra and especially the Guhyagarbha-tantra.225 He was also invited to Tibet by King Trisong Deutsen, but declined the invitation on the advice of his tutelary deity Mañjuśrī. 226 Some modern scholars have placed doubts on Buddhaguhya’s having been a disciple of Buddhajñānapāda’s, given that he appears to have been, if anything, Buddhajñānapāda’s senior contemporary rather than his junior, but the question remains unresolved.227 Of Buddhaśānti we know only of his association with Buddhaguhya as described in the Tibetan histories.228 Vaidyapāda There is considerable question about whether or not Vaidyapāda, Buddhajñānapāda’s most prolific Indian commentator, was his direct disciple. However, like with Buddhajñānapāda, the first issue we must address is that of his name. In almost all of the secondary literature mentioning this master, he is referred to as Vitapāda.229 This is no doubt due to the fact that in the preponderance of colophons of the Tibetan translations of his work this author’s name is given as Vitapāda.230 However, we also find his name in other colophons as Vitapāta,231 Vaidyapāda,232 Vidyapāda,233 Viryapata,234 and even Hahitapāda (!).235 When this name is Guarisco (Kongtrül 2005, 75n12) likewise notes that while traditional Nyingma scholarship identifies the author of these different works as a single figure, it is more likely that the author of the Guhyagarbha treatises is a different individual. Takahashi (2009, 198) however, opines that the claim of Nyingma authors that the author of the Yoga tantra and Mahāyoga tantra works are one and the same Buddhaguhya is worth taking seriously, especially given that both corpi seem to have been composed around the same period. She suggests that perhaps Buddhaguhya’s involvement with the Guhyagarbha teachings was less publicized by the author himself given their controversial nature, which would explain the absence of reference to the Guhyagarbha-related teachings in his other works. Van Schaik (2004, 187) simply notes that they “may be” one and the same. Germano (2002, 229-232) reports from Nyingmapa sources in which it is assumed to be one and the same Buddhaguhya who wrote both corpi, but does not express his own opinion one way or another. 225 Hodge 1995, 68-9. 226 ibid. 227 Hodge (2003, 22 and 2012, 68-9) and Weinberger (2003, 83) place doubts on the discipleship. 228 Rgya gar chos ‘byung, 269-71; Chimpa and Chattopadhyaya 1970, 280-83. Tāranātha relates several accounts in which Buddhaśānti appears as a companion of Buddhaguhya. Contrary to the letter in which Buddhaguhya declines to travel to Tibet, the two yogins are reported as companions traveling near Mount Kailash during the time of King Trison Deutsen in the Sba bzhed, where Buddhaguhya’s name is reported as Buddhagupta (Kapstein 2000, 26). The two names Buddhaguhya and Buddhagupta seem to refer to the same person (See Hodge 2003, 70; Weinberger 2003, 84). Germano cites Nyingmapa accounts stating that the exchange between Buddhaguhya and Trisong Deutsen took place precisely during Buddhaguhya’s journey to Kailash, and that he declined to visit central Tibet (Germano 2002, 229). 229 However, at least two modern scholars have addressed the issue of his name, one of whom has concluded that it is probably better rendered as Vaidyapāda. Szántó (2015, 540n6) reports a presentation given by Leonard van der Kuijp at Oxford in 2008 in which he suggested as much. I am unfamiliar with the details of van der Kuijp’s assessment, which has not been published. Kikuya (2012a, 1276n3) likewise reports that he himself has written about “the problems of Vitapāda’s transmission and his name,” but as the article referenced is in Japanese I have not been able to check it. 230 Bi ta pā da in the colophons of his Sukusuma (Tōh. 1866) (in D and P), Samantabhadrī-ṭīkā (Tōh. 1872) (P), Guhyasamājamaṇḍalopāyikā-ṭīkā (Tōh. 1873) (in D and P), Siddhisaṃbhavanidhi (Tōh. 1874) (in D and P), Yogasapta (Tōh. 1875) (in D and P), Mahābalividhi (Tōh. 1876) (in D and P), Ratnamati (Tōh. 1877) (in D and P), and Ātmārthasiddhikara (Tōh. 1878) (in D and P). 231 Bi ta pā ta in the colophon of his Samantabhadrī-ṭīkā (Tōh. 1872) (D). 232 Bai dya pā da, in the colophon to his Samyagvidyākara (Tōh. 1850) in the Derge Tengyur; the Peking here reads Bi dya pā da. 233 Bi dya pā da in the colophon to his Samyagvidyākara (Tōh. 1850) in the Peking Tengyur. 45 translated into Tibetan—in the Tibetan translations of Indic works and in indigenous Tibetan writings alike—it is consistently rendered as Sman pa’i zhabs, which fact already lends support to the Sanskrit Vaidyapāda, and perhaps even suggests that he may have been a physician (Skt. vaidya, Tib. sman pa).236 Gö Lotsāwa gives his name in translation as Sman pa’i zhabs, but in his Seven Transmissions Tāranātha—who mentions a number of Indic sources on which his writings relied—gives the name both in transliteration as Vaidyapāda (Bai dya pā da) and in translation as Sman pa zhabs.237 In the dedicatory verses of his Yogasapta (in the colophon of which his name, incidentally, is given as Vitapāda), this master writes [I] Vaidyapāda (Sman pa’i zhabs) have received This supreme nectar of the seven yogas Accomplished through practice in the presence of the gurus Of the ocean of the Glorious Samāja. Having drunk this nectar May the fatal illness of Mistaken conceptuality Be completely dispelled! Freed from that [[[Wikipedia:illness|illness]]] may all beings Perfectly unfold the genuine aggregates And attain the suchness that is the result: The supreme nature of the seven yogas!238 It seems to me that in these verses Vaidyapāda is loosely playing on his name (“Mr. Doctor”) in reference to conceptuality as a fatal illness that is healed by the nectar of the seven yogas. Of course, the use of a medical metaphor is not unique—medical metaphors have been used in Buddhist texts from the very earliest literature. My sense, though, is that Vaidyapāda is using it here as a way of integrating his name more smoothly into the dedicatory verse. Vaidyapāda frequently emulates Buddhajñānapāda’s writing,239 and as we will see below, Buddhajñānapāda himself often wove his own name cleverly into the dedicatory verses of his writings. For a master whose name was “Mr. Buddha Wisdom,” that was, however, an easier task than for his disciple “Mr. Doctor.” It seems to me that here in the Yogasapta, Vaidyapāda found his chance.240 Of course, the “Vitapāda” in so many of our Tibetan colophons could certainly be a 234 Martin (2011, 2078) notes that the Black Hat Tanjur catalogue his Yogasapta reports the author of the Yogasapta as Birya pa ta. 235 Ha hi ta pā da in the colophon of the Derge edition of his Multitilaka-vyākhyāna (Tōh. 1870). The Peking reads Bi ta pā da. 236 Another possible understanding of the term, one not taken up by the Tibetan translators, is vaidya as in someone well versed in the Vedas. Though none of our extant Indic sources report anything about Vaidyapāda’s life, Tāranātha states that he was born a brahmin and was not only learned in the non-Buddhist doctrines but had become powerful due to them, presumably through practice, before becoming a Buddhist (Bka’ babs bdun, 112; Templeman 78-9) 237 Neither Chögyal Phagpa nor Amye Zhab mention this master. 238 dpal ldan ‘dus pa’i rgya mtsho las// bla ma’i zhal snga (snga] P, sngas D) bsgrub pas na// sbyor ba bdun gyi bdud rtsi mchog// sman pa’i zhabs kyis thob pa’o// de ‘dra’i bdud rtsi de la ni// ‘thung bye de ni log pa yi (yi] P, yin D)// rtog pa yis ni rab ‘chi (‘chi] P, ‘cing D) ba’i// nad ni kun nas med gyur cig// de med pas na sems can kun// yang dag phung po rgyas ‘gyur te// sbyor ba bdun gyi rang bzhin mchog// ‘bras bu de nyid rtogs par shog// (Yogasapta, D 75b.1-3; P 89b.6-8) 239 See, for example, the opening verse dedicated to the buddha, dharma, saṇgha, and gurus in his Siddhisaṃbhavanidhi, which loosely emulates Buddhajñānapāda’s own opening verse, likewise dedicated to the three jewels and gurus, in the Dvitīyakrama. In his compositions Vaidyapāda likewise replicates much of the terminology that is especially characteristic of Buddhajñānapāda’s writings. 240 Vaidyapāda inscribes his own name into only one other of his ten extant compositions, his Siddhisaṃbhavanidhi, but in one of the introductory verses, rather than the conclusion. This verse actually has some parallels with the 46 vernacular rendering of the name Vaidyapāda.241 And though there is no single piece of definitive evidence, and thus the question still remains open, I feel there is sufficient reason to depart from the more common usage in the secondary literature and refer to this master as Vaidyapāda. To return to the question of whether or not Vaidyapāda was a direct disciple of Buddhajñānapāda’s, again there is no definitive piece of evidence one way or the other, but I believe it is likely that he was.242 Three of his ten surviving works are commentaries on Buddhajñānapāda’s compositions, and he directly mentions Buddhajñānapāda or his tradition in five of the remaining seven. In these references he makes statements such as having composed treatises, “in order to remember the stages of the pith instructions of my gurus who uphold the lineage, Buddhajñānapāda and so forth...,”243 “in order to remind myself and others of the stages of the pith intructions of my gurus who uphold the lineage, Buddhajñānapāda and so forth...,”244 and others which contain “the complete teachings that have come down in stages from Buddhajñāna,”245 or which are explained “in terms of the suchness of the instructions that have come down in stages from our great guru Buddhajñāna.”246 In one dedicatory verse he notes, “Since the appearance of the kindness of the unsurpassed great compassion of Śrībuddhajñanapāda has been veiled, although I am a fool, I have uncovered it slightly.”247 Admittedly none of these statements settles the case, but one does get the sense that Buddhajñānapāda was one among Vaidyapāda’s gurus with whom he had some direct personal connection, though certainly not his only master. This is upheld by Gö Lotsāwa248 and Tāranātha, who both clearly write that Vaidyapāda was Buddhajñānapāda’s direct disciple. But the way that Tāranātha states this is more telling: he lists Vaidyapāda in the lineage as a student of Dīpaṃkarabhadra, who he has already reported as Buddhajñānapāda’s disciple, but adds that Vaidyapāda, “trained in the tantras after having heard [teachings], beginning with the Prajñāpāramitā and continuing up to the outer and inner [[[tantras]] of the] secret mantra, at Nālandā in Madhya from both Dīpaṃkarabhadra and concluding verse from the Yogasapta, but here Vaidyapāda was working with a maritime, rather than a medical, metaphor: “From the great ocean of the Glorious Samāja/ By means of the ship of the gurus’ lineage/ [I] Vaidyapāda (Sman pa’i zhabs) have obtained/ The three wish-fulfilling gems! dpal ldan gsang ‘dus mtsho chen las/ bla ma’i brgyud rim gru gzings kyis/ yid bzhin nor bu rnam pa gsum/ sman pa’i zhabs kyis rnyed pa’o// (Siddhasaṃbhavanidhi, D 2a.1; P 2b.1-2).) 241 Could it have been written Vaidyapāda, but colloquially pronounced Vitapāda? 242 Modern scholars have asserted both positions. Tomabechi (2008, 172-3) and Klein-Schwind (2012. 17) hold Vaidyapāda to be Buddhajñānapāda’s direct disciple, while J. Dalton (2004, 17) and Kikuya (2012, 1264) assert that he was a later commentator, and Davidson (2002, 311) also seems to suggest that he was not a direct disciple. Szántó (2015, 547) hedges his bets, reporting only that Vaidyapāda was “supposedly Jñānapāda’s direct student.” In a recent article written together with Péter Szántó, (C. Dalton and Szántó, forthcoming) we suggest that Vaidyapāda was probably not a direct student, a position I have since come to revise. 243 rgyud don ‘dzin bdag bla ma la’ang/ sangs rgyas ye shes zhabs la sogs pa’i rim pa’i man ngag rang nyid kyis dran phyir/ (Ātmārthasiddhikara, D 84b.3-4; P 100b.2) 244 rgyud don ‘dzin bdag bla ma la’ang/ sangs rgyas ye shes zhabs sogs rim pa’i man ngag rang dang gzhan gyis dran pa’i phyir/ (Samyagvidyākara, D 180a.5”) 245 sangs rgyas ye shes zhal snga’i rim ‘ongs pa/ ma lus bstan pa (Mahābalividhi, D 75b.5-6; P 90a.3-4). 246 bdag cag gi bla ma chen po sangs rgyas ye shes kyi zhal snga nas kyi rim pa nas ‘ongs pa’i rlung gi de nyid gyi sgo nas... (Ratnamati, D 81a.5; P 96a.8). 247 This verse admittedly has some problems, and its transmission may be corrupt. dpal ldan ye shes zhabs ni bla na med pa thugs rje chen po yis// drin gyi (gyi] P, gyis D) snang ba (ba] sugg. em., bas D P) khebs phyir bdag ni blun yang cung zad tsam du phye// (Ātmārthasiddhikara, D 94a.7; P112a.5) 248 Deb ther sngon po, vol 1, 451; Roerich 1976, 271. 47 Dīpaṃkarabhadra’s guru Śrījñānapāda.”249 Given that Vaidyapāda wrote a commentary on Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, we know that he was certainly junior to this master.250 Moreover, in Vaidyapāda’s commentary on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana he gives a citation from Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi and identifies the author there simply as “the (my?) guru.”251 Tāranātha, elsewhere in the Seven Transmissions, lists Vaidyapāda as a disciple of Praśāntamitra, who was also a direct disciple of Buddhajñānapāda’s.252 Given all of these considerations, I believe it to be likely that Vaidyapāda was indeed a direct student of Buddhajñānapāda’s, but that he met him when the latter was already quite aged, while Vaidyapāda was likely still quite young. Thus several among Buddhajñānapāda’s disciples, including Dīpaṃkarabhadra and possibly Praśāntamitra, also count among Vaidyapāda’s gurus. This would place Vaidyapāda’s life squarely in the early to mid 9th century, a date which is further supported by the works he does and does not cite in his oeuvre, and the content thereof.253 Unfortunately we know nothing of Vaidyapāda’s life from Indic sources. Turning to the later Tibetan histories, Tāranātha gives a short biographical sketch in his Seven Transmissions in which he notes that Vaidyapāda was born a brahmin in a border region and was learned and accomplished in non-Buddhist practices before gaining faith in the Buddhist teachings. He is also said there to have been a practitioner of the wrathful deity Hūṃkāra, who receives mention in several of Buddhajñānapāda’s short tantric writings.254 Tāranātha goes on to note that it is 249 yul dbus nālandār slob dpon mar me mdzad bzang po dang/ mar me mdzad bzang po’i yang bla ma dpal sangs rgyas ye shes gnyis la pha rol tu phyin pa nas brtsam te/ gsang sngag phyi nang gi bar gyi thos pas rgyud sbyang/ (Bka’ babs bdun, 112; Templeman 1983, 79). 250 The text is his Guhyasamājamaṇḍalopāyikā-ṭīkā. 251 bla ma’i zhal snga nas (Samantabhadra-ṭīkā, D134b.7). Samantabhadra, another commentator on the Caturaṅga who seems to have been reading Vaidyapāda’s commentary (I write more about the relationship between their commentaries in Chapter Five), cites the very same passage from Dīpaṃkarabhadra at this point in his Caturaṅga commentary, but Samantabhadra identifies the author of the citation as Bhadrapāda, a commonly used name for Dīpaṃkarabhadra (Sāramañjarī, D5b.4). This may be an indication that this master was not Samantabhadra’s personal guru, whereas Vaidyapāda may indeed have had such a personal relationship with Dīpaṃkarabhadra. 252 Bka’ babs bdun, 117; Templeman 1983, 66. 253 Vaidyapāda knows several texts that Buddhajñānapāda does not, most crucially the Samājottara, the so-called eighteenth chapter of the Guhyasamāja-tantra, and Śākyamitra’s Anuttarasandhi, which was incorporated into Nāgārjuna’s Pañcakrama as its second stage, the sarvaśuddhiviśuddhikrama. I discuss the possibility of Śākyamitra's discipleship with Buddhajñānapāda below, and Buddhajñānapāda’s and Vaidyapāda’s relationship with the Samājottara in Chapter Eight. 254 Tāranātha’s short (and charming!) biographical sketch reads: “The master Vaidyapāda was born into the brahmin caste in a border region. He became learned in the tīrthika doctrines and gained power through their practice, but later he gave rise to faith in the Buddha’s teachings. He trained in the tantras after having heard [teachings] beginning with the Prajñāpāramitā and continuing up to the outer and inner [[[tantras]] of the] secret mantra at Nālandā in Madhya from both Dīpaṃkarabhadra and Dīpaṃkarabhadra’s guru Śrījñānapāda, as well. [They] bestowed initiation on him and gave him the complete instructions. In particular, during initiation into the maṇḍalas of the Samāja and Heruka, his flower fell on wrathful Hūṃkāra. After meditating for a long time he gave rise to a unique samādhi of the two stages. He knew that after practicing for six months he would attain siddhi, but that he needed to rely as a practice support on a vajra-family consort, a doṃbi girl who was blue in color like an utpala flower, so he searched for her in all directions and [finally] found her. When requesting her from her parents they said, “Are you crazy, brahmin ācārya!? Since we are of the doṃbi caste, won’t [this] bring punishment upon both of us?” He answered, “Since I need a practice support, the ordinary castigations of [my association with] lower castes and so forth will not apply.” They said, “Well, then, we need gold and silver equal in weight to the girl’s body,” upon which the ācārya immediately brought forth a treasure from below the earth and gave it to them. Then the ācārya together with his consort practiced in a cave for six months upon which on the eighth day of the waxing moon at dawn a great sound hūṃ resounded from the sky and [he] directly saw all the maṇḍalas of Śrīheruka and so forth. [He] also attained the state of the siddhi of supreme mahāmudrā. He benefited many beings, primarily by means of the path of the Samāja, composed many treatises, and finally, in that very body, set off for Buddha Akṣobhya’s pure 48 sometimes reported that Vaidyapāda is the same individual as the master Hūṃkāra renowned in the Nyingma School.255 There are ten compositions attributed to Vaidyapāda (under a variety of names, see above) that survive in Tibetan in the Tengyur.256 While I have read only five of these in full, I have surveyed all of them and feel confident that on grounds of both style (some points related to which have also been noted above) and content, all of these works can be attributed to a single author whose main interest was clearly the elucidation and propagation of Buddhajñānapāda’s Guhyasamāja tradition.257 I will address the contents of several of his works in the succeeding chapters. Śākyamitra The final disciple of Buddhajñānapāda’s we will consider here is Śākyamitra. He receives no mention in the Tibetan histories in connection with Buddhajñānapāda, but rather is identified in several sources as a disciple of (the tantric) Nāgārjuna.258 Yet one Śākyamitra is the author of the Mukhāgama (Tōh. 1854, not to be confused with Buddhajñānapāda’s Dvitīyakramatattvabhāvanā-mukhāgama (Tōh. 1853), which is often referred to as the Mukhāgama in the secondary literature!), which he claims in both its opening and closing verses to be a record of the oral instructions of Buddhajñānapāda. This text is identified by both traditional and modern scholars alike as Buddhajñānapāda’s composition, presumably because the colophon reports it to be the “Oral Instructions (mukhāgama) on the sādhana of Buddhajñānapāda, [the master] who came from Glorious Uḍḍiyāna.”259 But Śākyamitra clearly states within the text itself that it was he who composed the treatise,260 and in a passage that is admittedly difficult to interpret, he seems to claim to have met Buddhajñānapāda and received land, taking off in flight like the king of Garuḍas” (Bka’ babs bdun, 112-12). My translation here differs only slightly from that of Templeman (1983, 78-9). 255 Bka’ babs bdun, 112; Templeman 1983, 79. If this is indeed the case, Tāranātha continues, it means he was born in Nepal and visited Tibet in the time of king Senalek. 256 The ten are the Samyagvidyākara (Tōh. 1850), Sukusuma (Tōh. 1866), Muktitilaka-vyākhyāna (Tōh. 1870), Samantabhadrī-ṭīkā (Tōh. 1872), Guhyasamājamaṇḍalopāyikā-ṭīkā (Tōh. 1873), Siddhisaṃbhavanidhi (Tōh. 1874), Yogasapta (Tōh. 1875), Mahābalividhi (Tōh. 1876), Ratnāmati (Tōh. 1877), and Ātmārthasiddhikara (Tōh. 1878). 257 A glance at the full titles of Vaidyapāda’s works listed in the previous note may appear alarming to some, specifically given the presence of the Yogasapta-nāma-caturabhiṣekaprakaraṇa, The Seven Yogas: An Explanation of the Four Initiations. Certainly this is unexpected, given that the early Jñānapāda tradition as found in Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, and even up until the 11th-century commentary on that text by Ratnākaraśānti, is well known in modern scholarship to preserve a tradition of just three initiations, rather than four (See Isaacson 2010a, 269; Wedemeyer 2014, unpublished, and Wedemeyer forthcoming). Vaidyapāda’s position in this treatise, and indeed its entire content, are worthy of serious further study, but for now it will suffice to say that I do feel confident that this text was authored by him. I address the Yogasapta in some detail in Chapter Seven. 258 Gö Lotsāwa holds Śākyamitra to be one of the four main disciples of Nāgārjuna (Roerich 1976, 359-60) and Tāranātha mentions that “Ācārya Śākyamitra the great was certainly a disciple of Ācārya Nāgārjuna, though [I] have not seen or heard his story.” slob dpon śākya bshes gnyen chen po yang slob dpon klu grub kyi slob ma yin par nges mod kyi lo rgyus ma mthong zhing ma thos so// (Rgya gar chos ‘byung, 114; Chimpa and Chattopadhyaya, 128). 259 Mukhāgama, D 28b.6. 260 Unlike Tomabechi (2008, 174), who writes that Śākyamitra is the “compiler of the Mukhāgama,” which he attributes to Buddhajñānapāda, I believe it is clear in the pledge to compose at the beginning of the text Śākyamitra holds himself to be the author, rather than simply having compiled Buddhajñānapāda’s oral instructions. In the admittedly difficult verses at the beginning of the Mukhāgama, Śākyamitra writes that he will, in this text, explain with clear words and without rhetoric or philosophy the profound meaning of the sādhana of Buddhajñānapāda (Mukhāgama, D 27b.3-6). This claim by Śākyamitra, which seems to amount to a pledge to explain Buddhajñānapāda’s complicated teachings in more common language, is very much in support of Tomabechi’s argument in his 2008 article that Āryadeva in the Caryāmelāpakapradīpa is referring to the Jñānapāda School authors with his comments on the unsuitability of the rhetorical complexity of their compositions. 49 his oral instructions, and even to be his primary disciple.261 In any case, there is no doubt that the contents of the Mukhāgama derive from Buddhajñānapāda’s teachings, as this text, like Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, contains a summary and rephrasing of a good portion of the Caturaṅga/Samantabhadra-sādhana. Śākyamitra, however, departs from Buddhajñānapāda’s work significantly more than did Dīpaṃkarabhadra, in particular adding an extensive instruction on the protection circle that is absent in the Caturaṅga/Samantabhadra.262 Other compositions attributed to Śākyamitra include the Koṣalālaṃkāra, a commentary on the Sarvatathāgatatattvasaṃgraha and the Anuttarasandhi, which was integrated into Nāgārjuna’s Pañcakrama as its second stage.263 Both of these works seem to show the influence of Buddhajñānapāda’s writings. The Kośalāṃkāra contains an autobiographical section extremely reminiscient of Buddhajñānapāda’s in the Dvitīyakrama describing Śākyamitra’s travels and studies under gurus, including to the Koṅkan and Uḍḍiyāna, places where Buddhajñānapāda also reports traveling.264 This record makes no mention of Buddhajñānapāda,265 but several of the lines in the autobiographical section sound so strikingly similar to those in the Dvitīyakrama, that it seems likely that one account has inspired the other. If the author of the Mukhāgama and the Kośalālamkāra are indeed the same individual, it would appear that in his autobiographical account Śākyamitra was emulating his guru, 261 As Tomabechi (2008, 174) notes, the passage in question “presents some difficulty in interpretation,” to say the least. (In fact, unfortunately much of the Mukhāgama presents difficulty in interpretation, and my sense is that either the original manuscript that the translator was working with had problems or a number of difficulties arose in its translation. The colophon states that the Mukhāgama was translated by Rinchen Zangpo alone, without mention of the assistance of an Indian paṇḍita, which may have been part of the problem.) But I nonetheless agree with Tomabechi that these verses in question do suggest that the compiler of the Mukhāgama claims to have met Buddhajñānapāda and received instruction directly from him. Unlike Tomabechi, I will (perhaps unwisely!) hazard a tentative translation of the passage in question (I include here also the two verses that precede those cited by Tomabechi (2008)), “These instructions that I have composed/ I received with great faith from the lotus of my guru’s mouth/ Let scholars treat it as a spectacle [if they wish],/ [From my side] I wrote it in order to benefit all beings./ Just like [even] someone who has achieved something through lies/ Is [still] praised a bit by his [[[own]]] father/ Just like that, I have found a bit of merit—/ Through it may all the world become [like] Mañjuvajra!/ [He who was] born in glorious Uḍḍiyāna,/ Knower of the meaning of the countless tantras without exception,/ The glorious one, completely pacified and peaceful, spoke these words./ The supreme main disciple upon whom he bestowed initiation,/ Who was born [in a place located] in the direction of the Sindh from there [i.e. from Uḍḍiyāna],/ Who victoriously resided in the place called the abode of Vaiśravana,/ I, Śākyamitra, myself, awakened and/ [Recorded] his (i.e. Buddhajñānapāda’s) perfectly liberating oral instructions.” bdag gis man ngag brtsams ‘di rab dad pas// bla ma’i zhal gyi padma las rnyed de// mkhas pa rnams kyis ‘di la ltad mo (ltad mo] D, brtag mod P) gyis// ‘gro ba kun la phan phyir bdag gyis byas// ji ltar log pa’i rdzun (rdzun] P, brdzun D) gyis byas pa yis// pha yi drung nas cung zad rnyed pa ltar// de bzhin bdag gis bsod nams cung zad rnyed// de nas ‘jig rtan ‘jam pa’i rdo rjer shog// dpal ldan u rgyan yul du brten (brten] D, bstan P) skyes shing// dpag med rgyud don ma lus rab mkhyen pa// dpal ldan rnam par dul zhi’i zhal snga nas// de yis dbang bskur slob ma’i gtso bo mchog// de las sin dhū’i ngos su de skyes shing// rnam thos bsti gnas zhes bshad rgyal zhugs nas// shā kya’i bshes gnyen bdag nyid sangs rgyas shing// de yis yongs bkrol zhal gyi man ngag go// (Mukhāgama, D 28b.3-5; P 33a.7-33b.2). A second passage earlier in the text also clearly references Buddhajñānapāda as the source of the instructions contained therein (dpal ldan ye shes zhal bshad sgrub thabs kyi// man ngag...) (Mukhāgama, D 17b.5-6). 262 The Mukhāgama does appear to include what indeed may have been oral instructions on the practice of the Caturaṅga/Samantabhadra-sādhana, as it also parallels in some places comments that are made by Vaidyapāda in his commentary on the Samantabhadra. 263 The Anuttarasandhi is cited by Vaidyapāda in his Muktitilaka-vyākhyāna, adding further evidence to the connection between Śākyamitra and Buddhajñānapāda’s tradition (Tomabechi 2008, 173). 264 Kośalālaṃkāra, D 1b.5-2a.5. This short passage is translated in Davidson 2002, 159-60. 265 It does, however, mention a master named Dharmākara as one of Śākyamitra’s teachers in the Koṅkan. One “Dharmākara of the Koṅkana” is also identified by Śrīphalavajra as Buddhajñānapāda’s co-disciple (Samantabhadrasādhana-vṛtti, D 141a.5-6). 50 Buddhajñānapāda.266 The Anuttarasandhi likewise appears to have been influenced by another of Buddhajñānapāda’s works, the Muktitilaka. Here Śākyamitra recounts the story of Buddha Śākyamuni’s awakening by the river Nairañjana, but his account seems to be based not on that in the Lalitavistara (which he mentions), nor even the so-called tantric retelling of this account in the Sarvatathāgatatattvasaṃgraha, but on Buddhajñānapāda’s own version of this account, given in his Muktitilaka.267 The account from the Muktitilaka is central to understanding Buddhajñānapāda’s thought, and I will address it in detail in Chapter Three, but for our present purposes, it is enough to note that it is Buddhajñānapāda’s version of the account of Śākyamuni’s awakening that seems to serve as the basis for Śākyamitra’s in the Anuttarasandhi. The Mukhāgama, Kośalālaṃkāra, and Anuttarasandhi also all include dedicatory verses in which the author has inscribed his own name into the verse, a technique favored by Buddhajñānapāda. Four other works in the Tibetan canon are attributed to Śākyamitra, though Wedemeyer has shown one of these, a commentary on the Caryāmelāpakapradīpa (Tōh. 1834) to be a Tibetan composition.268 Further study of the works attributed to Śākyamitra is certainly necessary to determine more about this influential269 but somewhat elusive author (or authors; indeed, the question of whether the remainder of these works can even be attributed to a single author needs to be addressed). However, given what is currently known of the writings attributed to him, I think we can suggest as likely that one Śākyamitra, author of at least the Mukhāgama, the Kośalālaṃkāra, and the Anuttarasandhi, was a disciple of Buddhajñānapāda’s.270 Death We know little of Buddhajñānapāda’s later life, including his death, but in the Dvitīyakrama Mañjuśrī makes a prediction that has been interpreted by commentators to mean that Buddhajñānapāda would not attain an awakening that involved the full transformation of his bodily aggregates in his lifetime, but only at the time of his passing. Mañjuśrī said: 266 Moreover, I will argue in Chapter Two that Buddhajñānapāda’s autobiographical narrative, which culminates in a vision of Mañjuśrī from whom he directly recieved the instructions that constitute the primary contents of the Dvitīyakrama, serves an important legitimizing function for Buddhajñānapāda’s system of practice outlined therein. Śākyamitra’s account of his studies with human gurus in India does not serve such a function, and thus is more likely to have simply been added in emulation of Buddhajñānapāda’s autobiographical account. Such autobiographies are extremely uncommon in Indian texts of this period—these two are the only ones of which I am aware. 267 Tomabechi (2006, 139n157) notes this parallel and suggests that Śākyamitra’s passage is based upon Buddhajñānapāda’s. There is another half-verse in the Anuttarasandhi that parallels one in Buddhajñāpāda’s Dvitīyakrama, but in that instance both verses appear to be based on a passage in the Sarvabuddhasamāyoga-tantra (See note 123 in my Dvitīyakrama translation.) Nonetheless, the fact that Śākyamitra paraphrases the same passage that Buddhajñānapāda has paraphrased remains telling. 268 These are the Bhadracaryāpraṇidhānarājaṭīkā (‘Phags pa bzang po spyod pa’i smon lam gyi rgyal po’i rgya cher ‘grel pa) (Tōh. 4013), the Krodhamahābalasādhana (‘phags pa khro bo stobs po che’i sgrub thabs) (Tōḥ. 3636), the Mahāmudrāyogāvatārapiṇḍārtha (phyag rgya chen po’i rnal ‘byor la ‘jug pa’i man ngag tu bshad pa), and the Caryāsamuccayapradīpa-nāma-ṭīkā (Spyod pa bsdus pa’i sgron ma zhes bya ba’i rgya cher bshad pa) (Tōḥ. 1834). The latter is the one demonstrated by Wedemeyer (2009) to be a Tibetan composition. 269 Both the Kośalālaṃkāra and the Anuttarasandhi are seminal texts in their own fields. The Kośalālaṃkāra is one of three major Indic commentaries on the Sarvatathāgatatattvasaṃgraha and is important as its earliest word-byword commentary (Hopkins 2005, 19); it influenced another important commentary on the tantra by Ānandagarbha (Kwon 2002, 26). The Anuttarasandhi is important in first introducing the theory of prakṛti and āloka, which serves as the “ontological and epistemilogical foundation of the yogic practice of the Ārya School in its entirety” (my translation from the French) (Tomabechi 2006, 49-50). 270 Tomabechi (2008) has already suggested as much as regards to the author of the Anuttarasandhi and what he refers to as the “compiler” of the Mukhāgama (he does not address the Kośalālaṃkāra) and places this Śākyamitra at a critical juncture between the Jñānapāda and Ārya traditions. 51 However, because of [your] conduct regarding food, And holding a slight delusion with respect to me You will not, in this very life, Bring about a complete transformation of the state of Your body—the aggregates including form. |365| However, you will accomplish consciousness, Which is indestructible, as the mahāmudrā.271 |366| As we have seen in Chögyal Phagpa’s account translated above,Vaidyapāda was followed by the later Tibetan historians in interpreting this prediction to mean that due to Buddhajñānapāda’s earlier lack of faith in the monk who emanated the maṇḍala of Mañjuśrī and his refusal of some foods that the monk’s female companion served to him he would attain the final result of the path only at death.272 This position is stated most clearly in Tāranātha’s Seven Transmissions: If he had previously not given rise to any lack of faith at all with regard to the emanation of Mañjuśrī, who was a practitioner of the avadhūti, he would have transformed in that very [[[body]]] into the vajra rainbow body. However, since he had some minor disrespectful thoughts [towards him], at the age of eighty he left behind the body [produced by karmic] ripening and attained the body of unity.273 Here we have the only mention of which I am aware of Buddhajñānapāda having lived such a long life, passing away only at the age of eighty, no small feat in 8th-century India. Buddhajñānapāda’s Writings You should compose with a genuine intention a sādhana, homa, bali, gaṇacakra, summary, commentary, maṇḍala-vidhi, and so forth for the first stage of the tantra that is the gathering of all the buddhas, which is greatly secret, secret, and supremely secret—this great scripture, surpassed by none—to be like a scalpel for sentient beings who are obscured by the darkness of ignorance. - Mañjuśrī, addressing Buddhajñānapāda directly, Dvitīyakrama Fortunately for our study of his thought, Buddhajñānapāda was a fairly prolific author. Unfortunately, few of his works survive in their original Sanskrit, and our only access to most of his oeuvre is through their Tibetan translations, made primarily in the 11th century.274 The earliest list we have of Buddhajñānapāda’s writings is in the Sukusuma, in which Vaidyapāda comments on Mañjuśrī’s command to Buddhajñanapāda to compose a number of texts (see the 271 khyod kyang zad kyi spyod pa dang// nga la cung zad ‘khrul rtogs pas// khyod kyis (kyis] D C S, kyang P, kyi N) tshe ‘di nyid la ni// gzugs bcas phung po rang lus ni// gnas ni yang dag mi ‘gyur te// |365| rnam par shes pa mi shigs pa// phyag rgya chen por rab tu ‘grub// |366| (Dvitīyakrama, v 365-66). 272 Vaidayapāda’s own comments on the point of not transforming the aggregates of the body in this life are somewhat confusing, but their interpretation by later authors is clearer. See note 504 in my translation of the Dvitīyakrama for more detail on Vaidyapāda’s comments on this passage. 273 Bka’ babs bdun, 108. Here my translation differs significantly from that of Templeman (1983, 76), who I believe has misunderstood the passage. sngon ‘jam dbyangs sprul pa a ba dhū ti’i spyod pa can de la ma dad pa gtan nas ma skyes na/ de nyid du ‘ja’ lus rdo rje’i skur gnas ‘gyur ba yin pa las/ der ma gus pa’i rnam par rtog pa cung zad skyes pas dgungs lo brgyad cu lhan cig bzhes pa na rnam par smin pa’i sku lus bor te/ zung ‘jug gi sku brnyes pa yin no// 274 The translation of Buddhajñānapāda’s Sañcayagāthā-pañjikā, a Prajñāpāramitā commentary that is one of his early works, however, is attributed to the 8th-century Imperial Era translators Kawa Paltsek and Vidyākarasiṃha. 52 quotation above), by listing the master’s compositions that fulfill this command. The sādhana, Vaidyapāda says, refers to the “three Samantabhadrīs” (kun tu bzang mo gsum); the homa is (for?) the generation stage, and he notes that there are two such homa rituals; the bali ritual is that of Unfaltering Tārā (mi nub pa’i sgrol ma); the gaṇacakra text is the Mahāgaṇacakra (though it is unclear if this is meant to be the name of a text or simply stating that it is a ritual for the practice of the mahāgaṇacakra);275 the summary is the Blazing Gem (rin po che ‘bar ba); and the commentary “he did not compose.” As for the maṇḍala-vidhi, Vaidyapāda notes that this vidhi in two hundred and fifty verses was taken to Kaśmir and that he had not seen it. He explains that the “and so forth” includes the Great Root Wisdom (rtsa ba’i ye shes chen po) and the Treasure of Verses (tshigs su bcad pa’i mdzod),276 the Muktitilaka (grol ba’i thig le), and the Ātmasādhanāvatāra (bdag nyid grub par ‘byung ba),277 the *Bodhicittabindu (byang chub sems kyi thig le), the Great Commentary on Glorious Auspiciousness (dpal bkra shis kyi rnam par bshad pa chen po), The Method for Engaging in the Fourth (bzhi pa la ‘jug pa’i thabs),278 and three Jambhala sādhanas.279 Many of these texts cannot be identified among Buddhajñānapāda’s extant works,280 but some of them fortunately can: his Samantabhadra-sādhana,281 Muktitilaka, Ātmasādhanāvatāra, and the three Jambhala sādhanas all survive. 275 Although unfortunately this text is not extant, it does seem to have been translated into Tibetan, as a gaṇacakravidhi (tshogs kyi ‘khor lo’i cho ga) attributed to “Jñānapāda” is listed in the Black Hat Tanjur catalogue (p. 434): tshogs kyi ‘khor lo’i cho ga ye shes zhabs kyi o rgyan gyi yul nas spyan drangs nas mdzad pa smri ti’i ‘gyur (Martin 2011, 650). Smṛtijñānakīrti also translated Buddhajñānapāda’s Caturaṅgasādhana, so it seems likely that this may indeed have been Buddhajñānapāda’s gaṇacakravidhi that he translated, as well. 276 There is at least the outside possibility that this could refer to Buddhajñānapāda’s bsdud pa tshigs su bcad pa’i dka’ ‘grel, the Sañcayagāthā-pañjikā. 277 The title of this text is usually translated into Tibetan as bdag nyid grub pa la ‘jug pa, but presumably it refers to the same text here. 278 Among Buddhajñānapāda’s compositions mentioned by Vaidyapāda that are no longer extant, this title in particular is tantalizing. As noted above, the early Jñānapāda School is generally known to have preserved a system of three, rather than four, initiations, but Vaidyapāda composed the Yogasapta, a treatise on the “seven yogas” of the “the fourth.” We may guess that this Method of Engaging in the Fourth (if it did ever exist—but we have no good reason to doubt Vaidyapāda’s claim that it did) may have had something to do with Buddhajñānapāda’s own position regarding a fourth initiation. I discuss Vaidyapāda’s Yogasapta in Chapter Seven. 279 Sukusuma, D 133b.7-134a.3. Vaidyapāda then notes that these “fourteen teachings” were composed in accordance with Mañjuśrī’s prediction. The only way I have been able to make this list total fourteen is by counting each of the texts listed in the root text as one (7; ignoring the fact that Vaidyapāda says that the sādhana actually refers to three texts, and the homa to two), subtracting the commentary that Vaidyapāda says was not composed (-1), and adding the texts Vaidyapāda lists in as part of the “etc.” (+8; again ignoring the fact that the “three Jambhala sādhanas” counts only as one of the eight). Gö Lotsāwa also gives the list of Buddhajñānapāda’s compositions, clearly drawn from Vaidyapāda, and engages in a similar (but not identical) mathematical endeavor regarding this list of “fourteen”! (Deb ther sngon po, Vol I, 550). In the Blue Annals Roerich has mistakenly identified several of the members of the list that Gö gives with texts that are not Buddhajñānapāda’s. Regarding Roerich’s misidentification of the third “Samantabhadra” text, see note 281. Roerich also mistakenly identifies the “commentary on the tantra” (which Vaidyapāda and Gö both report that Buddhajñānapāda did not compose(!)) as the Candraprabhā (Tōh. 1852), which was actually composed by Pramuditākaravarman. Roerich later reports that this was not Buddhajñānapāda’s work, but it seems that his translation is confused here because he was not aware of Vaidyapāda’s passage from the Sukusuma, which Gö was clearly paraphrasing in this section. Roerich also reports Buddhajñānapāda’s 250-verse maṇḍala-vidhi to be the Guhyasamājamaṇḍalavidhi (Tōh. 1865), which was actually composed by Dīpaṃkarabhadra and has closer to 450 verses. The list of “fourteen” is likewise reported (obviously relying on Vaidyapāda) in Chögyal Phagpa’s biography (Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, 615-16). 280 There are twelve texts listed by Vaidyapāda here that we do not know: 1. the third among the “three Samantabhadrīs,” (on which see note 279) 2. and 3. the two homa rituals, 4. the bali ritual of Unfaltering Tārā, 5. the gaṇacakra rite, 6. the Blazing Gem, 7. the maṇḍalavidhi in 250 verses, 8. the Great Root Wisdom, 9. the Treasury of Verses, 10. the *Bodhicittabindu, 11. the Great Commentary on Glorious Auspiciousness, and 12. the Method for Engaging in the Fourth. 53 It seems unlikely that Vaidyapāda intends for the list of fourteen works given in the Sukusuma to encompass the entirety of Buddhajñānapāda’s oeuvre, as he is merely commenting on the specific set of Guhyasamāja-related texts that Buddhajñānapāda composed to fulfill Mañjuśrī’s command. Indeed, in addition to the six works from Vaidyapāda’s list that do survive, a number of other compositions in the Tibetan canon (a few of which also survive in their entirety or in part in Sanskrit) are attributed to Buddhajñānapāda under a variety of names. Many of these compositions do indeed appear to be the works of the 8th-century master, though at least one is certainly not his, and the attribution of others remains in question. The following sixteen works in the Tibetan canon are attributed to authors named Buddhajñānapāda, Buddhaśrījñāna, or one of the other above-mentioned variants of the name; I will present a very brief summary and assess the attribution of each of them below: the Dvitīyakramatattvabhāvanāmukhāgama (Tōh. 1853); Mukhāgama (Tōh. 1854); Samantabhadra-sādhana (Tōh. 1855); Caturaṅgasādhanopāyikā-samantabhadrī (Tōh. 1856); Śrīherukasādhana (Tōh. 1857);282 Muktitilaka (Tōh. 1859); Ātmasādhanāvatāra (Tōh. 1860); Bhaṭṭārakāryajambhalajalendrasādhana (Tōh. 1861); Guhyajambhalasādhana (Tōh. 1862); Vistarajambhalasādhana (Tōh. 1863);283 *Gativyūha (Tōh. 1864); *Trikāyavākcittādhiṣṭḥanoddeśa. (Tōh. 2085); Traisattvasamādhisamāpatti (Tōh. 2086); Mahāpratisarārakṣā (Tōh. 3124); Sañcaya-gāthāpañjikā (Tōh. 3798); and the Mahāyānalakṣaṇasamuccaya (Tōh. 3905).284 In assessing the 281 It is interesting that Vaidyapāda refers to this text(s) as the: “three Samantabhadrīs,” as two translations of the sādhana into Tibetan survive in the canon under two different names, the Samantabhadra-sādhana and the Caturaṅga-sādhanopāyikā-samantabhadrī, which upon comparison are not actually distinct texts. There is also a third “Samantabhadra” text, the Kun tu bzang po bsdus don listed in the Peking Tengyur catalogue, but the text itself, however, is strangely absent from the place where it should be in that Tengyur, and receives no mention whatsoever in the catalogues of the other Tengyurs. Gö Lotsāwa (Deb ther sngon po, Vol 1, 450) mentions the Kun tu bzang po bdus don as the third among the “three Samantabhadrīs,” and were the text to be found it would indeed be a good candidate, given its placement in the Peking Tengyur catalogue together with other Jñānapāda Schoool works. Roerich (1976, 370), however, inexplicably identifies the Kun tu bzang po bsdus don as Buddhajñānapāda’s Śrīherukasādhana (Tōh. 1857). I address the issue of the two different extant translations/versions of the Caturaṅga/Samantabhadra-sādhana in Chapter Five. 282 Also in Sādhanamālā, No. 243. 283 Also in Sādhanamālā, No, 285. This short sādhana contains no statement of authorship, but I have included it in the list because I believe we can attribute it to Buddhajñānapāda. I discuss this further below. 284 I have not included in this list three compositions from the canon (and a fourth that is a short prayer extracted from one of the other three) that are clearly attributed to the “Kaśmiri paṇḍita Buddhaśrījñāna,” who lived around 1200 and worked with the Tibetan Lotsāwa Nub Jampai Pal (Gnubs byams pa’i dpal, 1173-1236) on a number of translations preserved in the Tibetan canon, including those of his own writings. Makransky (1997, 268) and C. Dalton and Szántó (forthcoming) mention this Buddhaśrījñāna as the author of an Abhisamayālaṃkāra commentary, the Prajñāpradīpāvalī, and as a namesake of Buddhajñānapāda’s, and warn against confusing the two. In addition to the Prajñāpradīpāvalī (Tōh. 3800), two other works in the Tengyur—the Cittaratnaviśodhanamārgaphala (Tōh. 2465), and the Jinamārgāvatāra (Tōh. 3964) are clearly attributed to this Kaśmiri paṇḍita; indeed in the colophons of all three (and even in several of the colophons of the works he collaborated on as a translator) he is specifically referred to as “the Kaśmiri” Buddhaśrījñāna, presumably to distinguish him from Buddhajñānapāda who as we have seen was occasionally also called by that name. A fourth work, the Jinamārgāvatārodbhavapraṇidhāna (Tōh. 4391), attributed in its colophon to one Buddhaśrījñāna (Sangs rgyas dpal ye shes) (but with no specification that he is the Kaśmiri paṇḍita of that name, nor any translator's colophon) is in fact a prayer that has simply been extracted from his Jinamārgāvatāra (See Jinamārgāvatāra, D 234a.3-235a.1), with a single introductory verse added at the beginning. Buddhaśrījñāna is said in the colophon of his Citta-ratnaviśodhanamārgaphala to be of Kaśmiri blood but to have been born in Nepal, where several of his translations with Nub Jampai Pal were carried out. Besides the Kaśmiri Buddhaśrījñāna’s own three compositions, the two collaborated on the translation of a somewhat eclectic collection of other works in the canon: Ratnākaraśānti’s Vajratārāsādhana (Tōh. 1324); Kālapāda’s Śrīkālacakrasahajasādhana (Tōh. 1361); Maitrīpa’s Śrīcakrasaṃvarasādhana-ratnadīpa (Tōh. 1484); Śrīdhara’s Krodhavārāhīvajrayoginīsādhana (Tōh. 1586); Līlāvajra’s Śrīsahajaguhyasamājasādhana (Tōh. 1913) (That this short sādhana was composed by the later Līlāvajra and not Buddhajñānapāda’s guru Vilāsavajra, whose names in 54 authorship of these works, I am taking into account a number of factors: their presence in Vaidyapāda’s list of Buddhajñānapāda’s writings, their colophonic attribution, their style— including the presence of a dedicatory verse in which the author has inscribed his name, Buddhajñāna, in several of the works—and their content.285 In the course of my research I was able to read all of the sixteen works above in full, with the single exception of the Sañcayagāthāpañjikā, Buddhajñānapāda’s long Prajñāpāramitā commentary. Brief Summary and Assessment of Authorship of works Attributed to Buddhajñānapāda Non-Tantric Works Buddhajñānapāda’s Sañcayagāthā-pañjikā (Tōh. 3798), a commentary on the Sañcayagāthā Prajñāpāramita sutra, seems to have been composed quite early in his career. In the Dvitīyakrama Buddhajñānapāda mentions having composed some treatises at Nālandā in response to the request “the one of noble birth called *Guṇamitra/ā (yon tan bshes gnyen)”286 described by Vaidyapāda as a brahmin nun,287 prior to continuing his travels around the subcontinent studying with tantric gurus. The concluding verses of the Sañcayagāthā-pañjikā, in addition to including a dedicatory verse into which Buddhajñānapāda inscribes his own name, mention as its petitioner the very same *Guṇamitra/ā (here her name is transliterated as gu ṇa mi tra).288 In the Sañcayagāthā-pañjikā Buddhajñānapāda relates each section of the Abhisamayālaṃkāra to passages of the Ratnaguṇasañcayagāthā.289 Here he seems to be emulating his guru Haribhadra’s method of relating the Abhisamayālaṃkāra to one of the shorter Prajñāpāramitā texts; Haribhadra, in his well-known Abhisamayālaṃkārālokā, relates the Abhisamayālaṃkara to the Aṣṭasāhasrikā.290 The Sañcayagāthā-pañjikā was translated into Tibetan during the early translation period, very likely before 824, making it possible that this work may even have been translated during Buddhajñānapāda’s own lifetime.291 The Mahāyānalakṣaṇa-samuccaya (Tōh. 3905) was likewise composed early in Buddhajñānapāda’s career. As noted above, this text has the unusual distinction of having been cited by Buddhajñānapāda’s guru Vilāsavajra in his Nāmamantrārthāvalokinī.292 However, the Mahāyānalakṣaṇa-samuccaya is attributed to Buddhajñānapāda in Samantabhadra’s mid-late 9thcentury (?) Sāramañjarī,293 which cites portions of the text; thus we do have, besides the Tibetan Tibetan are often rather interchangable, is clear from a reference in the work to the four joys (dga’ bzhi). Buddhajñānapāda and even Vaidyapāda only spoke of three.); Niṣkalaṅka’s Śrībandhavimukta-śāstra (Tōh. 2463); Candrakumāra’s Śrīvajrasarasvatīdevyupāyikā (Tōh. 3699); and Vasudhara’s Āryajambhalajalendraviśeṣastotra (Tōh. 3747). 285 For another brief summary of Buddhajñānapāda’s surviving works see C. Dalton and Szántó forthcoming. 286 Dvitīyakrama, verse 4. 287 Sukusuma, D 89b.2. 288 This verse clearly gives the impression that the Sañcayagāthā-pañjikā is one of Buddhajñānapāda’s juvenilia. It reads, “When someone else, her gaze on the result (i.e. awakening), / entrusts one [with a task], why not try?/ [I] wrote this commentary on the difficult points/ for Guṇamitrā alone.” ‘bras bu la lta gzhan gyis kyang// bcol na ci phyir ‘bad mi bya// gu ṇa mi tra kho na’i ngor// dka’ ‘grel ‘di ni ‘dir byas so// (Sañcayagathā-pañjikā, D 189a.4-5). 289 Makransky 1997, 259-60. 290 ibid., 270. See Makransky 1997, 259-63 for a summary of Buddhajñānapāda’s position on the number of kāyas in the Sañcayagāthā-pañjikā, which suggests that here he did not follow the innovative four-kāya theory in his guru Haribhadra’s Abhisamayālaṃkārālokā. 291 See Tomabechi 2008, 175. 292 Tribe 1994, 16; Szántó 2015, 541; C. Dalton and Szántó, forthcoming. 293 Samantabhadra’s mid-late 9th-century dates are suggested by the fact that he mentions that he received the command to compose the Sāramañjarī by one Kīrtipāda, who is likely the same Śrīkīrti who was a disciple of 55 colophonic attribution, an early attestation of its attribution to Buddhajñānapāda by an Indian author who upheld his tradition not long after Buddhajñānapāda’s life.294 The text itself is a relatively short compilation of definitions of basic Buddhist terminology and important aspects of the Mahāyāna path like the aggregates, elements, links of dependent origination, pāramitās, bodhisattva bhūmis, sixteen emptinesses, and so forth.295 The format of the Mahāyānalakṣaṇasamuccaya, in which Buddhajñānapāda begins with a short invocation, then simply lists the topics he will discuss and proceeds to address each in turn, is also found in his *Gativyūha, a text that does bear a signature dedicatory verse. A fragment of the original Sanskrit of the Mahāyānalakṣaṇa-samuccaya survives and has been published, and more of the text can be reconstructed from its citations in the longest available recension of the aforementioned Sāramañjarī.296 Major Tantric Works The Dvitīyakramatattvabhāvanā-mukhāgama297 (Tōh. 1853), edited in Appendix I and translated in Part II of this dissertation, is one of Buddhajñānapāda’s most well-known compositions. It contains his autobiographical account, which serves as the text’s narrative frame, details his vision of Mañjuśrī, and reports the entire contents of the instructions Mañjuśrī gave him in the vision. The instructions given by Mañjuśrī to Buddhajñānapāda reported in the Dvitīyakrama are themselves quite eclectic and the text contains a diverse collection of doctrinal and ritual sections nestled within Buddhajñānpāda’s unique narrative framework. Almost all of the ritual material in the text pertains to the practices of the perfection stage, the “second” stage of tantric practice, as well as the initiatory rituals that permit the practitioner to engage in those practices. This work shares strong similarities and an overlap in vocabulary and content with the Muktitilaka, including a number of parallel passages, and to a lesser degree with the Caturaṅga/Samantabhadra-sādhana, and the Ātmasādhanāvatāra. The Dvitīyakrama is Buddhajñānapāda’s only major extant work that lacks a dedicatory verse with his name inscribed into it. I presume that in the case of the Dvitīyakrama, Buddhajñānapāda found this unnecessary given that the work itself contains his autobiographical details, and that in fact ninety percent of the content is, technically speaking, not actually Buddhajñānapāda’s own composition, but simply his report of Mañjuśrī’s direct instructions, recorded in Mañjuśrī’s own first-person speech. Chapter Two examines its structure and provides a summary of this unique composition, while different aspects of the Dvitīyakrama’s content—its doctrine, generation-stage ritual, perfection-stage ritual, and initiatory sequences—are explored in more detail in each of the remaining chapters. The Muktitilaka (Tōh. 1859) is another important work of Buddhajñānapāda’s, which is known, along with the Dvitīyakrama with which it shares much vocabulary and content and multiple parallel passages, as presenting the perfection stage practices according to Buddhajñanapāda’s Guhyasamāja tradition. While the Muktitilaka certainly does contain some perfection stage instructions—specifically, on the three bindu yoga and vajra recitation—those same practices are presented much more extensively in the Dvitīyakrama. Moreover, perfection Buddhajñānapāda’s guru Pālitipāda, and possibly a co-disciple of Buddhajñānapāda’s. He thus seems to have been one-and-a-half or two generations later than Buddhajñānapāda (See Szántó 2015a, 554). 294 Szántó 2015, 541; C. Dalton and Szántó, forthcoming. 295 Yonezawa 1998 summarizes its contents. 296 Yonezawa 1998; Szántó 2015a, 545. Szántó has generously shared with me his draft edition of the long recension of the Sāramañjarī. 297 I explain my departure from the title usually given for this work, the Dvikramatattvabhāvanā-mukhāgama, in Chapter Two, and also in note 3 of my translation of the Dvitīyakrama in Part II of the dissertation. 56 stage instructions make up only a small portion, maybe fifteen percent, of the Muktitilaka’s content. In fact, the text contains much more doctrinal than ritual material, some of which is presented in innovative narrative ways. The Muktitilaka places a special emphasis on nondual nonconceptual suchness, which can be known instantaneously through relying upon a realized lineage guru, and which subsumes all other outer Vajrayāna practices. Buddhajñānapāda has inscribed his name here in the concluding verses. Chapter Three examines the doctrinal content of the Muktitilaka in more detail, while Chapter Six focuses on the perfection stage practices found here and in the Dvitīyakrama. The Samantabhadra-sādhana (Tōh. 1855) and the Caturaṅgasādhana (Tōh. 1856) are in fact two translations of the same sādhana, which is undoubtedly Buddhajñānapāda’s composition.298 This important work details the rituals of the generation stage practice according to Buddhajñānapāda’s tradition, a nineteen-deity maṇḍala centered on Mañjuvajra, which became distinctive of his Jñānapāda School of Guhyasamāja practice. The Caturaṅga/Samantabhadra-sādhana shares significant parallels, including several parallel passages, with Buddhajñānapāda’s Ātmasādhanāvatāra, particularly in the philosopical section towards the end of the sādhana. It also bears one of his signature dedicatory verses. The sādhana was obviously popular, as it is the subject of five extant commentaries, one of which survives in three different recensions.299 Moreover, almost the entire sādhana was rephrased in his disciple Dīpaṃkarabhadra’s influential Guhyasamajamāṇḍalavidhi. The Samantabhadra survives in its entirety in Sanskrit, but is unfortunately not available for study.300 However, a short portion of the Sanskrit has been photographed and edited, and some has been reconstructed and published on the basis of the Sāramañjarī, an extant Sanskrit commentary on the sādhana.301 This sādhana, including the rituals of Buddhajñānapāda’s generation stage practice, the details of its two different surviving translations into Tibetan, and an overview of the extant commentaries and their relationships, is discussed in detail in Chapter Five. The Ātmasādhanāvatāra (Tōh. 1860) is Buddhajñānapāda’s only surviving tantric work written in prose, rather than verse. It is a complex treatise in which he sets a Yogācāra- Madhyamaka philosophical foundation for arguments that support and defend the tantric path, notably a defense of the practice of deity yoga. The treatise goes so far as to make the claim that full awakening is only possible through the tantric path of deity yoga. The work also deals, however, with a number of Mahāyāna topics approached without a tantric lens. It additionally seeks, as do several of Buddhajñanapāda’s other works, to homologize several important Mahāyāna concepts with tantric concepts or practices, and to identify all of these with, or subsume them within, suchness, which is also described here as the nature of the mind. This text is particularly important as it contains a number of citations of other Buddhist works, giving us 298 Some scholars have sometimes considered these two to be two different works, but more recent scholarly consensus (including Kikuya 2012a, Szántó 2015, and C. Dalton and Szántó forthcoming), which my own reading of the sādhana strongly supports, is that they are indeed two translations of the same work. 299 C. Dalton and Szántó (forthcoming). The commentaries are Vaidyapāda’s Samantabhadrī-ṭīkā, Śrīphalavajra’s Samantabhadrasādhana-vṛtti, Thagana’s Śrīsamantabhadrasādhana-vṛtti, Samantabhadra’s Sāramañjarī, and an unidentified commentary in Sanskrit mentioned in Kawasaki, 2004. The Sāramañjarī survives in three recensions, two in Sanskrit and one in Tibetan. 300 Kawasaki 2004 describes the manuscript that contains the Samantabhadra, but which remains unavailable to scholars to study. 301 Kanō (2014) has published an edition of the verses from a short section of this manuscript which was photographed. The manuscript is on display in the Tibet Museum at the Norbulingka in Lhasa. Other verses survive in several Nepalese ritual manuals (one at the Institute for Advanced Studies of World Religions, IASWR MBB-I- 11; one in the Nepal National archives, NAK 1/1697 = NGMPP A 936/1; and one at the Cambridge University Library Add. 1708.III, f. 2r4-5), and are edited in Tanaka 1996. Szántó (2015, 543n14) has published a revision of Tanaka’s edition of these verses. 57 an idea of the scriptural resources Buddhajñānapāda drew on in his thought. These include the Sarvatathāgatatattvasaṃgraha and the Sarvabuddhasamāyoga-tantra, but the Ātmasādhanāvatāra is unusual in lacking any reference to the Guhyasamāja-tantra, which serves as the central reference for most of his other tantric writings. This absence raises the question of whether the Ātmasādhanāvatāra may have been composed prior to Buddhajñānapāda’s Guhyasamāja-focused tantric works. Moreover, while the Ātmasādhanāvatāra’s only specific reference to Mañjuśrī is in the work’s homage “to the bodhisattva Mañjughoṣa,” as I noted above Buddhajñānapāda has reproduced a lengthy part of his guru Vilāsavajra’s Nāmamantrārthāvalokinī commentary on the Mañjuśrīnāmasaṃgīti towards the end of the Ātmasādhanāvatāra.302 The section in question equates the deities of the Mañjuśrīnāmasaṃgīti maṇḍala described in Vilāsasvajra’s commentary with a number of Mahāyāna principles. So while the Ātmasādhanāvatāra on the whole cannot be said to be a Mañjuśrīnāmasaṃgīti-centered work, the presence—albeit without any introduction or explanation—of the maṇḍala from that tantra in this work confirms Buddhajñānapāda’s familiarity with the Mañjuśrīnāmasaṃgīti, as well as suggesting that the work was indeed likely composed before the Guhyasamāja-tantra became the focus of his writings. In any case, the Ātmasādhanāvatāra is undoubtedly Buddhajñānapāda’s work, given both the presence of his “signature” in the dedicatory verse, as well as the parallels with his Caturaṅga/sSamantabhadrasādhana. Like the Caturaṅga/Samantabhadra, the Sanskrit of the Ātmasādhanāvatāra also survives in full but is likewise unavailable for study,303 but quite a large portion of the Sanskrit text can be reconstructed from the long recension of the Sāramañjarī where it is cited at length.304 The presence of the passage from the Nāmamantrārthāvalokinī, which is extant in Sanskrit, now provides us further access to another portion of the Ātmasādhanāvatāra in its original Sanskrit. The availability of a large portion of the Ātmasādhanāvatāra in Sanskrit is most fortunate, especially given that the Tibetan translation of the text is incomprehensible in a number of places.305 Chapter Three addresses some of the doctrinal points found in the Ātmasādhanāvatāra. Minor Tantric Works In addition to these important tantric compositions, Buddhajñānapāda wrote a few shorter works dealing with tantric practice, mostly sādhanas. The *Gativyūha (Stang stabs kyi bkod pa, Tōh. 1864) is the most unique among these works, as it is not a sādhana, but rather a short text detailing the postures and mudrās of deities, as well as postures and mudrās to be assumed in certain tantric ritual contexts. The first and last sections of the text are in verse, with a prose 302 A number of disparate short sections are taken directly from Chapter Five of the Nāmamantrārthāvalokinī, and incorporated into an almost continuous segment of the Ātmasādhanāvatāra, with very little interjection on Buddhajñānapāda’s own part. The Sanskrit edition of this section of the Nāmamantrārthāvalokinī is found in Tribe 2016 pp. 268-281 and corresponds with the Tibetan of the Ātmasādhanāvatāra found in the Derge recension of that text, D 57a.3-58b.6. The segments of the Nāmamantrārthāvalokinī incorporated into the Ātmasādhanāvatāra are: Nāmamantrārthāvalokinī, Chapter 5, lines 14-20, 36-43 71-76, 81-87, 104-113, 131-139, 156-159, 161-152, 178- 179, 194-195, and 200-203 (following Tribe’s Sanskrit edition, pp. 268-281). 303 See Kawasaki 2004, 51. 304 Szántó 2015, 545-46. Szántó has generously shared with me his draft edition of the Ātmasādhanāvatāra based on its citations in the Sāramañjarī. 305 The colophon states that the translation was done by Śāntibhadra (shānti bha dra) and Gö Khukpa Lhetse (‘go lhas btsas). The Sanskrit text is also difficult, so it is possible that the translators had some trouble with the text, or that they were working with a corrupt manuscript. In any case, the Ātmasādhanāvatāra as it stands in Tibetan translation alone is quite unapproachable in more than one place. 58 section in the middle. While the text lacks the traditional translator’s opening with a Sanskrit title and translator’s homage, as well as a translator’s colophon, it does have a dedicatory verse in which Buddhajñānapāda has inscribed his own name. The *Gativyūha also begins with an homage to Vajra Hūṃkāra, who is mentioned twice in Buddhajñānapāda’s Jambhala sādhanas as the self-visualization for the more wrathful methods of accomplishment.306 As noted above, a composition by Śrīkīrti, who seems to have been a fellow disciple of Buddhajñānapāda’s under their master Pālitapāda likewise focuses on postures among the other details of the initiatory ritual, so it is possible that this was something emphasized by their common guru.307 The Śrīherukasādhana (Tōh. 1857) is a short work which is, as its name suggests, focused on the figure of the wrathful Heruka. The general structure of the sādhana, albeit in an extremely condensed form, is the same as that of Buddhajñanapāda’s Caturaṅga/Samantabhadra-sādhana, and the causal deity (described here in the Śrīherukasādhana as the vajrasattva and in the Samantabhadra as Vajrabhadra308), is identical in terms of form, color, and implements, to the causal deity as described in the Caturaṅga/Samantabhadra. These shared features suggest that the attribution to Buddhajñānapāda (his name is given as the Ācārya Buddhajñāna (slob dpon sangs rgyas ye shes) in the colophon) may indeed be correct.309 Moreover, the sādhana is grouped with others of his works in the Tibetan canon, suggesting that the redactors of the canon may also have accepted the attribution.310 Although the gate guardians depicted in the Caturaṅga/Samantabhadrasādhana are certainly wrathful in aspect, the specifically cremation ground aesthetic of the Heruka in the Śrīheruka-sādhana—holding a skull garland, his body smeared with ash, garlanded by bones—is not reflected in Buddhajñānapāda’s other writings, and is more often associated with the later Yoginī tantras.311 The short commentary (Tōh. 1858) that follows this sādhana in the Tengyur clearly associates this work with the Guhyasamāja-tantra, and, as seen in the 9th-century Guhyasiddhi (a text that is probably slightly later than Buddhajñānapāda) that aesthetic was already associated with Guhyasamāja practitioners engaging in vrata practices, even if it was not representated in the aesthetic of the primary deities of the Guhyasamājamaṇḍala itself.312 What is more, the idea of and the term heruka were certainly in use in Buddhajñānapāda’s time, even in conjunction with the Guhyasamāja cycle, as both Vaidyapāda and Samantabhadra’s commentaries specify the herukas among the “others” in whose accumulations of merit Buddhajñānapāda rejoices in a verse from the Caturaṅga/Samantabhadra-sādhana.313 If the Śrīheruka-sādhana is indeed a composition of Buddhajñānapāda’s, it would thus seem to be an early example of the genre of a Heruka sādhana with a cremation ground aesthetic. Most of the sādhana, with the exception of four and a half 306 This practice is mentioned in both the Bhaṭṭārakāryajambhalajalendra-sādhana and the Vistarajambhalasādhana. 307 See Szántó 2015a, 552-3. 308 rdo rje bzang. A combination of the terms Vajradhara and Samantabhadra? The Caturaṅga reads rdo rje dam pa. 309 The Śrīherukasādhana is followed in the Tengyur by a short commentary (Tōh. 1858) on the work that is anonymous. The commentary clearly identifies the sādhana as pertaining to the Guhyasamāja tradition, and comments on the full sādhana as contained in the Tibetan canon, including the final four and a half verses on the Heruka vrata missing in the Sanskrit. There is nothing in the content of the commentary that would absolutely preclude its having been composed by Buddhajñānapāda, nor is there anything that strongly suggests that it was. 310 However, this was not always the case when works were included in the Tibetan canon as Wedemeyer 2009 has shown. 311 The wrathful deities of the protection cakra surrounding the maṇḍala in Śākyamitra’s Mukhāgama, which he says is based on oral instructions from Buddhajñānapāda, come closer to this aesthetic, garlanded by bones and snakes, but still there is no mention of cremation ground ash or skulls in that imagery. 312 See Krug 2018, 266. 313 Caturaṅga-sādhana-ṭīkā, D 135b.4; Sāramañjarī, D 6b.6. 59 verses at the end of the Tibetan translation, survives in Sanskrit, as No. 243 in the Sādhanamālā.314 The final verses that are absent in the Sanskrit are not part of the samādhi of the deity but rather briefly describe the Heruka vratas that the practitioner is enjoined to undertake. The Bhaṭṭārakāryajambhalajalendra-sādhana (Tōh. 1861), Guhyajambhala-sādhana (Tōh. 1862), and Vistarajambhala-sādhana (Tōh. 1863), are three short Jambhala sādhanas, the first two of which are attributed to Buddhajñānapāda, and all three of which share a common final colophon with the injunction that the three sādhanas may not be given to disciples who have not received initiation.315 While the Vistarajambhala-sādhana has no authorial attribution, nor a translator’s colophon, there are a number of factors that suggest it to be Buddhajñanapāda’s composition: Vaidyapāda mentions in the Sukusuma that Buddhajñānapāda composed three Jambhala sādhanas, the two Jambhala sādhanas immediately preceding the Visatarajambhalasādhana in the Tengyur are attributed to Buddhajñānapāda, and the three sādhanas, despite the first two having been translated by different teams of translators, all share a common colophon, as mentioned above. Moreover, the Vistarajambhala-sādhana shares some features with each of the two preceding sādhanas, and the translator’s colophon of the Bhaṭṭārakāryajambhalajalendra-sādhana suggests following the Guhyajambhala-sādhana and the Vistarajambhala-sādhana on some details regarding the visualization of the forms of the retinue deities. As noted above, Bhaṭṭārakāryajambhalajalendra-sādhana and the Vistarajambhala-sādhana both advocate the practice of Vajra Hūṃkāra, to whom homage is paid in the initial verses of Buddhajñānapāda’s *Gativyūha. Jambhala and Vasudharā, god and goddess of wealth, respectively, played important roles in Buddhajñānapāda’s life story: as noted above Buddhajñānapāda reports that Jambhala himself provided a daily stipend of sorts to Buddhajñānapāda and his disciples when he was living at the Parvata cave behind Vajrāsana, and Vaidyapāda explains that the daily provisions for Pālitapāda’s disciples were provided by the goddess Vasudharā. Vaidyapāda also notes that it was during his discipleship under the yoginī Guṇeru in Oḍḍīyāna that Buddhajñānapāda achieved the accomplishment of Jambhala. The Vistarajambhala-sādhana also survives in Sanskrit, as Sādhanamālā No. 285,316 though there are a number of places where the Tibetan translation includes passages not present in the Sanskrit, a few instances of the opposite, and one place where the two include divergent versions of the same passage.317 Works of Questionable Attribution, Unlikely to be Buddhajñānapāda’s Among the sixteen extant works attributed to Buddhajñānapāda under the variety of his names in the Tibetan canon, there are four that remain of questionable attribution, and one that we can definitively rule out as being his composition. Among the works whose attribution to Buddhajñānapāda it is difficult to be certain about is the Mahāpratisarārakṣā (Tōh. 3124). This is a short protective ritual centered around the goddess Mahāpratisarā, one of the figures in the well-known pan-Asian Pañcarakṣā tradition. While there is nothing in the ritual that would preclude its composition by Buddhajñānapāda, there is no suggestion that he composed such a work in any of the sources describing his life and his writings. What is more, the composition, which is written primarily in prose combined with short verses that the ritual officiant 314 Bajracharya 1928, No. 243. 315 Vistarajambhala-sādhana, D 66b.2. 316 Bajracharya 1928, No. 285. 317 The Tibetan translation of the text is also generally problematic, including instances where, for example, the Sanskrit phonetics for a word are included right in the middle of the two syllables of the word that is its Tibetan translation! For example, the Tibetan at one point reads sa bon bī ja pū ra gang ba, when in fact “sa bon gang ba” is itself the Tibetan translation of the Sanskrit word bījapūra! 60 presumably is to recite, along with a number of mantras, does end with dedicatory verses— precisely the kind of verses into which Buddhajñānapāda was wont to inscribe his name—but the author of the Mahāpratisarārakṣā has not done so. Of course Buddhajñānapāda did not always write such dedicatory verses, but in his extant works that do contain such dedicatory verses, he did always include his name.318 The Śrīraktayamārisādhana (Tōh. 2084), *Trikāyavākcittādhiṣṭhanoddeśa (Tōh. 2085), and Trisattvasamādhisamāpatti (Tōh. 2086) are three short Yamāri-related works ascribed in their colophons to Buddhajñānapāda.319 The colophon of the first of these is followed by a list of of its lineage masters which begins: Yamāri, Vajrayoginī, Buddhajñānapāda, Śrīdhara,320 Līlāvajra, and continues through a number of further masters up to “myself,” presumably one of the translators.321 In the authorial colophon of two of these short works Buddhajñānapāda is associated with Vajrayoginī or Vajravārāhī, and said to have had her vision or to have directly received her blessings, which is never reported in any of the life accounts of the 8th/9th century master. Two of the works also have short dedicatory verses at the end, but neither has Buddhajñanapāda’s characteristic signature within them. The vocabulary and style of the works (which is common among the three) is not reminiscent of that in Buddhajñānapāda’s other writings, and particularly the very casual use of the terms “generation stage” and “perfection stage”322 in the Trisattvasamādhisamāpatti gives the sense of terms that were in common usage, whereas we know that Buddhajñānapāda was one of the early masters to employ these terms and we find them introduced and used very deliberately, rather than casually, in his other writings. As mentioned above, though, Buddhajñānapāda is very clearly associated with Yamāri practices in the later Tibetan tradition, especially within the Ngor tradition of the Sakyapas, who also uphold his Guhyasamāja lineage, so it is not unexpected to find Yamāri-related texts attributed to Buddhajñānapāda. However, the Buddhajñānapāda who is the author of these particular Yamāri texts is, in my estimation and for the reasons noted above, unlikely to be the same as the Buddhajñānapāda who is the author of the Dvitīyakrama, the Caturaṅga/Samantabhadrasādhana and the other works mentioned above. 318 The single exception to this rule is the Dvitīyakrama, discussed above. In possible support of the attribution of the Mahāpratisarārakṣā to Buddhajñānapāda, the paṇḍita Sumatikīrti, who was involved in the translation of the Mahāpratisarārakṣā, also translated Samantabhadra’s Jñānapāda School Mañjuśrī-sādhana (Tōh. 1880), and we know that canonical translators often worked on multiple works by a particular author or that were connected to a particular tradition. The works that Sumatikīrti translated, however, appear to be a rather eclectic collection, so his involvement in these two works may not be related. 319 His name in the colophon reads bu ddha jñā na pā da in Tōh. 2084 and sangs rgyas ye shes zhabs in Tōh. 2085 and Tōh. 2086. 320 As noted above, in Amye Zhab’s Gshin rje chos ‘byung, Śrīdhara is the master listed after Buddhajñānapāda in one of the Yamāri lineages given there, but Amye Zhab’s lineage sequence begins with Mañjuvajra, rather than Yamāri and Vajrayoginī, and passes from Śrīdhara to Nāropa, not to Līlāvajra (Gshin rje chos ‘byung, 16a.3-4). But the lineage Amye Zhab writes about is a Kṛṣṇayamāri lineage, rather than a Raktayāmari lineage. In another lineage description in the same work Amye Zhab explains that Mañjuśrī passed the teachings on Vajrabhairava to Buddhajñānapāda, who gave them to Dīpaṃkarabhadra, who passed them to Śrīdhara, who then composed many Yamāri-related treatises (Gshin rje chos ‘byung, 48a.4-5). Śrīdhara does mention Buddhajñānapāda in at least three of his Yamāri-related works, but in all of these cases he cites Buddhajñānapāda specifically in his association with the Guhyasamāja tradition, rather than the Yamāri tradition. In one case the reference is to an iconograpical issue, an in two others to a doctrinal point (See Sahajāloka, D 86b.6-7, Kṛṣṇayamārisādhana, D 6b.5, and Śrīraktayamārisādhana, D 93b.6, respectively). 321 According to the translator’s colophon of all three of these short works—the initial translators were the paṇḍita from Madhyadeśa Ānandabhadra and the Tibetan lotsāwa Sonam Gyaltsen (bsod rnams rgyal mtshan), but the texts were later revised or retranslated at Nālandā by the Siddha Kaṛnaśrī and Neten Nyima Gyaltsen (gnas brten nyi ma rgyal mtshan). The three colophons are not entirely identical, but all convey the same information. 322 For example, the text is described in the colophon as skye rdzog zung du ‘jug pa’i man ngag. 61 Works not by Buddhajñānapāda While the Mukhāgama (Tōh. 1854), as mentioned above, is attributed to Buddhajñānapāda in both traditional histories and modern scholarship due probably to the presence of his name and the absence of Śākyamitra’s in the work’s colophon, the dedicatory verses of text itself clearly state that it was written by Śākyamitra, and I therefore do not consider it to be Buddhajñānapāda’s work. Moreover, in the initial verses of the text Śākyamitra goes so far as to distinguish his own style from Buddhajñānapāda’s. The Mukhāgama is, however, obviously based on Buddhajñānapāda’s Samantabhadra-sādhana and may very well be based on his oral instructions, as its title—and perhaps even its contents—indicate.323 Clearly Śākyamitra was referencing the Dvitīyakrama-tattvabhāvanā-mukhāgama when he selected the title of the Mukhāgama, and obviously wanted his own composition to be associated with Buddhajñānapāda and his tradition, despite stating in the concluding verses that he (Śākyamitra) himself composed it. Concluding Summary: Extant Works Reasonably Attributable to Buddhajñānapāda Out of the sixteen surviving works in the Tibetan canon attributed to Buddhajñānapāda under the variety of his names, I believe we can conclude with a reasonable degree of certainty that eleven are the writings of a single author. While five of these works also survive fully or in part in Sanskrit, as described in their summaries above, I know of no work attributed to Buddhajñānapāda that is extant in Sanskrit but not in Tibetan translation.324 The extant works we can confidently attribute to our 8th/9th-century author Buddhajñānapāda are: 1. Dvitīyakramatattvabhāvanā-mukhāgama (Tōh. 1853) 2a. Samantabhadra-sādhana (Tōh. 1855) 2b. Caturaṅgasādhanopāyikā-samantabhadrī (Tōh. 1856) 3. Śrīheruka-sādhana (Sādhanamālā No. 243, Tōh. 1857)325 4. Muktitilaka (Tōh. 1859) 5. Ātmasādhanāvatāra (Tōh. 1860) 6. Bhaṭṭārakāryajambhalajalendra-sādhana (Tōh. 1861) 7. Guhyajambhala-sādhana (Tōh. 186) 8. Vistarajambhala-sādhana (Sādhanamālā No. 285, Tōh. 1863) 9. *Gativyūha (Tōh. 1864) 10. Sañcayagāthā-pañjikā (Tōh. 3798) 11. Mahāyānalakṣaṇa-samuccaya (Tōh. 3905) A Life Remarkably Lived Buddhajñānapāda was a remarkable figure about whom we are able to glean a surprising amount of information for an individual who lived in early medieval India. In particular, his autobiographical narration in the Dvitīyakrama gives us a sense of the life of a well-educated yogin whose determination to attain awakening through tantric methods drove him to travel vast distances and serve many teachers in what appear to have been thriving tantric communities 323 See note 260. 324 If there were such an extant work, however, a very likely place for it to be found would be in the manuscript described in Kawasaki 2004. 325 Among the works in this list, this is the one about which I feel the most hesitation in making the attribution to Buddhajñānapāda, for the reasons outlined above. 62 across the breadth of the subcontinent. In many ways Buddhajñānapāda’s life was extraordinary—his transformative vision of Mañjuśrī, composition of popular and lasting tantric works, and possible connections with Pāla royalty and position of note in one of the large monasteries of his time. Yet his own account of his life also somehow gives a flavor of the ordinary lived experience of a Buddhist tantric practitioner in his time—studying with many different gurus, acknowledging his own lack of understanding early in his career, and persevering in his study and practice until he encountered what he felt to be genuine truth, which he then felt compelled to share with others. In the next chapter I will examine in more detail the work that contains this remarkable account, Buddhajñānapāda’s Dvitīyakrama, addressing the unusual structure of the text and how Buddhajñānapāda used that as a way to convey its equally unique contents. The great variety of topics addressed in the Dvitīyakrama will give us a helpful overview of Buddhajñānapāda’s thought before moving on, in the subsequent chapters, to an indepth assessment of some of its aspects. 63 Chapter Two Narrating Revelation: The Dvitīyakrama’s Unique Framing of Doctrine and Ritual On the eighth day of the seventh month, during the constellation Puṣya at the time when Mṛgaśīrṣa and Hasta are fading, in the early morning, right at dawn, towards the emanated maṇḍala-cakra of Mañjuśrī, I made a fervent supplication to understand the meaning. -Buddhajñānapāda, Dvitīyakrama The Dvitīyakrama is certainly the most unique of Buddhajñānapāda’s compositions, but it also stands out among Indian Buddhist writings as a whole. While descriptions of visionary encounters are not at all uncommon in Indian Buddhist literature, especially from the rise of the Mahāyāna onward, autobiographical descriptions of visionary encounters—and indeed any type of autobiographical writings at all—are. Buddhajñānapāda’s claims to visionary inspiration positioned him among important earlier Buddhist figures like Nāgārjuna and Asaṅga whose visionary experiences are regarded by the tradition as the source of new—or, more precisely, newly revealed—Buddhist doctrinal and practice systems. The accounts of the visionary experiences of these earlier authors, however, were recorded and passed down by subsequent members of their traditions, whereas Buddhajñānapāda documented his own, along with other details of his life. Such autobiographical writings are extremely rare; apart from the Dvitīyakrama and Buddhajñānapāda’s disciple Śākyamitra’s short autobiographical account in his Kośalālaṃkāra, which the Dvitīyakrama seems to have inspired, I know of no other autobiographical narratives in early Indian Buddhist literature.1 In the Dvitīyakrama Buddhajñānapāda crafts a narrative frame for the work’s sometimes innovative contents that puts him in a role that is both central and peripheral. That is, on the one hand, the Dvitīyakrama is Buddhajñānapāda’s account of his own life story, but on the other hand the primary content of the work is not his at all: a full ninety percent of the Dvitīyakrama is simply a record of Mañjuśrī’s direct (sometimes even first-person) speech, so it is he who is in some sense the author of—and perhaps more importantly the authority behind—that content. Buddhajñānapāda’s opening reference to the contents of the Dvitīyakrama as “the words of the guru Mañjuśrī,” Vaidyapāda explains, is meant to refute the idea that Buddhajñānapāda himself had composed the instructions.2 Throughout the history of Buddhist literature Buddhist authors—following a general trend in Indian traditions at large—have upheld this trope of “not having made up anything new,” explaining innovation as nothing more than the correct interpretation of what was already presented in the original scriptural sources, or in the compositions of lauded philosophers like Nāgārjuna (who themselves claimed simply to be correctly interpreting scripture). The production of new scriptures has thus been an important way in which Buddhist traditions have grown and developed over time. Paul Harrison’s work on the development of the early Mahāyāna scriptural corpus describes the pratyutpannasamādhi, advocated in the Pratyutpanna-sūtra, in which the practitioner engages in prolonged visualization of a buddha and his buddhafield in order to gain a visionary encounter in which he receives teachings from that buddha, that the practitioner 1 Janet Gyatso has written that “First-person discourse about one’s life is virtually nonexistent in Indian Buddhist literature; we can only mention the Therī- and Tharagāthā, which contain a few poems that may, be autobiographical, and occasional statements attributed to the Buddha.” (Gyatso 1998, 115). 2 rang bzo dgag pa (Sukusuma, D 89a.3). 64 subsequently remembers and shares with others.3 The Mañjuśrīnāmasaṃgīti, which Buddhajñānapāda’s guru Vilāsavajra wrote a commentary on, and which seems to have had some influence on Buddhajñānapāda and his work, likewise prescribes a similar, but tantricised buddhānusmṛti type of practice focused on Mañjuśrī: “He who...recites [from memory] this crest jewel called the Nāmasaṃgīti three times each day, or who recites it from a book [and] who, taking the form of the Fortunate One, Mañjuśrījñānasattva, reflects and meditates on that form...will before very long see him [i.e. Mañjuśrī] in his Form Body (rūpakāya).”4 The practice of revelation through visionary encounter was thus already well established in Buddhajñānapāda’s time, and, in fact, at the close of his teaching in the Dvitīyakrama, Mañjūśrī explicitly commands Buddhajñānapāda to compile his instructions and pass them on. Buddhajñānapāda was unique, however, in so directly narrating this personal encounter in his own writings. The title of the Dvitīyakramatattvabhāvanā-mukhāgama, Oral Instructions on the Meditation on the Reality of the Second Stage, already implies that its instructions were received from a source other than the author/compiler of the text—here, of course, Mañjuśrī. The term mukhāgama functions in the title as a sort of genre marker, in the place where one often sees such genre markers as sādhana, vidhi, ṭīkā, or vṛtti. It is not a popular genre in Buddhist literature; one finds just a few instances in the titles of Buddhist texts, where it seems to indicate instructions that have been received from an authority other than the author.5 As we saw above, Śākyamitra uses the term as the title of his own Mukhāgama, which purports to record Buddhajñānapāda’s oral instructions on the Caturaṅga/Samantabhadra-sādhana.6 In any case, the mukhāgama genre is an uncommon one, and Buddhajñānapāda seems to have been particularly unusual in using it to label instructions received directly from a buddha/bodhisattva teacher such as Mañjuśrī. The precise identity or nature of Mañjuśrī as the source of revelation in the Dvitīyakrama is not as straightforward as it might seem. In Chapter One I discussed the claims of some later Tibetan authors that the monk who emanated the maṇḍala of Mañjuśrī for Buddhajñānapāda in the Dvitīyakrama was the master Mañjuśrīmitra, a claim not found in any of the Indic texts (nor in many of the Tibetan works) relating to Buddhajñānapāda, nor in the Dvitīyakrama. Buddhajñānapāda and Vaidyapāda seem to understand this “emanated monk” simply as an emanation of the bodhisattva/deity who was the true source of Buddhajñānapāda’s revelation. But this still leaves the question of how to understand the identity of the visionary form of Mañjuśrī who taught Buddhajñānapāda directly. Buddhajñānapāda himself refers to him in the Dvitīyakrama as the “guru Mañjuśrī”7 and the “great bodhisattva Mañjuśrī.”8 Vaidyapāda, however, takes pains to clarify that Mañjuśrī is called a bodhisattva “because he is integrated 3 Harrison 1978, 54-5; 2003, 120. 4 Tribe 1997, 124. 5 These include the Nandyāvartatrayamukhāgama (Tōh. 2415) of Kaṅkālā and Mekalā, which states at the outset that it is based on “the guru’s oral instructions;” the Grub pa’i dbang phyug paṇḍita chen po shrī ba na ratna’i zhal lung rin po che’i snying po’i phreng ba (P 5096), and Zhal lung rin po che’i phreng ba (P 5099), which are two nearly identical short compilations of citations from Indic texts that the 15th-century Indian yogin Vaṇaratna often cited in his teachings (thanks to Ryan Damron for sharing with me his understanding of the content and function of the latter two works); and the Mgon po dmar po’i tshe bsgrub kyi zhal gdams (P 4927) of Śavaripa is a short longlife practice associated with red Mahākāla. The authorial attribution to Śavaripa is followed by a list of the instruction’s lineage, which begins with Vajradhara, and has its “author” Śavaripa next in the lineage. 6 In that text Śākyamitra does claim some agency, suggesting that he will deliver these instructions with more clarity and simplicity than Buddhajñānapāda did in the Caturaṅga/Samantabhadra. See Chapter One, note 258. 7 Dvitīyakrama, verse 2. 8 Dvitīyakrama, verse 19. 65 with awakening (bodhi), not because awakening is his goal.”9 This suggests that Vaidyapāda understands, and wants his readers to understand, Mañjuśrī as elevated above the level of an “ordinary” bodhisattva abiding on the bhūmis, and instead as representing a form of full awakening.10 Similarly, in his Muktitilaka-vyakhyāna Vaidyapāda identifies Mañjuśrī as the “foundation” of the unique qualities of the Bhagavan, both in terms of their cause and their effect.11 Such a portrayal is not at all outside of the range of Buddhist conceptions of the figure of Mañjuśrī in the 8th and 9th centuries. Early Mahāyāna sūtras associate him with wisdom, and often portray him as a tenth-bhūmi bodhisattva, but sometimes even as a fully awakened buddha or a teacher of buddhas.12 In the 8th-century Mañjuśrīnāmasaṃgīti, however, Mañjuśrī is named as the Ādibuddha.13 The Nāmamantrārthāvalokinī, a commentary on the Nāmasaṃgīti composed by Buddhajñānapāda’s guru, Vilāsavajra, likewise equates Mañjuśrī, here as Mañjuśrījñānasattva, with nondual wisdom itself, and thus the source of the buddhas’ awakening.14 Indeed, Vaidyapāda’s statement mentioned above distinguishing Mañjuśrī from a mere bodhisattva on the bhūmis echoes a similar statement made by Vilāsavajra about Mañjuśrījñānasattva in his commentary to the Nāmasaṃgīti: “The gnosis-being Mañjuśrī is not the bodhisattva who is the master of the ten stages (bhūmi). Rather, he is non-dual gnosis (advayajñāna), the perfection of wisdom (prajñāpāramitā) itself.”15 Such a perspective seems to underlie Buddhajñānapāda’s understanding of Mañjuśrī in the Dvitīyakrama. Though referring to him as a “bodhisattva,” when Buddhajñānapāda supplicates the visionary Mañjuśrī for instructions, he addresses him in quite elevated terms as “the father and the mother of all beings,” the “emptier of the three realms, greatest of the great,” as “beginningless, unvoiced, lacking the upper part of the bindu, the revered, the letterless, producer of nectar, the empty bliss of great joy,” and the “great protector.”16 These exalted epithets suggest that Buddhajñānapāda likewise identifies Mañjuśrī here as a fully awakened buddha, the ultimate source of the ultimate truth. In the homage at the beginning of his Muktitilaka, Buddhajñānapāda describes Mañjuśrī as the “emanation of boundless [[[buddha]]] families.”17 It is no wonder, then, that Mañjuśrī, in the form of Mañjuvajra, is the central figure in Buddhajñānapāda’s generation stage Guhyasamāja sādhana, the Caturaṅga/Samantabhadra, despite Mañjuśrī/Mañjuvajra’s peripheral role in the Guhyasamāja-tantra itself.18 In the Dvitīyakrama Buddhajñānapāda makes supplications “towards the emanated maṇḍala-cakra of Mañjuśrī,” thus implying he saw Mañjuśrī as the central figure in a maṇḍala of deities. Presumably this was the nineteen-deity maṇḍala of Mañjuvajra that Buddhajñānapāda describes 9 de nyid byang chub dang ‘dres pa’i phyir byang chub sems dpa’ ste/ byang chub la dmigs pa ni ma yin no/ (Sukusuma, D 93a.2). 10 In his Muktitilaka-vyākhyāna, when citing a passage from the Dvitīyakrama Vaidyapāda refers to the speaker of the instructions in the Dvitīyakrama simply as “the bodhisattva Mañjuśrī” (Muktitilaka-vyākhyāna, 46b.5). It is interesting, though, that when citing the passage he specifies Mañjuśrī rather than Buddhajñānapāda as the source of the quotation. 11 Muktitilaka-vyākhyāna, D 46b.6-7. 12 Tribe 2016, 15n27. Here Tribe, referencing the work of Lamotte and others, gives a useful short overview of Mañjūśrī’s role in the Mahāyāna sūtras. 13 ibid.; Tribe 1997, 109. 14 ibid. 15 Tribe 2016, 8. 16 In Dvitīyakrama, verses 12 and 13. 17 Muktitilaka, D 47a.1-2. This identification of Mañjuśrī with all five of the buddha families is found in a converse (but presumably complimentary) form in Vilāsavajra’s Nāmamantrāvalokinī. 18 Mañjuśrī receives two brief mentions in the Guhyasamāja-tantra, in Chapters 13 and 15, and Mañjuvajra likewise has two brief mentions in Chapters 12 and 16.

66 in his Caturaṅga/Samantabhadra-sādhana, and Mañjuśrī in the vision was Mañjuvajra, though neither Buddhajñānapāda nor Vaidyapāda makes that explicit.19 That it was Mañjuśrī, and not some other divinity, who appeared to Buddhajñānapāda and taught him reflects Buddhajñānapāda’s unique personal connection with that deity. Several factors may have helped forge that connection, including Buddhajñānapāda’s guru Vilāsavajra’s important role in the propagation of the Mañjuśrīnāmasaṃgīti, and the sādhana of Mañjuśrījñānasattva that is set forth in Vilāsavajra’s commentary on the Nāmasaṃgīti, a work that we know Buddhajñānapāda was familiar with because he reproduces a lengthy section from it in his Ātmasādhanāvatāra. As discussed in Chapter One, there are likewise a number of indicators that Buddhajñānapāda’s guru Pālitapāda may also have had a connection with Mañjuśrī, and perhaps even with the Nāmasaṃgīti as well. The Dvitīyakrama’s narrative structure quite literally frames the content of Mañjuśrī’s teachings therein. It includes both the autobiographical introduction to Buddhajñānapāda’s life leading up to his visionary encounter with Mañjuśrī, as well as an autobiographical conclusion subsequent to the vision. While such a framing narrative is rare among authored Indian Buddhist texts, it is structurally reminiscent of another important genre of Buddhist literature: the sūtras and the tantras.20 Buddhist scriptures, starting with the earliest sūtras and continuing up to the tantras, almost always have a narrative framework. The introductory nidāna gives the setting in which the teachings occurred, identifying the teacher, students, and location where the teaching was given, and often includes a request by the retinue for the teacher to teach. This is followed by the main content of the scripture—the acceptance of the request by the teacher and the teaching itself, which is usually then brought to a conclusion with a closing frame narrative. Sūtras and tantras are less authored texts than records of events written down by compilers who usually do not even receive mention within the scriptures themselves. Often the compiler is identified only in the exegetical tradition—typically Ānanda in the case of the sūtras, and Vajrapāṇi in the case of the tantras. Despite the anonymity of many compilers, however, Buddhist scriptures do traditionally begin with a first-person statement, “Thus have I heard...” (evaṃ mayā śrutaṃ...), before switching into the third-person narrative voice.21 The Dvitīyakrama does not fit this model exactly. Its frame narrative begins not with a focus on the main teacher and his surrounding setting, here Mañjuśrī (who appears only at the end of the opening narrative), but instead with the story of the student who is the recipient of his teachings, Buddhajñānapāda. The narrative then leads up to the encounter with Mañjuśrī in which the teachings were conveyed. The Dvitīyakrama makes use of the first-person voice at a number of points throughout the text, though the speaker does sometimes change; first it is Buddhajñanapāda, then Mañjuśrī,22 and finally Buddhajñānapāda again. Buddhajñānapāda’s first-person statements in both the first and later parts of the Dvitīyakrama are always autobiographical, apart from his supplication and request to Mañjuśrī to teach, which Mañjuśrī accepts. While most of Mañjuśrī’s teachings are delivered in a loose descriptive or imperative voice (“the practitioner does...” or “he should do ...”), at two points towards the end of the instructions he speaks in the first person, twice repeating the unusual declaration that “I abide in the bodies of practitioners” and thus receive offerings.23 Mañjuśrī also addresses 19 Some of the later Tibetan accounts do specify this. 20 I’m grateful to James Gentry for a conversation that sparked my investigation into the parallels between the Dvitīyakrama and Buddhist scriptural works. 21 Some of the later Yoginī tantras lack this traditional beginning. 22 Mañjuśrī’s first-person speech, and his direct address of Buddhajñānapāda in the second person, are striking attributes of the Dvitīyakrama. 23 Dvitīyakrama, verses 323-24 and 361. 67 Buddhajñānapāda in the second person as “you” both at the beginning and the end of his instructions. The concluding narrative of the Dvitīyakrama likewise returns its focus to Buddhajñānapāda’s continued autobiographical narrative, describing his life subsequent to the visionary encounter, rather than concerning itself further with Mañjuśrī. There are thus a number of differences between the narrative framework of the Dvitīyakrama and that found in the Buddhist scriptures. But the very presence of such a narrative frame enclosing ritual and doctrinal content in a śāstric work is so unusual that it is difficult not to see the structural parallels with a sūtra or a tantra. Buddhajñānapāda may or may not have been consciously evoking the model of Buddhist scripture in the Dvitīyakrama, but there is some evidence that suggests it may indeed have been intentional. The opening line of Mañjuśrī’s direct speech in the Dvitīyakrama is a series of Sanskrit syllables, which Vaidyapāda interprets as a summary of a cosmogonic narrative that is elaborated later in the Dvitīyakrama.24 The opening line of Śākyamuni’s instruction in the Mañjuśrīnāmasaṃgīti, which although it was spoken by Śākyamuni is nonetheless understood by commentators to be a verse related to Mañjuśrī, likewise begins with a set of Sanskrit syllables— the twelve vowels of the Sanskrit alphabet.25 The syllables in that verse are described by the Mañjuśrīnāmasaṃgīti itself, and clarified in Vilāsavajra’s commentary, as being the “coming forth of the non-dual.”26 The syllables in the Dvitīyakrama are also understood to reflect a coming forth or an emergence: the process by which duality emerges from the nonduality that is the inherent nature of all things. Though they admittedly have a different function in the Dvitīyakrama than they do in the Nāmasaṃgīti, presence of a set of such syllables at the outset of Mañjuśrī’s speech in the Dvitīyakrama may thus be meant to loosely evoke the Mañjuśrīnāmasaṃgīti, an important scriptural work that, like the Dvitīyakrama, focuses on Mañjuśrī, and with which it shares a number of doctrinal resonances.27 Whether or not the evocation of scripture was intentional, his decision to use a first-person narrative framework for reporting the content of Mañjuśrī’s teachings, and the fact that this allowed the content of the Dvitīyakrama to be clearly attributed to Mañjuśrī (rather than by, for example, adding a note in a colophon about its having been revealed to Buddhajñānapāda by Mañjuśrī), represents a skillful legitimizing tool for the work, and by extension the entirety of Buddhajñānapāda’s oeuvre, particularly his other Guhyasamāja-related works. While it certainly contains references to sexual yogas and brief descriptions of practices like the sūkṣma-yoga, which had received mention already in the Sarvatathagatatattvasaṃgraha, the Guhyasamāja root tantra itself does not contain detailed instructions on perfection stage practices, nor does it even mention there being two separate stages of tantric practice.28 The 24 The syllables are a vi yaṃ raṃ vaṃ laṃ hūṃ. I discuss the cosmogonic narrative found in Buddhajñānapāda’s works in Chapter Three. 25 The syllables occur in verse 26 of the Mañjuśrīnāmasaṃgīti. See Davidson 1982 for a Sanskrit edition and English translation of the Mañjuśrīnāmasaṃgīti. See Tribe 1997, 118 for a discussion of the speaker of verse 26 of the Mañjuśrīnāmasaṃgīti. Tribe 2016 gives a translation of Chapters 1-5 of Vilāsavajra’s commentary, including a translation of the root verses commented upon. 26 See Tribe 2016, 135. 27 The Mañjuśrīnāmasaṃgīti, though it does not carry the word tantra in its title, follows a traditional scriptural model and has been treated by commentators from the very earliest times (including Vilāsavajra), as well as the redactors of the Tibetan canon, as a tantric scripture. Indeed the text presents itself as a part of the Māyājāla-tantra. Thanks to Harunaga Isaacson for bringing the latter point to my attention. 28 The two stages are famously mentioned, along with more detailed perfection stage practices, in the Samājottara, a commentarial tantra on the Guhyasamāja that was first circulated separately and then appended to the Guhyasamāja as its eighteenth chapter. I discuss this text and Buddhajñānapāda’s relationship to it in more detail below in Chapter Eight, but it appears that Buddhajñānapāda did not know the Samājottara and its circulation seems to post-date his writings. 68 Dvitīyakrama—along with Buddhajñānapāda’s Muktitilaka, which describes many of the same practices, though in much less detail—thus serves as the primary source, and authority, for the perfection stage practices of the Jñānapāda School of Guhyasamāja practice. The Ārya School, which developed later, derives the authority for its perfection stage practices, outlined in texts such as the Pañcakrama and the Caryāmelāpakapradīpa, from scriptural sources, in particular the later explanatory tantras (vyākhyānatantra) of the Guhyasamāja, such as the Vajramālā. Since the Jñānapāda School works do not look to such scriptural sources, it was important that the practices described in the Dvitīyakrama have some kind of legitimate source and authority. The Dvitīyakrama’s use of the mukhāgama genre, with the bodhisattva/buddha Mañjuśrī as the narrator of most of the work’s content, places the Dvitīyakrama at an unusual juncture between scripture and treatise, and provides precisely the kind of authority necessary to support the presentation of such a set of practices.29 Viewing the Dvitīyakrama as a work that sits at the juncture between scripture and treatise is one helpful way to understand the text and its function in Buddhajñānapāda’s oeuvre, and indeed its function in his tradition on the whole. Another is to look more closely at the Dvitīyakrama’s use of autobiography. Janet Gyatso’s writings on Tibetan autobiography, though they examine autobiography in a different Buddhist culture—one in which autobiography, unlike in India, is quite common—and in a later period than Buddhajñānapāda’s writings, nonetheless provide some insights that are useful in helping us to further understand Buddhajñānapāda’s autobiographical frame narrative in the Dvitīyakrama.30 Among the autobiographical writings that Gyatso considers, those that are most similar to what we find in the Dvitīyakrama are the autobiographical narrations of the events of a treasure (gter ma) revelation, the (sometimes) visionary revelation of (usually, but not always) scriptures that were said to have been hidden by earlier masters, and which are later discovered by destined treasure revealers for the benefit of modern disciples.31 Treasure texts in Tibet function as important sources of newly-revealed doctrines and practices, and it is therefore not surprising that we should find some similarities in the narratives that surround these texts and the narrative surrounding Mañjuśrī’s instructions as recorded in the Dvitīyakrama. Tibetan treasure revealers were also, perhaps not incidentally, among the most prolific autobiographers in Tibet.32 With respect to the autobiographical narratives surrounding treasure revelation, Gyatso writes, The point of narrating the events of a Treasure revelation is to demonstrate its authenticity. Since these scriptures were accused of being apocryphal, the “story” of a Treasure, which argues to the contrary, had long been a standard section of the published Treasure corpora. The discoverers frequently position such a narrative as a prolegomenon to the rest of the Treasure…Often explicitly labeled “story that engenders 29 Szántó (2012a, 456), has also briefly noted the Dvitīyakrama’s unusual position between scripture and treatise, and suggested that this was an ideal choice for introducing innovative material, such as the technique of utkrānti. 30 Gyatso 1998. 31 The topic of revealed treasures is a very large one that is very much beyond the scope of what I can address here. Not all treasure revelations are visionary, and not all treasures are scriptural—some are revealed in the presence of a crowd of viewers (“public treasures,” khrom gter), and some are revealed as physical objects (“earth treasures,” sa gter), a category that can include objects like statues as well as physical manuscripts of texts. The process of the visionary revelation of the contents of the Dvitīyakrama is of course different in many ways from a treasure revelation (it would perhaps be most similar to what is, in the treasure traditions, called a “pure vision” (dag snang), which is sometimes distinguished from the revelation of a treasure proper, though a pure vision can also be considered a type of treasure revelation); nonetheless, I do think we can benefit from considering Gyatso’s comments on the function of autobiographical writing in the context of the treasure tradition. I happily direct the interested reader to Janet Gyatso’s (1986, 1992, 1993, 1994, 1996) and Andreas Doctor’s (2005) excellent work on Tibetan treasure traditions. 32 Gyatso 1998, 104. 69 confidence” (nges-shes-skyes-pa’i lo-rgyus), accounts of this kind consist in a narrative either of the Treasure’s ultimate origin, from its original preaching by a buddha down to its transmission by Padmasambhava, or of how the Treasure was later revealed by the Tibetan discoverer.33 The autobiographical narrative in the Dvitīyakrama is precisely a prolegomenon—as well as a conclusion, we might add—to the instruction conveyed by Mañjuśrī during the visionary encounter that makes up the majority of the Dvitīyakrama’s content. We may recall here, as well, Vaidyapāda’s comments on the autobiographical narrative in the Dvitīyakrama. He introduces the autobiographical section of the text with the statement, “Then, in order to inspire faith in beings, the revered master speaks about the story of his own encounter with suchness with the verse beginning with, In a town called....”34 Vaidyapāda is explicit here about the function of the autobiographical account in bringing about faith, or confidence, in the reader. I think that we can understand the faith or confidence that Vaidyapāda suggests is brought about through Buddhajñānapāda’s autobiographical narrative in multiple ways, two of which are also mentioned by Gyatso in her writings on treasure revealers’ autobiographical narratives. Gyatso continues, The discoverer’s personal account engenders confidence in several ways. Most obviously, the story of the Treasure’s revelation is meant to demonstrate that the revelation actually happened at a particular time and place. The more detailed accounts add a sense of awesomeness to this historicity, giving the revelation’s precise circumstances and showing that it occurred in a marvelous way…35 We again see a very strong resonance here with Buddhajñānapāda’s account in the Dvitīyakrama: he gives not only the location of the vision, in “a forest called Kuvaca” behind Vajrāsana, but also provides an extremely precise date and time for the moment of his visionary experience—“On the eighth day of the seventh month, during [the constellation] Puṣya/ At the time when Mṛgaśīrṣa and Hasta are fading,36 in the early morning, right at dawn”37—giving the reader a very specific sense of the historicity, as it were, of his experience. And Buddhajñānapāda’s description of Mañjuśrī’s response to Buddhajñānapāda’s supplication for teachings certainly conveys a sense of amazement and awe: Then, the great bodhisattva Mañjuśrī Looked upon me with a smiling face and said, “Excellent” three times. With this vajra song, like an echo, he taught to me The playful dance and the suchness of all phenomena. |19|38 Gyatso continues: But even an autobiographical account that exceeds the particular event of the revelation…still has everything to do with the legitimation of the Treasure cycle. This is because it is finally an ad hominem argument regarding the discoverer that is the most important sign of a Treasure’s authenticity.39 33 ibid., 9. 34 da ni ‘gro ba rnams dad par bya ba’i phyir/ rje btsun bdag nyid kyis de kho na nyid mnyes pa’i lo rgyus gsungs pa/ dbus kyi yul chen zhes pa la sogs pa’o// (Sukusuma, D 89a.7; P 107a.3). 35 Gyatso 1998, 9. 36 Puṣya is the eighth lunar mansion in Indian astrology; Mṛgaśīrṣa is the fifth; Hasta is the thirteenth. 37 mgo dang lag gnyis yol dang tshes brgyad rgyal la bab// ston zla ra ba’i tho rangs skya rengs shar dus su// (Dvitīyakrama, verse 11ab). 38 de nas ‘jam dbyangs byang chub sems dpa’ chen po yis// bdag la ‘dzum pa’i bzhin bltas legs zhes lan gsum gsungs// rol pa’i gar dang chos kun de bzhin nyid// sgra brnyan lta bur rdo rje glu yis bdag la bstan// |19| (Dvitīyakrama, verse 19). 39 Gyatso 1998, 9. 70 Buddhajñānapāda’s autobiographical account, though not lengthy, does indeed go beyond the particular events of his visionary experience to include both his early life as a student and his later life as a teacher. Here we may again return to Vaidyapāda’s comments on the autobiographical narrative, cited above, in which he explains that the purpose of the autobiographical account was to inspire faith in disciples with regard to Buddhajñānapāda’s “own experience of suchness.” In showing that Buddhajñānapāda himself encountered suchness through Mañjuśrī’s instructions, the reader is made to feel confident not only in Mañjuśrī’s words—which certainly are a valid source of truth—but also in Buddhajñānapāda as their messenger, since he, himself, directly encountered suchness on the basis of these instructions, making him a reliable source for conveying the instructions on that teaching to others. Buddhajñānapāda’s account of his later life likewise includes further confidence-engendering details, like the fact that at the conclusion of the vision Mañjuśrī “sang and praised” Buddhajñānapāda, and that in his later life Buddhajñānapāda and his students received patronage directly from the wealth-deity Jambhala. At the conclusion of the autobiographical account, Buddhajñānapāda writes, “Thus, in this way everyone, having come to know the detailed accounts [of my life]…”40 and Vaidyapāda comments, Thus, having generated faith in that way (i.e. by means of telling the story of his own encounter with suchness), he teaches about the training in nondual wisdom and its result with the verse beginning, Thus... Having come to know the detailed accounts means the detailed accounts about the great master: the taming of Nālandā, making offerings at Vajrāsana, the [account of] the consecration and the others. Through these accounts the faith of those who have fortune is further increased.41 Buddhajñānapāda and Vaidyapāda both appear here to be making reference to further accounts from Buddhajñānapāda’s life, upon which Buddhajñānapāda himself does not elaborate, and which Vaidyapāda mentions only in brief, as if these would have been well-known lore within the community of disciples whom Vaidyapāda expected to be reading the Dvitīyakrama and his commentary on it. 42 But, yet again, the function of sharing or knowing these life accounts is explicitly mentioned as inspiring faith in the fortunate. As noted above, Gyatso’s comments on the autobiographical narratives of treasure revealers pertain to a different Buddhist culture and to a time period many centuries removed from Buddhajñānapāda’s account in the Dvitīyakrama. Yet we can see that many of the features of Buddhajñānapāda’s autobiographical account, and especially its function as a confidence- or faith-inspiring narrative—even self-consciously so, within the tradition itself—very closely parallel many of the features and functions of the autobiographical narratives of treasure revealers—and particularly the accounts of the revelation of a treasure—that we find several centuries later in Tibet. This is especially interesting given the fact that, since it occurs in an Indian Buddhist text, Buddhajñānapāda’s autobiographical account is extremely unusual, whereas autobiography is a very common genre in Tibetan literature. It seems, then, that this 40 de bas de ltar kun gyis gtam rgyud rgyas par shes byas te// (Dvitīyakrama, verse 378a). 41 da ni de lta bus dad pas byas te/ gnyis su med pa’i ye shes bsgom pa ‘bras bu dang bcas pa gsungs pa/ de bas zhes pa la sogs pa’o// gtam rgyud rgyas par shes byas nas/ zhes pa ni bla ma chen po’i gtam rgyud rgyas pa na landa ([landa] P, lendra D) ‘dul ba dang/ rdo rje gdan gyi mchod pa byas pa dang/ rab tu gnas pa byas pa la sogs pa’i lo rgyus kyis skal ba dang ldan pa cher dad par byas nas/ (Sukusuma, D 135b.3-4; P 163a.6-8). 42 These same accounts are described in the later Tibetan histories in much more detail, though unfortunately only one such supportive detail is, to my knowledge, found in an extant Indian source, Atīśa’s

• Bodhipathapradīpapañjikā, which I discussed in Chapter One. As I noted in Chapter One, some of the Tibetan

historians who provide the more detailed accounts of Buddhajñānapāda’s life, like Tāranātha, do list Indian sources that are no longer known to us. 71 confidence-inspiring or legitimizing function of Buddhist autobiography in relation to works that reveal new doctrines and practices is a feature that spans Buddhist cultures and traditions. The Dvitīyakrama’s Contents The sphere of the Buddha’s nirvāṇa, the unborn vajra, manifest awakening, the supreme essence of all sugatas—this great nondual nonconceptual reality is explained as the second stage. -Mañjuśrī instructing Buddhajñānapāda, Dvitīyakrama The Dvitīyakrama’s formal use of a narrative frame that is structurally reminiscent of Buddhist scripture, and the mukhāgama genre are not its only features that call to mind Buddhist scripture; its contents are likewise structured similarly to those of the tantras, in the sense that they cover a broad and eclectic range of topics both doctrinal and ritual. Because of their sheer breadth, the topics covered in the Dvitīyakrama provide an excellent overview of Buddhajñānapāda’s thought. Though most of the topics covered in the text deserve a full treatment in and of themselves, I constrain myself here to a brief summary of the Dvitīyakrama’s contents, to give a general sense of how they are arranged and related to one another. In the subsequent chapters I will tease out further details of the doctrinal and ritual content in both the Dvitīyakrama and Buddhajñānapāda’s other works. Before addressing the contents of the work, however, a discussion of its title is in order. The Sanskrit title is given in the Tibetan translations as the *Dvikramatattvabhāvanāmukhāgama. 43 The Tibetan translation of the title is Rim pa gnyis pa’i de kho na nyid sgom pa zhes bya ba’i zhal gyi lung.44 The primary discrepancy between the two is that the Sanskrit title, as given in the Peking, Narthang, and Sertrima Tengyurs, reads dvikrama, the “two stages” (the Derge and Cone give the nonsensical dvakrama), whereas the Tibetan translation consistently reads rim pa gyis pa, the “second stage.” In modern scholarship the title of this text is usually left untranslated, and the Sanskrit title is given as Dvikramatattvabhāvanā-mukhāgama, following its rendering in the Tibetan canonical translations. Where it is translated into English, the title is often rendered in the shortened form sometimes seen in the commentarial literature, as the Oral Instructions of Mañjuśrī (Skt. Mañjuśrī-mukhāgama, Tib.‘jam dpal gyi zhal gyi lung).45 In the instances in modern scholarship where the title has been translated in full, it has, so far as I am aware, always been taken to mean the training in both of the two stages, rather than just the second stage.46 However, both the Tibetan translation of the title, as well as the work’s contents, suggest that it is just the “second stage” that is intended in the title; the text is almost exclusively focused on instructions and practices relating to the perfection stage, with just three brief mentions of the “first stage,” as well as two very short summaries of generation stage practices, the latter comprising a total of just six of the work’s three hundred ninety-nine verses. Moreover, 43 dvi (dvi] P N C S, dva D C) kra ma ta ttva bhā (bhā] D C, bha S P N) wa (wa] D C, ba S P N) na (na] D C, sa S P N) nā ma mu khā (khā] D C S, khī P N) ga (ga] D C, saddhya S P N) ma (ma] D C, nya S P N) 44 Rim pa gnyis pa’i (pa’i] S P N, ba’i D C) kho na nyid sgom (sgom] D C, bsgom S P N) pa zhes bya ba’i zhal gyi lung 45 See, for example Kongtrül 1994, 126; Thurman 2010 689; Kongtrül 2005, 205; Kilty 2012. 46 Davidson (2002, 313) translates the title as Direct Revelation of the Cultivation of the Reality of the Two Processes. Guarisco (Kongtrül 2005, 535) translates it as Oral teachings of Manjushri/ Meditation on the Reality of the Two Stages. Richard Barron (Kongtrül 2011) translates it as Oral Transmission of Mañjuśrī: Meditation on Suchness according to the Two Stages. Roberts (2010) translates it as Oral Transmission Entitled Meditation on the True Nature of the Two Stages. 72 in verse 34 of the text, suchness/reality is directly equated with the second stage: “This great nondual nonconceptual reality/ Is explained as the second stage.47 In several other passages the use of the phrase “the suchness of the second stage,” (rim pa gnyis pa’i de kho na nyid) suggests that suchness is understood to pertain specifically to the second stage of practice. For instance, verse 283 reads, Therefore, with the mind that has already [generated] faith, Genuinely maintain the nature of all phenomena, The profound, luminous, nondual great reality, The suchness of the second stage, |283| Which has been taught by the guru.48 Similarly, verse 315 reads, “In this way, as for the suchness/ Of the second stage…”49 Vaidyapāda even uses the full phrase “the training in the suchness of the second stage” (rim pa gnyis pa’i de kho na nyid sgom pa) several times in his commentary on Buddhajñānapāda’s Muktitilaka. In one instance he writes, “Upholding, in this way, the samayas and vows, in order to [be able to] train in the reality of the second stage, [the text first] teaches, by means of example, the aspects of the first stage that are the basis for this...”50 At the end of that section of his commentary Vaidyapāda again repeats the phrase: “Having [first] remained in the generation stage, [now] in order to teach the training in the reality of the second stage...”51 Given the Tibetan translation of the title; the contents of the Dvitīyakrama, including the use of the term “the suchness of the second stage” within the work itself; and Vaidyapāda’s use of the term “training in the suchness of the second stage” in his commentary on Buddhajñānapāda’s other writings, it seems indeed quite likely that the Sanskrit title of the work is given incorrectly in the Tibetan translations, and that the correct title Sanskrit of the text is not the *Dvikrama-, but the

• Dvitīyakramatattvabhāvanā-mukhāgama. There are many cases in the Tibetan canon where the

titles of Sanskrit works have been given incorrectly—likely on the basis of a later and mistaken back-translation from the Tibetan title—so such a confusion of the ordinal (dvitīya) and cardinal (dvi) numbers in the Sanskrit title as given in the Tibetan translation is not terribly unusual or even unexpected. (And indeed, the appearance of the nonsensical “dvakrama” in D and C may perhaps suggest something in the direction of dvitīya, rather than just dvi, and at least points to there being some confusion with the issue.) Since the preponderance of the evidence, then, suggests that the title *Dvikramatattvabhāvanā-mukhāgama is simply based on a mistake in the rendering of the Sanskrit title in the Tibetan canonical translations, I depart here from the convention of referring to the work as such, and instead refer to it as the

• Dvitīyakramatattvabhāvanā-mukhāgama, which I translate into English as Oral Instructions on

Training in the Suchness of the Second Stage. The Dvitīyakrama begins with the traditional śāstric homage and commitment to compose the treatise, though already in this pledge Buddhajñānapāda writes that he will explain suchness, “so that beings can realize it through the words of guru Mañjuśrī,” displacing his own 47 gnyis med rtog bral don chen te// rim pa gnyis par rab tu bshad// (Dvitīyakrama, verse 34de) This verse could equally well be translated “This great nondual nonconceptual reality / Is explained in the second stage,” in which case this verse would fit into the category of the next two verses (283 and 315) I will discuss, which describe suchness as pertaining to the second stage. In either case, my argument here remains the same. 48 de bas dad pa sngon ‘gro ba’i// sems kyis chos kun de bzhin nyid// zab gsal gnyis med don chen po// rim pa gnyis pa’i de kho na// |283| bla ma’i gsung ni yang dag gzung// (Dvitīyakrama, verse 283-284a). 49 de ltar rim pa gnyis pa yi/ de bzhin nyid ni… (Dvitīyakrama, verse 315a-first half of pāda b). 50 de ltar dam tshig dang sdom pa la rnam par gnas pas rim pa gnyis pa’i de kho na nyid bsgom pa’i phyir de’i gzhi’i rim pa dang po rnams dpe’i sgo nas bstan pa... (Muktitilaka-vyākhyāna, D 51b.1-2; P 337b.3). 51 da ni de ltar bskyed pa’i rim pa la gnas pas rim pa gnyis pa’i de kho na (kho na] D, P om.) nyid bsgom pa bstan pa’i phyir... (Muktitilaka-vyākhyāna, D 52b.3-4; P 339a.1-2). 73 agency in favor of Mañjuśrī’s.52 This is followed by the autobiographical narrative detailed above that describes Buddhajñānapāda’s travels throughout the subcontinent, studying and practicing with different gurus, and culminating in his visionary encounter with Mañjuśrī, to whom he makes a supplication for teachings on “the supreme suchness of all phenomena.”53 The supplication is specific to the tantric nature of the instructions requested and, even without naming it specifically, to the perfection stage; it amounts to a request for instruction on sexual yogic practices and the realization of suchness that is their outcome. In his supplication to Mañjuśrī for teachings, Buddhajñānapāda equates suchness with the “moon (i.e. bodhicitta) which is born from the vajra and petals,” as a result of “playfully dancing the great dance” and “open[ing] the eight soft lotus petals and insert[ing] the vajra, the cause of nondual bliss.”54 Mañjuśrī expresses his pleasure at the request and agrees to give Buddhajñānapāda a teaching that is taught by the vajra holders of the past, present, and future, but only to some disciples, in order that they might realize the genuine truth.55 The instructions begin on a doctrinal note. The first topic that Mañjuśrī addresses is the nature of all phenomena, which he equates with the wisdom that knows the nonduality of “the profound and the luminous,” which he also refers to as great Vajradhara. The text explains that the identity of everything—the three worlds, the elements, literally everything—is the essence of the mind, and that realizing this amounts to awakening to perfect buddhahood. This nondual, nonconceptual reality is here explicitly identified with the “second stage” of tantric practice.56 The next major topic in the text is a cosmogonic narrative, describing the way in which saṃsāra arose out of nondual wisdom. This narrative, which is paralleled in the Muktitilaka,57 describes the “arising in the manner of the great thought” from the pure reality of suchness, which has been taking place since beginningless time. It is from this that the elements arise, and thus the world and all beings. The presence of all of these phenomena obscures, in the perception of beings, the nondual nonconceptuality that is their nature, and it is due to this lack of recognition of the nature of reality that beings cycle in saṃsāra.58 Next the Dvitīyakrama describes the qualities of the ideal disciple who will come to realize the nature of this reality and the guru who has the capacity to show it to him. The disciple should have merit, joy, respect, honor, proper intention, and be stable, generous, compliant, and free from doubts. The guru must hold the lineage of the oral instructions, be intent upon the Mahāyāna path, know the secrets of the ten suchnesses (daśatattva), and be willing to teach. A description of what the student should offer the guru follows, and includes a wide variety of precious items ranging from lands, houses, elephants, gold and rubies, to one’s wife, sons, daughters, sisters, and nieces.59 The text then moves on to present a typology of women. Here, following a verse that strongly parallels one in the Sarvabuddhasamāyoga-tantra on women as the superior illusion among all illusions,60 it is explained that women are of four families or types, corresponding with the four buddha consorts from the Guhyasamāja-tantra: Locanā, Māmakī, Pāṇḍaravāsiṇī, and 52 These traditional śāstric introductory verses are found in the Dvitīyakrama, verses 1-2. 53 The opening autobiographical narrative is presented in the Dvitīyakrama, verses 3-11. 54 The supplication to Mañjuśrī is presented in Dvitīyakrama, verses 12-18. 55 Mañjuśrī accepts the request and makes his own pledge to teach to Buddhajñānapāda in the Dvitīyakrama, verses 19-22. 56 The topic of the nature of suchness/reality is presented in the Dvitīyakrama, verses 23-34. 57 Muktitilaka, D 50b. 58 The cosmogonic narrative is presented in Dvitīyakrama verses 35-42. 59 The description of the disciple, guru, and offerings is presented in Dvitīyakrama verses 43-49. 60 Dvitīyakrama, verse 50. See also note 123 in my translation of the Dvitīyakrama. 74 Tārā. The characteristics of each “type” of woman—the four types are called kamalī,61 śaṅkhinī, citriṇī, and hastinī—are described in some detail, including both the physical features and the character of women who belong to each type, and each is then equated in terms of her “pure form” (viśuddhi) with one of the four buddha consorts.62 The appropriate and inappropriate behaviors for a female consort are discussed, followed by a description of the desirable and undesirable characteristics of her “secret place.” The text then advocates a yogin’s finding an appropriate female partner, since it is by means of relying upon her that accomplishment is possible.63 Here the Dvitīyakrama switches to ritual content, describing a series of initiatory sequences for the sexualized second and third tantric initiations—the guhya and prajñājñāna initiations.64 This section of the Dvitīyakrama includes several verses that are to be recited by the guru bestowing the initiation, as well as a dialogue between the yogin and his partner. Parallel passages of several of the verses in this section of the Dvitīyakrama appear in quite a number of later texts, both scriptural and authored (but in no earlier works of either type of which I am aware).65 The section on the third initiation includes several sequences of practices that Vaidyapāda identifies as meant to arouse mental, verbal, and finally physical passion in the initiate couple, culminating in their ritual sexual union. The passage on arousing passion physically includes a description of several postures and sexual acts and reads very much like a work of kāmaśāstra. This passage is followed by the “delighted” consort instructing the yogin to search for the cakra inside her secret place, which the text states that he must find with his fingers “by means of the oral instructions from the guru.” The yogin is instructed to thus discern the location of the three main subtle channels in his partner’s body, and only then to join with her in sexual union. The stages of union are here equated with the processes of sevā, upasādhana, sādhana, and mahāsādhana—the so-called “four branches” (caturaṅga) that are commonly associated with generation stage practice. In the sexual act itself the yogin is instructed to control the inner winds to bring about “blazing” and “dripping” in what seems to be a very early instance of what later comes to be described as the caṇḍālī yoga, and is found commonly in the later Yoginī tantras.66 The culmination of the third initiation, which involves the practitioner(s?)67 observing suchness directly while in sexual union, results in emission of sexual fluids, and the 61 This is an unexpected form for the type that is normally called padminī. See note 127 in my translation of the Dvitīyakrama. 62 The four types of women described here are the classical four types from kāmaśāstra, and Buddhajñānapāda’s use of this classificatory schema is a very early one, in fact the earliest of which I am aware, not just in Buddhist literature but in extant Indian literature on the whole. I discuss this in more detail in Chapter Six in the context of perfection stage practices. 63 The topic of an appropriate consort is presented in Dvitīyakrama, verses 50-82. 64The rituals for the guhya and prajñājñāna initiations are described in Dvitīyakrama, verses 83-125. 65 I discuss these initiatory sequences in Chapter Seven. Parallel verses from this section of the Dvitīyakrama appear in a number of later sources including the Samājottara, the Caṇḍamahāroṣaṇa-tantra, Vaidyapāda’s Guhyasamājamaṇḍalopāyikāṭīkā, Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, Nāgabodhi’s Maṇimālā, Advayavajra’s Saṃkṣiptābhiṣekaprakriyā, Kṛṣṇācārya’s Śrīguhyasamājamaṇḍalopāyikā, Prajñāgupta’s Abhiṣekaratnāloka, Prajñāśrī’s Abhiṣekavidhi, Vagīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, Kṣitigarbha’s Daśatattvasaṃgraha, Ratnākaraśānti’s Ratnāvalī, Abhayākaragupta’s Vajrāvalī, and Kuladatta’s Kriyāsamgraha. See also notes to verse 87-95 in my translation of the Dvitīyakrama. 66 The practice is described only very briefly in the Dvitīyakrama, and the term caṇḍālī is not applied to it here. Tsongkhapa, in his much later Tibetan commentary on the practice of the “five stages” of Ārya School Guhyasamāja practice does identify this passage in the Dvitīyakrama as referring to the practice of caṇḍālī (Kilty 2015, 324). 67 The sexual practices in the Dvitīyakrama are, as is predominately (always?) the case in Indic tantric Buddhist writings, described from the perspective of the male partner. There are a number of instances, though, in Buddhajñānapāda’s and Vaidyapāda’s writings, that suggest the female partners were not supposed to be just passive participants but also to be educated and trained in these yogas. 75 yogin is subsequently instructed to “take up the liquid nectar that abides in the lotus with his mouth and drink it.” Here the Dvitīyakrama returns to doctrinal topics, describing briefly the “final identity of all things” as “profundity and clarity.” What follows is a doxography of philosophical views, which presents the various ways in which different groups of beings mistakenly fixate on reality due to their ego clinging. The list begins with descriptions of the mistaken views held by unspecified “non-Buddhists,” with Vaidyapāda noting in his commentary the identity of the groups who hold each of the specific views listed (Sāṃkhyas, Vaiśeṣikas, etc.). Then the mistaken views of specified Buddhists groups—the Kaśmiri Vaibhāṣikas, Sautrāntikas, Yogācārins, and Mādhyamikas—are listed, with the clear indication that the views of each group listed are more accurate than those of the group before. These are all subordinated to the view of one who has realized the *adhideva, which is later in the Dvitīyakrama equated with the result of tantric practice, and which is here said to be realized “through spontaneously arisen wisdom in reliance on the words of the guru.”68 The Dvitīyakrama briefly praises nondual wisdom as the purview of the vidyādharas alone—beyond the realization of the śrāvakas, pratyekabuddhas, Yogācārins, Mādhyamikas, bodhisattvas, and even the “non-superior buddhas”69—and emphasizes that it can be transferred to the qualified disciple “[even] without words.” This is followed by a condemnation, paralleled in the Muktitilaka, of the yogin’s being too focused on practices involving action since they “are in contradiction to the unelaborate.” Most of the practices listed here—the maṇḍala, homa, bali, recitation, and so forth—are generally connected to the generation stage of tantric practice, though this fact is not explicitly mentioned. The text does, however, give an injunction to engage instead in “the second stage” of tantric practice.70 Detailed instructions on perfection stage practices fill the next more than one hundred verses, making up over a quarter of the Dvitīyakrama’s total length. The practices in this section are also described in the Muktitilaka, though their presentation there is significantly less detailed. The Dvitīyakrama’s presentation begins with what amounts to a very short overview of generation stage practice—just five verses—and the injunction to accustom oneself to these practices through training, which makes it clear that the perfection stage practices are meant to take place within the framework of the generation stage. The perfection stage practices that are given here are three different bindu yogas: the yoga of the indestructible bindu, the secret bindu, and the emanated bindu. The first two yogas are versions of the practice often called sūkṣmayoga, and the latter is essentially a form of the practice of vajrajāpa, a term also used to describe this practice within the Dvitīyakrama itself. At the conclusion of the instructions on the three bindu yogas comes a short explanation of the fact that the indestructible bindu is the only thing that will remain at the end of an aeon, and that from which all other phenomena subsequently arise. This is followed by a short summary of the previously described perfection stage practices with a focus on the dissolution of the visualization within which the practices are performed. Then, the process of the dissolution of the psychophysical elements into one another along with the accompanying five signs are described. These are processes that are generally taught to take place naturally at death, or in an intentional and controlled way through perfection stage practice. The Dvitīyakrama states that having “genuinely trained in this, one attains great non-abiding nirvāṇa.”71 68 The doxographical section is in the Dvitīyakrama, verses 126-143. 69 I address this unusual term in Chapter Three. 70 This short section of the Dvitīyakrama encompasses verses 144-155. 71 The perfection stage practices are detailed in the Dvitīyakrama, verses 156-271. 76 The next section of the Dvitīyakrama gives a rather extensive list of the various names of suchness which, under countless names, is said to be the single intention of all sūtras and tantras.72 Subsequently the benefits and results of the practice of the second stage are mentioned. The Dvitīyakrama also here mentions the attainment of the three blisses—bliss, middling bliss, and the bliss of cessation—that are the progressive attainments of the perfection stage practice in this system. In addition to practicing with a female partner (human or otherwise), it advocates the undertaking of the unmatta-vrata, a practice in which the yogin feigns madness as a test of the stability of his yogic practice, and other time-constrained practices (the period prescribed here is six months).73 What follows is essentially an equating of the stages of tantric sexual practice with the ten bodhisattva bhūmis, with the text suggesting that it is through relying upon a female partner while “endowed with the ten bhūmis,” that the practitioner attains the supreme result. Here a series of sexual acts that bring forth various stages of bliss are described, one-by-one, as constituting each of the ten bhūmis. The Dvitīyakrama goes on to explain that the very same thing has been taught as the traditional bodhisattva bhūmis, “Perfect Joy” (the first bhūmi) and the rest, for “those disciples who are unable to authentically engage in this great truth.” Such a subordination of the traditional bodhisattva path to tantric sexual practices is made even more explicit with the statement that those who train in the former path “gain realization—though there is still something higher.” The final result of the tantric path is here called the *adhideva, and identified with the thirteenth bhūmi.74 The Dvitīyakrama goes on to describe the greatness of the yogin who engages in these practices; he is even worshipped by the pure deities of the ten directions. This section contains two unusual verses in which Mañjuśrī speaks in the first person and appears to say that he will abandon anyone who deprecates a practitioner of these practices, while someone who praises and worships such a yogin, because Mañjuśrī “abides in his body,” will have his physical obscurations cleared away. The text then insists that the practitioner who trains in the second stage “with the goddess acting as the condition”—thus implying the practice of sexual yogas— will undoubtedly attain the mahāmudrā in this very life.75 If, the Dvitīyakrama explains, a disciple has pleased the guru and received initiation, samayas, and vows, “obtained suchness...through the guru’s words,” and realized the “secret and the supreme secret,” but he has been unable to train in the way explained above, he should practice the yoga of utkrānti, the ejection of consciousness from the body, at the time of death. The text goes on to give instructions on how to practice this yoga, including descriptions of the resulting rebirths that ensue from consciousness departing from the body’s various orifices at death. All of these apertures are therefore blocked with visualized syllables during the practice. It is asserted that through this practice the yogin will realize “that which is luminous and perfectly joyful, like the sky,” which Vaidyapāda clarifies is the dharmakāya. The subsequent attainment of the two form kāyas is also referenced. The ritual of utkrānti is praised as being able to bring accomplishment to “even one who has committed the gravest sin, a deluded being, or a brahmin-slayer.”76 72 The names of suchness are set forth in the Dvitīyakrama, verses 272-284. 73 The benefits and results of practice, and the injunction to take up various vratas are described in the Dvitīyakrama, verses 285-297. 74 The equating of tantric sexual practice with the bodhisattva bhūmis is described in the Dvitīyakrama, verses 298- 315. 75 These verses on the benefits and result of the practice of the sexual yogas of the second stage are in Dvitīyakrama, verses 316-326. 76 The instructions on utkrānti are found in Dvitīyakrama, verses 327-359. 77 The instructions on utkrānti are followed by another curious passage in which Mañjuśrī again assumes the first-person voice. Here Mañjuśrī cautions that he and the tathāgatas will not join with or bless an individual who teaches these secret practices without having realized them. Mañjuśrī then repeats his previous statement about abiding within the bodies of certain practitioners and in that way receiving offerings, by means of which those who make the offerings are able to purify their obscurations. He then explains that as long as these particular teachings remain in the world, the Buddha’s teaching will remain; when this lineage is broken, the Dvitīyakrama contends, that will signal the disappearance of the Buddha’s teaching.77 Mañjuśrī then instructs Buddhajñānapāda directly to compile and pass on the instructions. This is followed by a prediction given by Mañjuśrī to Buddhajñānapāda, discussed above in Chapter One. Here Mañjuśrī states that because of several mistakes that Buddhajñānapāda made with respect to him—holding delusion about his identity upon first meeting the emanated monk in the forest, and his “conduct regarding food,” which Vaidyapāda explains as Buddhajñānapāda’s refusal of foods offered to him by that monk’s female companion, he will not attain a complete transformation of his aggregates within this very life. Buddhajñānapāda will, however, Mañjuśrī assures, “accomplish consciousness, which is indestructible, as the mahāmudrā,” a statement interpreted by later commentators to mean that Buddhajñānapāda would attain full accomplishment only in the intermediate state following death (antarābhava, bar do). Mañjuśrī then commands Buddhajñānapāda to compose a number of specific texts pertaining to the generation stage practice of the Guhyasamāja-tantra, presumably as a complement to the perfection stage instructions that Mañjuśrī had just given, which Buddhajñānapāda had already been instructed to compile and pass on. Mañjuśrī completes his instructions with an injunction for yogins of the future to please a guru who knows the truth, and having received instruction, to train in these practices. He notes that when their minds have completely abandoned conceptuality, practitioners accomplish the state of the vajra holder, “due to which they will then genuinely accomplish buddhahood.” His final words of instruction are an injunction to endeavor fervently towards the accomplishment of supreme suchness, or at the very least to make aspirations toward accomplishing it.78 The Dvitīyakrama here returns to Buddhajñānapāda’s first-person recounting of the episode, noting that Mañjuśrī “sang and praised me,” and then disappeared “like a cloud into the sky,” at which point the monk and two gurus, whom he had met in the forest at the outset of the visionary experience, likewise disappeared. Buddhajñānapāda then details, in a few verses, some of the events of his life subsequent to the vision: he lived, wrote, and taught at the Parvata cave behind Vajrāsana together with his disciples, and was supported by the patronage of the wealth deity Jambhala himself. Buddhajñānapāda also reports traveling to meet his former guru Pālitapāda and writing a sādhana at his request, a meeting which Vaidyapāda says took place at that master’s residence in the Koṇkan.79 Following his autobiographical account, Buddhajñānapāda encourages listening to and contemplating these special teachings on the second stage of tantric practice, as well as putting them into practice in order to attain awakening. This section of the text contains a number of particularly poetic and evocative verses in which Buddhajñānapāda cautions against wasting the 77 Some later commentators, such as the 14th-century Tibetan master Tsongkhapa (Kilty 2012, 56), have interpreted this passage to be referencing the Guhyasamāja teachings, while others such as the 16th-century Tibetan master Dakpo Tashi Namgyal (Dwags po bkra shis rnam rgyal), have interpreted the passage to be referencing the tantric teachings more generally (See Roberts 2010, 484). 78 The prediction, command to Buddhajñānapāda to compose texts, and Mañjuśrī’s final instructions are in the Dvitīyakrama, verses 360-374. 79 The concluding short autobiographical narrative is in the Dvitīyakrama, verses 375-378. 78 precious opportunity to practice and gain accomplishment in this life, declares that through training repeatedly in these practices there is no doubt that wisdom will arise, and again clearly indicates the superiority of the tantric path to that of the traditional Mahāyāna. It is in one of these verses that Buddhajñānapāda uses the term “the great perfection” (rdzogs pa chen po) to describe the final result of the practice of the second stage, which he also describes as “the universal form of wisdom” and equates with Great Vajradhara. The final five verses of the Dvitīyakrama are Buddhajñānapāda’s dedication of the merit from his “compiling these oral instructions,” and his aspiration that yogins will take up this practice and attain the dharmakāya, and then fill the three realms “with awakened body, speech, and mind, and uncountable emanations, liberating all beings from existence!”80 In the authorial colophon the Dvitīyakrama is described as “the oral instructions of the bodhisattva Mañjuśrī, a lineage passed from mouth to mouth, which were compiled by the great maṇḍalācārya, Buddhaśrījñānapāda.” The Place and Role of the Dvitīyakrama in Buddhajñānapāda’s Oeuvre Buddhajñānapāda had already composed several non-tantric works in his youth—notably the Sañcayagāthā-pañjika and the Mahāyānalakṣaṇa-samuccaya. He may also have written some of his other tantric works not specifically related to the Guhyasamāja-tantra prior to the encounter with Mañjuśrī described in the Dvitīyakrama.81 But it is this visionary encounter, the practice instructions he received from Mañjuśrī therein, and the command from Mañjuśrī to compile those instructions and compose additional related literature that seem to have defined and guided the remainder of Buddhajñānapāda’s career. However, the precise temporal relationship between the compilation of the Dvitīyakrama and the composition of Buddhajñānapāda’s other Guhyasamāja-related tantric writings remains unclear. Given that the composition of a sādhana, identified by commentators as the Caturaṅga/Samantabhadrasādhana, on the occasion of a visit to his guru Pālitapāda, is mentioned by Buddhajñānapāda in the second part of his autobiographical narrative at the end of the Dvitīyakrama, it would seem that the Caturaṅga/Samantabhadra was composed prior to the compilation of the Dvitīyakrama. The question of the temporal relationship between the Dvitīyakrama and the Muktitilaka, which share quite a significant amount of material, is less clear. The fact that the Muktitilaka introduces and sets forth the two categories of the generation and perfection stages, while the Dvitīyakrama essentially assumes them, suggests the possibility that the Muktitilaka may have been the earlier composition.82 This would mean that the more detailed perfection stage instructions in the Dvitīyakrama, together with their narrative frame, may have been written later in 80 Buddhajñānapāda’s dedication and aspiration are in the Dvitīyakrama, verses 395-399. 81 As noted in Chapter One, I wonder if the Ātmasādhanāvatāra may have been composed prior to the visionary encounter with Mañjuśrī (perhaps even prior to Buddhajñānapāda’s discipleship under Pālitapāda?) given its complete absence of any reference to the Guhyasamāja-tantra. The three Jambhala sādhanas likewise lack any such reference, as does the *Gativyūha. In Buddhajñānapāda’s other tantric writings—the Dvitīyakrama, the Muktitilaka, the Caturaṅga/Samantabhadra-sādhana and the Śrīheruka-sādhana, the Guhyasamāja-tantra plays a central role. 82 I have argued in an earlier conference paper (C. Dalton 2014), and will address further in Chapter Eight, that Buddhajñānapāda’s verse in the Muktitilaka that sets forth the two stages of tantric practice appears to have been the source for its parallel—and much better known—verse in the Samājottara that became the scriptural locus classicus for the two stages, referenced throughout later tantric Buddhist literature. I am certainly not suggesting that Buddhajñānapāda invented this distinction between the two stages of tantric practice—the two stages are referenced in other works from this period, including Padmasambhava’s Man ngag lta ba’i ‘phreng ba and Padmavajra’s Guhyasiddhi—just that the verse in which he articulates it in the Muktitilaka was the source for that in the Samājottara, rather than the other way around. There are likewise two pādas found in both Dvitīyakrama and the Samājottara, where again the Dvitīyakrama’s appear to be the earlier of the two. I discuss this in Chapter Eight. 79 Buddhajñānapāda’s career as a method of encapsulating his life and teachings into a single coherent narrative and providing a strong and clear legitimation of the source of his teachings.83 At this point, though, it is impossible to determine this with any certainty. What we can say with certainty is that the Dvitīyakrama, including the narrative of Buddhajñānapāda’s encounter with Mañjuśrī described in the text and its perfection stage practice instructions, became definitive of his Guhyasamāja practice tradition. And while it is not his most commented-upon work—the Caturaṅga/Samantabhadra-sādhana holds that honor— the Dvitīyakrama still stands out as Buddhajñānapāda’s most unique. What is more, much of the doctrinal and ritual content therein is reflected in his other works, making the Dvitīyakrama both central to, and an excellent anchor for, understanding Buddhajñānapāda’s thought on the whole. But despite its uniqueness, the Dvitīyakrama is not Buddhajñānapāda’s only work in which we find narrative used to doctrinal ends. The Muktitilaka, as well, contains a short and unusual narrative passage, modeled on Buddhist scripture, that is used to make a pointed doctrinal statement. It is to the topic of the doctrinal aspects of Buddhajñānapāda’s thought, found across the range of his writings including the Dvitīyakrama, that we now turn. 83 This theory would seem to go against Vaidyapāda’s understanding, however, as he asserts both in Sukusuma and his Muktitilaka-vyākhyāna that the Muktitilaka is included in the texts that Mañjuśrī, in the Dvitīyakrama, commands Buddhajñānapāda to compose. That would not, however, necessarily preclude the Muktitilaka’s being composed first, in response to the command to compile the perfection stage instructions given by Mañjuśrī, and then detailed later and following Mañjuśrī’s own words in the Dvitīyakrama. In fact, the passage from the Dvitīyakrama that Vaidyapāda cites in the Muktitilaka-vyākhyāna as the reason for the Muktitilaka’s composition, begins with Mañjuśrī’s command to Buddhajñānapāda to “compile [these instructions]” (Muktitilaka-vyākhyāna, D 48b.1) 80 Doctrine 81 Chapter Three Following the Tantric Path to the Reality of Nondual Wisdom: Buddhajñānapāda’s Doctrinal Positions The nature of phenomena, from form and the rest up to omniscience, is the perfectly pure wisdom of the nonduality of the profound and the luminous, which is like the center of space...That itself is the Victors, their offspring, and their fields of influence, the three existences comprised of everything animate and inanimate. That, the identity of everything, is the very essence of one’s mind, supreme among all things. -Mañjuśrī instructing Buddhajñānapāda, Dvitīyakrama We saw in the Dvitīyakrama evidence of Buddhajñānapāda’s great creativity in formulating the unique narrative structure of that work. Looking to the doctrinal content of his writings we likewise see the creativity with which he, like other early tantric Buddhist authors, presented doctrinal positions that contextualized and supported the innovative ritual structures of the tantric soteriological traditions whose practices they advocated and personally upheld. As Anthony Tribe has noted with reference to Vilāsavajra’s Nāmamantrārthāvalokinī, tantric Buddhist writings of the 8th and early 9th centuries show both a tendency to encode tantric structures within Mahāyāna doctrinal categories, as well as a tendency to modify those doctrinal categories to accommodate the structures of tantric practice.1 Buddhajñānapāda’s work shows evidence of both of these trajectories. Widely framed within the Mahāyāna Buddhist tradition as it stood in the late 8th century—in particular the Yogācāra-Madhyamaka philosophical synthesis advocated by masters like his teacher Haribhadra— Buddhajñānapāda’s writings display resonance and confluence with doctrinal trends in several different tantric Buddhist textual communities of the period, including those springing from the Mañjuśrīnāmasaṃgīti and the Guhyasamāja-tantra, as well as contemporary early (or proto-) Great Perfection literature. His work was also influenced by other early tantric systems, including those of the Sarvatathāgatatattvasaṃgraha and the Sarvabuddhasamāyoga-tantra, both of which he cites in his Ātmasādhanāvatāra, and his writings additionally reflect familiarity and engagement with non-Buddhist systems. In reading widely across Buddhajñānapāda’s work, and in particular the tantric writings that have been the focus of my study, we do get the sense that he was articulating a unified and coherent system of tantric Buddhism. Buddhajñānapāda’s vision of reality and Buddhist practice brings together an overarching emphasis on nondual wisdom as both the nature and the source of the phenomenal world; the idea that this reality of nondual wisdom can be directly indicated to a disciple by his guru; a prioritizing of the tantric path, including the deityyoga- oriented, transgressive, and especially the sexual elements of Mahāyoga tantra; and a rhetoric of non-action standing in contrast to the elaborate ritual structures of the very tantric path he so strongly advocates. In this chapter I will explore a number of the doctrinal positions that stand out across Buddhajñānapāda’s tantric oeuvre, and note, where possible, some places of influence or confluence between his thought and those of other authors and systems, both Buddhist and non-Buddhist. But before we begin, I must make clear that what follows is just a preliminary exploration of Buddhajñānapāda’s thought. First, the focus of my study of his doctrinal positions has been on his tantric writings. A full study of Buddhajñānapāda’s doctrinal and philosophical positions 1 Tribe 2016, 6. 82 would necessarily also include a detailed analysis of his Mahāyānalakṣaṇa-samuccaya, and especially his Samcayagāthā-pañjikā, the Prajñāpāramitā commentary written towards the beginning of his career prior to his studies with tantric teachers, which I have not done.2 In what follows, I attend to and examine some of the doctrinal positions that are either stated directly or implied by statements made in his mature tantric writings, primarily the Dvitīyakrama and the Muktitilaka, which are composed in verse and also include ritual content, and the Ātmasādhanāvatāra, Buddhajñānapāda’s only surviving prose work that includes tantric material. I also make reference to several philosophical and doctrinal passages in his Caturaṅga/Samantabhadra-sādhana. But it is not only the limited scope of my inquiry that makes this study into Buddhajñānapāda’s doctrinal positions preliminary; it is also quite simply the first such study. With the single exception of the Dvitīyakrama, which I have edited and translated in this dissertation, not one of Buddhajñānapāda’s works has been fully edited (in either its Sanskrit original or Tibetan translation) or translated into any modern language, and no attempt has been made in modern scholarship to present a full overview of any single one of his writings, let alone his thought on the whole. Indeed the most in-depth modern scholarship on Buddhajñānapāda’s works are short pieces that have focused not on doctrine, but on his ritual systems.3 Most of the brief references to Buddhajñānapāda’s doctrinal positions in secondary scholarship are either gleaned from sources other than Buddhajñānapāda’s writings themselves (i.e. they are based on other traditional authors’ assessments of his thought), or they have been made with reference to just a brief passage from one of his compositions.4 Thus while I have endeavored here to draw attention to the doctrinal features of his writings that I have found most prominent or remarkable, my presentation here is just a preliminary sketch, and I hope that this introduction will act as a gateway for more research that further illuminates Buddhajñānapāda’s thought. I. Nondual Wisdom as The Nature of the Mind and of all Phenomena Defining Nondual Wisdom The nature of phenomena, from form and the rest up to omniscience, is the perfectly pure wisdom of the nonduality of the profound and the luminous, which is like the center of space. -Mañjuśrī instructing Budhajñānapāda, Dvitīyakrama 2 Buddhajñānapāda himself tells us that the Sañcayagāthā-pañjika was composed at Nālandā, prior to his travels over the subcontinent during which he studied with his tantric gurus. 3 These studies are Ryuta Kikuya’s short article (in Japanese) on the three bindu yogas of the perfection stage in Buddhajñānapāda’s works (Kikuya 2008), and Kimiaki Tanaka’s brief descriptions of the structure of the generation stage practice in the Samantabhadra-sādhana: (Tanaka 1996, 176-194 (in Japanese) and 257-271 (in English); Tanaka 2007b (in Japanese); Tanaka 2010 (in Japanese); Tanaka 2017 (in English); Tanaka 2018 (in English). 4 The single exception I am aware of is a few paragraphs in John Makransky’s book, Buddhahood Embodied, that examine Buddhajñānapāda’s position on the doctrine of the kāyas as expressed in his Sañcayāgāthā-pañjikā in relation to his teacher Haribhadra’s position on that topic. Again, the Sañcayāgāthā-pañjikā is one of Buddhajñānapāda’s two non-tantric works that I have chosen not to address in the present study. Given that there is more scholarship on Buddhajñānapāda in Japanese than in Western languages, it is possible that some of his doctrinal positions have received brief attention there, as well, but in familiarizing myself with that scholarship to the best of my ability with the assistance of several Japanese friends and colleagues, I have not become aware of any extensive treatments of his doctrinal positions. As I noted in the Introduction, Tanaka (2018, 29) mentions a threepage 1985 article by Chizuko Yoshimizu that “argues the Jñānapāda school from the side of philosophy for the first time.” According to Tanaka’s bibliography, the article is titled “On the Yogācāra-Mādhyamika Theory in the Jñānapāda School,” but as the article is in Japanese I have been unable to consult it, and I am unsure whether it focuses on Buddhajñānapāda’s own writings or those of later authors in his tradition. 83 Probably the most pervasive and overarching aspect of Buddhajñānapāda’s thought is his emphasis on nondual wisdom (advayajñāna, gnyis med ye shes), which he identifies as the nature of mind and of all phenomena. While the use and development of the term “nondual wisdom” in Buddhist systems deserves a full study in and of itself, very broadly speaking the term is found sparingly in exoteric Mahāyāna texts, including some sūtras and the śāstric writings of both Madhyamaka- and Yogācāra-oriented authors, but it achieved much greater currency in tantric literature—both the tantras themselves and their commentaries—especially from the late 8th century onwards. However, even within the tantric tradition, the earlier tantras—all the way up through the Guhyasamāja root tantra itself, which was probably completed some time in the mid-to-late 8th century, make almost no use of the term nondual wisdom. A very brief survey of its use in the early tantras reveals that the term “nondual wisdom” (advayajñāna, gnyis med ye shes) does not appear in the Mahāvairocanābhisambodhi, the Sarvatathāgatatattvasaṃgraha, the Sarvabuddhasamāyoga, the Sarvadurgapariśodhana, nor even in the Guhyagarbha-tantra or the Guhyasamāja root tantra. It does appear once in the Mañjuśrīnāmasaṃgīti (as jñānam advayarūpadhṛk, “wisdom that holds the form of nonduality”), and one time in the Samājottara. From the turn of the 9th century—precisely in the time Buddhajñānapāda was writing—and onwards, the term nondual wisdom begins to find more currency in the tantras, and especially in tantric commentarial literature. The general term “nondual” or “nonduality” (advaya, gnyis med/gnyis su med pa) does appear in the earlier tantras, but again not with the kind of frequency that it is found in the later tantric tradition.5 Ronald Davidson has written that, “Buddhism has traditionally defined reality in terms of multiplicity and its resolution into nonduality,”6 and indeed we see this tendency reflected in Buddhist texts from many different historical periods and traditions. However, unsurprisingly, what is intended by “nondual wisdom,” where we do find this term used, is not uniform throughout Buddhist literature, neither with regard to the type of nonduality expressed—that is, what specific two things any given work claims to be ultimately “nondual”—nor in terms of its scope. Many times the term is given without further elaboration, and it is thus possible to understand the term “nondual wisdom” in Buddhist writings to refer to wisdom that goes beyond any and all types of duality. Some authors do, however, elaborate the point, specifying the particular aspect of nonduality they intend.7 Among earlier usages of the term nondual wisdom in systems of thought with which Buddhajñanapāda was certainly familiar, one such elaboration 5 The term “nondual” (gnyis med/gnyis su med pa) unattached to the word wisdom appears thirteen times in the Mahāvairocanābhisambodhi, including as a description of awakening and in the compounds “nondual mind” (yid gnyis med) and “nondual yoga” (gnyis su med pa’i rnal ‘byor). It does not appear at all in the Sarvatathāgatatattvasaṃgraha nor in the Sarvadurgatipariśodhana-tantra, and is found just twice in the Sarvabuddhasamāyoga-tantra, both times as part of the term “nondual vehicle” (advayaṃ yānaṃ). The term “nondual” alone appears five times in the Mañjuśrināmasaṃgīti, including two uses of the term “nondual dharmatā” (advayadharmatā). It appears six times in the Guhyagarbha-tantra, several times in the compound “nondual dharmadhātu” (gnyis med chos kyi dbyings), and twice in the compound “nondual mind” (gnyis med blo/ yid gnyis med pa). In the Guhyasamāja root tantra the term appears seven times including in the compound “nondual vajra” (advayavajra, gnyis su med pa’i rdo rje); the phrase “neither dual nor nondual, like space,” which forms part of the instructions on how to meditate upon nonorigination; and the compound “the ultimate nondual suchness (paramārthādvayatattva), which is part of the title of Chapter 17. The Samājottara also includes one use of the term “nondual” (advaya) apart from the term nondual wisdom in the phrase, “phenomena are nondual, but they are marked by duality” (advayāḥ sarvadharmās tu dvayabhāvena lakṣitāḥ|, Samājottara, 126ab). 6 Davidson 1995, 104. 7 Perhaps it is more precise to say the aspect of nonduality that they intend to emphasize. That is, in the instances when one specific aspect or type of nonduality is directly mentioned, it is often the case that other aspects of nonduality seem to be implied. 84 specifies nondual wisdom to be the wisdom of subject-object nonduality, or as it is more frequently described in Yogācāra writings, of the absence of the duality of the perceiver (grāhaka, ‘dzin pa) and perceived (grāhya, gzung ba).8 In some Madhyamaka writings we find references to nondual wisdom that emphasize other aspects of nonduality, including the nonduality of appearance and emptiness; that is, of the relative and ultimate perspectives.9 In that context, nondual wisdom refers to the knowledge that while phenomena appear, they are by nature empty, and that the appearance of these phenomena is not different from their emptiness. The nondual wisdom that Buddhajñānapāda asserts to be the nature of the mind and of all phenomena is not exactly identical to either of these presentations, but seems to be informed by, and draw from, both. Moreover, unlike these exoteric presentations of nondual wisdom, Buddhajñānapāda’s includes elements that are decidedly tantric. We are most fortunate that while Buddhajñānapāda does at times simply use the term “nondual wisdom” (gnyis med ye shes) without further elaboration, he also employs the more descriptive phrase “the wisdom of the nonduality of the profound and the luminous” (zab gsal gnyis med ye shes)10 in both the Dvitīyakrama and the Muktitilaka. It is through this phrase, which specifies “the profound” and “the luminous” as the two aspects that are realized as nondual, that we are able to gain a clearer sense of precisely what he means by nondual wisdom. Buddhajñānapāda elaborates upon these two terms at several places in his writings. In a passage on the supremacy of the view of the nondual nature, the Muktitilaka describes “the profound” and “the luminous” as follows: Completely free of conceptuality, It is far beyond the reach of thought or speech, Stainless like the sky, it is the source of everything, Beyond imputation: thus, it is called the profound. Because it is the purification of the mind of oneself and others As illusory and rainbow-like In the form of the mahāmudrā 8 For example Sthiramati, in his Sūtrālaṃkāravṛttibhāṣya references the “wisdom of the nonduality of subject and object” (gzung ‘dzin gnyis su med pa’i ye shes) (Sūtrālaṃkāravṛttibhāṣya, D 140a.6), and Haribhadra, in his Abhisamayālaṃkārālokā writes of “nondual wisdom, free from subject and object” (gzung ba dang ‘dzin pa dang bral pa’i gnyis su med pa’i ye shes), though as a Yogācāra-Mādhyamika author, he takes care not to ascribe any ultimate nature to such a nondual wisdom (Abhisamayālaṃkārālokā, D 256b.5). 9 For example Candrakīrti’s Triśaraṇasaptati uses the term “nondual wisdom” to refer to the knowing of the nonduality of the appearance and emptiness of objects, or of their relative and ultimate natures, using the traditional metaphor of a reflection. He writes, “The Victors, knowing these entities/ To be like reflections/ [With] nondual wisdom bring ignorance to an end.” rgyal bas dngos po de rnams ni// gzugs brnyan dang yang ‘dra mkhyen pas// gnyis med ye shes ma rig zad//. (Triśaraṇasaptati, D 251b.2-3). While, as Kevin Vose has shown, Candrakīrti’s writings and his Prasāngika Madhyamaka position became more popular in Tibet than was the case in India, it seems that his work was known in Buddhajñānapāda’s circles, given Vilāsavajra’s citation of Candrakīrti’s Madhyamakāvatāra in his Nāmamantrārthāvalokinī (Vose 2009; Tribe 2016, 376). 10 The term zab gsal gnyis med ye shes could certainly be translated more concisely as “profound, luminous nondual wisdom.” Yet that phrase, in English, suggests the terms “profound” and “luminous” to be adjectives describing nondual wisdom, which I do not believe to be Buddhajñānapāda’s intent. I have therefore opted for the more lengthy and awkward translation, “the wisdom of the nonduality of the profound and the luminous,” because I feel it reflects the nuance of Buddhajñānapāda’s understanding of the terms zab [mo] and gsal [ba] nominally rather than adjectivally. This is borne out both in his own clarifications on these terms (on which I elaborate below), and in Vaidyapāda’s gloss of the phrase, which parses the compound exactly as I have translated it: “the wisdom of the nonduality of the profound and the luminous” (zab mo dang gsal ba gnyis su med pa’i ye shes) (Sukusuma, D 94b.2- 3). 85 It is called genuine luminosity.11 A shorter but very similar presentation is found in a section of the Dvitīyakrama that elaborates on nondual wisdom as the nature of all pheonomena: [It] is totally free from all constructs The cause of excellence, difficult to fathom. [And yet] appears as the mahāmudrā, Whose light rays ripen oneself and others. |29| That is the supreme nondual nature...12 Vaidyapāda’s commentary explains that the first two of the lines cited here from the Dvitīyakrama express “the profound,” while the second two express “the luminous,” which, given the great similarity between this passage and the one from the Muktitilaka, is a compelling analysis.13 The references to “freedom from conceptuality” and the “purification of the mind” suggest that, just as specified in the Yogācāra presentation of nondual wisdom, Buddhajñānapāda intends a state free from the subject-object duality that involves the conceptual mind. In addition, the descriptions of “the profound” share much in common with exoteric Mahāyāna descriptions of the aspect of emptiness, or the ultimate nature: it is beyond conceptuality or constructs, beyond speech or thought, free from imputation, difficult to fathom. Perhaps drawing on the tathāgatagarbha literature, this profundity is additionally described as “stainless.” But the reference to “the profound” as the “source of everything” uses positive language emphasizing its generative aspect, which is much less typical of the deconstructive Madhyamaka rhetoric surrounding emptiness, like Nāgārjuna’s arguments in the Mūlamadhyamikakārikās that specifically refute origination by any means,14 and begins to give the description its tantric flavor. “Luminosity” as described in the Muktitilaka can be understood to correspond loosely with the aspect of “appearance” or the “relative” described above in the Madhyamaka analysis of nonduality. However, here luminosity is not the relative or apparent aspect of an outer object, but of the purified mind itself (beyond subject and object), similar to the Yogācāra concept of the “transformation of the basis” (āśrayaparāvṛtti), the notion that various aspects of consciousness are transformed into aspects of wisdom at the time of awakening. But it is with what follows that the description becomes undeniably tantric: the luminous aspect is the mind purified “in the form of the mahāmudrā,” a term that in the common 8th- and 9th-century usage refers to the form of the deity.15 Thus for Buddhajñānapāda “the profound” is the empty aspect of the mind—an 11 rnam par rtog pa kun bral bas// bsam brjod yul las shin tu ‘das// mkha’ bzhin dri med kun ‘byung ba (ba] D, bas P)// brtags bral zab mo nyid ces bya// phyag rgya chen po’i gzugs ‘chang ba// sgyu ma ‘ja’ tshon lta bu ru// rang dang gzhan gyi rgyud sbyong bas// yang dag gsal ba zhes bya’o// (Muktitilaka, D 47b.2-4). 12 brtag pa kun las rab dben pa// phun sum tshogs rgyu dpag dka’ ba// phyag rgya chen por rab snang ba’i// zer gyis rang gzhan smin byed pa// |29| de gnyis med pa’i rang bzhin mchog// (Dvitīyakrama, verse 29a-30a.) NB: In this and all subsequent citations from the Dvitīyakrama in the footnotes, I have cited the edited Tibetan from my own edition, and have not provided all of the variant readings. To see the details of the variant readings in these passages, please refer to the appropriate passage in the edition itself. 13 Sukusuma, D 95b.3-5. 14 Mūlamadhyamakakārikās, I.1. 15 In his commentary on the passage of the Dvitīyakrama just cited Vaidyapāda clearly identifies the mahāmudrā in terms of its usual 8th and 9th century usage: as the form of the deity “with a face, hands, and so forth.” phyag rgya chen por (por] P, po D) rab snang ba/ zhes te/ zhal dang phyag la sogs pa dang ldan pa ni phyag rgya chen po’o// (Sukusuma, D 95b.6; P 114b.7-8). In the Muktitilaka-vyākhyāna Vaidyapāda expresses more closely the relationship of “the profound” and “the luminous.” He states that “the luminous” refers to “Śāśvata (i.e. Vairocana) and the rest, who hold the form of the mahāmudrā, [and] who emerge from within that state [of the profound]...” (de’i ngang (ngang] P, dang D) las byung ba’i rtag pa (rtag pa] P, rtags D) la sogs pa’i phyag rgya chen po’i gzugs ‘chang ba/ (Muktitilaka-vyākhyāna D 49b.6; P 335a.1-2). (Vaidyapāda frequently uses the epithet Śāśvata (rtag pa) to refer to Vairocana.) 86 emptiness that is the source of everything; luminosity is the expression or manifestation of that emptiness in the “illusory and rainbow-like” form of the deity that has the capacity to “ripen oneself and others;”16 and nondual wisdom is a nonconceptual state that knows or is the nonduality of these two aspects. The Scope of Nondual Wisdom It is not only in terms of identifying “nondual wisdom” as the nonduality of the mind’s generative emptiness and its manifest expression as deity that Buddhajñānapāda departs from earlier usage of the term. He also broadens its scope significantly by declaring this nondual wisdom to be the very nature of the mind and of all phenomena. In both the scriptural and śāstric writings of the earlier Mahāyāna traditions that informed the thought of Buddhajñānapāda and his contemporaries, the term “nondual wisdom” was often employed to describe one among the causes for awakening, such as the mind-states of realized beings like bodhisattvas, and in occasional instances to describe the goal of the Buddhist path itself, the awakened state of a buddha.17 Like the term “the perfection of wisdom” (prajñāpāramitā), with which it was sometimes equated,18 the term nondual wisdom thus indicated either a state on the path leading to awakening or the final state of awakening itself. In the writings of so-called Yogācāra- Madhyamaka authors like Buddhajñānapāda’s guru Haribhadra and Haribhadra’s elder codisciple, Kamalaśīla, with which Buddhajñānapāda was surely familiar,19 nondual wisdom— there specified as the wisdom of subject-object nonduality—is described as an important aspect of wisdom on the path, but a state that must still be transcended. Both authors take care to avoid identifying a reified nondual wisdom with the ultimate result.20 While in certain more 16 These two aspects of the profound and the luminous parallel to what, in later traditions, is frequently termed “emptiness and clear light” (śūnyatā and prabhāsvara). 17 For example, in the Gaṇḍavyūha-sūtra the term “one whose experiential sphere is that of nondual wisdom” (advayajñānagocara) is one among a long list of adjectival compounds used to qualify the bodhisattvas (Gaṇḍavyūha-sūtra, 31). In the Akṣayamatinirdeśa-sūtra knowing the truths of suffering, the origin of suffering and the cessation of suffering, along with “entering into nondual wisdom” (gnyis su med pa’i ye shes su ‘jug pa) are what define the truth of the path that leads to the cessation of suffering; that is, engaging in nondual wisdom is described as a necessary cause of awakening (Akṣayamatinirdeśa-sūtra, D 191b.6-7). Bhāviveka identifies nondual wisdom as the mind that knows the ultimate, writing that “the ultimate is the experiential sphere of nondual wisdom” (don dam pa ni gnyis su med pa’i ye shes kyi spyod yul yin...) (Prañjāpradīpamūlamadhyamakavṛtti, D 240b.7). Sthiramati writes that on the first bodhisattva bhūmi when the bodhisattva realizes the all pervasiveness of the dharmadhātu he abandons subject-object grasping and “obtains nondual wisdom” (gnyis su med pa’i ye shes thob) (Sūtrālaṃkāravṛttibhāṣya, D 168b.5-6). Dignāga, in the first verse of his Prajñāpāramitāpiṇḍārtha, which is cited in Vilāsavajra’s Nāmamantrārthāvalokinī, identifies nondual wisdom with the state of awakening, writing that that “The perfection of wisdom is nondual wisdom (jñānaṃ advayaṃ); it [the perfection of wisdom] is [also] the Tathāgata.” (Prajñāpāramitāpiṇḍārtha, D 292b.1-2; Engle 2009, 456n831 cites the Sanskrit of the passage). 18 Such as in the quote from Dignāga cited in the previous note. Lopez describes how the term prajñāpāramitā was used by different authors to indicate both “the wisdom, possessed only by a Buddha, which has gone beyond all forms of saṃsāra and as the wisdom that goes beyond, which occurs on the path to enlightenment” (Lopez 1988, 24). 19 We know from his autobiography that Buddhajñānapāda studied under Haribhadra, and Haribhadra quotes and paraphrases from Kamalaśīla’s writings in his major Prajñāpāramitā work, the Abhisamāyālaṃkārālokā (Sparham 1989, 2). 20 This is generally the case with Mādhyamika authors who use the term, and specifically with Yogācāra- Mādhyamika authors with reference to nondual wisdom, understood as the wisdom that is free from subject-object duality. When he cites Dignāga’s statement mentioned in note 17 that equates the perfection of wisdom with nondual wisdom and the Tathāgata, Haribhadra carefully notes that the perfection of wisdom is the Buddha, which he equates with “the non-dual wisdom which is like an illusion” (Sparham 1989, 277; emphasis mine). Haribhadra also makes an objection to the position of holding the dharmadhātu, which he equates with pure nondual wisdom, to be the ultimate on the grounds that this would entail holding to, or settling upon, a dharma (Sparham 1989, 75). 87 philosophical passages of his writings Buddhajñānapāda likewise appears to uphold a similar Yogācāra-Madhyamaka stance,21 in other contexts he clearly places such a view in a subsidiary position, and prioritizes nondual wisdom, which he both relates specifically to the tantric path and characterizes with decidedly positive language. In the works of Buddhajñānapāda’s guru Vilāsavajra we see a more all-encompassing emphasis on nondual wisdom similar to that expressed in Buddhajñānapāda’s writings, and which may have served as an inspiration for Buddhajñānapāda’s own position. Vilāsavajra, in his Mañjuśrīnāmasaṃgīti commentary, the Nāmamantrārthāvalokinī, states nondual wisdom (advayajñāna) to be the main subject matter (abhidheya) of the Nāmasaṃgīti, and in fact nondual wisdom functions as the “conceptual anchor” of Vilāsavajra’s whole work.22 Moreover, he explains the “names” of the Nāmasaṃgīti to be both the names of all objects of experience, mundane and supramundane, as well as the names of Mañjuśrījñānasattva, whom he identifies with the nondual wisdom abiding in the heart of the tathāgatas.23 Thus, although he does not express it directly, Vilāsavajra essentially identifies all mundane and supramundane objects of experience with nondual wisdom. Buddhajñānapāda, in both the Dvitīyakrama and the Muktitilaka, states this explicitly: he declares nondual wisdom to be the nature of all phenomena and the nature of mind, “suchness”24 itself. In fact, this is the very first topic addressed in Mañjuśrī’s instructions to Buddhajñānapāda in the Dvitīyakrama. He explains that, The nature of phenomena, From form and the rest up to omniscience, Is the perfectly pure wisdom of the nonduality of the profound and the luminous, Which is like the center of space. |23| 25 The Dvitīyakrama clarifies further that “The identity of everything/ Is the very essence of one’s mind,/ Supreme among all things,”26 and that simply realizing this nature is the awakening of buddhahood itself. What is meant by “all phenomena” and “everything” is explored only briefly in the Dvitīyakrama, where it is described with the traditional all-encompassing phrase “all phenomena from form up to omniscience,” explicitly including the “three worlds” and the “great Kamalaśīla similarly insists that meditators must eventually go beyond holding nondual wisdom as existent, noting that “when the yogins abide in the cognition where [even] nondual cognition (advayajñāna) [itself] does not appear, then due to [their] abiding in the ultimate suchness, they see the Mahāyāna. [The seeing of] the Mahāyāna means just the seeing of the ultimate thusness” (Keira 2004, 76). 21 For example, in the Ātmasādhanāvatāra after showing, in a Yogācāra-style fashion, that both the perceiver and the perceived are mind alone, Buddhajñānapāda then makes a more Madhyamaka-style move, declaring that this very mind is neither one nor many, and thus emphasizing the emptiness of even that nondual mind. I explore Buddhajñānapāda’s apparent Yogācāra-Madhyamaka tendencies further below in the section on his prioritizing of tantra over non-tantric paths. 22 Tribe 2016, 44. Moreover, if Vilāsavajra is indeed also the author of the Spar khab commentary on the Guhyagarbha-tantra, a similar emphasis on nondual wisdom is found in that work, where, when commenting on the word guhya in the title, Wayman reports that the author describes the “utmost secret” to be “non-dual, selforiginated wisdom” (Tribe 2016, 31n36). 23 Tribe 2016, 50; 47. 24 The term “suchness,” here is my translation of both the Tibetan de kho na nyid and de bzhin nyid (as well as their shorter forms de kho na and de nyid) in Buddhajñānapāda’s writings—terms that presumably render the Sanskrit terms tattva and tathatā, respectively. Like in many places in Buddhist literature, these terms are used synonymously in Buddhajñānapāda’s work. In general, the term “suchness” is used in Buddhist literature to be synonymous with whatever a given work takes to be the ultimate nature of things. In Buddhajñānapāda’s case, that is nondual wisdom. 25 chos rnams gzugs la sogs pa rnams// kun mkhyen bar gyi rang bzhin ni// nam mkha’ dkyil ltar rnam dag pa’i// zab gsal gnyis med ye shes te// |23| (Dvitīyakrama, verse 23). 26 thams cad kun gyi bdag nyid de// dngos kun gtso bo rang sems kyi// ngo bo nyid de.... (Dvitīyakrama, verse 31). 88 elements,” and, in another traditional phrase used also by Vilāsavajra, as “[all things] animate and inanimate.” The Muktitilaka—also in the very first topic addressed in that work, directly following the author’s preliminary homage and pledge to compose the treatise—endeavors to make this same point but does so with much more detail as to what constitutes “all phenomena.” In the Muktitilaka Buddhajñānapāda does not merely give a few stock phrases meant to indicate the entirety of phenomena. Instead, as if to allay any possible doubts as to his intentions, he provides an extensive list of all it includes: he lists each of the traditional Abhidharmic Buddhist existences within the Desire, Form, and Formless realms, starting at the very bottom with the hell realms of the Desire Realm, continuing up through each of the Desire and then the Form realms, and culminating with the realm of Neither Perception nor Non-perception, the highest abode within the Formless Realm. Then he mentions the Five Paths that constitute the Mahāyāna journey to awakening. Given that the fifth path, which he here calls the Path of Uninterrupted Complete Liberation (bar mi chod rnam grol chen po’i lam), encompasses the result of awakening itself, Buddhajñānapāda has effectively provided an all-encompassing list of the categories of phenomena included within saṃsāra, nirvāṇa, and the path leading to awakening— the full scope of phenomena within the traditional Buddhist worldview.27 After providing this all-encompassing list, Buddhajñānapāda then declares that The nature of all of these Is the wisdom of the nonduality of the profound and the luminous, Which is perfectly pure like space, And free from all conceptual fabrication.28 In this way, Buddhajñānapāda goes beyond holding nondual wisdom to be simply a mind-state on the path or the state of awakening—although for him it is also both of these things. But beyond just this, Buddhajñānapāda identifies nondual wisdom as the very nature of all the phenomena of saṃsāra, nirvāṇa, and the path. Stepping Beyond Yogācāra In a passage frequently cited by many Yogācāra (and Yogācāra-Mādhyamika) authors— in fact Buddhajñānapāda himself cites the passage in his Ātmasādhanāvatāra—the Daśabhūmika-sūtra states, “These three worlds are mind alone.”29 Buddhajñānapāda’s assertion that the nature of all phenomena and the mind is nondual wisdom comes very close to the Yogācāra position that the three worlds (i.e. the whole of phenomenal existence) are mind, and that this very mind, when transformed through practice, becomes a wisdom transcending subjectobject duality, which has always been the mind’s true nature. However, in stating not just that the three worlds are mind, but that all the phenomena of saṃsāra, nirvāṇa and the path have the nature of nondual wisdom, Buddhajñānapāda at the very least goes rhetorically beyond the Yogācāra position, which is in itself not insignificant. What is more, by identifying the nondual wisdom that is the nature of all phenomena and the mind not just as a wisdom of subject-object nonduality (though it is also that), but as a wisdom of the nonduality of the mind’s generative emptiness and its manifest expression as deity, he makes a clear step beyond the Yogācāra position. In fact, with some of the language that Buddhajñānapāda uses to describe this ultimate nature or “suchness,” the nondual wisdom that is the nature of all phenomena and of the mind, he makes steps that could be construed as going beyond Buddhist doctrine entirely. Immediately 27 Muktitilaka, D 47a.4-7. 28 de dag kun gyi rang bzhin ni// nam mkha’ dbyings ltar rnam dag pa’i// spros pa’i rnam pa kun bral ba (ba] P, pa D)// zab gsal gnyis med yes shes so// (Muktitilaka, D 47a.6-7). 29 Daśabhūmikasūtra, D 220b. 89 following the two verses from the Muktitilaka on “the profound” and “the luminous” cited above, he writes, This supreme nondual nature, The self that pervades all things Is beyond the purview of saṃsāra— It is called the dharmadhātu.30 To be sure, we do not have the original Sanskrit for this passage, but the word “self” is generally not used lightly in Buddhist literature, and one might assume the translators would probably have chosen a Tibetan term besides bdag if they were not reading ātman in the Sanskrit.31 What is more, the very same phrase, the “self that pervades all things” (dngos po kun la khyab pa’i bdag), is used in the Dvitīyakrama to describe the “indestructible bindu” (mi shig pa’i thig le) of the perfection stage.32 If “self” (ātman) is indeed the term that Buddhajñānapāda used here, which does seem to be the case, it certainly suggests the influence of non-Buddhist ideas on his thought.33 As we will see below, this is not the only instance in his oeuvre where evidence of such influence appears. However, as regards the concept of self in particular, there is no question that Buddhajñānapāda maintains the traditional Buddhist denunciation of the false conception of a personal self. The Dvitīyakrama states, Thus, the final identity of all things Is profundity and luminosity, [But] since beginningless time ordinary beings Have fixated upon it as “me” and “mine,” Thus, without examining, they grasp to the self. |126|34 30 de gnyis med pa’i rang bzhin mchog// dngos po kun la khyab pa’i bdag// ‘khor bas rab tu ma zin pa// chos kyi dbyings zhes bshad pa’o (zhes bshad pa’o] D, V (D and P), kyi shes pa’o P) // (Muktitilaka, D 47b.4). 31 The translation of Vaidyapāda’s commentary uses the term bdag nyid rather than bdag in his paraphrase of the verse, but both bdag and bdag nyid seem to have been regularly used to translate the Sanskrit ātman (Muktitilakavyākhyāna, D 49b.7). For example, the translators of the Sarvabuddhasamāyoga-tantra consistently translated ātman as bdag nyid. 32 This itself [becomes] the precious jewel/ That produces the qualities of all buddhas,/ The self that pervades all things, / The great indestructible bindu.” de nyid sangs rgyas thams cad kyi// yon tan kun bskyed rin po che// dngos po kun la khyab pa’i bdag// mi shigs pa yi thig le che// |167| (Dvitīyakrama, verse 167). The term self (ātman) is also used in the name of Buddhajñānapāda’s unusually titled work, the Ātmasādhanāvatāra, literally “Entering into the Practice of the Self,” which is attested in Sanskrit in the extensive manuscript of Samantabhadra’s Sāramañjarī. Szántó has translated this title as “An Introduction into Accomplishment in the Body,” perhaps inspired by the fact that in that text Buddhajñānapāda denies the existence of a personal self and strongly advocates the tantric path, especially the practice of deity yoga (Szántó 2015b, 756). But I would suggest that we need to take Buddhajñānapāda’s use (and that of other tantric authors) of the term ātman seriously, especially given the resistance to its usage throughout much of earlier Buddhist literature. See also notes 31 and 33. 33 This is, however, far from the first or the only instance of the word “self” used in a positive sense in Buddhist tantric literature. For example, the term ātman occurs in multiple places in the Sarvabuddhasamāyoga-tantra, including its first verse, where, like in Buddhajñānapāda’s writings, it is used to describe a universal or all-pervasive self. In fact, the use of the term ātman in the Sarvabuddhasamāyoga-tantra may have inspired Buddhajñānapāda’s own use of the term; we know he was familiar with this tantra since he cites it in his Ātmasādhanāvatāra and incorporates several pādas from its first chapter into at least three different passages in the Dvitīyakrama. The use of the term ātman in tantric Buddhist literature at large is a topic deserving of further study. 34 de ‘dra’i dngos pos thams cad kyi// mtha’ yi de nyid zab gsal ba// thog med dus nas so so yi// skye bo nga dang ngar ‘dzin pas// ma brtags par ni bdag tu bzung// (Dvitīyakrama, verse 126). It is worth noting that the single reference to nondual wisdom in the Samājottara appears quite similar to Buddhajñānapāda’s use of the term to describe the nature of phenomena, and also specifically references the problem of ego-clinging in relation to that nature. The two pādas read, “Holding a sense of ‘I” with reference to the phenomena of nondual wisdom/ Is called “confusion.” advayajñānadharmeṣu ‘haṃkāro moha ucyate/ (Samājottara, 50ab). I was unable to make sense of Matsunaga’s reading here (advayajñānadharmerṣyā), and thus re-edited the line in accordance with one of 90 The subsequent verses go on to give a doxographical presentation of a variety of non-Buddhist and Buddhist systems of thought that spring from such an incorrect perception of reality, all of which are thus ultimately declared to be mistaken systems. In the verse cited above, though, Buddhajñānapāda references nondual profundity and luminosity as the ultimate identity of all things and asserts that it is this ultimate nature upon which beings mistakenly fixate (i.e. not (just) upon the five psychophysical aggregates, as is traditionally stated in Buddhist writings35), and which becomes the basis for their false idea of a self. Thus, for Buddhajñānapāda, nondual wisdom—the nonduality of profundity and luminosity—is the true nature of all that there is. Yet despite having this nondual wisdom as their ultimate identity, beings find themselves in the confused state of saṃsāric existence. The process by which such confusion with regard to this nature—the “fixat[ion] upon it as “me” and “mine”—unfolds is also explained in his writings. II. The World Arises out of Nondual Wisdom: A Gnostic Cosmogony The reality, which is like that, is present pervading all things. Yet, from beginningless time, even from this there was arising in the manner of the great thought. -Mañjuśrī instructing Buddhajñānapāda, Dvitīyakrama Cosmogony Narrated Another doctrinal position articulated at multiple places in Buddhajñānapāda’s writings is the idea that the illusory, dualistic phenomenal world not only has the nature of nondual wisdom, but that it emerges out of nondual wisdom; nondual wisdom is its source. In the Dvitīyakrama, Mañjuśrī’s presentation of nondual wisdom as the nature of all things is immediately followed by a description of this emergence. The process is presented in the form of a cosmogonic narrative. Vaidyapāda introduces these verses: “Now, in order to indicate the way in which saṃsāra arises out of nondual wisdom [the text] begins with The reality...”36 The reality which is like that Is present pervading all things. Yet,37 from beginningless time, even from this There was arising in the manner of the great thought. |35| And from that also the great elements [arose]: The maṇḍala of wind arose, And from that also, the great element of fire Arose and spread. |36| From that, the great element of water also Arose and spread, and from that also earth. From the essence of the gathering of the four [elements] Matsunaga’s Sanskrit manuscripts that reads advayajñānadharmeṣu, which corresponds with the Tibetan translation, and, according to Matsunaga’s notes, also with the Chinese and Bhattacarya’s edition. 35 Of course the five aggregates themselves, like all phenomena, have this same nondual nature in Buddhajñānapāda’s analysis. But again, the rhetorical difference between identifying the aggregates to be the basis of the false clinging to self and identifying the nondual nature to be that basis constitutes, I believe, an important distinction. 36 da ni gnyis su med pa’i ye shes las ‘khor ba ji ltar ‘byung ba bstan pa de ‘dra’i zhes pa la sogs pa’o// (Sukusuma, D 97a.2-3; P 116b.1-2). 37 Vaidyapāda explains further, “Although that kind of nonduality pervades and remains [as the nature of] all things, the reason that this is not apparent is explained with the lines beginning, Yet, from beginningless time...” de yang gnyis su med pa de lta bus dngos po kun rnam par khyab ste gnas kyang de mi gsal ba’i rgyu ni thog med dus nas zhes te/ (Sukusuma, D 97a. 4; P 116b.2). 91 Mountains, and so forth, and all sentient beings also [arose] |37| In all their variety: subtle and gross: Men, women, and hermaphrodites, The young and old, Gods, nāgas, and yakṣas, |38| Evil spirits, planets, Yāma, The Lord of Water, Indra, hell beings, Pretas, animals, and those who abandon all of this,38 Beings who rely upon consciousness alone,39 Such beings remain, spread far and wide.40 |39| Therefore, the nondual nonconceptuality That is higher than that is completely obscured. Because of not realizing it, all beings Cycle around in saṃsāra. |40|41 This narrative, which is paralleled in the Muktitilaka,42 proceeds from declaring reality to be nothing other than the nondual wisdom that “pervade[s] all things,” precisely as it had been described in the passages discussed above, and then sets forth a cosmogonic narrative of the world’s emergence out of nondual wisdom that is timeless, both internal and external, and therefore simultaneously microcosmic and macrocosmic, personal and universal. First, the arising of conceptuality (and therefore duality) out of “nondual nonconceptuality” is declared to be timeless: it has been happening “from beginningless time,” a phrase which allows the narrative to be both historical as well as applicable to the present moment. That is, following Buddhajñānapāda’s presentation, it can be understood that beings continually participate in this devolution, as it were, from the innate state of nondual nonconceptuality to a state in which there is “arising in the manner of the great thought.” This somewhat opaque phrase—“arising in the manner of the great thought”—receives clarification from Vaidyapāda: “From that time, just as clouds arise within space, “the great thought,” the 38 Vaidyapāda identifies these as the śrāvakas and so forth. de kun spangs pa ni nyan thos la sogs pa’o// (Sukusuma, D 97b.1; P 117a.1-2). 39 Vaidyapāda identifies these as those beings of the realm of Limitless Space, and so forth—inhabitants of the Formless Realm—since they have abandoned form. nam mkha’ tha’ yas la sogs pa ste/ gzugs spangs pa’i phyir ro// (Sukusuma, D 97b.1-2; P 117a.2). 40 Vaidyapāda comments that the statement that these beings live far and wide means that, “having been produced by conceptuality, they appear in the ten directions.” de kun rgyas par gnas zhes pa ni rtog pas bzo byas nas phyogs bcu kun du snang ba’i phyir ro// (Sukusuma, D 97b.2; P 117a.2-3). 41 de ‘dra’i don des dngos po kun// rnam par khyab ste rnam gnas kyang// thog med dus nas de las kyang// rnam rtog chen po tshul byung ste// |35| de las yang ni ‘byung ba che// rlung gi dkyil ‘khor nyid byung ste// de las kyang ni me yi khams// chen po byung nas khyab mdzad de// |36| de las chu khams chen po yang// byung ste khyab mdzad de las kyang// sa byung bzhi bsdus ngo bo las// ri sogs sems can thams cad kyang// |37| sna tshogs phra ba sbom po dang// skyes pa bud med ma ning dang// gzhon nu dang ni rgan po dang// lha dang klu dang gnod sbyin dang// |38| gdon dang skar ma gshin rje dang/ chu bdag rgya byin dmyal ba dang/ yi dags dud ‘gro dang de kun/ spang dang shes tsam rab brten pa’i/ / ‘gro ba kun du rgyas par gnas// |39| de bas de yi gong ma yi// gnyis med rtog bral rab bsgribs te// ma rtogs pas na ‘gro ba kun// ‘khor bar rab tu ‘khor bar ‘gyur// |40| (Dvitīyakrama, verses 35-40). 42 The Muktitilaka’s narrative is much shorter, a single verse. There reality is described not with the term nondual wisdom, but as a “maṇḍala of self-awareness.” “From such a maṇḍala of self-awareness/ Emerged concept, and from that wind, and then fire,/ Water, and earth, and from these/ Mistaken [things like] the aggregates, and so forth were established,/ Thus obscuring this maṇḍala.” de ‘dra’i rang rig dkyil ‘khor las// rtog las rlung byung me dang ni// chu dang sa ste de dag las// phyin (phyin] P, phyir D) log phung sogs rab grub pas// de’dra’i dkyil ‘khor bsgribs par ‘gyur// (Muktitilaka, D 50b.4-5). 92 mind alone, arises in a manner [in which it appears] as if it were endowed with conceptuality.”43 At least in Vaidyapāda’s reading of Buddhajñānapāda’s text, the mind arises from nondual wisdom in a way that only makes it appear as if conceptuality is present, with the implication that conceptuality itself is just an illusory manifestation of what is in fact still nondual wisdom. We will revisit this point below. Up until this point in the narrative, the process of devolution from nondual wisdom is an internal, microcosmic, or personal one, concerned with the arising of the conceptual mind from the state of nondual wisdom. But then, perhaps within the context of the Yogācāra assertion that “the three worlds are mind alone,” the narrative goes on to describe the arising of the external world from this mind: from “the great thought” the four elements are produced—first wind, then fire, then water, then earth—and from their combination gross objects like mountains and sentient beings. All of this, it is then said, obscures the “nondual nonconceptuality” that is its true nature. This narrative effectively describes a cosmogonic process by which both sentient beings and their world emerge—or devolve—from the true nature of nondual wisdom into a confused and dualistic state of saṃsāric existence. The remainder of Mañjuśrī’s instructions to Buddhajñānapāda in the Dvitīyakrama focus on methods for reversing this process in order to bring about awakening, which is at times defined as the simple recognition of nondual wisdom as the nature of the mind and all phenomena,44 but is also described as occurring via a process that is precisely the opposite of the devolution that brought about saṃsāra in the first place.45 A “Syllabic Cosmogony” According to Vaidyapāda’s interpretation of the Dvitīyakrama, this same narrative was already expressed earlier in the text, in the very first syllables that Mañjuśrī uttered to Buddhajñānapāda following the latter’s request for teachings: “a vi yaṃ raṃ vaṃ laṃ hūṃ a la la la ho.” Vaidyapāda interprets these syllables as constituting in brief the entirety of Mañjuśrī’s instructions in the Dvitīyakrama. He reads the syllables a vi yaṃ raṃ vaṃ laṃ hūṃ as referring to the cosmogonic process just described above, and a la la la ho as referring to the path that leads to awakening, in particular the path of the perfection stage.46 Vaidyapāda’s analysis of the 43 de’i dus nas nam mkha’ la sprin ‘byung pa bzhin du rnam rtog chen po zhes te sems tsam rtog (rtog] P, rtogs D) pa dang ‘brel pa lta bu’i tshul du byung ste zhes bya ‘o// (Sukusuma, D 97a.4; P 116b.3-4). Vaidyapāda’s commentary on the parallel passage in the Muktitilaka uses the same metaphor of clouds emerging within space to describe the way that conceptuality arises out of nondual wisdom (Muktitilaka-vyākhyāna, D 59a.7-59b.1). 44 “When that is realized, this is the awakening of buddhahood.” …de rtogs na// sangs rgyas byang chub de nyid do// (Dvitīyakrama, verse 31d-32a). 45 The process paralleling that of death that is intentionally re-enacted as the culmination of the perfection stage is described as the gradual dissolving of the great elements back into one another: “And the mind should be placed upon the bindu./ When one’s faculty is held there,/ The earth maṇḍala enters into water,/ That water likewise enters into fire,/ |266| The fire then enters into wind,/ And the wind enters into mind./ As an indication that the mind has to some degree/ Entered nondual wisdom/ |267| There are five signs that will appear..../” sems ni thig ler gzhag par bya// rang gi dbang po der bzung nas// sa yi dkyil ‘khor chu la zhugs// chu de me la de bzhin zhugs// |266| me yang rlung la rab tu zhugs// rlung yang sems la zhugs par gyur// sems ni gnyis med ye shes su// cung zad zhugs pas rtags gnas pa// |267| (Dvitīyakrama, verses 266-67). I discuss this point further in Chapter Six. 46 Sukusuma, D 93a.5-94a.4. An interesting feature of Vaidyapāda’s commentary on this section is that throughout the description of the process of devolution he makes statements indicating that each of the aspects of the world that emerges in this process is nothing other than the essential nature itself, and that simply coming to realize this—at any point in the process—constitutes awakening. For instance, after the description of the syllable vi, the first “mere knowing” that appeared from the essence of the nondual state (prior to the emergence of the elements) he remarks, “Some, starting only from there, come to realization,” which suggests that in that case the subsequent processes would not unfold (kha cig de ‘ba’ zhig las brtsams te rtogs par byed de) (Sukusuma, D 93a.7). Such an analysis has strong parallels in Great Perfection narratives of cosmogonic emergence. I discuss some parallels and differences between Buddhajñānapāda’s cosmogonic narrative and those found in Great Perfection works in Chapter Four. 93 syllables a vi yaṃ raṃ vaṃ laṃ hūṃ as indicating the same cosmogonic process described in verse the Dvitīyakrama is particularly compelling given the way that these particular Sanskrit syllables are commonly employed and understood in tantric Buddhist literature. He explains the first two syllables as follows: A is the nature of all things because they are unarisen. It is said that “a is the gateway to all phenomena.” If we examine that statement, [we can understand that a is] the gateway through which all [phenomena] emerge. Moreover, it should be known as the nature of the nonduality of the profound and the luminous, which is like the maṇḍala of space— not arisen from any sort of conceptual imputations, [but] primordially and spontaneously present. That which appeared from its essence as mere knowing is vi, the first named syllable, which is called “awareness.”47 The syllable a is the first syllable of the Sanskrit alphabet and the vowel inherent in all Sanskrit consonants, and thus began to be linked in tantric Buddhist texts with the idea of source or origin.48 Here that syllable is identified with the “nonduality of the profound and the luminous,” the source of all phenomena according to Buddhajñānapāda’s system. The next step in the cosmogonic process, the devolution from this “primordially and spontaneously present” state that is “free from any sort of conceptual imputations,” into a state of conceptuality is expressed here with the syllable vi, described by Vaidyapāda as “that which appeared from its essence as mere knowing (shes pa tsam).” Though Vaidyapāda does not comment upon this point, the syllable vi functions as a Sanskrit prefix that is commonly used to express distinction or division—in short, duality. Vi here represents the first moment of knowing, which involves dualistic conceptual distinctions. Thus, the Sanskrit syllables a vi, the first two pronounced by Mañjuśrī in the Dvitīyakrama, do indeed succinctly but clearly symbolically express the process of duality emerging from a nondual state. The subsequent syllables yaṃ, raṃ, vaṃ, and laṃ are used throughout tantric Buddhist literature, including at a later point in the Dvitīyakrama itself, to symbolize the great elements, and they appear here in precisely the order in which their (d)evolution is described in the Dvitīyakrama’s verses of cosmogonic narrative: yaṃ represents wind, raṃ fire, vaṃ water, and laṃ earth.49 Vaidyapāda’s commentary describes the process by which each of these elements unfolds.50 Finally, Vaidyapāda explains that hūṃ is the seedsyllable of the five aggregates, which emerge through the coming together of the four great elements and mind, the final parallel between this “syllabic cosmogony” and the cosmogonic 47 de la a zhes pa ni dngos po thams cad kyi rang bzhin te/ ma skyes pa’i phyir a ni chos thams cad kyi sgo’o zhes pa’i gsung la dpyad na/ thams cad ‘byung ba’i sgo ste/ de yang gsal zab gnyis su med pa’i rang bzhin nam mkha’ dkyil ‘khor lta bu brtags pa thams cad kyis ma skyes pa dang po nas lhun gyis grub pa nyid du shes par bya ba’o// de yi ngo bo las shes pa tsam lta bur snang ba ni bi ste/ rig pa zhes pa’i ming gi yi ge dang po’o// (Sukusuma, D 93a.5-7; P 111b.7-112a.2). 48 Though this function of the syllable a to symbolize origin is found in tantric Buddhist literature more widely, it is specifically employed in this way in Vilāsavajra’s Nāmamantrārthāvalokinī, which we know to have influenced Buddhajñānapāda’s work, as he reproduces quite a few passages from it in his Ātmasādhanāvatāra (See Tribe 2017, 62-66 on the use of the syllable a in the Nāmamantrārthāvalokinī). Tribe points to the value of the syllable a in the earlier Prajñāpāramitā literature as a symbol of ultimate truth as expressed through negation, given the syllable a’s function as a negating prefix in Sanskrit. He also notes that the Nāmamantrārthāvalokinī specifically emphasizes the syllable a’s function as source or origin, as the first syllable in the Sanskrit alphabet (Tribe 2017, 65). The Nāmasaṃgīti itself announces the syllable a to be “the foremost of all syllables, the great good, the supreme sound” and indicates that Mañjuśrījñānasattva himself is “born from the syllable a” (Tribe 2017, 404). 49 Slight variants of these same syllables (ślāṃ, ṣvāṃ, hyāṃ etc.) are used to represent the great elements in the bhūtaśuddhi practices in some Śaiva texts, as well, so it appears that their use is common to multiple tantric systems (See Flood 2002, 30). 50 Sukusuma, D 93b.1-3. 94 narrative elaborated in verse later in the text.51 Thus the initial syllables of Mañjuśrī’s teaching in the Dvitīyakrama are cogently explained to convey precisely this same cosmogony.52 Cosmogony Enacted While I am unaware of other narrative presentations in Indian tantric Buddhist literature of the particular cosmogonic account presented in the Dvitīyakrama, aspects of the account do appear as a standard feature of quite a number of tantric ritual texts, specifically sādhanas pertaining to the generation stage. In generation stage sādhana both the practitioner’s world and personal identity are recreated or generated through meditative visualization via a process explained in commentarial literature as mimicking the features of saṃsāra, but reimagined in a purified form. The process of producing such a pure universe begins with a recollection of the emptiness that is its nature, and which functions in the ritual as its source. This recollection is accomplished variously in different generation stage sādhanas. In Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana it is done in two ways: first by the recollection of the socalled “three gates of liberation,” by means of which the practitioner calls to mind the fact that phenomena are emptiness “because of lacking essence” (ngo bo nyid dang ‘bral phyir), without characteristics “because they are naturally causeless” (rang bzhin rgyud dang bral bas), and beyond aspiration “because they are free from conceptualization” (rtog pa rnams dang bral phyir).53 This recollection is followed by a second affirmation, not just of emptiness, but of the empty wisdom that is the practitioner’s nature: she recites the mantra oṃ śunyatāvajrajñānasvabhāvātmako ‘ham, “oṃ, I have the nature of emptiness, vajra wisdom.”54 In what has become a standard feature of generation stage sādhanas, shortly after this point in the ritual, during the process of generating the pure universe that becomes a support for the celestial palace of the practitioner-as-deity, the great elements are gradually produced precisely in the very same order in which they emerge in the Dvitīyakrama’s cosmogonic narrative. This syllabic generation of the elements is not found in the Caturaṅga/Samantabhadra-sādhana (nor his three very short Jambhala sādhanas), but we have reason to believe Buddhajñānapāda was familiar with the process (which I discuss below); indeed it may have formed a part of the visualization that the practitioner was instructed to follow, although not mentioned directly in the text.55 In this process the syllable yaṃ is used first to generate a maṇḍala of wind, raṃ to generate fire, vaṃ to generate water, and finally laṃ to produce earth, as the practitioner 51 ibid., D 93b.2-3. 52 As mentioned above, the remainder of Mañjuśrī’s initial syllabic address, “a la la la ho,” is taken by Vaidyapāda to refer to the method through which one comes to realization, perfection stage practice with a consort (indicated by a), and the realization that emerges from that practice (indicated by la la la ho) (Sukusuma, D 93b.5-94a.4; P 112b.1-113a.1). Indeed, we find the set of syllables a la la ho (sometimes with just two repetitions of la, other times with three) in other texts as well, often corresponding to the emission and offering of bodhicitta in perfection stage practices. See, for example, PT841, 2v.2-3: byang chub kyi sems babs na/ a la la ho zhes brjod de/ lha mo mnyes par bsam…; Chapter Eight of the Thabs kyi zhags pa (ITJ321, 30r.2-3): byang chub kyi sems bab na a la la la ho zhes dkyil 'khor thams cad la dgyes par mchod nas...; Chapter 15 of the Guhyagarbha-tantra may likewise show a connection between the syllables a la la ho and the emission of bodhicitta (D 126b.7-127a.1) de dag kun dgyes par rol pa’i dkyil gyi ‘khor sprin ‘byung ba zhes bya ba’i ting nge ‘dzin la snyoms par zhugs nas sku dang gsung dang thugs rdo rje las ‘di dag phyung ngo// oṃ āḥ (āḥ] sugg. em., āṃ D) hūṃ vajra pra be sha a la la hoḥ zhes brjod pas ‘byung mo’i rgyal mo rnams shin tu chags pa’i yid g.yos nas padma’i dkyil ‘khor sdud cing rgyas par gyur nas…. Thanks to Jacob Dalton for bringing these additional references to the syllables a la la ho to my attention. 53 Samantabhadra-sādhana, verse 18, D 29b.4 (= Caturaṅga-sādhana, D 36b.7-37a.1). 54 ibid., D 29b.4 (= Caturaṅga-sādhana, D 37a.1). 55 In fact, it is quite common in modern Tibetan generation stage sādhanas for this process to go unmentioned in the sādhana itself but still be indicated by the commentaries as part of the visualization for practicing the sādhana. 95 meditatively brings forth the deity’s pure world. The syllable hūṃ can have a number of functions, but one of them is as a generative syllable, or “seed-syllable” that symbolizes the awakened mind of all buddhas; in generation stage sādhana such a seed-syllable is used to bring forth the form of the deity.56 In this way, then, the standard process of generation stage sādhana practice involves a ritual re-enacting of the very same cosmogonic process narrated in the Dvitīyakrama. This gradual generation of the great elements does appear in Śākyamitra’s Mukhāgama, which, Śākyamitra tells us, was based on Buddhajñānapāda’s oral instructions. The Mukhāgama is essentially a rewriting of the Caturaṅga/Samantabhadra-sādhana with some additional features, including details on generating the protection circle that are absent in the Caturaṅga/Samantabhadra; it is in that section of the Mukhāgama that the gradual generation of the elements appears in its standard progression, precisely as described above.57 The standard process of the unfolding of the elements is also found in other generation stage sādhanas from around this period or slightly later, including the Ārya School’s Piṇḍīkṛta-sādhana. It does not, however, appear in the sādhana-like sequence in chapter four of Vilāsavajra’s commentary on the Mañjuśrīnāmasaṃgīti.58 The absence of this sequence in Vilāsavajra’s work, and its presence in Śākyamitra’s, which was supposedly based on Buddhajñānapāda’s oral instructions, along with its presence in sādhanas from Dunhuang,59 and in the slightly later Piṇḍīkṛta-sādhana, and its proliferation in later sādhanas suggests that as a generation stage ritual procedure, the visualized gradual unfolding of the elements may have been a relatively recent development in Buddhajñānapāda’s time, but one of which he was almost certainly aware. In its narrative describing the same cosmogonic process, the Dvitīyakrama even uses precisely the language found in many generation stage sādhanas to describe the production of the elements: the arising of a “maṇḍala” of wind, fire, water, and earth. It seems, then, that the passages from the Dvitīyakrama and the Muktitilaka put into narrative form precisely the same cosmogonic sequence that was beginning to be employed in generation stage ritual practice around this same period. While its ritual form became a standard feature of generation stage sādhana, however, the corresponding narrative presentation does not seem to have become such a popular feature of later tantric literature. Given the relative proliferation of this cosmogonic sequence in generation stage sādhanas and its singular narrative representation in Buddhajñānapāda’s writings, we can suspect that in the Dvitīyakrama and the Muktitilaka Buddhajñānapāda was likely articulating in narrative form—in effect, doctrinalizing—a more widely known ritual sequence. 56 The particular seed-syllable or generative syllable used varies depending on the deity, but as the general seedsyllable of awakened mind, hūṃ can function as representative of any and all such seed-syllables that would give rise to the deity. Vaidyapāda’s commentary on the “syllabic cosmogony” in the Dvitīyakrama describes hūṃ as representative of the five aggregates, but also explains that the five buddhas are the purified form of those aggregates, thus linking the cosmogonic devolution with the possibility of its purification through recognizing or recreating it in a pure form (Sukusuma, D 93b.3-4). 57 Vaidyapāda, in his commentary on the Caturaṅga/Samantabhadra-sādhana also notes that there are more details to the protection circle than found in the Caturaṅga/Samantabhadra-sādhana itself. Rather than adding those details in his commentary, though, Vaidyapāda states that they should be learned from the oral tradition (de’i cho ga ni man ngag gis shes par bya’o) (Samantabhadrī-ṭīkā, D 139b.3). Other features of Śākyamitra’s Mukhāgama that are absent from the Caturaṅga/Samantabhadra-sādhana do appear in Vaidyapāda’s Samantabhadrī-ṭīkā, suggesting that both Vaidyapāda and Śākyamitra were indeed basing their respective works on oral explanations of the practice, even if they differed (which they apparently did!) on what points should or should not be committed to writing. 58 Tribe (2017, 55-60) gives a brief outline of the sādhana-like ritual from the Nāmamantrārthāvalokinī. The same volume also includes Tribe’s Sanskrit edition and English translation of the first five chapters of that work. The sādhana material occurs primarily in Chapter Four. 59 e.g. ITJ 331/2. Thanks to Jacob Dalton for bringing this to my attention. 96 A Buddhist Cosmogony? Since the Buddhist tradition holds that there is no omnipotent creator God, Buddhist literature on the whole displays a dearth of cosmogonic narratives when compared to theistic traditions. The standard Buddhist cosmology, outlined in classical works like the Abhidharmakośa and Abhidharmasammucaya, essentially follows that of the broader Indic worldview of the time. It seems that elements of the cosmogony narrated in Buddhajñānapāda’s works and ritually enacted in generation stage sādhanas likewise owe a debt to systems outside of the Buddhist fold: the cosmogonic narrative outlined in the Dvitīyakrama bears a striking, if significantly abbreviated, resemblance to the unfolding of the tattvas of the Sāṃkhya tradition, which were later adopted and adapted into Śaiva and Vaiṣṇava traditions, as well.60 Like Buddhism, Sāṃkhya is a non-theistic tradition, and thus an expedient source for a nontheistic cosmogony that could be adapted to fit a Buddhist narrative. The Classical Sāṃkhya61 system is a decidedly dualist one, and outlines twenty-five tattvas that comprise the entirety of existence.62 Sāṃkhya upholds the hard duality of puruṣa, the individual, self,63 or consciousness, and prakṛti, the primordial substance, which constitute two among the tattvas.64 The remainder of the twentyfive tattvas are manifestations that unfold from prakṛti, but crucially they all share the feature of appearing to be what they are not: all of the remaining tattvas appear in their various guises, and yet they are actually nothing other than transformations of prakṛti, the primordial substance itself.65 Of the twenty-five tattvas, only six correspond to categories in Buddhajñānapāda’s cosmogonic narrative, but for those that do correspond the parallels are close. All of these correspondences fall upon the prakṛti side of the Sāṃkhya duality divide; the puruṣa, corresponding to the self, is notably absent. The first tattva66 of the Classical Sāṃkhya system is the primordial substance, prakṛti, the source of all manifestations that appear to be other than it, but are actually not. This “mūlaprakṛti,” the “fundamental primordial substance,”67 is described in the Sāṃkhyakārikās as uncreated;68 it functions as a type of “ultimate first principle.”69 In Buddhajñānapāda’s cosmogonic narrative, this corresponds with nondual wisdom, which he describes as “the supreme nondual nature” and “the self that pervades all things.”70 Expressed in the 60 I’m grateful to Ryan Damron for a number of conversations on this topic that turned my attention to parallels with the Sāṃkhya system. See Goodall 2016 on the adoption of additional tattvas into Śaiva tantra. 61 The Sāṃkhya system underwent a series of developments from early systems to the so-called “Classical Sāṃkhya” of the Sāṃkhyakārikās, composed by Īśvarakṛṣṇa around the fifth century CE (Kent 1982, 259). The system I am comparing with Buddhajñānapāda’s cosmogonic narrative is the Classical Sāmkhya system. 62 Larson’s Classical Sāṃkhya gives a good overview of the tattvas, including a very clear chart outlining their progression (Larson 1969, 236). 63 The puruṣa is the “self” in an individual but not a personal sense; the aspect of personal ego appears as the fourth tattva, ahaṃkāra (ibid. 1969, 170-71). 64 ibid. 1969, 172. 65 ibid. 1969, 173-4. The manifest world—beginning right from the third tattva, buddhi—is just an emergence from or evolution of prakṛti, described as a “transformation or modification of itself” (ibid. 1969,176-7). 66 In his earlier work Larson (1969) numbers prakṛti as the second tattva, while puruṣa is stated to be the first. In a later article (Larson 1983) in which he considers the tattvas in terms of their numerical value, he assigns prakṛti as number one, and places puruṣa as number twenty-five. This makes more sense given mūlaprakṛti’s function as the primordial or first substance, so I have followed that schema here. In any case, the crucial aspect is that puruṣa stands on the opposite side of a dualistic system in contrast to prakṛti and all of the remaining tattvas, which remains the case whichever way the numbering is done. 67 The term mūlaprakṛṭi is specific to the Classical Sāṃkhya system and is used in place of term avyakta, used for the second tattva in earlier Sāṃkhya systems (See Larson 1969, 160). 68 ibid. 1969, 160. 69 ibid., 161. 70 Muktitilaka, D 47b.4. 97 Dvitīyakrama’s “syllabic cosmogony” as the syllable a, Vaidyapāda comments that it is “the nature of all things because they are unarisen (i.e. uncreated).”71 The next Sāṃkhya tattva, the first to emerge from the prakṛti as its manifestation, is buddhi, the “mind,” or “intellect.” The Sāṃkhyakārikās describe buddhi as characterized by ascertainment or determination.72 In the Sāṃkhyakārikabhāṣya, two of the synonyms given for buddhi are “great” (mahat), and “thought” (mati).73 This tattva corresponds in the Dvitīyakrama’s narrative with the “great thought” (rnam tog chen po), in the Muktitilaka’s briefer narrative called simply “thought” (rtog), which is the first aspect of duality to emerge from the nondual nature. As noted above, Vaidyapāda’s commentary makes it clear that “the great thought, the mind alone, arises in a manner [in which it appears] as if it were endowed with conceptuality”74 (emphasis mine), thus indicating that this manifestation is actually nothing other than the nondual wisdom from which it arose. This is parallel to the Sāṃkhya view that the manifestations of prakṛti only appear to be something else, while in fact they are transformations of the prakṛti itself. The parallels between Buddhajñānapāda’s cosmogonic narrative and the unfolding of the tattvas now fall away, as the Sāṃkhya system continues to set forth an extensive series of tattvas that follow from buddhi.75 However, the parallels pick up again in the final four of the tattvas that emerge as manifestations of prakṛti: wind, fire, water, and earth, which parallel perfectly the order of the emergence of the great elements from the “great thought” in Buddhajñānapāda’s narrative. What is more, this is not the only appearance of Sāṃkhya ideas in Buddhajñānapāda’s writings. In the section of the Dvitīyakrama outlining perfection stage practices, when describing the descent of the so-called “indestructible bindu” downwards from the practitioner’s heart, the text specifies that it descends “in the form of rajas, tamas, and sattva,”76 which are the names of the three gūṇas, or characteristics, outlined in the Sāṃkhya system. This is particularly significant in light of the observation made above about the correspondence between nondual wisdom in Buddhajñānapāda’s system and prakṛti in the Sāṃkhya. The indestructible bindu is said in the Dvitīyakrama to be the “relative form” of nondual wisdom (a point that is explored further in the section below), and the three gūṇas are, in the Sāṃkhya system, the constituent aspects of prakṛti.77 The statement on the descent of the bindu “in the form of rajas, tamas, and sattva,” then, further strengthens the correspondence between nondual wisdom—of which the 71 Sukusuma, D 93a.5-6. 72 Larson 1969, 181. 73 ibid. 74 rnam rtog chen po zhes te sems tsam rtog (rtog] P, rtogs D) pa dang ‘brel pa lta bu’i tshul du byung ste zhes bya ‘o// (Sukusuma, D 97a.4; P 116b.3-4). 75 These are: ahaṃkāra (ego), manas (mind), the five buddhīndriyas (sense faculties; hearing, feeling, seeing, tasting, smelling), the five karmendriyas (faculties of action; speaking, grasping, walking, excreting, generating), the five tanmātras (subtle elements; sound, touch, form, taste, smell), and finally the five mahābhūtas (gross elements; space, wind, fire, water, earth) (Larson 1969, 236). Perhaps it seems odd that the correspondences with the tattvas would break off precisely at the point that seems to hold the greatest similarity to Buddhist thought: with the tattva of ahaṃkāra, ego. But we have to remember the ritual function of this cosmogonic sequence in Buddhist tantric practice: the cosmogony is re-enacted in the generation stage as a way to purify the mistaken process of emergence into saṃsāra. Ahaṃkāra is innately negative in the Indic worldview, and does not have a pure form (apart from just not doing it!). What is more, in sādhana practice the generation of the mind, sense faculties, capacities for activity, and so forth, occur with respect to the practitioner-as-deity, not with respect to the practitioner’s ordinary identity. In the generation stage this process of the generation of the deity happens only after the production of the outer world, made up of the elements, which form the support for the deity’s celestial palace, at the center of which the deity is generated. 76 rdul dang mun pa dang// snying stobs tshul du....babs// (Dvitīyakrama, verse 188d-89a). 77 For our purposes here it is enough to simply note that the three guṇas are understood to constitute prakṛti. For more details on the nature and character of these three and their function in the Classical Sāṃkhya system see Larson 1969, esp. pp. 162-67. 98 bindu is the “relative form”—and the idea of prakṛti. While the correspondences between the two systems are certainly far from comprehensive, the aspects that do correspond are close enough that it does appear that the cosmogony described in Buddhajñānapāda’s writings, and thus also the enactment of that cosmogony in generation stage sādhana practice, have roots in the Sāṃkhya system of the tattvas. The “Relative Form” of Nondual Wisdom as Cosmic Source Not only does Buddhajñānapāda declare nondual wisdom to be the source of the phenomenal world, he also makes the striking statement that even its relative form, the indestructible bindu, acts as that source. This claim is found in yet another description in the Dvitīyakrama of the emergence of the world, in addition to the cosmogonic narrative and “syllabic cosmogony” found at the beginning of the text. This third narrative, however, specifically references the world’s re-emergence after its destruction at the end of a cosmic aeon. The idea of cyclic time and cosmological cycles of arising and destruction is part of the wider Indic worldview and, like other features of the Indic cosmology, was adopted into Buddhist cosmology from an early date. This passage from the Dvitīyakrama, which immediately follows instructions on the three types of perfection stage bindu meditation, explains that the cosmos reemerges not just form nondual wisdom, but from its “relative form.”78 Although nondual wisdom itself Has taken on a relative form, Even when the inanimate, and so forth, along with the animate, Brahmā and the others, the gods, asuras, and the rest |242| Completely disappear, That bindu will not cease. Everything animate and inanimate79 Will again be made to emerge from it. |243| But because they do not realize What is genuine, Beings are confused, [believing everything] to be arisen from Brahmā’s egg. That [bindu], which cannot be moved by any phenomena, |244| Which cannot be destroyed by anything at all, As long as it remains embodied Brings about [engagement in virtuous] activity and non[-virtuous] acts.80 Therefore81 the meditation on the indestructible bindu, Stable and beyond destruction, is explained. |245|82 The “relative form” (kun rdzob gzugs) assumed by nondual wisdom is identified by Vaidyapāda as the so-called “indestructible bindu” (mi shig pa’i thig le), described earlier in the Dvitīyakrama as a sphere about the size of a chickpea, radiant with five-colored light at the 78 I present further details of Buddhajñānapāda’s perfection stage system in Chapter Six. 79 rgyu] sugg. em. based on V (D and P), rgyur D C S P N 80 Here I rely upon Vaidyapāda’s commentary, which notes that abandoning killing is an example of “activity” and killing is an example of “non-acts” (Sukusuma, D 122a.2-3). 81 pas] sugg. em. based on V (D and P), pa D C S P N 82 da ni gnyis med ye shes nyid// kun rdzob gzugs la brten nas kyang// mi g.yo la sogs g.yo dang bcas// tshangs sogs lha dang lha min sogs// thams cad rab tu mi snang yang// thig le ‘gog par mi ‘gyur te// |242| rgyu dang mi rgyu bcas pa kun// slar yang de las ‘byung bar byed// de las skye bo rnams kyis ni// yang dag ma rtogs pa yi phyir// |243| tshangs pa’i sgo nga las skyes par// ‘khrul pas chos rnams kyis mi bsgul// gang gis kyang ni mi shigs pa// ji srid ‘di ni lus gnas pas// |244| las dang las min rab byed pas// brtan po ‘jigs pa rab spangs pa’i// mi shigs thig le bsgoms par bshad// |245| (Dvitīyakrama, verses 242-45) 99 center of the practitioner’s heart.83 This bindu and others are visualized and manipulated during the perfection stage practices outlined in the Dvitīyakrama and the Muktitilaka. But the indestructible bindu is here described not just as a feature of the subtle body connected to perfection stage practices, but as the very source from which the entirety of the animate and inanimate cosmos re-emerges after its destruction at the end of an aeon;84 in the face of cosmic destruction, it alone does not cease. Vaidyapāda explains this to be the case because “it has the nature of vajra wisdom, so fire and the rest cannot destroy it.”85 The use of the term “relative” (kun rdzob) to describe the form of nondual wisdom as a bindu of light makes it clear that this form is not nondual wisdom’s ultimate (don dam) nature, which is presumably beyond form; wisdom simply assumes (brten, literally “relies upon”) this form on the relative level. And yet this bindu—and since the term bindu is used, this statement presumably does refer specifically to the relative form, not just its ultimate nature—“does not cease” at the end of an aeon when all else is destroyed. This is a striking statement in a Buddhist text, especially in reference to a phenomenon declared to pertain to the relative level. What follows is a condemnation of the views of confused individuals who “believe everything to be arisen from Brahmā’s egg.” The belief in arising from Brahmā’s egg references a non-Buddhist theistic cosmogonic narrative in which the universe is said to have emerged from a cosmic egg produced by the god Brahmā.86 The passage in the Dvitīyakrama appears to suggest that the confusion about Brahmā’s egg occurs because certain individuals have mistaken the actual emergence of the cosmos from the indestructible bindu to instead be emergence from Brahmā’s egg; that is, Buddhajñānapāda explains the theistic narrative of emergence from Brahmā’s egg as simply a mistaken apprehension of the genuine cosmic emergence from the indestructible bindu. Vaidyapāda makes this point explicit. He writes, “Because they do not realize genuine reality, they do not know it to be the indestructible bindu and are confused, calling [it] Brahmā’s egg. They think that due to the karma of beings, even though Brahmā does not appear, he left behind an egg and everything arises from that, so they say that Brahmā is their ancestor.”87 In referencing the confusion of beings who believe the world to be born from Brahmā’s egg, Buddhajñānapāda expresses familiarity with a theistic cosmogony that was apparently similar enough to his narrative of cosmic emergence to warrant mention, while simultaneously subordinating the other account to his own assertions about the true source of the cosmos. Religious texts from multiple traditions in medieval India used a variety of strategies of subordination to distinguish themselves from and assert their own supremacy over competing systems. Such narratives are particularly common in cases where traditions shared significant commonalities. For example, the Buddhist tantra, the Mañjuśrīmūlakalpa, states that the Buddha taught the Śaiva, Vaiṣṇava, Gāruḍa and other tantras in order to help outsider, non-Buddhist 83 Sukusuma, D 121b.2; Dvitīyakrama, verse 168. Vaidyapāda repeats the precise description of this bindu from earlier in the Dvitīyakrama at this point in the Sukusuma. 84 Vaidyapāda’s commentary here specifies that this is in the context of the destructive fires and so forth at the end of an aeon (Sukusuma, D 121b.4). 85 ye shes kyi rdo rje’i rang bzhin pas na me la sogs pa gang gis kyang zhig tu med pa... (Sukusuma, D 121b.4-5). 86 There are multiple versions of this narrative in Purānic literature. One of these, set forth in the Brahmāṇḍa Purāṇa (the “Brahmā’s Egg Purāṇa”), is articulated within a strikingly Sāṃkhya-influenced system. Yet unlike the Sāṃkhyakārikās of Classical Sāṃkhya, the Sāṃkhya principles set forth in the Brahmāṇḍa Purāṇa are adapted to the theistic context of its narrative (See Tagare 1958, esp. pp. 28-35). The Brahmāṇḍa Purāṇa, like other purāṇas, has many compositional layers; it has been dated to 400-1000 CE (White 2003, 310n136). 87 des yang dag pa’i don ma rtogs pa’i phyir mi shigs pa’i thig ler ma shes te/ tshangs pa’i sgo nga zhes ‘khrul te/ sems can gyi las kyis tshangs pa mi snang yang de’i sgo nga lus pas de las thams cad skyes pa’o zhes ‘dogs shing tshangs pa la mes po zhes smra ste/ (Sukusuma, D 121b.6). 100 practitioners.88 From the opposite perspective, there is a Śaiva account in the Haracaritacintāmaṇi in which Bṛhaspati, in order to make it possible for Śiva to destroy some demons who were following a Śaiva guru, tricked them into doing “Buddhist” practice, which he simply made up using Śaiva elements. The disciples all converted to Buddhism, including accepting the views of emptiness and no self, which then allowed Śiva to destroy them.89 Buddhajñānapāda’s assertion that the Brahmā’s egg cosmogony is merely a mistaken apprehension of cosmic emergence from the indestructible bindu has the same function as the narratives just described—the subordination of an alternative tradition. However, in doing this, also like in those other cases, it openly acknowledges a similarity between the two. But Buddhajñānapāda’s claim perhaps also has the function of preventing potential attacks on the doctrine of the emergence of the cosmos from the indestructible bindu as unbefitting of a Buddhist system by preemptively asserting the Brahmā’s egg narrative to be precisely a non- Buddhist misapprehension of a Buddhist truth. The cosmogonic accounts in Buddhajñānapāda’s writings thus show the influence of more than one non-Buddhist system on his thought. They also, however, provide doctrinal basis and support for multiple aspects—those relating to both the generation and perfection stages—of the tantric Buddhist ritual systems that were emerging and growing in prominence in his time. Anchored in the nondual wisdom that is at the core of the view of reality and Buddhist practice espoused in his writings, these narratives of cosmic emergence explain the world, and the central Buddhist problem of saṃsāra, to be an illusory devolution from the true reality of nondual wisdom. In these accounts we have again seen doctrine ensconced in narrative, similar to Buddhajñānapāda’s creative framing of the content of the Dvitīyakrama within his own personal narrative. In the Muktitilaka we find yet another use of narrative—the story of Śākyamuni Buddha’s awakening—to powerfully convey a doctrine about the ways in which nondual wisdom can be communicated and realized. III. The Transmission of Nondual Wisdom, Its Cultivation, and Instantaneous Perfection: The Structure of the Higher Tantric Path Any yogin who wishes to realize this sublime reality of suchness should please a guru who knows this and genuinely receive the reality, just as it is. -Mañjuśrī instructing Buddhajñānapāda, Dvitīyakrama Buddhajñānapāda indicates at multiple places in the Dvitīyakrama and the Muktitilaka that the experiential realization of nondual wisdom, or “suchness,” is transmitted directly from a guru to his or her disciple. This true reality of suchness is often said to be “received” (blang) or “obtained” (thob) from the guru’s words. However, Buddhajñānapāda is also careful to clarify that despite this language of “receiving,” as if suchness were something transferred from an outside source (i.e. the guru), in fact it is simply realized within the disciple himself in reliance on the guru’s words. Several references to the receiving or obtaining of suchness in Buddhajñānapāda’s works occur in passages that involve or suggest the context of tantric initiation, and all such references indicate that receiving suchness takes place prior to the disciple’s training in suchness on the second stage of the tantric path. Other statements speak to the immediacy or instantaneous dawning of wisdom, both in the context of the receiving of suchness, and more frequently in the context of post-initiatory training, including at the final moment of awakening. Through examining these assertions, made at various points in 88 Sanderson 2009, 130. 89 ibid., 222. 101 Buddhajñānapāda’s writings, we are able to see the basic features and structure of the higher tantric path—the path of the perfection stage—according to his thought.90 Many of these aspects are brought together in perhaps the most powerful and striking way in which Buddhajñānapāda advocates this higher tantric path: the narrative of Śākyamuni Buddha’s awakening, recounted in the Muktitilaka. In Buddhajñānapāda’s retelling of the story, that quintessential Buddhist event is re-cast as a direct transference, or pointing out, of nondual wisdom to Śākyamuni by the sugatas, followed by his (brief!) training in that wisdom, and finally resulting in its instantaneous full realization at the moment of his awakening. Let us first examine some of the ways in which Buddhajñānapāda tells us about the direct communication of nondual wisdom from a guru prior to its cultivation and full realization by the disciple, and then we will look at how he shows this process to us in his recounting of Śākyamuni Buddha’s awakening. Receiving Suchness A number of passages in both the Dvitīyakrama and Muktitilaka reference the “receiving” or “obtaining” of suchness from the guru, usually through his words. Introducing the yoga of utkrānti (Tib. ‘pho ba),91 which it advocates for a practitioner who has not had the opportunity to perfect other types of training during his lifetime, the Dvitīyakrama states: Someone who has pleased the guru And received the vase [initiation] and the others Together with the samayas and vows given by him And thus obtains the suchness92 |327| That is found through the guru’s words... 93 The Muktitilaka, describing the “inner yoga,” which Buddhajñānapāda identifies as the “supreme suchness” (de nyid mchog),94 states: [When] this inner yoga Is obtained from the mouth Of the sublime guru, one has no doubts.95 At the conclusion of the Muktitilaka Buddhajñānapāda suggests that: ...Any yogin who wishes to realize This sublime reality of suchness Should please a guru who knows this And genuinely receive the reality, just as it is.96 While all of these passages use the language of “obtaining” or “receiving” to describe the process by which the disciple comes into direct contact with “suchness” or “reality,” the first two clearly indicate a teaching context where the guru’s words are the “source” of the suchness that 90 I examine the yogic practices of the perfection stage according to Buddhajñānapāda’s writings in Chapter Six. This section focuses on the doctrinal aspects that frame the structure of that path. 91 I discuss this practice as it appears in Buddhajñanapāda’s works in more detail in Chapter Six on the perfection stage practices in Buddhajñānapāda’s writings. 92 Vaidyapāda specifies that this refers to having received the instructions on suchness together with the sādhana for accomplishing suchness via the seven yogas (Sukusuma, D 130a.3). I discuss the seven yogas in brief below, and in more detail in Chapter Seven. 93 gang zhig bla ma mnyes byas nas// des gnang dam sdom bcas ba ru// bum pa la sogs rab thob ste// bla ma’i zhal las rnyed pa yi// |327| de bzhin nyid ni rab thob cing// (Dvitīyakrama, verses 327a-327a). 94 Muktitilaka, D 50b.3. Vaidyapāda also indicates that the “inner yoga” is the “suchness of all phenomena” (Muktitilaka-vyākhyāna, D 55a.7). 95 de yi nang gi sbyor ba ni// bla ma dam pa’i zhal nas ni// thob kyi de la the tshom med// (Muktitilaka, D 50a.5; P 60a.6). 96 de bas de nyid don dam pa// rnal ‘byor gang gis rtogs ‘dod pas// de shes bla ma mnyes byas te// ji bzhin don ni yang dag blang// (Muktitilaka, D 52a.4; P 62b.5-6) 102 is received. Other references in Buddhajñānapāda’s writings also mention the guru’s words as functioning in some way as the source of the disciple’s realization: Luminous and perfectly joyful like the sky, The self-arisen great *adhideva97 Is realized from the words of the guru By means of innate wisdom. |143|98 Vaidyapāda’s commentary on this verse from the Dvitīyakrama further emphasizes the guru’s role in a process of “transference:” “From the words of the guru means that due to what is transferred by the words of the great causal ācārya,99 the bliss that actually arises within oneself should be realized.”100 The mention of something being “transferred” by the guru to the disciple is repeated at several other places in Vaidyapāda’s commentaries, and Buddhajñānapāda uses the term once in that regard, as well. These statements show us clearly that Buddhajñānapāda held suchness, reality, or nondual wisdom to be something that can be “received” by the disciple from the guru or “transferred” from the guru to the disciple. Most of the references to this process also specifically mention that it takes place through the guru’s words. Although at the outset of the Dvitīyakrama Buddhajñānapāda clearly asserted nondual wisdom to be the very nature of all phenomena, and further specified it to be the nature of the mind, repeated statements to the effect that suchness or wisdom is “transferred” or “received” might nonetheless be construed to describe a situation in which something enters the disciple from outside. In the Muktitilaka Buddhajñānapāda is careful to make it clear that this is not the case. Again, in reference to the “inner yoga” (that Buddhajñānapāda and Vaidyapāda identify as suchness), the Muktitilaka explains: It is only said to come from elsewhere Though [in fact] it is realized by self-aware bliss, It is thus explained as “bestowing initiation.”101 Clarifying that “from elsewhere means from the guru,” Vaidyapāda writes, “Such an essence only nominally appears as if it comes from the guru, but it is [actually] realized by means of self-aware bliss. This is thus also called bestowing initiation.”102 In another passage from the Muktitilaka-vyākhyāna, Vaidyapāda likewise indicates that while the guru’s words are the method through which the realization takes place, the wisdom that is realized is the disciple’s own. Referencing, presumably, the nondual wisdom that according to Buddhajñānapāda’s system is the disciple’s own true nature, Vaidyapāda writes of “having realized one’s own wisdom through the words of one’s guru...”103 Thus we see in these writings, on one hand, the use of terms like “transferring” and “receiving,” and on the other hand, claims that it is not 97 Buddhajñānapāda uses the term *adhideva to refer to the final result of awakening at several points in his writings. I address this point below. 98 gsal zhing rab dga’ nam mkha ‘dra// rang byung lhag pa’i lha chen po// lhan cig skyes pa’i ye shes kyis// bla ma’i kha las rtogs par bya// |143| (Dvitīyakrama, verse 143). 99 The “causal ācārya” is one of three types of ācārya or guru (he uses the terms guru and ācārya interchangably to reference these three) mentioned in Buddhajñānapāda’s writings. The causal ācārya is explained in a verse cited by Vaidyapāda earlier in the Sukusuma, as well as in the Muktitilaka-vyākhyāna, to refer to the guru who bestows initiation upon the disciple. I address the topic of the three types of gurus in more detail below. 100 bla ma’i kha las zhes te// de (de] P, de’i D) rgyu’i (rgyu’i] P, rgyud D) slob dpon chen po’i kha las rnam par ‘pho ba las mngon sum du bde ba rang la (la] D, las P) ‘byung ba rtogs par bya’o zhes so// (Sukusuma, D 111b.4-5; P 134a.8). 101 ming tsam gyis ni logs (logs] P, log D) ‘byung yang// rang rig bde bas rtogs byas na// dbang bskur shes ni bshad pa yin// (Muktitilaka, D 50b.2; P 60b.4) 102 De lta bu’i ngo bo de ming tsam gyis bla ma las ‘byung ba lta bur snang yang/ rang rig pa’i bde bas rtogs pas na dbang bskur ba zhes kyang de la bya’o// (Muktitilaka-vyākhyāna, D 55b.7-56a.1; P 143a.4). 103 rang gi bla ma’i kha nas rang gi ye shes rtogs... (Muktitilaka-vyākhyāna, D 47a.1; P 331b.7-332a.1). 103 actually a transfer that takes place, but a realization within the disciple that is catalyzed by the guru’s words. These statements seem intended to balance the necessity of relying upon a guru to “receive” suchness with the fact that the suchness “received” must be understood as something already present within the disciple as her innate nature. The Transference of Suchness in the Context of Initiation Like the passage just discussed, several of the references to receiving suchness in Buddhajñānapāda’s writings mention or suggest an initiatory context, and all such references make clear that the transference of suchness was meant to—or, indeed, had to—take place prior to the disciple’s training in it. In Buddhajñānapāda’s time a series of three tantric initiations had developed, and the early Jñānapāda School, even several centuries after Buddhajñānapāda’s life, is generally known in modern scholarship for representing a system in which just three initiations, rather than the four of later traditions, are bestowed (although this narrative may be challenged by Vaidyapāda’s Yogasapta).104 While there is evidence in Buddhajñānapāda’s work indicating that he understood the disciple to come into direct contact with suchness in some way through both of the “higher” tantric initiations related to perfection stage practices—the second, the guhyābhiṣeka, and the third, the prajñājñānābhiṣeka—his writings and those of his immediate disciples suggest that the disciple’s “obtaining” of suchness took place in the context of the third initiation, or in an instruction given just afterwards. The Muktitilaka contains a reference to the realization of suchness in an initiatory context, using language that suggests the third initiation, in particular. This passage, already cited above, is found in a section where a great number of tantric practices, not just initiation, are homologized with the “inner yoga” of suchness. The text states: It is only said to come from elsewhere Though [in fact] it is realized by self-aware bliss, It is thus explained as “bestowing initiation.”105 It is the description of suchness as realized by the disciple’s “self-aware bliss,” that points to the third initiation. In this initiation, the prajñājñānābhiṣeka, the disciple entered into sexual union with a partner and the blissful experiences of union were, under the guidance of the guru, used to evoke, point to, or act as an example of the nonconceptual experience of suchness itself.106 A section of the Dvitīyakrama that describes the prajñājñānābhiṣeka107 states, In the space of the lotus, the jewel of the vajra and the heart of the lotus join, and in vajra posture The mind is observed, [within] up to the jewel. The bliss that arises is ascertained [and] that itself is wisdom.108 104 Isaacson 2010b, 269; Wedemeyer 2014. Vaidyapāda’s Yogasapta, however, shows evidence that runs contrary to the narrative that the early Jñānapāda School on the whole did not know or advocate a “fourth,” since in the Yogasapta Vaidyapāda does precisely that. However, it appears that at this time this “fourth” was still not considered a separate initiation. I address this topic in brief below and in more detail in Chapter Seven. 105 ming tsam gyis ni logs (logs] P, log D) ‘byung yang// rang rig bde bas rtogs byas na// dbang bskur shes ni bshad pa yin// (Muktitilaka, D 50b.2; P 60b.4) 106 See Isaacson 2010b, 268 on the function of the third initiation. Precisely what the third initiation entailed, in particular in relation to the “consort observance” (vidyāvrata), is a topic that has been briefly explored by Wedemeyer (2014 and forthcoming), and which I address further in reference to Buddhajñānapāda’s and his immediate disciples’ writings in Chapter Seven. 107 The terms guhya- and prajñājñāna-abhiṣeka are not used in the Dvitīyakrama. But the text does describe in some detail the rituals to which those names are normally ascribed, and Vaidyapāda uses the names of the abhiṣekas in his commentary. The rituals pertaining to the second and third initiations are detailed in Dvitīyakrama verses 83-125. 108 padma’i mkha’ la rdo rje nor bu pad snying gnyis la ‘byor dang rdo rje skyil krung sems// nor bu’i bar du mthong byas gang de bde ba ‘byung ba nges par de nyid ye shes te// (Dvitīyakrama, verse 124c-d). Several later 104 In this passage, using the coded language of the tantras, the “vajra,” or penis, joins with the “lotus,” the vagina, and while in union the “mind,” or citta—in this case referring to the resulting bodhicitta, or drop of semen—is to be observed having traveled up to the “jewel,” the tip of the penis. This produces a blissful experience, the ascertainment of which is identified in the verse as wisdom. The observation of suchness during the third initiation is also mentioned in Buddhajñānapāda’s disciple Dīpaṃkarabhadra’s initiation manual, which states in its description of the prajñājñānābhiṣeka that “While joined with the consort, the splendid one should observe suchness.”109 Vaidyapāda’s commentary on the lines from the Dvitīyakrama cited above describes the process by which the bindu of bodhicitta is first brought into the heart, and then by means of the “downward clearing wind” it is brought to the tip of the vajra, at which point the disciple should “come to know the seven yogas by means of the guru’s oral instructions.”110 These “seven yogas,” mentioned in Buddhajñānapāda’s Muktitilaka and elaborated in Vaidyapāda’s Yogasapta, are seven aspects of the resultant state of awakening.111 It seems that in this system it was in terms of these seven yogas that suchness was communicated by the guru to the disciple. Vaidyapāda’s Yogasapta identifies the seven yogas with “the fourth”112 in a move that certainly goes beyond the way initiation is described in Buddhajñānapāda’s surviving writings, which make no such reference to a “fourth.”113 However, the practice of the guru’s authors have incorporated part of this verse into their presentation of the prajñājñānābhiṣeka. Cf. Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, verse 366c-d. vajraparyaṅkataś cittaṃ maṇyantargatam īkṣyan. Cf. also Vajrāvalī (Mori 2009 Vol. 2, 444), which incorporates these two pādas into the section on the prajñānābhiṣeka immediately after the incorporation of Dvitīyakrama 88a-c, and Daśatattva V.15, which follows the Vajrāvalī in incorporating these pādas after the incorporation of Dvitīyakrama 88a-c. 109 prajñāsamparkataḥ śrīmān tattvaṃ samupalakṣayet | (Guhyasamājamaṇḍalavidhi, 366a-b.) 110 de’i dus su bla ma’i man ngag gis (gis] P, gi D) sbyor ba bdun shes par bya ste (Sukusuma, D 109a.6; P 131b.3). 111 The seven yogas are mentioned by Buddhajñānapāda in the Muktitilaka, where they are described as the “perfection stage of the perfection stage,” (Muktitilaka, D 52a.2), and are also said to be realized instantaneously by a yogin engaged in post-initiatory practice (cārya) (Muktitilaka, D 51b). These seven yogas, which are mentioned but not listed in the Muktitilaka, thus seem to refer in Buddhajñānapāda’s work to practices that are to be carried out by the yogin subsequent to initiation. In Vaidyapāda’s Yogasapta, however, the seven yogas are explained in much greater detail as seven states or experiences that the student is meant to undergo in the context of initiation— specifically during what is called “the fourth” (See note 112). The seven are perfect example-less bliss (dpe med bde rdzogs), nonduality (gnyis su med pa), great bliss (bde ba chen po), lacking nature (rang bzhin med pa), unfolding compassion (thugs rjes rgyas pa), unbroken continuity (rgyun mi chad pa), and non-cessation (‘gog pa med pa). The same seven factors are addressed in Vāgīśvavarakīrti’s later Saptāṅga and his Tattvaratnāvaloka and its autocommentary, where they are called the seven aṅgas of mahāmudrā, with reference to which see Isaacson (2010b, 271, 271n27) and, with a bit more detail, Isaacson and Sferra (2014, 271), where they are mentioned with reference to a citation from the Saptāṅga in Rāmapāla’s Sekanirdeśapañjikā. The seven aṅgas are listed in Vāgīśvarakīrti’s work as sambhoga, sampuṭa, mahāsukha, niḥsvabhāva, kāruṇyanirbhara, nirantara, anirodhaḥ. I discuss the seven yogas in more detail in Chapter Seven. 112 Vaidyapāda’s Yogasapta-nāma-caturabhiṣekaprakaraṇa (Sbyor ba bdun pa zhes bya ba dbang bzhi’i rab tu byed pa) mentions the “four initiations” in the title, but throughout the work itself the term “fourth initiation” is never used; the first three initiations are clearly called initaiton but “the fourth” is only ever referred to as simply “the fourth” (bzhi pa). Given this fact, along with the unreliability of the Sanskrit titles in the Tibetan canon, some of which (like Dvikrama for the *Dvitīyakrama!) appear to be incorrect Sanskrit reconstructions made by the redactors of the Tibetan canon, we may be inclined to raise doubts about the “fourth initiation,” mentioned in the title of the Yogasapta. However, the Tibetan translation of the title likewise makes reference to “four initiations (dbang bzhi). Moreover, in his Guhyasamājamāṇḍalopāyikā-ṭīkā Vaidyapāda does indeed refer to a “precious fourth initiation” (dbang bskur ba rin po che bzhi pa) (Guhyasamājamaṇḍalopāyikāṭīkā, D 211b.3-4; P 539b.6-7). The issue of whether Vaidyapāda considered “the fourth” an initiation or not therefore appears to be a slippery one; I discuss initiation in Buddhajñānapāda’s system and in Vaidyapāda’s writings further in Chapter Seven. 113 Vaidyapāda’s list in the Sukusuma of Buddhajñānapāda’s compositions, however, references a work of Buddhajñānapāda’s called The Method for Engaging in the Fourth (bzhi pa la ‘jug pa’i thabs). Most unfortunately this work seems not to be extant in either its original Sanskrit nor in Tibetan translation (as we saw in Chapter One, 105 communicating suchness to the disciple in the context of tantric initiation does indeed feature in Buddhajñānapāda’s writings. A passage from the Dvitīyakrama, cited earlier, describes, Someone who has pleased the guru And received the vase [initiation] and the others Together with the samayas and vows given by him, And thus obtains the suchness |327| That is found through the guru’s words...114 Here the text mentions the “vase [initiation] and the others,” referring to the set of initiations of which the vase (kalaśa, bum pa) is the first. But it is only after mentioning the initiations, as well as samayas and vows given by the guru to the disciple in that context, that the text brings up obtaining “the suchness that is found through the guru’s words.” It is not completely clear whether Buddhajñānapāda is referencing something that is ritually separate from the other aspects of initiation. But it is clear that he wants to single out the topic of “obtain[ing] the suchness that is found through the guru’s words” as something that is particularly important within the initiatory context. Vaidyapāda makes the distinction between the receiving of initiation and samayas and the obtaining of suchness stronger by commenting on the two topics separately, adding the phrase “and then...” between his commentary on initiation and vows and his comments on obtaining suchness. With regard to the latter, Vaidyapāda writes, “And then, And thus obtains the suchness/ That is spoken by the guru,115 means that the suchness of the seven yogas, together with the method for accomplishing that, is received.”116 This passage in the Dvitīyakrama seems, then, to refer to a moment in the context of, or possibly immediately subsequent to the third (and final) initiation, when the guru verbally communicates the nature of suchness to the disciple, who then “receives” that suchness. Such a moment is also mentioned in Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi: “having bestowed the guhya and prajñā [initiations], suchness should be fully pointed out.”117 In fact, a verbal communication of suchness by the guru to the disciple subsequent to the third initiation became the predominant one among a number of ways in which the so-called “fourth initiation” (caturthābhiṣeka) came to be understood. Debates on the topic of a “fourth initiation,” which took place over a number of centuries, seem to have centered on—and perhaps sprung from—the meaning of a reference to “the fourth” (caturtha) in a passage in the many of the works in this list are extant), but the fact that Buddhajñānapāda may have composed a work on “the fourth” remains a very interesting possibility. I discuss this point further in Chapter Seven, but, in brief, Vaidyapāda seems to be quite a reliable witness of Buddhajñānapāda’s system, and I see little reason to doubt him with respect to his list of Buddhajñānapāda’s compositions. Yet, the fact remains that we have no reference at all in any of Buddhajñānapāda’s surviving works to a “fourth.” 114 gang zhig bla ma mnyes byas nas// des gnang dam sdom bcas ba ru// bum pa la sogs rab thob ste// bla ma’i zhal las rnyed pa yi// |327| de bzhin nyid ni rab thob cing// (Dvitīyakrama, verse 327a-328a). 115 Bum pa la sogs pa’i dbang gong du gsungs pa ltar rab tu thob par byas te/ de nas/ bla ma’i zhal nas gsungs pa yi// de bzhin nyid ni rab thob cing// zhes pa ni sbyor ba bdun gyis de kho na nyid sgrub pa’i thabs dang bcas pa rnyed pa… (Sukusuma, D 130a.3; P 156b.4-5). Vaidyapāda’s commentary preserves a slight variant on one line from the Dvitīyakrama. The pādas as found in the Sukusuma read: bla ma’i zhal nas gsungs pa yi// de bzhin nyid ni rab thob cing// (Sukusuma, D 130a. 3; P 156b.4) as opposed to bla ma’i zhal nas rnyed pa yi// de bzhin nyid ni rab thob cing// in the Dvitīyakrama. However, in his comments on these two lines Vaidyapāda mentions the word “received” (rnyed pa) which is absent in the verse as translated in his commentary but present in the verse as translated in the Dvitīyakrama, so I suspect the variant arose in the context of translating the Sukusuma into Tibetan, rather than in the citation of the verse in Vaidyapāda’s commentary itself. 116 de nas/ bla ma’i zhal nas gsungs pa yi/ de bzhin nyid ni rab thob cing/ zhes pa ni sbyor ba bdun gyi (gyi] P, gyis D) de kho na nyid bsgrub pa’i thabs dang bcas pa rnyed pa [/] (Sukusuma, D 130a. 3; P 156b.4). 117 The full verse reads, maṇḍalaṃ devatātattvam ācāryaparikarma ca | saṃkathya guhyaprajñābhyāṃ siktvā tattvaṃ samuddiśet | (Guhyasamājamaṇḍalavidhi), 367.) 106 Samājottara on the initiatory sequence.118 Vaidyapāda, one of the earliest commentators on the topic, referencing the passage from the Samājottara in his Yogasapta, describes “the fourth,” as the verbal communication of suchness to the disciple following the third initiation, consisting of the seven yogas. Buddhajñānapāda, however, appears not to have known the Samājottara, so the passage from the Dvitīyakrama that describes “obtain[ing] the suchness that is found through the guru’s words” as part of or subsequent to initiation, seems to be an early indication of the practice of what came to be called “the fourth,” perhaps prior to its designation as such.119 Since Vaidyapāda’s Yogasapta, however, does refer to the practice of verbally communicating suchness to the disciple as “the fourth,” (although, importantly, not as “the fourth initiation”) if that designation post-dates Buddhajñānapāda’s life, it does not do so by very long.120 Harunaga Isaacson has suggested that the line cited above from Dīpaṃkarabhadra’s popular ritual manual—“having bestowed the guhya and prajñā [initiations], suchness should be fully pointed out”121—may have, “while not knowing a Fourth empowerment itself, provided the element that was re-interpreted as being the Fourth.”122 He also notes that in this pāda, an alternative version of which is found in some other later sources, it is possible that Dīpaṃkarabhadra was following an earlier, possibly scriptural, literary source.123 While I cannot comment on the literary source for that particular pāda of Dīpaṃkarabhadra’s text, I would suggest that what we have seen in Buddhajñānapāda’s writings makes it extremely likely that Dīpaṃkarabhadra’s statement in the Maṇḍalavidhi was drawing on the practice already found in his guru Buddhajñānapāda’s ritual system, of verbally indicating suchness to the disciple at (or as?) the conclusion of the third initiation. Indeed, whatever their relationship to the use of the 118 Isaacson 2010b, 268-271. Isaacson describes this particular position on what constitutes the fourth initiation as something that “is given verbally to the initiand by his guru and...in some way should also be seen as being or as containing, ideally at least, the goal (lakṣya or phala)” (Isaacson 2010b, 271). The suchness or nondual wisdom that Buddhajñānapāda says is conveyed by the guru to the disciple here is understood simultaneously to be the goal and the very nature of all things, importantly including the disciple’s own mind. 119 In fact, Isaacson and Sakurai argue, partly on the basis of the absence of a fourth initiation in early Jñānapāda School and early Ārya School works, that a separate fourth initiation was not actually intended in the Samājottara at all but arose later out of the debate on the passage therein referring to “the fourth” (Isaacson 2010b, 269). As I will discuss further in Chapter Seven, Vaidyapāda’s Yogasapta, does, however, provide a relatively early reference “the fourth” in the work of at least one lineage-holder of the Jñānapāda School, though as I noted above, within the text itself he does not refer to it as “the fourth initiation,” just “the fourth.” (The full title of the Yogasapta-nāmacaturabhiṣekaprakāraṇa does, however, mention “four initiations” and Vaidyapāda also references a “fourth initiation” in his commentary to Dīpaṃkarabhadra’s Maṇḍalavidhi (Guhyasamājamaṇḍalopāyikāṭīkā, D 211b.3-4; P 539b.6-7. See also note 112 on this point.) In any case, whether a fourth initiation was intended in the Samājottara or was only conceived subsequent to it, given that Buddhajñānapāda does not seem to have known that text, the Dvitīyakrama provides early evidence of the practice of communicating suchness verbally to the disciple at or as the conclusion of initiation, however such a practice may have been designated at the time. 120 I do not know whether Dīpaṃkarabhadra—who, let us recall, was a direct disciple of Buddhajñānapāda’s and senior to Vaidyapāda—knew the Samājottara, but it seems that either he did not know it, or felt no need in his Guhyasamājamaṇḍalavidhi to hew to its systems. In that text Dīpaṃkarabhara makes no attempt to account for “the fourth,” nor does he include a separate ritual for the bestowal of the “consort observance” (vidyāvrata), as the Samājottara does. Vaidyapāda’s commentary on Dīpamkarabhadra’s manual both mentions the bestowal of a “precious fourth initiation” that consists of the guru’s oral instructions with respect to union, and it also introduces a separate ritual for the bestowal of the vidyāvrata subsequent to the sequence of initiations, following the model of the Samājottara (on “the precious fourth initiation” see Guhyasamājamaṇḍalopāyikāṭīkā, D 211b.3-4; P 539b.6-7). The passage on the ritual for bestowing the vidyāvrata in Vaidyapāda’s commentary includes several liturgical verses, ten pādas of which are taken directly from the vidyāvrata section of the Samājottara (125c-127d), but without attribution (Guhyasamājamaṇḍalopāyikāṭīkā, D 211b.6-212a.1; P 540a.4-7) 121 The full verse reads, maṇḍalaṃ devatātattvamācāryaparikarma ca | saṃkathya guhyaprajñābhyāṃ siktvā tattvaṃ samuddiśet | (Guhyasamājamaṇḍalavidhi), 367. 122 Isaacson 2010b, 275-6. 123 ibid., 276n42. 107 term “the fourth,” or “the fourth initiation,” to describe the procedure of communicating the nature of suchness to a disciple, Buddhajñānapāda’s writings clearly show that he advocated the position that such a verbal communication of suchness from the guru to his disciple was not only possible, but actually necessary as a prerequisite to the disciple’s training in the second stage of tantric practice. And the Dvitīyakrama does suggest that this communication took place at or as the conclusion of the third initiation. Initiation also makes sense as the context for the “transference” of suchness to the disciple, given that throughout the references to the “receiving” of suchness in his writings, it is very clear that its communication is meant to serve as the starting point of the disciple’s training in the “suchness of the second stage” (rim pa gnyis pa’i de kho na nyid).124 The Muktitilaka, in a section of the text that Vaidyapāda explains as setting forth “the preliminary foundation” for “the training in the suchness of the second stage,”125 states: Therefore, having obtained that suchness Observe all beings with compassion, Remain in seclusion or some similar place, And constantly train in the truth of non-duality.126 Here, just like in the instances where initiation is specifically mentioned, we see “obtaining suchness” taking place in the context of the disciple’s initial instruction in the practice of the perfection stage. Such a “transfer” provides the basis for the initial cultivation, and eventually the full realization, of that suchness—the “truth of non-duality”—which results in awakening. The fact that this “receiving” takes place prior to the training in suchness—indeed, it seems to be a prerequisite for doing so—indicates that “receiving” suchness, while it is said to constitute an instantaneous “knowing” of reality, as we will see below, certainly does not constitute the full freedom of awakening. Though not explicitly stated in Buddhajñānapāda’s writings, what this suggests is that when the disciple “receives” suchness, what she receives is a mere glimpse of the reality of nondual wisdom, which must then come to be known more fully through training. The function of the higher tantric initiations in providing such a glimpse of reality is explicitly articulated in the writings of later tantric authors.127 That it is possible to have an initial glimpse of the ultimate truth that is further cultivated through training is not a position held uniquely in tantric Buddhist traditions. The so-called “Path of Seeing” (darśanamārga, mthong lam), the third of the Five Paths in the traditional Mahāyāna system, is entered when the bodhisattva has an initial glimpse of emptiness on the first bhūmi. The bodhisattva's realization of this ultimate truth of emptiness deepens as she traverse thes remaining nine bodhisattva bhūmis on the “Path of Cultivation” (bhāvanāmārga, bsgom lam), culminating in ultimate truth being seen in its fullness in the vajra-like samādhi that marks the moment of awakening.128 In fact, the very names of these two paths of Mahāyāna training—the Path of Seeing and Path of Cultivation—convey precisely this progression from initially seeing 124 This term is used throughout Buddhajñānapāda’s writings, including in the title of the Dvitīyakrama. I discuss it below in the context of Buddhajñānapāda’s privileging of the tantric path. 125 rim pa gnyis pa’i de kho na nyid bsgom pa’i phyir de’i gzhi’i (gzhi’i] D, gzhi P) rim pa dang po rnams dpe’i sgo nas bstan pa gsung pa de/ (Muktitilaka-vyākhyāna, D 51b.1; P 337b.3) 126 de bas de nyid thob byas nas// snying rjes sems can kun dmigs nas// dgon pa la sogs gnas nas ni// rtag tu gnyis med don bsgom bya// (Muktitilaka, D 48b.1; P 58a.3-4) 127 This is one interpretation of the function of the prajñājñāna initiation set forth in the Abhiṣekanirukti, attributed to the 11th-century master Ratnākaraśānti (but, according to Isaacson, more likely composed by his disciple Jinasujayaśrīgupta) (Isaacson 2010b, 266-68). I explore the relationship of this function of the third initiation and that of the instructions that came to be designated as “the fourth” further in Chapter Seven. 128 In fact, these two paths were already set forth in the system of the Sarvāstivādin-Vaibhāṣikas, as described by de La Vallée Poussin (1936-7). 108 ultimate reality and gradually cultivating that until its full realization in the moment of awakening. What is unique about this process on the tantric path described in Buddhajñānapāda’s and later tantric writings is the idea that suchness can be directly communicated by a guru to the student, that this initial moment of insight is somehow precipitated by a communication from a qualified guru, such that the disciple comes to know reality in an instant—a sort of kick-start to the path of its cultivation through the yogas of the perfection stage. Instant Knowing, Gradual Training, Sudden Realization A number of passages in Buddhajñānapāda’s work describe the arising or dawning of wisdom as “instantaneous.” However, when we look into the context of these references to sudden insight, they seem to point to two (or maybe three) specific moments in a disciple’s experience: at the beginning of and towards, or at, the end of an otherwise gradual path of training in suchness. The majority of the references to the “instantaneous” dawning of wisdom in Buddhajñānapāda’s writings refer not to the initial “receiving” of suchness from the guru prior to training, but to an experience that takes place subsequent to this, as the disciple is in the process of cultivating the suchness he has previously “received.” But one passage from the Muktitilaka does appear to describe the “knowing” of suchness obtained at the outset of training as instantaneous, as well. Immediately following a description of the supreme non-dual nature, the “great secret of all the buddhas,” and prior to a description of how to train in this state, the text explains that, By constantly revering a lineage guru Who knows this reality Occasionally, like the [appearance of] the uḍumbara flower, Those with merit will know it in an instant.129 Like the references above that mention “pleasing” a guru in the context of “receiving” suchness, here the text indicates that “revering” a master who knows reality can result in the disciple’s coming to know it “in an instant.” As noted above, it is the context of the passage—after a description of suchness and prior to the description of the methods for training in suchness—that suggests that this is a reference to the guru’s communicating suchness directly to the disciple at the outset of his training in the perfection stage practices, although initiation is not specifically mentioned. This initial “receiving,” “obtaining,” or “knowing” of suchness thus appears, according to Buddhajñānapāda, to be something that takes place instantaneously. Before we move on to look at passages that describe the “instantaneous” arising of wisdom for a disciple who has already “received” it and is in the process of training in suchness, two other points in this first passage are worth noting. First, the verse makes it clear that the guru who is to be revered is one who holds a lineage; he must be, as Vaidyapāda explains, “a sublime guru who knows nondual reality and possesses the great pith instructions of the lineage that has been passed from ear to ear.”130 The aspect of lineage is also emphasized at other places in both the Muktitilaka and the Dvitīyakrama. In the latter, holding “the lineage of supreme oral instructions” is listed as one of the qualifications for a guru.131 As we shall see below, 129 ‘di yi don shes brgyud pa yi// bla ma dam pa rtag bkur bas// u duṃ bar (u duṃ bar] P, u dumbār D) ltar brgya lam na// bsod nams can gyis skad cig shes// (Muktitilaka, D 48b.6; P57a.7-8). 130 gnyis su med pa’i don shes shing rna ba nas rna bar brgyud pa’i man ngag chen po dang ldan pa’i bla ma dam pa (Muktitilaka-vyākhyāna, 50a.5-6; P 336a.2-3). 131 mchog gi gdams ngag rgyud la ldan (Dvitīyakrama, verse 45b). The guru should also be genuine and venerable, be intent upon the conduct and training of the Mahāyāna path, know the secret, great secret, and exceptional secret of the ten suchnesses (daśatattva), and be willing to teach those to whom reality is concealed (see Dvitīyakrama, verses 45-6). 109 Buddhajñānapāda traces this lineage of oral instructions back to Śākyamuni Buddha. A second point of note in the passage is that here instantaneous knowing of the truth is possible only for “those with merit.” The uḍumbara flower is used in Buddhist literature as an example of something rare, and a number of other passages in Buddhajñānapāda’s works similarly refer to the passing on of wisdom to just “a few” disciples, suggesting the rarity of such an occurrence— and perhaps also the rarity of disciples who have the merit necessary to receive it—seemingly indicating that this particular Buddhist path was not meant for everyone.132 In fact, these two aspects of the guru’s holding the lineage of suchness and the disciple’s having the proper merit to be able to receive it appear to be points that Buddhajñānapāda understands as essential to the transmission and realization of this wisdom. Near the conclusion of the Dvitīyakrama he explains, Through relying upon a genuine lineage teacher, And one’s own previously gathered accumulation of merit— One will come to realize this [reality]. |389| Apart from133 these [circumstances], those with little merit Even after countless aeons will not realize this.134 Vaidyapāda’s commentary describes these points—relying on a genuine lineage teacher and having previously accumulated merit—as the “two causes” without which the disciple is unable to connect to reality.135 We will see both of these points come up again in the Muktitilaka’s narrative of Śākyamuni Buddha’s awakening, which I examine below. Let us return now, not to the disciple’s initial connection with suchness, which, as we saw above, comes about through “an instant” of knowing under the guidance of a lineage-holding guru, but to subsequent instances of “instantaneous” wisdom that occur as the practitioner trains in suchness. In several passages, training in the suchness that disciple first “received” through his guru’s guidance is said to bring about an instantaneous dawning of wisdom. Referencing the concentration on the indestructible bindu in the yogin’s heart-center during perfection stage practice, the Muktitilaka states: Through concentrating upon this, Great wisdom will immediately blaze Within the yogin—of this there is no doubt.136 Also regarding the training in suchness, the text says: [When] maintaining samaya and training in suchness, Suddenly, wisdom blazes in an instant Like lightning dispelling the darkness Within one who is [still, at present] bound in existence!137 132 For example: “I explain this for the sake of a few yogins...” rnal ‘byor ‘ga’ zhig don du bshad par bya (Muktitilaka D 47a.4); “From that time onward this supreme reality/ Was [passed on] to a few fortunate ones,/ From mouth to mouth, from ear to ear—/ That lineage teaches genuine reality.” de nas bzung ste don mchog ‘di// skal bar ldan pa ‘ga’ zhig la// zhal nas zhal dang snyan nas snyan// brgyud de nges pa’i don bstan pa’o// (Muktitilaka D 51a.2-3); “This will be explained to a few yogins/ who are fortunate due to their actions...” bya bas bskal pa ‘ga’ zhig la// yang dang tu ni bshad par bya// (Dvitīyakrama, verse 327c-d). 133 ma gtogs] sugg. em., ma rtogs D C S P N 134 brgyud pa’i bla ma yang dag rab bsten dang// rang gi bsod nams tshogs ni sngon bskyed pas// rtogs par rab tu ‘gyur ba ma gtogs par// |389| bsod nams chung ba’i mi yis bskal ba dpag med par// ‘di ni rtogs par mi ‘gyur... (Dvitīyakrama, verse 389b-390b). 135 Sukusuma, D 137b.1. 136 de la ltas bas ye shes che// skad cig tsam gyis rnal ‘byor la (la] P, pa D)// ‘bar ba ‘di la the tshom med// (Muktitilaka, D 49a.3-4; P 59a.1). 137 dam tshig dang ldan de nyid bsgom// myur du srid pa’i ‘cing ba la// glog gis mun pa bsal ba bzhin// skad cig tsam gyis ye shes ‘bar// (Muktitilaka, D 48a.7-48b.1; P 58b.3). The phrase “one who is bound within existence” is 110 Later in the text, referencing the disciple’s taking up post-initiatory practices (caryā), the Muktitilaka states: The one who engages in these [types of] practices Realizes the seven yogas in a single instant. And for as long as existence persists He will have the eight characteristics of the taste.138 The first of these three passages occurs in the middle of the instructions on the perfection stage practice of the indestructible bindu. Since this is only the first of the three bindu yogas taught in the Muktitilaka and the Dvitīyakrama, the wisdom that “immediately blazes” within the yogin who concentrates on this bindu is presumably not the realization of the final result of practice. It is simply a moment of sudden insight that takes place within the course of practice. In the second passage, it is unclear whether the wisdom that suddenly “blazes” while the yogin trains in suchness as he upholds his samayas refers to a moment of insight along the path or to the final result. Vaidyapāda takes it to refer to “the instant blazing of nondual wisdom due to the observance of post-initiatory practices (gtul zhugs kyi spyod pa, vratacaryā) at the time when one [has reached the state of being a] third [-level] yogin.”139 His interpretation of this passage, which connects it with post-initiatory practice, corresponds closely with the third passage from the Muktitilaka, cited above, that mentions the instantaneous realization of the seven yogas through engaging in post-initiatory practices and being endowed with the “eight characteristics of the taste,” listed in the Dvitīyakrama as eight characteristics of the awakened state.140 Moreover this verse is immediately followed by a number of descriptors of one who has “gone to the far side of saṃsāra,”141 making it clear that the instantaneous and full realization of the seven yogas described here does, according to Buddhajñānapāda’s understanding, mark final awakening. The methods of “training” in suchness referenced in all three of these passages—the bindu yogas, as well as the post-initiatory practices (caryā) that, according to Vaidyapāda, somewhat curious, given that wisdom is meant precisely to break through the bondage to existence. I have understood this to indicate that a direct glimpse of the final result is possible even within someone who is, as yet, still bound in saṃsāra, and will not become fully free from those bonds until she completes her training in, or cultivation of, such wisdom. 138 de ‘dra’i spyod pa la gnas pa// skad cig gis ni sbyor ba bdun// rtogs nas ji srid bar du ni// ro myang mtshan nyid brgyad ldan par// (Muktitilaka, D 51b. 5-6; P 62a.5-6). 139 rnal ‘byor pa gsum par ‘gyur pa’i tshe/ brtul zhugs kyi spyod pas skad cig tsam la gnyis su med pa’i ye shes ‘bar bas... (Muktitilaka-vyākhyāna, D 51b.1; P 337b.2-3). Such a “yogin of the third level” is explained later in the Muktitilaka-vyākhyāna, commenting on a verse in the Muktitilaka that likewise indicates “the occasion of being a third[-level] yogin” as the time when one should engage in the various post-initiatory practices described in the text. Vaidyapāda there explains the “third[-level] yogin” as someone who has not only gone beyond being a beginner (the first level), but also having gone beyond the intermediate level of a yogin who has “control over limited wisdom,” (the second level) to the level of a yogin who has “control over the wisdom that brings oneself and others to behold the illusory maṇḍala” (Sukusuma, D 57b.2-3). This three-level schema of the development of a yogin’s meditative progress, mentioned at several places in Buddhajñānapāda’s and Vaidyapāda’s works, seems to either have been later expanded into four levels of progress in later works of the Jñānapāda School or perhaps more likely simply be mentioned in an abbreviated form in these earlier works. Sabine Klein Schwind writes of the distinctive practice instructions included in Kṣitigarbha’s Daśatattvasaṃgraha that are connected with generation stage pratice for the yogins on each of four levels, and also references the four-fold schema also in Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, and in Ratnākaraśānti’s works (Klein-Schwind 2012, 87-92). 140 The eight are listed in the Dvitīyakrama as: permanent, free from torment, cool, singular, blissful, stainless, joyful, and mentally joyful (Dvitīyakrama, verses 292-3). Vaidyapāda elaborates them in the Sukusuma (D 127a.4-7) and gives a similar presentation in his Yogasapta (Yogasapta, D 74a.1-3; P 88a.2-5). 141 ‘khor ba’i mtha’ yi pha rol son (Muktitilaka, D 51b.6). 111 include the “consort observance” (vidyāvrata)142—involve the practice of sexual yoga. I explore the perfection stage yogas described in Buddhajñānapāda’s writings in more detail in Chapter Six, but for now it will suffice to note that according to his system, not only did “receiving” suchness take place in the context of initiations that had sexual elements, but the process of training in suchness once it had been “received,” also involved sexual yogas. It was thus in the context of such training that the practitioner experienced the moments of sudden insight and realization described above. We see in Buddhajñānapāda’s writings, then, references to instantaneous “knowing,” “wisdom,” and “realization” occurring at multiple occasions during the disciple’s training: upon the initial “receiving” of suchness from her guru, in certain moments of insight that occur while she is training on the path, and also at the final moment of awakening. Through the positions identified and examined in these passages from his writings, we can piece together an outline of the higher stage of the tantric path as Buddhajñānapāda conceived of it. That path looks something like this: first the disciple “receives” suchness from her guru, most likely in an oral instruction conveyed in the initiatory context, at the end of or following the third tantric initiation. This “receiving” involves the disciple’s coming to an instantaneous “knowing” of the suchness that was “transferred” by her guru, and is also intimitely connected to the blissful experience of sexual union with a partner in the third initiation. Then the disciple trains in, or cultivates, this suchness via perfection stage practices, during which she may have moments of sudden insight. Finally, having taken up post-initiatory observances (vrata) and practices (caryā), including the observance of the consort-vow (vidyāvrata) and the sexual practices it entails, she “[fully] realizes the seven yogas in an instant” and is awakened. While we do not find all of these various pieces of receiving, training in, and perfecting the realization of suchness laid out in a clear step-by-step procedure in Buddhajñānapāda’s writings, we do find in the Muktitilaka a striking passage that draws most of these elements together into a single brief narrative—that of Śākyamuni Buddha’s awakening. Let us turn now to that narrative and its function in Buddhajñānapāda’s oeuvre as a powerful statement of advocacy for the higher tantric path to awakening. Śākyamuni Buddha’s Awakening Through the Higher Tantric Path In a remarkable retelling of the story of Śākyamuni Buddha’s awakening in the Muktitilaka, Buddhajñānapāda shows that it was through the higher tantric path of receiving suchness, training in it, and coming to instantaneously realize it fully that Śākyamuni Buddha himself attained perfect awakening. The account is narrated as follows: Why did Śākyamuni, though he gathered the requisites143 For countless [aeons], not realize this reality? At the Nairañjanā [River] He remained in the samādhi of “Nothing Whatsoever.” At that time, all of the sugatas Cast away the conceptuality [regarding emptiness] that had befallen him144 And showed him the non-dual profundity and luminosity 142 Muktitilaka-vyākhyāna, D 57b.6-7. 143 tshogs bsags. This refers to the two “provisions” of merit and wisdom that must be gathered in order to attain awakening. 144 thal byung blo can. I discuss this perplexing phrase in some detail in note 148. A slightly different version of this line is found in the citation of this passage from the Muktitilaka in Vaidyapāda’s Siddhisaṃbhavanidhi, where it reads thal ‘byung blo chags de bzlog ste/ “Cast away his attachment to the suddenly befallen conceptuality [regarding emptiness]” (Siddhisaṃbhavanidhi, D 3a3). 112 That is perfectly pure like the expanse of space. At midnight, just like the [previous] Victors, He trained in suchness, And at dawn, in an instant, he realized it genuinely. Abiding thus in the essence of awakening He conquered the great armies of māras. Then, in order to care for beings, He turned the wheel of dharma. From that time onward, this supreme reality Was passed on through the lineage to a few fortunate ones From mouth to mouth, from ear to ear— Thus genuine reality has been taught.145 The question posed at the outset of the passage as to why Śākyamuni had not previously attained awakening despite having gathered the requisite conditions for doing so is not answered directly; the narrative of his awakening itself serves as the answer. That is, although Śākyamuni had gathered all of the other requisites for awakening—the traditional “two requisites” (saṃbhāradvaya, tshogs gnyis) are merit and wisdom—he lacked one essential factor: a qualified guru who could show him suchness directly. We may recall here the two circumstances without which, Buddhajñānapāda states in the Dvitīyakrama, a disciple is unable to gain access to reality: “Relying upon a genuine lineage teacher/ And one’s own previously gathered accumulation of merit.”146 Thus, although he had gathered great stores of merit, it was only when the sugatas— here functioning as his gurus—appeared and showed Śākyamuni nondual profundity and luminosity that he was able to train in this suchness, and thereby to fully awaken. This recounting of Śākyamuni’s awakening thus includes the very same key elements of the higher tantric path that we have seen articulated at other places in Buddhajñānapāda’s writings: Śākyamuni encountered his “gurus;” they showed him nondual profundity and luminosity; he then trained in that suchness; and finally, through that training, he gained instantaneous realization. The conclusion of the narrative is also remarkable, essentially claiming Śākyamuni Buddha himself as the source of the oral lineage of teachings on suchness. Like several passages in Buddhajñānapāda’s writings that mention “receiving” suchness from a guru without specifying an initiatory context, tantric initiation is not mentioned in this account of Śākyamuni’s awakening. While much of the evidence from the Dvitīyakrama and Muktitilaka does point to an initiatory context (or, as noted above, an immediately post-initiatory context) as the one in which a “transference” of suchness from the guru occurred, the absence of a reference to initiation in this account of Śākyamuni’s awakening suggests that Buddhajñānapāda was most interested in highlighting the aspect of the guru’s showing or pointing out suchness to his disciple. This is further corroborated by the emphasis on lineage in the conclusion of the narrative, and Vaidyapāda’s commentary on the passage also very clearly gives that sense. Vaidyapāda’s comments are worth citing here in full: 145 ci’i phyir shā kya thub pa yis// grangs med bar (bar] D, par P) du tshogs bsags kyang// ‘di don ma rtogs nai (nai] D, ni P) ra nydzar// ci yang ma yin ting ‘dzin gnas// de tshe bde gshegs kun gyis kyang// thal byung blo can de bzlog ste// nam mkha’i dbyings ltar rnam dag pa’i// zab gsal gnyis (gnyis] P, gsal D) med rab bstan te// nam phyed (phyed] P, gyed D) dus su rgyal (rgyal] D, brgyal P) ba ltar// de nyid bsgoms pas tho rangs su// skad cig gis ni yang dag rtogs// byang chub snying por bzhugs byas nas// dpung chen bdud rnams rab bcom ste// sems can gzung (gzung] D, gzud P) bar bya ba’i phyir// chos kyi ‘khor lo bskor ba ’o// de nas bzung ste don mchog ‘di// skal bar ldan pa ‘ga’ zhig la// zhal nas zhal dang snyan nas snyan // brgyud (brgyud] D, rgyud P) de nges pa’i don bstan pa’o// (Muktitilaka, D 50b.7-51a.2; P 61a.4-7) 146 brgyud pa’i bla ma yang dag rab bsten dang// rang gi bsod nams tshogs ni sngon bskyed pas// (Dvitīyakrama, verse 390b-c). 113 Then, in order to show that even Śākyamuni and others, despite having gathered limitless requisites, were not sublime, genuine buddhas without coming to know nondual wisdom from the mouth of their gurus, the text states, “Why did Śākyamuni...” etc. Did not realize this reality means he had not realized this stage (rim pa) that is passed on though the oral lineage. [The samādhi of] Nothing Whatsoever is remaining in [a state of] suddenly befallen emptiness (thal byung gi stong pa nyid). To indicate the gurus who showed him [suchness], the text states At that time, all of the sugatas... Cast away the conceptuality [regarding emptiness] that had befallen him, means that this samādhi [of Nothing Whatsoever] occasions falling [into an extreme], and because of that he did not realize the ultimate suchness; so [they] turned his mind away from it. Well, how did he realize [suchness], then? They showed him the nature of nondual profundity and luminosity that is perfectly pure like the expanse of space. Then, at midnight, blessed by the sahaja ācārya, he attained suchness like the [previous] Victors. Due to the strength of that [attainment], at dawn, abiding within the essence of awakening, he conquered the armies of māras. Then, he turned the wheel of dharma for beings. Since that time, this genuine reality has been taught by transferring this great reality from mouth to mouth to a few fortunate ones who rely as their foundation upon genuine beings. In this way he [i.e. Buddhajñānapāda] also indicates the source of the lineage.147 As we can see, Vaidyapāda emphasizes that awakening was not possible, even for Śākyamuni, without receiving a direct transmission of nondual wisdom from a guru, and he explicitly identifies the sugatas as Śākyamuni’s gurus. Precisely what incorrect view “all the sugatas” cast from Śākyamuni’s mind in order to show him the nondual profundity and luminosity that precipitated his awakening is difficult to interpret, even with Vaidyapāda’s commentary, but it appears to be a kind of fixation on the quiescent aspect of meditative experience.148 In any case, 147 da ni śākya thub pa la sogs pa las kyang tshogs dpag tu med pa bsags kyang gnyis su med pa’i ye shes bla ma’i zhal (zhal] D, kha P) las ma shes bar dam pa yang dag pa’i sangs rgyas ma yin par bstan pa’i phyir/ ci’i phyir śākya thub pa yis/ zhes (zhes] D, shes P) pa la sogs pa’o// ‘di’i don ma rtogs zhes pa ni zhal las brgyud pa’i rim pa ‘di ma rtogs bar zhes so// ci yang ma yin zhes pa ni thal byung gi stong pa nyid la gnas zhes so// de ston pa’i bla ma gsungs pa/ de tshe bde gshegs kun zhes so// thal byung blo can de bzlog ste// zhes pa ni bsam gtan ‘di ni ltung ba’i (ba’i] D, P om.) skal can te/ ‘dis mtha’i de bzhin nyid mi rtogs pa’o zhes de las blo bzlogs pa’o// ‘o na ji lta bur rtogs she na/ nam mkha’i dbyings ltar nam par dag pa’i zab gsal gnyis su med pa’i rang bzhin bstan te/ nam gyi phyed dus su lhan cig byed pa’i slob dpon gyis byin gyis brlabs nas rgyal (rgyal] sugg. em., brgyal D P) ba ltar de nyid thob nas/ de’i mthus tho rangs kyi dus su byang chub kyi snying po la zhugs pas dpung chen po’i bdud rnams bcom ste/ de nas sems can rnams la chos kyi ‘khor lo bskor te/ dus de nas bzung ste don chen po ‘di nyid skyes bu yang dag pa’i gzhi la brten pa’i skal ba dang ldan pa ‘ga’ la zhal nas zhal du ‘pho bar byas nas/ dnges pa’i don ‘di bstan to zhes (‘di bstan to zhes] D, P om.) ‘di ltar brgyud pa’i khungs kyang bstan to// (Muktitilaka-vyākhyāna, D 56b.5-57a.3; 148 This view is described in the Muktitilaka with a term—thal byung blo can, the meaning of which is difficult to decipher. I have provisionally translated it as “conceptuality [regarding emptiness] that had befallen [him].” The difficulty is in understanding precisely what is meant by the term thal byung, which usually has the sense of “sudden” or “suddenly arisen.” Kilty (2013, 467) in translating this passage as cited in Tsongkapa’s Lamp has translated the term thal ‘byung as “unimpeded,” which does not seem to me the most likely meaning of the term. In looking at the way the term is used in Buddhajñānapāda’s writings, and especially how it is interpreted in Vaidyapāda’s commentaries, I believe the term thal ‘byung here may perhaps translate the Sanskrit āpatita, which has the sense of something that has suddenly or unexpectedly happened, and, as we shall see, the conceptually that seems to have “(suddenly) befallen” a practitioner in the instances where the term thal byung blo can is used, seems to be connected specifically with a conceptualization with regard to emptiness. My conclusions with regard to this term, however, remain tentative, as I have had to be somewhat relaxed with grammar in order to make this reading work in some of the phrases where it appears. I am grateful to Harunaga Isaacson for his suggestion regarding a possible Sanskrit term. 114 Buddhajñānapāda uses the term thal byung zab mo, “a (suddenly) befallen profundity” earlier in the Muktitilaka in a passage where he rejects exclusively profundity or exclusively luminosity as being a cause for awakening and emphasizes the need for both. It is in that context that Buddhajñānapāda states, “I have no faith in the assertion that a suddenly befallen profundity is liberation” (...thal byung zab mo ni// grol ‘dod de la nga mi dad// (Muktitilaka, 47b.2). As I discussed earlier, the term profundity, in Buddhajñānapāda’s usage (as in the case of “the nonduality of profundity and luminosity”), indicates emptiness. Vaidyapāda’s comments on both of these terms, thal byung zab mo and thal byung blo can, suggest he understands the thal byung in the phrase thal byung blo can to refer to thal byung gi zab mo. In his comments on the passage describing Śākyamuni’s awakening Vaidyapāda clarifies that, “The samādhi of] Nothing Whatsoever is remaining in [a state of] suddenly befallen emptiness (thal byung gi stong pa nyid). To indicate the gurus who showed him [suchness], the text states At that time, all of the sugatas... Cast away the conceptuality [regarding emptiness] that had befallen him, means that this samādhi [of Nothing Whatsoever] occasions falling [into an extreme], and because of that he did not realize the ultimate suchness; so [they] turned his mind away from it.” (thal byung blo can de bzlog ste/ zhes pa ni bsam gtan ‘di ni ltung ba’i skal can te/ ‘dis mtha’i de bzhin nyid mi rtogs pa’o zhes de las blo blzog pa’o// (Muktitilaka-vyākhyāna, D 56b.6-7). Indeed, since Buddhajñānapāda mentions that Śākyamuni was remaining in the “samādhi of Nothing Whatsoever” before the sugatas cast away this state, we can indeed presume that this samādhi is equivalent to, or can perhaps be described as, thal byung blo can. There are two ways in which we might understand a samādhi with that name. One is as a reference to the third of the four formless concentrations, called “Nothing Whatsoever.” (This concentration is usually rendered in Tibetan as ci yang med pa’i skye mched (ākiñcanyāyatana), rather than ci yang ma yin pa’i ting ‘dzin, but we are dealing here with an Indic text in translation, so this discrepancy may not present much of a problem.) In some early Buddhist texts progression through the form and formless concentrations is said to be the process through which awakening takes place (See Vetter 1988, 63-71). In reference to thal byung zab mo mentioned earlier in the Muktitilaka, Vaidyapāda connects this position to that of the śrāvakas. He writes, “Since exclusively cultivating a suddenly befallen profundity (thal byung gi zab mo) brings about the circumstance of falling into [a state] that is like that of the śrāvakas, as for the assertion that this is liberation [Buddhajñānapāda states], I have no faith at all [in that.]” thal byung gi zab mo kho nar bsgom pa ni nyan thos lta bur ltung ba’i skal pa can pas/ de grol bar ‘dod pa de la nga ni shin tu yang ma dad do zhes so// (Muktitilakavyākhyāna, D 49b.3). In his commentary on the Caturaṅga/Samantabhadra-sādhana Vaidyapāda gives a further clue to his understanding of the term. Commenting on the phrase “Clinging to complete purification” he writes, “Complete purification is nirvāṇa. If one clings to that, then the mind is extremely attached to a suddenly befallen profundity” rnam par byang la mngon par zhen zhes gsung la rnam par byang ba ni mya ngan las ‘das pa’o// de la mngon par zhen pa ni thal byung gi zab mo la blo shin tu chags pa’o// (Samantabhadrī-ṭīkā, D 148a.2-3). Taken together, these references suggest that Vaidyapāda understands this state to be a view of emptiness that involves a sort of fixation on the quiescent experience of meditation—i.e. falling into the “extreme” of the peace of nirvāṇa, rather than cultivating a dynamic and compassionate realization that results in continued compassionate action for the benfit of beings, a common criticism of the śrāvakayāna by the Mahāyāna. Additionally, as I will discuss below, the passage in the Muktitilaka is clearly modeled on the awakening narrative from the Sarvatathāgatatattvasaṃgraha-tantra, and in that narrative the tathāgatas rouse Śākyamuni not from the “samādhi of Nothing Whatsoever” like in the Muktitilaka, but from the so-called āsphānakasamādhi, a term that remains etymologically obscure, but is usually interpreted to refer to a type of unmoving breath-holding concentration (See Yamada 1981, 7; Todaro 1985, 168-9; Kwon 2002, 51and 51n96; Tomabechi 2006, 140n160. Thanks to Jacob Dalton for bringing this point on the āsphānakasamādhi to my attention.). Understood as such, this samādhi, which is rejected by the Sarvatathāgatatattvasaṃgraha, likewise suggests the rejection of a non-dynamic approach to the awakened state, and thus correlates with Vaidyapāda’s reading of the rejection of the“samādhi of Nothing Whatsoever” and thal byung blo can in the Muktitilaka, as entailing a rejection of a śrāvaka position. While the term thal byung does not seem to be used frequently with reference to a perspective on emptiness, as we find in Buddhajñānapāda’s and Vaidyapāda’s writings, the Vajrahṛdayālaṃkāra-tantra, one of the explanatory tantras to the Guhyasamāja-tantra (probably somewhat later than both Buddhajñānapāda and Vaidyapāda) lists stong pa thal byung la rtog pa, “conceptualizing in regards to a suddenly befallen [experience of] emptiness,” as the eleventh of the fourteen root downfalls that constitute breaking tantric commitments (Vajrahṛdayālaṃkāra-tantra, D 446b.1), though the text offers no further clarification of what is intended by thal byung. The lists of the fourteen tantric samayas do vary from tradition to tradition, but there is generally some correlation between such lists with respect to each particular committment. One later formulation of the eleventh commitment is that it entails applying “discursive thoughts to transcendent reality,” (Kongtrül 1998, 263). An alternative understanding of the term thal byung gi stong pa is found in the writings of the 15th-century Tibetan master Śākya Chokden, who states that, “In Indian texts of quintessential instructions (man ngag, upadeśa) 115 once he had been turned away from such a mistaken perception and directly shown the “nature of nondual profundity and luminosity,” Śākyamuni trained in that suchness. Vaidyapāda here adds one further—but rather consequential—detail of his training that is not mentioned in the Muktitilaka: Śākyamuni was “blessed by the sahaja ācārya.” The “sahaja ācārya” (or sahaja guru—the terms are used interchangeably in Buddhajñānapāda’s and Vaidyapāda’s writings) is one of “three gurus” (bla ma gsum) mentioned at several points in Buddhajñānapāda’s works.149 While Vaidyapāda does not provide a definition of the sahaja ācārya in his commentary on Śākyamuni’s awakening narrative, at several other places Vaidyapāda identifies the sahaja ācārya as the tantric consort,150 and in the Sukusuma he writes that receiving her “blessing” means engaging in sexual union with her.151 Thus while neither Buddhajñānapāda nor and in sūtras, the emptiness explained by Niḥsvabhāvavādins was called the “inanimate emptiness” (bems po’i stong pa nyid), “nihilistic emptiness” (chad pa’i stong pa nyid) and “overextended emptiness” (thal byung gi stong pa nyid)” (Komarovski 2007, 284). I have been unable to find other instances of the term thal byung gi stong pa nyid in Indic sources, so it is not impossible that Śākya Chokden was actually referring to Buddhajñānapāda’s and Vaidyapāda’s works—which could definitely be qualified as “Indian upadeśa texts”—and interpreting the “overextended profundity” or “emptiness” that they refer to as a critique of the Niḥsvabhāvavādin (i.e. Mādhyamika) position on emptiness as “overextended” in the sense of tending towards nihilism. (Vaidyapāda also use the term chad pa’i stong pa nyid (“nihilistic emptiness”) in his Yogasapta, making it even more likely that Śākya Chokden was referencing this particular corpus with his remarks.) The samādhi of “Nothing Whatsoever” that the sugatas turned Śākyamuni away from could certainly also be interpreted to refer to such a (mis-)conception of emptiness. But Śākya Chokden’s understanding of the term thal byung gi stong pa nyid as a critique of the Madhyamaka position does appear at odds with Vaidyapāda’s interpretation of it as a critique of the śrāvaka view, and as we will see below, Buddhajñānapāda in fact held the [Yogācāra-] Madhyamaka philosophical position to be the highest among the various Buddhist philosophies. In any case, Vaidyapāda’s testimony certainly carries more weight in this case given his close proximity to Buddhajñānapāda, so I have privileged his understanding in my translation of the term in the Muktitilaka. Buddhajñānapāda’s own writings unfortunately give us no substantial context for what he meant by the term thal byung—though he did find that perspective important enough to mention—and reject—twice in the Muktitilaka. 149 The “three gurus” (bla ma gsum) are mentioned in the first verse of the Dvitīyakrama. Vaidyapāda explains that these three are the causal, conditional, and sahaja ācāryas (de yang gsum ste/ rgyu dang rkyen dang/ lhan cig byed pa’i slob dpon no/) (Sukusuma, D 88a.6; P 106a.1). The sahaja ācārya is mentioned by Buddhajñānapāda himself in verse 142 of the Dvitīyakrama, and Buddhajñānapāda mentions the “three gurus” as a set, but without listing them, in the Muktitilaka, as well. Vaidyapāda gives the same gloss on the identity of these three in the Muktitilakavyākhyāna as he does in the Sukusuma. In both the Sukusuma and the Muktitilaka-vyākhyāna Vaidyapāda additionally provides a citation of a passage about the three gurus from a work that he identifies in the Muktitilakavyākhyāna as The Precious Garland (rin chen phreng ba (phreng ba] P, phrod pa D); I have been unable to identify this source). In the Sukusuma Vaidyapāda mentions that this passage was cited by Buddhajñānapāda himself on this topic (possibly in the context of oral instructions, since the citation is not found in any of Buddhajñānapāda’s surviving writings). There are some slight variations in the transmission of the verse in the Sukusuma and Muktitilaka-vyākhyāna, but in summary the verse identifies the causal ācārya as the master who gives vows and commitments and who purifies one’s mind through the stages of initiation, beginning with the water initiation; the conditional ācārya as the “great goddess” with whom one engages in play and who purifies the field of one’s mind by means of the “sixteenth part;” and the sahaja ācārya as the one from whom one receives that (bindu?) and by means of whom and through whose blessing one realizes innate joy. Vaidyapāda further adds that these three are supreme because they are superior to other gurus (Muktitilaka-vyākhyāna, D 47b.5-7). The difference between the conditional and the sahaja ācāryas is difficult to understand from the passage that Vaidyapāda cites, as both seem to refer to the tantric consort. However, in his Yogasapta, Vaidyapāda states that the kalaśābhiṣeka is bestowed by the causal ācārya, the guhya initiation is bestowed by the causal and the conditional ācārya, and the prajñājñāna is bestowed by means of the causal, conditional and sahaja ācāryas (Yogasapta, D 70a.4; D 70a.7; D 70b.4). This suggests that the “conditional” ācārya may be the consort in the role as the partner of the guru for the guhya initiation, while the sahaja ācārya is the consort in her role as the disciple’s partner in the prajñājñāna initiation. Later in the Sukusuma, Vaidyapāda clearly states that the sahaja ācārya is the practitioner’s consort (shes rab, prajñā), and that uniting with her entails receiving her “blessing” (Sukusuma, D 111b.3-4; P 134a.6-7). 150 See Muktitilaka-vyākhyāna, D 47b.5-7 and Sukusuma, D 88a.6-7, D 111b.3-4, D 139a.1. See also previous note. 151 Sukusuma, D 139a.1. 116 Vaidyapāda make any explicit reference to a sexual initiatory context—or indeed any initiatory context at all—Vaidyapāda’s commentary on Śākyamuni’s awakening narrative in the Muktitilaka does essentially claim that after being shown nonduality by the sugatas, Śākyamuni trained in suchness by practicing sexual yogas with a consort, which led to his full awakening. And though Buddhajñānapāda himself did not explicitly include sexual yoga in his account of Śākyamuni’s awakening, sexual yoga does constitute an essential component of both the initiatory context in which suchness is “transferred,” and the higher tantric path of perfection stage practice that Buddhajñānapāda’s writings espouse. Thus, despite the fact that such sexual elements are not explicitly highlighted by Buddhajñānapāda here, it is not difficult to imagine that they were intended. In either case, reading this awakening narrative as a condensed presentation of the perfection stage path is crucial to understanding its function in Buddhajñānapāda’s oeuvre. To more fully appreciate the Muktitilaka’s narrative of Śākyamuni’s awakening, we need to look to earlier tantric literature, for Buddhajñānapāda’s retelling of the awakening narrative is not the first reimagining of Śākyamuni’s awakening. It is in reference to an earlier such account, found in the Sarvatathāgatatattvasaṃgraha-tantra, that we can best understand Buddhajñānapāda’s own retelling of this story. The Sarvatathāgatatattvasaṃgraha-tantra’s account of Śākyamuni’s awakening appears to be the earliest specifically tantric re-envisioning of the awakening narrative, in which certain tantric methods—in this case the practice of the “five manifestations of awakening” (pañcākarābhisambodhi152), through which the practitioner generates herself in the form of the deity—are shown to be an essential component of the path to awakening.153 In that narrative, as well, the tathāgatas first appear and rouse the bodhisattva from a samādhi that will not lead him to awakening. They ask how he will attain perfect awakening without knowing the “suchness of all the tathāgatas” (sarvatathāgatatattva), and when he responds by asking what suchness is and how it is accomplished, they guide him through the stages of the five manifestations of awakening, leading to his full awakening. We know that Buddhajñānapāda was familiar with the narrative from the Sarvatathāgatatattvasaṃgraha-tantra because he cites a passage from precisely this section of the tantra in his Ātmasādhanāvatāra. It thus appears that in the Muktitilaka Buddhajñānapāda intentionally followed the Sarvatathāgatatattvasaṃgraha-tantra’s model of rewriting the narrative of Śākyamuni’s awakening in order to emphasize a specific method or process that leads to perfect awakening. The tantric method introduced as essential in the Sarvatathāgatatattvasaṃgraha-tantra’s awakening narrative is, as noted above, the practice of deity yoga. In the late 7th century when the Sarvatathāgatatattvasaṃgraha-tantra was first circulated, the techniques of deity yoga were at the cutting edge of Buddhist tantric practice. By the time Buddhajñānapāda was writing, probably in the early 9th century, tantric practice had developed further, and deity yoga was now relegated to the “first stage” of tantric practice, the so-called “generation stage.” The newest techniques on the scene in Buddhajñānapāda’s time were the practices of what was being described as the “second stage,” the “higher stage,” or the “perfection stage” of tantric practice. 152 While this term itself does not occur in the Sarvatathāgatatattvasaṃgraha-tantra, the practices that it describes do. Thanks to Harunaga Isaacson for pointing out to me the fact that the term itself is absent in this tantra. 153 See Snellgrove 2002, 120-21, Weinberger 2003,185-89, and Onians 2003, 80-81 for a description of and analysis of the Sarvatathāgatatattvasaṃgraha’s awakening narrative, and Lessing and Wayman 1978, 25-35 for a Tibetan edition and English translation of the Tibetan scholar Khedrub Je’s (1385-1438) summary of the Yoga Tantra position on Śākyamuni’s awakening (i.e. the narrative from the Sarvatathāgatatattvasaṃgraha-tantra), including the positions found in the commentaries on the Sarvatathāgatatattvasaṃgraha-tantra’s narrative by three Indian commentators, Ānandagarbha, Śākyamitra, and Buddhaguhya. See Onians 2003, 78-80 for a summary of Khedrub Je (via Lessing and Wayman’s translation). 117 The Sarvatathāgatatattvasaṃgraha-tantra’s narrative of Śākyamuni’s awakening had already shown that the generation stage was essential to bring about to full awakening. The framework of Buddhajñānapāda’s narrative in the Muktitilaka is essentially the same as that of the Sarvatathāgatatattvasaṃgraha-tantra, but while the tantra’s emphasis is on Śākyamuni’s being taught and following the method of the five manifestations of awakening associated with the deity yoga of the generation stage, the emphasis in the Muktitilaka’s narrative is on the sugatas directly showing Śākyamuni the “nondual profundity and luminosity” so that he is then able to train in it. Buddhajñānapāda’s narrative in the Muktitilaka—in describing the process of Śākyamuni’s awakening as precipitated by his being shown nondual profundity and luminosity by the sugatas, who acted as his gurus, and achieving awaking through training in that suchness—thus shows the perfection stage to be essential to the attainment of full awakening. Given the centrality of sexual practice in both the perfection stage initiations during (or after) which suchness was communicated to the disciple, as well as in the sexual yogas of the subsequent training in suchness, the absence of any explicit reference to sexual elements in Buddhajñānapāda’s account of Śākyamuni’s awakening is notable. As I mentioned above, this may be in part due to Buddhajñānapāda’s wish to focus on the very fact of the transference of suchness from a guru and the disciple’s subsequent training in it, rather than the specific methods by which such a transference and cultivation took place, as being the most essential aspects of the perfection stage path. Again, the focus on lineage at the conclusion of the narrative emphasizes the great importance of this aspect of the account. However, the absence of sexual elements in the narrative may also be due to the potentially controversial nature of explicitly claiming that sexual practice was involved in the awakening of the founding figure of the Buddhist tradition, who is well known to have left behind the householder’s lifestyle for that of a celibate renunciate in search of awakening, and to have founded an order of celibate monastic practitioners after his awakening. And yet, apart from the sexual elements that are omitted in the narrative, the description of Śākyamuni’s awakening in the Muktitilaka hews so closely to the structure of the higher tantric path, which, when elaborated in more detail in Buddhajñānapāda’s works undeniably involves sexual practice, that it would be no stretch at all for a reader familiar with Buddhajñānapāda’s practice system to understand those sexual elements as constituting an unstated part of the narrative. Indeed, we saw that Vaidyapāda has clearly understood the narrative in this way, but even he makes the sexual aspects of the account only slightly more explicit, simply stating, “Then, at midnight, blessed by the sahaja ācārya, he attained suchness like the [other] Victors.” Without familiarity with the term “sahaja ācārya,” a reader would be unaware of the reference to a tantric consort. In fact, even though at three points in the Sukusuma and once in the Muktitilaka-vyākhyāna Vaidyapāda expressly defines the sahaja ācārya as the consort, he leaves the term unexplained here. Perhaps, despite initiations involving sexual elements and the practice of sexual yogas being otherwise unabashedly advocated in Buddhajñānapāda’s and Vaidyapāda’s writings, to very explicitly associate these kinds of practices with Śākyamuni Buddha would have been too radical at the time they were writing. But the use of Śākyamuni’s awakening narrative to advocate a specific tantric path, which had already been done in the Sarvatathāgatatattvasaṃgraha-tantra, apparently was not. Steven Weinberger has argued that the Sarvatathāgatatattvasaṃgraha-tantra’s account of Śākyamuni’s awakening, in depicting tantric techniques as essential to the attainment of awakening, is a crucial moment in the development of Buddhist tantra, representing “tantra’s coming out party, its “declaration of independence” as its own tradition, distinct from earlier Buddhist traditions.”154 But while the Sarvatathāgatatattvasaṃgraha-tantra’s narrative may 154 Weinberger 2003, 189. 118 have been the first specifically tantric re-imagining of Śākyamuni’s awakening, and seems clearly to have served as the impetus for Buddhajñānapāda’s own retelling of that narrative, Buddhists had been emphasizing specific doctrines via differing accounts of the awakening narrative since the earliest days of the tradition. De La Vallée Poussin, Schmithausen, and Vetter have all examined a number of different doctrinal positions in early Buddhist sūtras and commentaries regarding the content of Śākyamuni’s and his disciples’ awakening and the process by means of which it is said to have occured.155 These various doctrinal positions are not our concern here, but it is worthwhile to note that while none of these scholars draws particular attention to the way in which such doctrines are conveyed, a number of the sources they draw upon are indeed narrative accounts of Śākyamuni’s awakening, within which the differing doctrines on the content and process of his awakening are ensconced.156 Thus while the Sarvatathāgatatattvasaṃgraha-tantra’s awakening narrative may have been the first tantric version, it was certainly not the first occasion in which Śākyamuni’s awakening narrative was used to set forth a particular doctrinal system. Likewise, Buddhajñānapāda was not at all the last to rework the account in accordance with his own doctrinal positions. The Anuttarasandhi composed by Buddhajñānapāda’s disciple Śākyamitra and included as the second chapter of Nāgārjuna’s Pañcakrama, includes its own such account— apparently modeled on Buddhajñānapāda’s—which describes Śākyamuni’s awakening in terms of the prabhāsvara doctrine of the Ārya School.157 Āryadeva’s Caryāmelāpakapradīpa likewise contains a brief such account that alludes to the third initiation and references the Ārya tradition’s prabhāsvara doctrine.158 By the time of the much later Caṇḍamaharoṣaṇa-tantra, which emphasizes the worship of women along with the crucial nature of bliss as a cause of awakening, attributing Śākyamuni’s awakening to sexual practices was apparently no longer too radical, as seems to have been the case in Buddhajñānapāda’s time. The retelling of the awakening story in the Caṇḍamaharoṣaṇa-tantra explains that the traditional narrative of Śākyamuni’s abandoning his wives and harem and going to the banks of the Nairañjanā to manifest full awakening did not actually take place “from the perspective of the absolute truth...since it was in the female quarters, that the Buddha, experiencing pleasure in the company of Gopā, became accomplished,”159 because “awakening is attained through pleasure, and there is no pleasure in being separated from women.”160 Such a reading of the awakening narrative goes so far as to render Śākyamuni’s renunciation irrelevant from the perspective of his personal accomplishment (though of course in such a narrative the act of renunciation—like the entire “display” of awakening itself according to many Mahāyāna accounts in which Śākyamuni is said to have already been liberated prior to even taking birth in this world—can be understood to have an extrememly important didactic function for his disciples). It is likely that there are also other 155 De la Vallée Poussin 1936-7; Schmithausen 1981; Vetter 1988. 156 See, for example, the accounts of the awakening in the Dhammacakkapavattana-sutta, Mahāvagga, Mahāsaccaka, and Aṅguttara IV, referenced in their works. 157 See Tomabechi 2006, 140-141 for the Sanskrit edition and French translation of this passage in the Anuttarasandhi. See Kilty 2012, 463-65 for an English translation of a citation of part of this passage from the Anuttarasandhi in Tsongkhapa’s Lamp to Illuminate the Five Stages. As discussed in Chapter One, I agree with Tomabechi’s assessment that Śākyamitra based his account in the Anuttarasandhi on Buddhajñānapāda’s (Tomabechi 2006, 139n158). See also note 161 with reference to another Ārya School presentation of Śākyamuni’s awakening in the Caryāmelāpakapradīpa. 158 Wedemeyer 2007, 262. See also note 161 below. 159 Caṇḍamahāroṣaṇa-tantra, Chapter Ten, verses 26d-27b. The fuller account, part of which I have summarized here, is recounted in Chapter Ten, verses 25-30. See also Onians 2003, 73-77 for an analysis of this passage. 160 Caṇḍamahāroṣaṇa-tantra, 10.28a-b. 119 such accounts of the awakening narrative in tantric Buddhist literature that similarly recast the mode of Śākyamuni’s awakening to correspond with their own doctrinal and ritual systems.161 Following the awakening narrative proper, the conclusion of the Muktitilaka’s account in which it is explained that from the time of Śākyamuni onwards the lineage of “this supreme reality was passed on...to a few fortunate ones from mouth to mouth, from ear to ear,” is also remarkable. In making this statement Buddhajñānapāda appears to be connecting the special oral lineage of the teachings on the reality of “non-dual profundity and luminosity” directly to Śākyamuni Buddha, which might not seem unusual in a Buddhist text. But it comes precisely in a period in which the tantras themselves were becoming distanced from the historical Buddha and instead connected to cosmic buddhas like Bodhicittavajra, one of the more frequently used epithets for the buddha who is the teacher of the Guhyasamāja-tantra. In that context, such a reassertion of a direct lineal connection with Śākyamuni is indeed somewhat unusual.162 The tantras’ gradual distancing from the historical Śākyamuni and connection with other buddhas as teachers was presumably due at least in part to the increasily antinomian nature of the practices advocated therein. The opening narrative of the Guhyasamāja-tantra locates the buddha who is 161 Khedrub Je’s Introduction to Tantric Systems (Rgyud sde spyi’i rnam par gzhag pa rgyas par brjod, translated, with annotation, in Lessing and Wayman, 1978) gives a fascinating summary of the varying accounts of Śākyamuni’s awakening according to the various Buddhist vehicles, including the distinctions in the account according to different tantric systems, and even differentiating among the position of different Indian Yoga tantra commentators (Lessing and Wayman 1978, 17-40). In this account, Khedrub explains that the position of the Ārya School of Guhyasamāja is found in the Caryāmelāpakapradīpa, and that of the Jñānapāda school is found in the Dvitīyakrama and the Mukhāgama, though Buddhajñānapāda’s account, in fact, does not appear in either of these, but rather in the Muktitilaka. Khedrub further states there is no divergence between the position of the two schools (ibid., 35). The account of Śākyamuni’s awakening in the Caryāmelāpakapradīpa is described in the context of the third initiation (the account begins just after the Caryāmelāpakapradīpa’s own presentation of that initiation and the author begins the presentation of Śākyamuni’s awakening with the words “by this process,” suggesting that he also received the third initiation at this point) and includes the Ārya School practices of prabhāsvara and the māyopama samādhi. It also concludes with a statement, like Buddhajñānapāda’s, of the teachings being passed on in a lineage from “mouth to mouth” since that time (Wedemeyer 2007, 262). And while we saw that Buddhajñānapāda’s account in the Muktitilaka does not explicitly reference the third initiation, Vaidyapāda has understood it to include that practice, and as such Khedrub’s claim that the two traditions are in accord on this account is indeed substantiated. The account that Khedrub gives, however, is much more detailed than that found in either the Muktitilaka or the Caryāmelāpakapradīpa, and includes the bestowal of the third initiation with a consort whose name is given as Tilottamā. This account also focuses on the development of the prabhāsvaras of the Ārya school (ibid., 37). The same account is also found in Ngawang Palden’s Illumination of the Texts of Tantra (Gyatso 2006, 114). In his Lamp for the Five Stages, Tsongkhapa cites Abhayākaragupta’s Āmnāyamañjarī where it is mentioned that, “having received the wisdom initiation given by the perfect buddhas abiding in the skies above, the Śākya master relied upon it in order to attain enlightenment while sitting in front of the Bodhi tree. Also, the Bhagavan, by this alone, was initiatiated into the vajra essence by Tilottamā” (Kilty 2013, 468; Kilty (213, 596n849) has identified the citation in Tōh 1198, 34b.1). Whatever the original source of this account in which Śākyamuni’s partner in his the final stepts towards awakening is named Tilottamā, this is quite an interesting choice of name for Śākyamuni’s tantric consort; Tilottamā is a famous apsaras from the Mahābhārata and other purāṇic myths, known for her seductive beauty (see Ludvik 2007, 120n17 and Nihom 1995) and, along with the apsaras Rambhā, her propensity for “distract[ing] advanced sages from their development of tapas” (English 2002, 92). How fitting that she should act as precisely the condition for Śākyamuni’s awakening, rather than an obstacle to his path. Tilottamā appears as an apsaras in both the Hevajra and Sampuṭa-tantra (Regarding the Hevajra see Snellgrove 1958, 94; Farrow and Menon 2001, 281 and Nihom 1995; Regarding the Sampuṭa-tantra see 84000: Translating the Words of the Buddha forthcoming), and in a sādhana of a form of Vajrayoginī, called the Guhyavajravilāsinī-sādhana, described by Elizabeth English in which Tilottamā and Rambhā are involved in the consecration process of the tantric yogin and his partner, in a similar inversion of their ordinary role as obstacles to a yogin’s practice (English 2007, 92). 162 In fact, in this particular regard it seems to share more in common with the lineage narratives from the Chan tradition which, precisely in the 8th century, were setting forth lineages connecting the teachings of Bodhidharma to a chain of masters going back to Śākyamuni. On 8th-century Chan lineages see Yampolsky 1967, Foulk 1992, Adamek 2007, Sharf forthcoming. 120 its teacher “in the bhagas of the vajra goddess” rather than in a geographical location in India as the earlier sūtras had done; it would likely have been awkward for such a narrative to feature Śākyamuni. As noted above, the practices advocated in Buddhajñānapāda’s system certainly did include antinomian elements; in the present discussion I have drawn attention to their sexual elements, but his writings also include references other important antinomian aspects of the Mahāyoga tantras, including the use and consumption of impure substances (sexual and otherwise). But, as we have seen, Buddhajñānapāda omitted any direct reference to such features in his account of Śākyamuni’s awakening. What is more, as I will demonstrate in the next section of this chapter, Buddhajñānapāda appears to have held the position that it was only through tantric practice that full awakening could be obtained, and that the realization attained through the tantric path surpassed that of all lower paths, including even the traditional Mahāyāna path of the bodhisattvas. Given these claims, it seems that Buddhajñānapāda’s linking of the special oral lineage of the teachings of the suchness of the perfection stage directly to Śākyamuni must take into account his statement that this lineage has, since the time of Śākyamuni, been passed only to “a few fortunate ones.” That is, given his claims of the superiority of the tantric perfection stage path over other Buddhist paths, linking its lineage directly to Śākyamuni and stating that it has been passed only to “a few fortunate ones,” Buddhajñānapāda is essentially claiming that this tantric path came from Śākyamuni himself, but that unlike the śrāvaka vehicle or the general Mahāyāna, it was not taught publicly to the majority of disciples and was instead passed on quietly in an unbroken lineage up to the present time. While not specifically referencing Śākyamuni, the first few verses spoken by Mañjuśrī in the Dvitīyakrama also link the oral teachings on the perfection stage that Mañjuśrī passes on to Buddhajñānapāda to the sugatas, and emphasize that in both the past, present, and future its teachings are only given to “some” worthy disciples.163 In avoiding an emphasis on the antinomian aspects of the tantric practices that Buddhajñānapāda held as necessary for the attainment of awakening, he was able to link the oral lineage of the suchness of the perfection stage to Śākyamuni directly, but without making overtly controversial statements. And in linking that oral lineage directly to Śākyamuni but stating that it was passed on only to a few disciples, Buddhajñānapāda was able to connect the tantric perfection stage teachings to the historical Buddha, and still maintain his position on the superiority of tantra over non-tantric paths. Buddhajñānapāda’s account of Śākyamuni’s awakening in the Muktitilaka, then, draws together the most important elements—or at least those that were not too controversial to mention in conjunction with Śākyamuni—of the higher tantric path according to his practice system, and demonstrates that even Śākyamuni Buddha achieved awakening by these means. A stronger statement of advocacy for this system is difficult to imagine, even within the extremely rich Buddhist imaginaire. The presence of this narrative in his ouevre indicates that it was important for Buddhajñānapāda to advocate for the system of the higher stage of tantric practice, and thus suggests that his writings may have been among the earlier works to set forth such a system. The Dvitīyakrama and the Muktitilaka are, in fact, some of the earliest surviving 163 Mañjuśrī states: a163 vi yaṃ raṃ vaṃ laṃ hūṃ a la la la ho!/ The great compassionate ones, / Who have realized this, / Those vajra holders of the past, present, and future/ |20| Who obtained the excellence of the sugatas,/ Have taught, teach, and will teach [this truth]/ To [only] some worthy [disciples]./ In order that they may realize the genuine meaning,/ |21| I will teach this to you— / Concentrate your mind and listen!/ |22| . a bi yaṃ raṃ baṃ laṃ hūṃ a la la la ho// rje btsun thugs rje che rnams kyis// ‘di rtogs pa yis bde gshegs kyi// phun sum tshogs pa ‘thob ‘gyur pa’i// |20| ‘das dang da ltar ma byon pa’i// rdo rje ‘dzin pas snod ‘ga’ la// gsungs shing gsung dang gsung ‘gyur ba// yang dag don rab rab rtogs phyir// |21| nga yis khyod la bstan par bya// yid gcig bsdus la mnyan par gyis// |22| (Dvitīyakrama, verses 20-22). 121 Buddhist texts to describe the stages of a perfection stage practice system.164 Attributing the essential components of the perfection stage path to the awakening of Śākyamuni himself, this narrative indicates the importance of tantric practice, and specifically the practice of the perfection stage in Buddhajñānapāda’s thought. In fact, there are a number of other passages in his writings that show that Buddhajñānapāda held the position that not only is tantric practice primary, but it is actually an essential component of the path to full awakening. Let us move on to examine some of these statements now. IV: Privileging Tantric Practice: The Superiority of the Tantric Path and Its Result In order to accomplish great awakening you must experience great bliss with the girl who liberates and gives joy. Nothing else can bring about buddhahood; this girl is the genuine supreme. -Mañjuśrī instructing Buddhajñānapāda, Dvitīyakrama Historically, tantric Buddhist scriptures and authors have held quite a number of different positions on the ways in which tantra is superior to non-tantric paths. One common position is that the result of the tantric path is identical to that of non-tantric Mahāyāna, but tantra remains superior because its special methods more blissfully produce a swifter result. A well-known and often-cited (at least in the later Tibetan tradition) quotation from Trivikrama Bhaṭṭa’s165 Nayatrayapradīpa conveys this position. Here I cite the verse along with the surrounding comments in Onians’ translation: Since this [Mantranaya] has the same goal as the Pāramitānaya, it should be explained what its distinction is: Although the goal is the same, the Tantric teaching is superior to [the Pāramitānaya] because it is not confused, has many means, is not difficult, and is appropriate for those of keen faculties. Thus, although there is no difference between the two Mahāyānas, that of mantras and that of perfections, with regard to the result, described [already above for example] as nondual omniscience, nevertheless, the mantramahāyāna is superior to the pāramitāmahāyana by virtue of those four distinctions.166 An alternative position is that in addition to its superior methods, the final result of tantric practice itself surpasses that of the non-tantric path. Buddhajñānapāda’s oeuvre indicates that he held the latter view. While, as we have seen above, Buddhajñānapāda does suggest that the tantric path outlined in his writings is meant for just “a few” fortunate disciples—presumably those of “keen faculties,” as Trivikrama Bhaṭṭa states—he also asserted the tantric Buddhist perspective, path, and its result to be superior to that of non-tantric Buddhism, including even the 164 Tanemura (2015, 329) even suggests that Buddhajñānapāda “was probably the first person who integrated the two systems of meditation [i.e. the generation and perfection stages] into Buddhist tantric practice.” As I will discuss further in Chapter Five, it seems to me difficult to attribute such a significant development to a single individual, but Buddhajñānapāda was undoubtedly one of the first authors to structure tantric practice in terms of the two stages. 165 The name of this author is given in Onians (2001) as Tripiṭakamala and Sanderson (2009, 233n536) as Tripiṭakamalla, presumably based on the name as given in the colophon of the Tibetan translation of the Nayatrayapradīpa. However, since then a manuscript of the Nayatrayapradīpa that gives the author’s name as Trivikrama Bhaṭṭa has turned up. Thanks to Harunaga Isaacson for bringing this to my attention. 166 Onians 2001, 97-98. While Onians studies Trivikrama Bhaṭṭa’s work in some detail in her dissertation, she does not discuss his dates, presumably because that information is unavailable. The Nayatrayapradīpa was translated into Tibetan by Padmākaravarman and Rinchen Zangpo, so Trivikrama Bhaṭṭa must have lived during or prior to the 11th century, but without a more detailed study of what appears to be his single surviving work, I am unable to say anything more about his dates. 122 traditional bodhisattva path of exoteric Mahāyāna. Buddhajñānapāda also appears to have held that within tantric practice the practice of the perfection stage is not only superior to that of the generation stage, but actually essential for the attainment of full awakening. In addition to several rather direct statments of the fact, the ways in which Buddhajñānapāda demonstrates tantric superiority are wide-ranging. He indicates the superiority of tantric and perfection stage ritual practices over non-tantric or non-perfection stage practices through drawing comparisons between sets of practices, describes the tantric view as superior to non-tantric philosophical positions, frequently refers to the ultimate nature of reality not just as suchness, but specifically as the “suchness of the second stage,” and uses unique terminology to describe the final result of awakening achieved through the tantric path. By means of these diverse statements and indications we can see clearly that Buddhajñānapāda held the higher stage of the tantric path to be an essential—and indeed the only—means to the attainment of complete awakening. Tantric Pefection Stage Practice is the Only Means to Awakening In Buddhajñānapāda’s writings we find several rather direct statements to the effect that it is only through the practice of certain tantric methods that one is able to attain complete awakening. One of these is found in his Ātmasādhanāvatāra, which contains a passage defending the practice of deity yoga against an unnamed interlocutor. That passage includes the following verse:167 Therefore, he who meditates upon himself as Samantabhadra, shining with full radiance, And concept-free, he alone partakes in awakening.168 This statement is followed by a citation of a block of verses from the Sarvabuddhasamāyogatantra advocating the practice of deity yoga, though unlike Buddhajñānapāda that tantra does not specify that such a practice is the only way to awakening, simply that reliance upon this practice will bring about awakening in this very life, even for someone who has failed for billions of aeons to attain that state.169 Given the reference in Buddhajñānapāda’s verse to meditating upon oneself as the deity170 and its context in a passage defending deity yoga, it seems that in the Ātmasādhanāvatāra Buddhajñānapāda is advocating the tantric practice of deity yoga as essential to attaining awakening. However, three passages in the Dvitīyakrama suggest that even the practice of the deity yoga of the generation stage is not sufficient for the attainment of full awakening. All three of these passages focus on the necessity of the sexual yogas connected to the perfection stage for 167 The Sanskrit appears to be in verse, but the Tibetan translators did not render it as such. 168 Emphasis mine. tasmān nirastasaṃkalpaṃ samantaspharaṇatviṣam | Samantabhadram ātmānaṃ bhāvayann eva bodhibhāk || (Szántó unpublished, 147). de bas na kun tu rtog pa spangs pa ‘od zer ma lus pa spro ba can gyi kun bzang po’i bdag nyid sgom (sgom] D, bsgom P) par byed pa/ de kho na byang chub kyi snod yin te/ (Ātmasādhanāvatāra, D 56a.1-2; P 67b.2-3). Thanks to Harunaga Isaacson for pointing out that this verse seems to deliberately echo the first verse of Dharmakīrti’s Pramāṇavārttika, a point I discuss further in Chapter Eight. 169 Buddhajñānapāda does not name his source. I am grateful to Péter Szántó for sharing his draft Sanskrit edition of the Ātmasādhanāvatāra in which the source of these verses is identified. The verses cited are from the Sarvabuddhasamāyoga’s first chapter, verses 1. 118a-119a followed by yāvat (nas) and verse 1.124c-d (Sarvabuddhasamāyoga, D 152a). 170 It is unclear what precisely is meant by “Samantabhadra” here. In Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana—which features the central deity Mañjuvajra, not Samantabhadra— Samantabhadra (when not acting as the bodhisattva Samantabhadra, in which role he also appears in that sādhana(!)) seems to be equivalent to Vajrasattva as a generic name for, or source of, the yidam deity (see Caturaṅga-sādhana D 37a.2 where Samantabhadra is the source of the dharmodaya; Vaidyapāda and Samantabhadra (the commentator, not the deity!) identify Samantabhadra here with Vajrasattva), as well as a name/term referencing the innate nature (see Caturaṅga-sādhana, D 42b.3). 123 gaining complete realization. The first of the two, which precedes a discussion on the four types of tantric consort says, That which is luminous and joyful, equal to space— One will not know it any other way. Thus, a woman, the illusory mudrā, Is superior among all illusions.171 |50|172 Vaidyapāda explains that the subject of the verse is nondual wisdom,173 and that the only way to come to know it is through the seven yogas, which as we saw above involve sexual practice.174 Whether or not Buddhajñānapāda meant to specify the seven yogas, he is clearly stating that engaging in practice with a woman is necessary to gain the realization of “that which is luminous and joyful, equal to space.” Yet another passage from a section of the Dvitīyakrama that lays out the ritual procedure for the third initiation includes several verses that form part of a liturgy to be spoken by the guru to the disciple during initiation. In these lines, the guru states that buddhahood can only be attained through practicing with the consort who is given to the disciple as part of the third initiation. It seems that these lines became part of a popular initiation liturgy, as several lines from these verses—and indeed quite a bit of the section on the third initiation from the Dvitīyakrama—have been adapted into quite a number of later tantric texts, including the Samājottara and the Vajrāvalī.175 As he gives the consort to the disciple at the beginning of the ritual the guru says, “This goddess is suitable for you. Great being,176 all of the buddhas have given177 This delightful girl to you to enjoy178 By means of your desire |87| Through the ritual for the maṇḍala-cakra.179 171 bud med sgyu ma’i phyag rgya ni// sgyu ma kun las khyad par ‘phags//. These two lines have strong parallels with the first two lines of the Chapter One, verse 4 of the Sarvabuddhasamāyoga-tantra, which read (in Tibetan) sgyu ma dag ni thams cad pas// bud med sgyu ma khyed par che// (D 151a.3). The Sarvabuddhasamāyoga also mentions the woman as a mudrā in the last two lines of the immediately preceding verse: bud med kun gyi sgyu ma’i rgya// ‘di ni gnyis med theg pa’i mchog// (D 151a.2). Thanks to Ryan Damron for bringing the Sarvabuddhasamāyoga reference to my attention. The two lines from the Dvitīyakrama are also strongly paralleled in Śākyamitra’s Anuttarasandhi, included as the second stage in Nāgārjuna’s Pañcakrama which reads: sarvāsām eva māyānāṃ strī-māyaiva viśiṣyate/ (Mimaki and Tomabechi 20); sgyu ma dag ni thams cad las// bud med sgyu ma khyad par ‘phags// (Pañcakrama, D 49a.7; Mimaki and Tomabechi 20). Tomabechi (2006, 132n128) has already noticed these parallels and additionally notes that a passage identical to the one from the Pañcakrama is found in the Vajramaṇḍālaṃkāra. 172 gsal shing rab dga’ mkha’ mnyam pa// gzhan du rig par mi ‘gyur bas// bud med sgyu ma’i phyag rgya ni// sgyu ma kun las khyad par ‘phags// |50| (Dvitīyakrama, verse 50). 173 gnyis su med pa’i ye shes (Sukusuma, D 99a.7). 174 Sukusuma 99a.7-b.2 175 See notes from the citation below for more details. In an earlier conference paper (C. Dalton 2014) I argued that the Samājottara is later than, and incorporates elements from, Buddhajñānapāda’s writings. I will lay out the arguments from that paper, along with some additional evidence, in Chapter Eight. 176 sems chen] S P N V(P), sems can D C V(D); Vaidyapāda’s commentary also suggests that sems chen is the better reading: “Great being means someone who has the intention to liberate sentient beings.” sems chen (chen] P, can D) zhes pa ni sems can bsgral ba’i sems gang la yod pa’o// (Sukusuma, D 105b.7-106a.1; P 107b.1). 177 gnang] D C V(D and P), snang S P N; The Peking edition of Vaidyapāda’s commentary cites the line from the verse with snang, but then in the explanation of the verse uses the correct spelling, gnang. 178 ~Cf. Vajrāvalī (Mori 2009 Vol. 2, 444). This line could also be understood as “to practice with.” Vaidyapāda writes, “Given by the buddhas to enjoy means that the unsurpassed buddhas give [a woman] to some suitable disciples to enjoy.” sangs rgyas kun gyis spyod du gnang (gnang] D, snang P) / zhes pa’i bla na med pa’i sangs rgyas rnams kyis snod du rung ba’i gang zag ‘ga’ la spyod du gnang ba’o// (Sukusuma, D 106a.1; P 127b.1-2). 124 In order to accomplish great awakening You must experience great bliss180 [With] the girl who liberates and gives joy. |88| Nothing else can bring about buddhahood This girl is the genuine supreme181 Thus, throughout endless saṃsāra You must never separate from her.” |89|182 As the passage makes very clear, engaging in sexual practice together with a consort is an essential part of the path, since “nothing else can bring about buddhahood.” Later in the Dvitīyakrama sexual yoga is described as superior to the traditional means of cultivation of the ten bodhisattva bhūmis, as well as the practice of the generation stage. Like the passage above in which the guru tells the disciple that practice with a consort is necessary for the attainment of buddhahood, the passage on the ten bhūmis of sexual practice again indicates that any of the “lower” practices not involving sexual yogas will not produce the final result of perfect awakening. Following a passage in which each of the ten bodhisattva bhūmis is, one by one, homologized with various stages of sexual practice,183 the Dvitīyakrama states: By means of these ten 179 Vaidyapāda seems to understand this as referring to the rituals of the generation stage practice, as he specifies that this means the ādiyoga-samādhi—the first of the three samādhis of generation stage practice—“and so forth.” The Sukusuma reads, “By means of the ritual of the maṇḍala-cakra means the ritual of the ādiyoga[-samādhi] and so forth. Thus, by means of reversing the ordinary, one attains liberation in one life.” dkyil ‘khor ‘khor lo’i cho ga yis/ zhes te/ dang pa’i rnal ‘byor (D adds pa) la sogs pa’i cho ga ste/ tha mal pa bzlog pas tshe gcig gis grol ba’o// (Sukusuma, D 106a.1-2; P 107a.2-3). This is an indication of the possibility of sexual yogas in the generation stage practices, rather than only in perfection stage practices as became the case in the later tradition. In any case the distinction between generation and perfection stage practices was probably still being developed at this point so some overlap is to be expected. 180 ~Cf. Vajrāvalī (Mori 2009 Vol. 2, 444). Pādas a and c of this verse are represented in the Vajrāvalī, but here in Buddhajñānapāda’s text—at least in the Tibetan translation—there is an intervening pāda b that is elided in the Vajrāvalī version. Vaidyapāda’s Sukusuma corresponds to the inclusion of pāda b in the Dvitīyakrama’s verse. The subsequent two pādas in the Vajrāvalī correspond with the second half of Dvitīyakrama 124 c and the first half of Dvitīyakrama 124d, which were also incorporated into Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, presumably the source from which Abhayākaragupta draws them. Cf. Guhyasamājamaṇḍalavidhi, verse 366c-d. vajraparyaṅkataś cittaṃ maṇyantargatam īkṣayan. 181 These two pādas are nearly identical with Samājottara 125 c and d. The verse in the Samājottara, however, uses the term vidyā rather than “girl” (*kanyā?). This suggests that Buddhajñānapāda’s verses may be earlier. In an earlier study (C. Dalton 2014) I have argued in some detail that the verse on the two stages of tantric practice in the Samājottara is likely modeled on Buddhajñānapāda’s verse in the Muktitilaka, rather than vice versa. In that instance, it appears that the term “buddhas” from Buddhajñānapāda’s earlier verse in his Muktitilaka was transformed into “vajra holders” in the Samājottara. Just like in this verse with the use of the term vidyā in the Samājottara rather than “girl” (*kanyā?) in the Dvitīyakrama, a move towards increased tantrification is much more likely than the reverse. In this case, moreover, the second two pādas of this verse in the Dvitīyakrama are also found in the Samājottara, though with an intervening two pādas about the nondual nature of reality. Again, the fact that there are two intervening pādas in the Samāmjottara’s version suggests that if one text is based upon the other (i.e. if they are not both drawing from some separate earlier source) the Samājottara’s is likely later than Buddhajñānapāda’s verse, as it would be unlikely that Buddhajñānapāda would cite from a scriptural source—even unattributed—and not provide the complete citation. I discuss both of these points in Chapter Eight. 182 lha mo ‘di ni khyod dang mthun// sems chen khyod kyis ‘dod pa gyis// yid ‘ong bu mo ‘di nyid ni// sangs rgyas kun gyis spyod du gnang// |87| dkyil ‘khor ‘khor lo’i cho ga yis// bu mo sgrol byed dga’ byin ma// byang chub chen po bsgrub pa’i phyir// khyod kyis bde chen myong bar gyis// |88| gzhan kyis sangs rgyas mi nus pa// bu mo ‘di ni yang dag mchog// de bas mtha’ med ‘khor ba’i bar// khyod kyis ‘di dang bral mi bya// |89| (Dvitīyakrama, verses 87- 89). 183 I discuss Buddhajñānapāda’s use of homologizing non-tantric and non-perfection stage practices with tantric and perfection stage practices, and examine the passage on the ten bhūmis of sexual practices below. 125 The first and the later supreme result Are attained, just as explained above. But for those disciples |310| Who are unable to authentically engage in this great reality The tathāgatas have taught it in terms of characteristics Like “Perfect Joy” and the rest. Through engaging in that reality, and by means of [its practice] |311| They gain realization—though there is still something higher. That itself has [also] been taught, To the yogins of the first [stage] 184 [As] the support and supported maṇḍala-cakra. |312| Engaging in and relying upon that, one may gain realization, But those who do not know this reality Are not genuine buddhas. |313|185 This passage is important in explicitly stating that the result of tantric practice is superior to that of a non-tantric path. The text states that by means of the ten bhūmis of sexual practices, the disciple is able to attain “the first and the later supreme result.” Vaidyapāda explains that “the first” indicates “foolish individuals obtaining unfailing suchness from the words of the guru, like a blind person finding a jewel in a heap of rubbish,”186 likely a reference to the disciple’s first “obtaining” suchness from the guru at the outset of the perfection stage path. He goes on to explain that “the later supreme result occurs when, having followed the unique methods, one is just like a destitute child who, with a single utterance, takes posession of his father’s wealth and enjoys it. This is because it is something that was received from someone [i.e. the guru] who is like a father.”187 Thus the disciple’s first encounter with suchness and its complete realization— the “supreme result” of perfect awakening—are both said to be attained by means of sexual yogas. Referencing the first bodhisattva bhūmi of Perfect Joy (rab tu dga’ ba), the Dvitīyakrama suggests that it was only on account of “those disciples who are unable to engage in authentically in this great reality” by means of the ten bhūmis of sexual yoga that the ten bhūmis were taught in their exoteric Mahāyāna iteration. Vaidyapāda makes the point more explicitly: “This great reality means the path that perfects the ten bhūmis in a single instant. Those disciples who are unable to enter into it are those [who practice] the six perfections, since they see [sexual practices] as acts that are at odds with purity.”188 The cultivation of the ordinary ten bhūmis is still said to bring about realization, but it is not the full realization of perfect awakening because, as the Dvitīyakrama makes clear, “there is still something higher.” Again Vaidyapāda is more explicit: “The sugatas of the past and others have taught [this reality], in the sūtra piṭaka and 184 Vaidyapāda clarifies that this refers to yogins who are at the generation stage level of practice (Sukusuma, D 128b.2-3). 185 bcu po de yis dang po dang// phyis kyi ‘bras bu mchog ‘gyur ba// gong du gsung pa rab thob ste// de bas de ‘dra’i don chen la// |310| yang dag ‘jug par mi nus pa’i// gdul bya rnams la bde gshegs kyis// rab tu dga’ sogs mtshan nyid du// bstan nas de yi don zhugs pas// |311| de yis rtogs kyang bla dang bcas// de nyid dang po’i rnal ‘byor la// rten dang rten can dkyil ‘khor gyi// ‘khor lo rab tu bstan byas nas// |312| de zhugs de la brtan byas pas// rtogs kyang don ‘di ma shes na// yang dag sangs rgyas ma yin no// |313| (Dvitīyakrama, verses 310d-313d). 186 dang po zhes pa ni mi slu (slu] D, bslu P) ba’i de kho na nyid dmus long gis phyag dar khrod phung du nor bu rnyed pa ltar gang zag blun pas bla m’i zhal nas de las thob pa’o// (Sukusuma, D128a.5-6; P 154b.2) 187 phyis kyi ‘bras bu mchog gyur ba (ba] D, pa P) / zhes pa ni thabs kyi khyad par du gyur bzhin par byas na (na] D, bas P) byis pa (byis pa] D, P om.) dbul pos pha’i nor tshig gcig gis bdag gir byas nas spyod pa bzhin du pha lta bu de las thob pa’i phyir ro// (Sukusuma, D 128a.6-7; P 154b.2-3). 188 de ‘dra’i don chen zhes pa ni sa bcu skad cig gis rdzogs pa’i lam la ‘jug par mi nus pa’i gdul bya ni pha rol tu phyin pa rnams te/ rnam par dag pa ‘gal ba’i las nyid du mthong ba’i phyir ro// (Sukusuma, D 128a.7-128b.1; P 154b.4-5) 126 other places, in terms of the characteristics of “Perfect Joy” and the rest. Through engaging in that reality by means of the wisdom of learning and the rest via those [bhūmis], even if [a disiciple] realizes its ultimate character, that is still an “ultimate” beyond which there is something higher. This is because the unsurpassed result is [only] realized by means of the unique path.”189 Even without Vaidyapāda’s helpful clarifications, however, this passage in the Dvitīyakrama clearly asserts the superiority of the result of tantric pratice. But Buddhajñānapāda does not leave it just at this; the next lines of the passage indicate that “that itself”—which Vaidyapāda specifies is still the ten bhūmis—was taught to generation stage practitioners as the “support and supported maṇḍala-cakra,” a term referring to the maṇḍala and the deities residing within it. Engaging in deity yoga practice the practitioner may gain realization, but even this is not sufficient because, as the text specifies, without knowing “this reality”—the reality that is known through the ten bhūmis of sexual yogas—one cannot become a genuine buddha. Again, Vaidyapāda makes the point even more explicitly: he states that focusing the mind upon the appearance of the support maṇḍala and supported deities was merely taught as the provisional truth (drang pa’i don), and that “without relying upon the practice of the post-initiatory observences (vrata) and the rest, one will not know this nondual reality just as it is, and will thus not become a sublime authentic buddha.”190 As discussed above, the post-initiatory observences in this context certainly involve the vidyāvrata, the consort observance. Vaidyapāda himself specifies this elsewhere,191 and the context of the discussion makes it rather certain that the vidyāvrata was intended by the term vrata here. Thus while the citation from the Ātmasādhanāvatāra that we examined above suggests only that the tantric path of deity yoga is a crucial method for attaining awakening, all three of the passages we have seen from the Dvitīyakrama make it clear that Buddhajñānapāda held not only deity yoga, but the sexual yogas of the perfection stage, as essential for the attainment of full awakening. The results of other paths, as the Dvitīyakrama states explicitly, can still be surpassed. In another passage from the Dvitīyakrama Buddhajñānapāda explains that nondual wisdom is unknown to the practitioners and scholars of quite a wide range of non-tantric Buddhist paths and philosophies: What can the rain do To someone with an umbrella in his hand? Likewise, when carrying the umbrella Of nondual wisdom, |147| Even if a rain of concepts should fall How could they do any harm? That kind of perfect supreme wisdom How could it be known by an ordinary being? |148| It is not known by the śrāvakas Nor by the pratyekyabuddhas The Yogācārins, Mādhyamikas, And bodhisattvas do not know it. |149| 189 de la ‘das pa la sogs pa’i bde bar gshegs pa rnams kyis sa dang po rab tu dga’ ba la sogs pa’i mtshan nyid du mdo sde la sogs par bstan nas/ de rnams kyis thos pa la sogs pa’i shes rab kyis de’i don la zhugs pas mthar thug pa’i mtshan nyid rtogs kyang bla dang bcas pa’i mthar thug pa ste/ bla na med pa’i ‘bras bu ni lam khyad par can gyis rtogs pa’i phyir ro// (Sukusuma, D 128b.1-2; P154b.5-7). 190 ...brtul zhugs la sogs pa’i spyod pas ji bzhin pa’i gnyis su med pa’i don ‘di ma shes na dam pa yang dag pa’i sangs rgyas ma yin no zhes pa’o// (Sukusuma, D 128b.5-6; P 155a.3-4) 191 Muktitilaka-vyākhyāna, D 57b.6-7. 127 Even all of the non-superior buddhas192 Do not know this at all. [But] by pleasing the future vajra-holders, Who know this reality, |150| Due to the power of one’s great merit It will be transferred [even] without words.193 Vaidyapāda’s commentary identifies the “umbrella of nondual wisdom” as the unique path of the second stage of tantric practice,194 and the context of this passage in the Dvitīyakrama, immediately following a doxography that places the view of the perfection stage above all philosophical positions, supports this reading. Indeed, Buddhajñānapāda claims directly in this passage that many of the lower paths lack knowledge of nondual wisdom. Certainly, ordinary beings, śrāvakas, pratyekabuddhas, Yogācarins, Mādhyamikas, and bodhisattvas “do not know it,” a statement that clearly places the wisdom of the tantric path above that of both the exoteric Buddhist practice traditions and philosophical systems.195 But the subsequent statement that “even all of the non-superior buddhas do not know it” is more cryptic. While unfortunately we do not have the original Sanskrit of this passage, the term that I have translated as “non-superior buddhas” (bla bcas sangs rgyas) literally means “surpassed buddhas,” and—in Tibetan at least— is a quite unusual phrase that appears to have been coined in contradistinction to the very common term “unsurpassed buddha” (or “unsurpassed buddhahood”) (bla med sangs rgyas). Here in this passage the “surpassable” categories mentioned, beginning with the śrāvakas, appear in a hierarchical order and include what is, for Buddhajñānapāda, the highest philosophical position, that of the Mādhyamikas, as well as the highest non-tantric form of practice, that of the bodhisattvas.196 Thus the “non-superior buddhas,” who are placed above even all of these, appear here to constitute a separate category that is higher than the bodhisattvas but lower than those who realize the nondual wisdom of the perfection stage. Vaidyapāda identifies these “nonsuperior buddhas” as “the buddhas of the Kriyā tantras, the buddhas of the Caryā tantras, and the buddhas of the Yoga tantras.”197 While we cannot be sure that this is precisely what Buddhajñānapāda intended, the Dvitīyakrama is clearly referring to a category superior to the exoteric Mahāyāna but inferior to the teachings on nondual wisdom found in Buddhajñānapāda’s perfection stage practice system; the lower tantras—including even the Yoga tantras—are an 192 Tib. bla bcas sangs rgyas. Literally “those buddhas who are surpassed by something else.” Vaidyapāda identifies these as the buddhas of the Kriyā, Caryā, and Yoga tantras (Sukusuma, D 112a.4-5). I discuss this term below. 193 gang zhig lag na gdugs thogs la// de la char pas ci byar yod// de bzhin gnyis med ye shes kyi// gdugs thogs la ni rtog pa yi// |147| char pa rab tu ‘bab ‘gyur yang// de la de yis ji ltar gnod// de ‘dra’i rab mchog ye shes ni// so so skye bos ga la shes// |148| nyan thos rnams kyis mi shes so// rang sangs rgyas kyis kyang mi shes// rnal ‘byor spyod dang dbu ma pa// byang chub sems dpas mi shes so// |149| bla bcas sangs rgyas kun gyis kyang// ‘di ni cung zad mi shes so// ‘di yi don shes ma ‘ongs pa’i// rdo rje ‘dzin pa mnyes byas nas// |150| rang gi bsod nams chen stobs kyis// yi ge med par rnam par ‘pho// (Dvitīyakrama, verses 147-151b). 194 Sukusuma, D 112a.2. 195 I discuss Buddhajñānapāda’s philosophical position in relation to his view on the superiority of the tantric view, training, and result below. 196 The term “surpassed” (bla dang bcas) is used at three other places in the Dvitīyakrama, always indicating a state or level of realization that is lower than that attained by means of the perfection stage. In the three other passages that use the term, the point of comparison (i.e. that which is “surpassed” or “surpassable”) is always an exoteric Mahāyāna position or system: in the first instance it is non-Buddhist and Buddhist philosophical positions culminating in the Madhyamaka; in the second, the traditional Mahāyāna bodhisattva bhūmis; and in the third, the awakening that is attained by means of the “path involving suffering.”See Dvitīyakrama verses 141, 312, and 393, respectively. 197 bla bcas sangs rgys zhes pa ni bya ba’i rgyud kyi sangs rgyas dang spyod pa’i rgyud kyi sangs rgyas dang rnal ‘byor rgyud kyi sangs rgyas so// (Sukusuma, D 112a.4-5). 128 excellent candidate for such a category. Thus, in this passage the lower tantric systems appear to be overtly subordinated to tantric systems in which perfection stage practices are taught.198 This is further indicated by the concluding verses of the passage, which state that “through pleasing the future vajra-holders who know this reality” it will be “transferred [even] without words.” This seems to be yet another reference to the process of the guru’s transferring suchness directly to the disciple, which, as discussed above, is the crucial initial moment of the perfection stage path. However, this verse is unusual in seeming to assert the possibility of a nonverbal communication of reality from the guru to the disciple.199 The context of such an assertion within Buddhajñānapāda’s system remains perplexing given the predominance of references to the communication of suchness specifically by way of the guru’s words in both Buddhajñānapāda’s and Vaidyapāda’s writings.200 In any case, this passage’s reference to the transference of reality, or supreme wisdom, from the guru to disciple certainly constitutes part of the perfection stage system taught in Buddhajñānapāda’s works, and it is thus this system that is asserted to be superior. Outside of Buddhajñānapāda’s own writings, we also find later Indian and Tibetan authors attributing to him the position that tantra, including its result, is superior to non-tantric paths. In his Sarvasamayasaṃgraha Atīśa writes, “The great ācārya Jñānapāda wrote, ‘The tantric vehicle is extraordinary in three ways: the practitioner, the path, and the result.’”201 While this statement itself is not found in any of Buddhajñānapāda’s surviving works, the positions that it mentions do certainly fit with what we see in Buddhajñānapāda’s own writings, including the superiority of the tantric result. The same statement cited by Atīśa is also attributed to Buddhajñānapāda in the works of several later Tibetan scholars.202 Moreover, the Tibetan scholar Jamyang Sheypa attributes to the Dvitīyakrama the position that “without the addition of a Mantra path, final Buddhahood is not attained through the Perfection Vehicle alone.”203 The passages from Buddhajñānapāda’s works, including the Dvitīyakrama, that we have examined here do indeed either state or intimate that the path of tantra, and especially that of the perfection stage, is necessary in order to attain perfect awakening. Yet another way in which Buddhajñānapāda demonstrates the superiority of the tantric path, and especially the path of the perfection stage, is by drawing direct equivalents between non-tantric practices or tantric non- 198 Regarding precisely what type of tantra might be superior to lower tantric systems, both Buddhajñānapāda (see Dvitīyakrama, verse 388) and Vaidyapāda (Sukusuma, D 107a.6-7) use the term Mahāyoga to describe what they seem to consider the highest class of tantra (though Vaidyapāda also uses other terms; see Chapter One, note 100). We may recall, however, that even the Guhyasamāja root tantra—the most important tantra for Buddhajñānapāda, the ostensive basis of his practice systems, and thus presumably the tantra he would have considered to fall into the Mahāyoga class—does not distinguish between the two stages of tantric practice, although sexual yogic practices that were further developed and eventually classified as the pefection stage are found in the Guhyasamāja-tantra. As I noted in Chapter Two, the specific perfection stage practices in Buddhajñānapāda’s writings, since they were received in a visionary encounter, have Mañjuśrī himself as their source. 199 The phrase in question, yi ge med par rnam par ‘pho, which I have translated as “transferred [even] without words,” uses the term yi ge—“letters,” “syllables,” or “words”—so it is possible that this could mean “without writings/texts.” Vaidyapāda says it means that “without relying upon external words (writings?) and so forth, the guru transfers [it] into the mindstream of the disciple.” phyi rol gyi yi ge la sogs pa la brten pa med par bla mas slob ma’i rgyud du rnam par ‘pho zhes so (Sukusuma, D 112a.6; P 135a.3). 200 It does, however, fit well with later claims in Great Perfection and Mahāmudrā traditions of the guru’s being able to communicate suchness to a disciple by means of a gesture, or some other nonverbal means. 201 slob dpon chen po ye shes zhabs kyi zhal snga nas sngags kyi theg pa ni rnam pa gsum gyis thun mong ma yin te/ ‘di ltar sgrub pa po dang lam dang ‘bras bu gsum mo zhes gsungs so// (Sarvasamayasamgraha, D 44a.7). 202 For example, Butön, and Jamgon Kongtrül cite the passage (See Hopkins 2008, 241 and Kongtrül 2005, 80, respectively). 203 Hopkins 2003, 637. 129 perfection stage practice and the practices of the perfection stage. Let us examine some of these equivalences now. Homologizing Tantric Practices with Non-tantric Ones In his early work on the Guhyasamāja-tantra, Alex Wayman wrote that the Jñānapāda School, and Buddhajñānapāda’s writings specifically, “adopted an interpretive position in which at each point the explanations of the Guhyasamāja are tied in with Mahāyāna Buddhism, particularly of the Prajñāpāramitā type,” an assertion that has been repeated by a number of later scholars.204 Indeed, we do see Buddhajñānapāda homologizing Mahāyāna doctrines and practices with tantric ones, but again it is always perfection stage doctrines and practices with which the non-tantric doctrines and practices are related, and in drawing those relationships the perfection stage is consistently privileged. Moreover, there are other passages where Buddhajñānapāda connects not just the Mahāyāna, but tantric generation stage practices and terminology, with those of the perfection stage, always privileging the latter. This suggests that he was not merely concerned with relating tantra to the Mahāyāna, but rather showing that perfection stage practices encompassed and indeed superceded the exoteric Mahāyāna as well as the lower stages of tantric practice. This is not to say that Buddhajñānapāda held the foundational tradition of Buddhism, the Mahāyāna, or the first stage of tantra in disregard. The Muktitilaka contains an extensive section on the importance of the foundational practice of cultivating compassion, joy, and equinimity, three among the four boundless attitudes (apramāṇa, tshad med pa) found in early Buddhist traditions. The Mahāyāna practice of generating bodhicitta and its doctrine of emptiness are emphasized in his works, and Buddhajñānapāda wrote more than one generation stage sādhana, including the Samantabhadra/Caturaṅga-sādhana, his well-known Guhyasamāja sādhana centered on the maṇḍala of Mañjuvajra. However, his writings do suggest that while he understood all of these practices as important—even essential—they were understood as such specifically in their role as foundational supports to the practice of the perfection stage. In a passage in the Ātmasādhanāvatāra Buddhajñānapāda equates mind’s nature, or nondual wisdom, with a number of Mahāyāna principles, which are then equated with specific deities.205 The principles invoked include the five wisdoms, the four (!) gates of liberation, the sixteen emptinesses, and eight among the ten pāramitās.206 However, this is precisely the passage that I identified in Chapter One as a rather lengthy series of unattributed quotations from Chapter Six of Vilāsavajra’s Nāmamantrārthāvalokinī. Buddhajñānapāda has thus borrowed these correspondences of nondual wisdom with the various Mahāyāna principles and the deities of the Mañjuśrīnāmasaṃgīti maṇḍala from his guru’s work. As we will recall, Buddhajñānapāda, in the Dvitīyakrama and the Muktitilaka, frequently identifes nondual wisdom as something known only through the perfection stage path, but such a claim is not found in the Ātmasādhanāvatāra. This, I believe, constitutes further evidence—in addition to the conspicuous absence of any reference to the Guhyasamāja-tantra, so central to Buddhajñānapāda’s other tantric writings—that the Ātmasādhanāvatāra is an earlier work. In any case, like the emphasis on nondual wisdom found in the Nāmamantrārthāvalokinī, it seems that Buddhajñānapāda’s practice of making equivalences between doctrines and practices from different Buddhist systems may be due in part to the influence of his guru Vilāsavajra. In the Dvitīyakrama we see another set of Mahāyāna categories homologized with the sexual practice of the perfection stage. This passage, which I already mentioned above, draws a 204 Wayman 1977, 94. 205 Ātmasādhanāvatāra, 57a.4-58b.6. 206 As Tribe (2016, 6 and 14n15) has noted, some of the sets of principles have been modified to more effectively hew to tantric categories or sets of deities that they are said to represent. 130 direct correspondence between each of the ten bodhisattva bhūmis and ten different stages of sexual union, and concludes by explicitly privileging the tantric perfection stage “version” of the practice over the Mahāyāna one.207 [Adorned] with garlands, necklaces, anklets,208 and more, Her complexion, breasts,209 and the rest, Knowing the bliss of examining the lotus— This should be known as the first. |299| Praising with melodious song Like the shajarishanisha210 and others, And delighting with the sweet sound ṣīt —211 This should be known as the second. |300| At the time of anointing the body With sandalwood and other scents, The genuine bliss which is so produced— This should be known as the third. |301| Having sucked the honey from [her] lower lip The bodhicitta that abides in the head melts, Tasting it brings pleasure,212 thus delighting oneself— This should be known as the fourth. |302| Through anointing the body and a variety of acts At the time of playing Genuine bliss is brought about through touch— This is known to be the fifth. |303| By means of this the three wisdoms Are known, and one’s mind Is made to experience great bliss— This should be known as the sixth. |304| By means of the hardness that results From one’s relying on her body213 Genuine bliss is produced— This should be known as the seventh. |305| The dew from her lotus and The wetness of bodhicitta Bring about great delight to the mind— 207 The verses on the ten bhūmis correspond with first the six sensory experiences—visual, auditory, olfactory, gustatory, tactile, and mental—and then with the four elements—earth, water, fire, and wind—respectively. Thanks to Harunaga Isaacson for pointing out these correspondences, which were obviously intended in the text. 208 Tib. ha ra nu pur. This seems to be a Tibetan transliteration of hāranūpura, necklaces and anklets. Thanks to Harunaga Isaacson for his assistance with this point. 209 ku tsa. This may be a Tibetan transliteration of kuca, breasts. Thanks to Harunaga Isaacson for this suggestion. 210 This term is rendered in four different ways in the five recensions of the root text and two further ways in the two recensions of the commentary I am looking at. I have randomly selected one to attempt to phoneticize here, but I remain at a loss as to what the word should actually say. Vaidyapāda’s commentary indicates that it is an erotic melody from the *devīśāstras (lha mo’i bstan bcos) (Sukusuma, D 128a.3). 211 sid sgra] P V (D), sing sgra D C S N V (P). 212 I am not entirely sure of this line, but this seems to be the meaning. Vaidyapāda writes, “Melting the bodhicitta that abides in the head, means that regarding the path of the bodhicitta that resides in the head, it is by means of that [path], that one drinks this elixir.” mgor gnas byang chub sems ‘ju bas/ zhes pa ni mgor gnas pa’i byang chub kyi sems kyi lam ni des te des ro ‘thung ba’o// (Sukusuma, D 128a.3-4; P 154a.7-8). 213 de yi lus ni bdag gi ni// rten du gnas pa sra pa yis// I am unsure about the translation of these two lines. 131 This is known as the eighth. |306| Due to heat—the warmth and so forth of the secret place— One’s mind is brought to the supreme Genuine delight— This should be known as the ninth. |307| Then, through stirring, the wisdom fire Burns the aggregates, elements, and the rest Through this the mind becomes genuinely blissful— This should be known as the tenth. |308| By means of these ten The first and the later supreme result Are attained, just as explained above. But for those disciples |309| Who are unable to authentically engage in this great reality The tathāgatas have taught it in terms of characteristics Like “Perfect Joy” and the rest.214 Through engaging in this truth, and by means of [its practice] |310| They gain realization—though there is still something higher.215 The passage is interesting in that it focuses more on the erotic acts leading up to the bliss of orgasm rather than explicitly referring to particular sexual yogic techniques. However, the line about melting and tasting the bodhicitta abiding in the head and, of course, the final verse on the blazing of the wisdom fire that burns through impure appearances do clearly indicate the yogic context and purpose of these acts. As such, it does not appear to be any particular sexual yogas, but simply the act of coitus carried out with a soteriological intention and focus, that is here equated with the ten bodhisattva bhūmis. As I discussed in the preceeding section, the final verses of this passage indicate that the sexualized version of these ten bhūmis are of higher value than their exoteric Mahāyāna iteration, since the latter are said to be unable to bring about full awakening. Homologizing Generation Stage Practices with Perfection Stage Practices Moving on from equating exoteric Mahāyāna practices with those of the perfection stage, other statements in Buddhajñānapāda’s writings homologize other “lower” tantric practices with perfection stage practice. A passage from the Muktitilaka equates the “inner yoga,” which 214 Rab tu dga’ ba, “Perfect Joy,” is the name of the first bodhisattva bhūmi. 215 phreng ba ha ra nu phur sogs// mdog dang ku tsa la sogs pa// padma rtags pa’i dga’ shes pa// dang por rab tu shes par bya// |299| ṣa dzdze rī ni ṣā na sogs// glu byangs bstod dang sid sgra yi// snyan pa’i dbyangs kyis dga’ ‘gyur bas// gnyis pa ru ni shes par bya// |300| tsandan la sogs sna tshogs dris// lus byugs lhan cig rtsen byed tshe// dga’ ba yang dag thob byed pa// gsum pa ru ni shes par bya// |301| ma mchu sbrang rtsi gzhib byas pas// mgor gnas byang chub sems ‘ju bas// ro ‘thung dga’ bas bdag mnyes pas// bzhi par rab tu shes par bya// |302| lus la byug cing sna tshogs kyi// spyod pas rtsen tshe reg bya yis// yang dag dga’ bar byed pas na// lnga pa ru ni shes par bya// |303| de yis ye shes rnam pa gsum// rig par byed cing rang gi yid// yang dag dga’ bar rab byed pa// drug par shes bya rnal ‘byor pas// |304| de yi lus ni bdag gi ni// rten du gnas pa sra ba yis// yang dag dga’ bar byed pas na// bdun pa ru ni shes par bya// |305| de yi padma’i zil sogs dang// byang sems rlan gyis rang gi sems// rab tu dga’ bar byed pas na// brgyad par rab tu shes par bya// |306| gsang gnas drod sogs tsha ba yis// bdag gi yid ni yang dag par// dga’ byar byed pa’i mchog yin pas// dgu pa ru ni shes par bya// |307| de nas bskyod pas ye shes mes// phung po khams sogs sreg byed pas// yid ni yang dag dga’ ‘gyur pas// bcu par rab tu shes par bya// |308| bcu po de yis dang po dang// phyis kyi ‘bras bu mchog ‘gyur ba// gong du gsungs pa rab thob ste// de bas de ‘dra’i don chen la// |309| yang dag ‘jug par mi nus pa’i// gdul bya rnams la bde gshegs kyis// rab tu dga’ sogs mtshan nyid du// bstan nas de yi don zhugs pas// |310| de yis rtogs kyang bla dang bcas// (Dvitīyakrama, verses 299a-311a). 132 Buddhajñānapāda identifies as the “supreme suchness” (de nyid mchog),216 with an extensive list of tantric practices that are otherwise primarily associated with the more external ritual practices of the generation stage or initiatory ritual: [When] this inner yoga Is received [directly] from the mouth Of the sublime guru one has no doubts [about this]. Because it is the union of the profound and the luminous It is called “yoga.” Because of being one-pointed it is [called] samādhi Because of pleasing and satisfying It is called bali. Since it pacifies evil deeds, it is [called] peaceful [activity]. Because of enriching all merit [it is called enriching activity]. Because it is the great passion, it is asserted to be magnetizing [activity]. Since it kills everything, it is [called] wrathful [activity].217 Since it draws in the great sign [of accomplishment], it is [called] summoning [activity]. [Since] it pacifies, for the duration, the dharma Of the śrāvakas and so forth, it is [called] the driving out of hindrances. Since can be realized by a mantrin, a yogin, abiding anywhere, It is [called] protection.218 [Since] knowing this, one realizes the essence of all things As a singular nonduality it is [called] sharp eyes. Since it transforms all concepts without exception Into the essence of wisdom It is [called] the poison-destroying ritual. [Because] it burns with the fire of wisdom The lifeless219 matter of the aggregates and so forth it is [called] homa. Because it equalizes all phenomena By means of the profound and the luminous it is [called] the line measuring.220 Because the luminous rainbow-like body Naturally upholds all of the twenty-one phenomena It is [called] applying color.221 Because222 its essence abides everywhere It is [called] the entry [into the maṇḍala] of everyone: the ācārya and the disciples. Since it is only said to come from elsewhere223 But, in fact, it is realized by blissful self-awareness, 216 Muktitilaka, D 50b.3. Vaidyapāda also indicates that the “inner yoga” is the “suchness of all phenomena” (Muktitilaka-vyākhyāna, D 55a.7). 217 These four lines are a list of the four tantric activities—peaceful, enriching, magnetizing, and wrathful. 218 I am unsure of the meaning of these two lines. sngag pa gang gnas rnal sbyor pas// rtogs par nus pas brsung ba’o// 219 Tib. blun pa; Skt. *jaḍa 220 Tib. thig gdab pa. This refers to the process of drawing the lines of the maṇḍala in their proper ratio. 221 This presumably refers to applying color to the drawn maṇḍala. I am unclear on the precise meaning of these three lines. Vaidyapāda specifies that the twenty-one phenomena refer to form and so forth, which are naturally upheld by Vairocana and so forth (Muktitilaka-vyākhyāna, D 55b.6-7) 222 Emending nas pa to gnas pas, following Vaidyapāda (Muktitilaka-vyākhyāna, D 55b.7). 223 logs] logs P, log D. Vaidyapāda specifies that the place where it is only said to come from is the guru (Muktitilaka-vyākhyāna, D 55b.7). 133 It is called bestowing empowerment. [Since] it bears the names, like these and others, Of224 all phenomena it is [called] supreme suchness.225 In homologizing the inner yoga of suchness with all of these rituals, Buddhajñānapāda subsumes the more external ritual practices within the perfection stage practice of cultivating supreme suchness. While he does not claim directly that such outward practices are not necessary, he does seem to imply that all of their functions are fulfilled by the cultivation of the suchness of the perfection stage, thus placing the perfection stage practice of suchness above these other kinds of practices. Another passage in which Buddhajñānapāda equates a set of terms generally associated with the lower stage of tantric practice with the perfection stage is found in the Dvitīyakrama. In this passage the practices in question are the four branches of sevā, upasādhana, sādhana, and mahāsādhana, which are mentioned in the twelfth chapter of the Guhyasamāja-tantra. In the tantra itself the practices are used to describe the process of generating oneself in the form of the deity, and indeed these same four practices are the “four branches” that are used to structure Buddhajñānapāda’s Guhyasamāja generation stage sādhana of that name, the Caturaṅgasādhana. 226 However, in the Dvitīyakrama the four branches are used to describe the sexual yogas of the perfection stage—in this specific instance, in the context of tantric intitiation: When the vajra touches the lotus This is explained to actually be sevā. The vajra entering the lotus Is actually upasādhana. |119| Then, through moving and stirring a bit, The heart quivers and attentiveness wanes The hair on the crown falls loose and garments are cast off Sweat227 covers the body and it takes on a reddish hue, |120| And with reddened eyes [she] looks at one. Moving228 repeatedly brings about sādhana. Thus, without concern229 224 Emending kyis to kyi following Vaidyapāda (Muktitilaka-vyākhyāna, D 56a.1). 225 de yi nang gi sbyor ba ni// bla ma dan pa’i zhal nas ni// thob kyi de la the tshom med// de ni zab gsal gnyis sbyor bas// sbyor ba shes su bshad pa’o// rtse gcig pas ni ting nge ‘dzin// mnyes byed tshim par byed pa ste// gtor ma zhes ni rab tu bya// sdig pa zhi phyir zhi ba’o// bsod nams kun gyis rgyas pa’o// chags pa chen pos dbang du ‘dod// thams cad gsod phyir mngon spyod do// rtags chen dgug phyir dgug pa’o// nyan thos la sogs chos rnams ni// ring du zhi byed bskrad pa’o// sngags pa gang gnas rnal ‘byor pas// rtogs par nus pas bsrung ba’o// de shes chos rnams ngo bo kun// gnyis med gcig rtogs mig rnon no// rnam par rtog pa ma lus rnams// ye shes ngo bor de ‘gyur pas// dug gzhom pa yi cho ga’o// phung po la sogs bslun pa’i rdzas// ye shes mes bsregs sbyin sreg go// zab pa dang ni gsal ba yis// chos kun mnyam pas thig gdab po// gsal ba’i ‘ja’ tshon lta bu’i skus// nyi shu rtsa gcig chos kun la// rang bzhin gyis bzung tshon (tshon] P, mtshon D) btab po// de yi ngo bo kun gnas (gnas] sugg. em based on Vaidyapāda’s commentary nas D P) pa // slob dpon slob ma kun zhugs pa’o// ming tsam gyis ni logs (logs] P, log D) ‘byung yang// rang rig bde bas rtogs pas na// dbang bskur zhes ni bshad pa yin// de la sogs pa mtha’ yas kyi (kyi] sugg. em. based on Vaidyapāda’s commentary, kyis D P) chos kun ming can de nyid mchog// (Muktitilaka, D 41a.5-41b.2; P 60a.6- 60b.5) 226 As mentioned above, this same sādhana is also known by the name Samantabhadra-sādhana. I address the issue of the two titles of the sādhana and describe the generation stage practices of Buddhajñānapāda’s system in more detail in Chapter Five. 227 rngul] sugg. em., rdul D C S P N V (D and P). This emendation is based on the line from the parallel verse in Vaidyapāda’s Yogasapta which reads rngul chu thigs pas lus kun khyab// (Yogasapta, D 71a.5; P 84b.7) 228 bsgul] sugg. em. based on V, bskul D C S P N. Buddhajñānapāda’s text here reads bskul, but given the fact that earlier the text read bsgul ba, as well as the fact that this is glossed in Vaidyapāda’s commentary as yang dang yang du bskyod pa suggests that it is bsgul that is is meant. 134 The diligent vow-holder, by means of moving that which is bow-shaped230 |121| Causes the blazing of the triangular wisdom fire Thereby the elements melt and the sixteenth part, Which is like a jasmine flower,231 Should be offered by unifying the winds. |122| Naturally perfectly pacified The suchness that is the pacification of all phenomena, That bliss itself, dwells at the jewel [for] an instant. Free from recollection, [it] is made to move This itself is mahāsādhana.232 |123| Just as we have seen in the passages describing sexual practices above, the text here uses quite a few coded tantric terms: the “lotus” is the vagina, the “vajra” the penis, the “sixteenth part” the drop of semen, and the “jewel” the tip of the penis. The wording used in this passage, with statements like “this is explained to actually be sevā...,”233 seems to indicate that while perhaps the fourfold set of terms—sevā and the rest—was more commonly used to describe generation stage practices, Buddhajñānapāda wishes to assert that is the sexual practices of the perfection stage that constitute the actual identity of sevā, upasādhana, sādhana, and mahāsādhana.234 Regarding the final line of this passage Vaidyapāda writes, “This itself is mahāsādhana because it is the essence of the accomplishment of the mahāmudrā,”235 a statement that further supports Buddhajñānapāda’s claim that sexual yogic practices constitute the actual identity of the generation stage processes of sevā and the rest. That is, the term mahāmudrā, in its 8th- and 9thcentury usage, refers to the form of the deity, so Vaidyapāda is stating that the mahāsādhana of sexual practice described in the Dvitīyakrama constitutes mahāsādhana insofar as it is the very essence of the accomplishment of deity yoga of the generation stage of tantric practice.236 229 Vaidyapāda seems to suggest that this means something like “effortlessly.” He writes, “Without concern means without having to search for it. Since the causes have already come about, have no doubt that the fourth tattva will arise.” sems khral med pa ru zhes pa ni btsal (btsal] D, brtsal P) dgos pa med de/ rgyu sngon du song ba’i phyir te de kho na nyid bzhi pa skye ba la the tsom mi bya’o// (Sukusuma, D 109a.5; P 131a.7-8). 230 This is a reference to the wind element, the “maṇḍala” of which is represented in the traditional sādhana visualizations as a bow-shape. Vaidyapāda makes it clear that this refers to the wind maṇḍala. (Sukusuma, D 109a.5; P 131a.8). 231 i.e. the bindu of bodhicitta, or semen. Jasmine is frequently used as a metaphor for semen in tantric texts. 232 rdo rje padmar reg pa ni// bsnyen pa’i de nyid yin par bshad// rdo rje padmar zhugs pa ni// nye bar sgrub pa’i de nyid do// |119| de nas bsgul zhing bskyod tsam gyis// snying ni ‘dar zhing dran pa nyams// spyi bo’i skra grol gos kyang ‘dor// rngul gyis lus khyab mdog dmar te// |120| mig dmar phra bas bdag la blta// yang du bsgul bas sgrub pa’o// de bas sems khral med pa ru// sdom brtson gzhu dbyibs gyo ba yis// |121| sum mdo ye shes me sbar nas// khams bzhus nas ni bcu drug char// gyur ba me tog kunda ‘dra// rlung gi sbyor bas phul bar bya// |122| rang bzhin gyis ni rab zhi ba// chos kun zhi ba de kho na// bde ba de nyid nor bur ‘dug// skad cig dran med g.yo bar byed// sgrub pa chen po de nyid do// |123| (Dvitīyakrama, verses 119-123). 233 bsnyen pa’i de nyid yin par bshad. This phrase could also be translated something like, “this is explained as the identity of sevā.” In either case I understand the text to be asserting that the perfection stage practices constitute the actual identity of sevā etc. 234 I will address the perfection stage practices described in Buddhajñānapāda’s writings in more detail in Chapter Six, and it would be too much of a digression at this point to examine the practices described in this passage in much detail. However, we should note that unlike the passage above that discussed the ten bhūmis in terms of sexual practice, the Dvitīyakrama here does include more specifics regarding sexual yogic practices. In fact, the description in this passage of moving the winds to cause the blazing of wisdom fire and the dripping of the elements is a very early instance of what later comes to be called caṇḍalī yoga, a practice found commonly in the later Yoginī tantras. 235 de nyid la sgrub pa chen po zhes te phyag rgya chen po dngos grub kyi ngo bo nyid kyi phyir ro// (Sukusuma, D 109b.3; P 131b.7). 236 While the identification of the four branches of sevā and the rest not only with generation stage practice, but also with perfection stage practice, is not found in the Guhyasamāja root tantra (which does not distinguish between the 135 In these passages from his writings we have seen the ways in which Buddhajñānapāda homologizes non-tantric and generation stage doctrines and practices with those of the perfection stage as a way of not only justifying perfection stage practice, but also indicating that it is superior to, or constitutes the very essence of, the “lower” stages of Buddhist practice. Thus, while he certainly understood exoteric Mahāyāna and generation stage practices as fundamental in their role as foundational for perfection stage practice, Buddhajñānapāda’s comparisons of aspects of these traditions with the perfection stage serve to subordinate them to the higher stage of tantric practice. Our observations on Buddhajñānapāda’s techniques for advocating the superiority of tantra have up to this point primarily focused on his articulation of tantric practice and its result as superior, but we also find passages in his writings that set the perfection stage view above the views of Mahāyāna philosophical positions, a point to which we will now turn our attention. Philosophical Doxographies and Tantric Views Recent scholarship has begun to draw attention to the philosophical underpinnings of the perspectives of tantric Buddhist authors, which is a very welcome trend.237 To make a full assessment of Buddhajñānapāda’s philosophical position(s) would require a detailed examination of his earlier non-tantric writings, which I have not done in this study. However, in the tantric works I have examined in more detail, a philosophical perspective nonetheless does emerge, and I will make some preliminary remarks on that position here. In general, Buddhajñānapāda seems to favor the Yogācāra-Madhyamaka philosophy advocated by his guru Haribhadra, and which is also evident in the work of another of his gurus, Vilāsavajra.238 This is substantiated by the later Tibetan tradition which also seems to generally classify Buddhajñānapāda as a Svātantrika Mādhyamika who upheld the Yogācāra view of denying external objects and affirming selfawareness. 239 There are several passages in Buddhajñānapāda’s writings in which he seems to two stages) such a distinction of the four branches in terms of both generation and the perfection stage practice is found in the Samājottara. As I noted above, Buddhajñānapāda does not seem to have known the Samājottara, but Vaidyapāda did. This use of the four branches in the context of both tantric stages, then, may be an instance—of which there are in fact several—of the influence of Buddhajñānapāda’s thought on the Samājottara. I examine instances of and argue for the influence of Buddhajñānapāda’s thought and writings on the Samājottara in Chapter Eight. The association of the four branches with perfection stage practice is also found in Chapter 11 of the Guhyagarbha-tantra, though in a more general sense and without the specifics found here in the Dvitīyakrama. Buddhajñānapāda gives no indication of knowing this tantra, though the only surviving Indic commentary on the Guhyagarbha-tantra, the Spar khab, is attributed to Buddhajñānapāda’s guru Vilāsavajra. I explore some of the commonalities between Buddhajñānapāda’s writings and those of the early Great Perfection tradition below. 237 See Isaacson 2013, McNamara 2017, and Yiannopoulous 2017. All of these studies focus on the philosophical positions of the 11th-century polymath Ratnākaraśānti. Isaacson and Sferra (2014) introduce, edit, and translate the Sekanirdeśapañjikā, a tantric commentary by Rāmapāla that is itself very much concerned with issues of philosophy. 238 Tribe states that a closer examination of the philosophical dimension of Vilāsavajra’s writings is necessary to make any clear conclusions on his philsophical orientation, but in the Nāmamantrārthāvalokinī, Vilāsavajra cites both Madhyamaka and Vijñānavāda works, and even directly acknowledges both traditions as sources for his work, though Tribe reports that Vijñānapāda “terms, structures, and perspectives predominate in Vilāsavajra’s approach to praxis” (Tribe 2016, 11). 239 The 15th-century Sakya scholar Taktsang Lotsawa describes Buddhajñānapāda (along with Śāntarakṣita, Kamalaśīla, Haribhadra, Dīpaṃkarabhadra, Vaidyapāda, and Thagana) as a Svātantrika Mādhyamika who affirms self-awareness and denies the existence of outer objects (Grub mtha’ kun shes’grel pa, 212.5-213.2. Thanks to Thomas Doctor, in a fortuitous conversation with whom I became aware of this passage). Jamgön Kongtrül also asserts that Śāntarakṣita, Kamalaśīla, Vimuktisena, Haribhadra, Buddhajñānapāda, Dīpaṃkarabhadra (who Callahan erroneously equates here with Atīśa), Vaidyapāda, and Thagana are Yogācāra-Svātantrika-Mādhyamikas (Kongtrül 2007, 219). We know that Haribhadra was Buddhajñānapāda’s teacher, and interestingly, the list of masters given in both these sources, proceeding from Buddhajñānapāda, is more or less a direct lineage list of Jñānapāda School 136 uphold such a Yogācāra-[Svātantrika]-Madhyamaka position. However, he also indicates in multiple passages that even the Madhyamaka philosophical perspective—and indeed every philosophical perspective—is superceded by the perspective, or view, afforded through tantric practice. So in his works, yet again, even in the realm of philosophy tantra reigns supreme. A look at the non-tantric works cited in Buddhajñānapāda’s Ātmasādhanāvatāra—his only tantric prose work and therefore the only tantric work to include textual citations240— already gives the sense that he has a preference for Yogācāra and Prajñāpāramitā works. In the Ātmasādhanāvatāra Buddhajñānapāda cites the Daśabhūmika-, Laṅkāvatāra-, Samādhirāja-, Mahāsāṃghikaprātimokṣa-, and Ratnaguṇasamcayagāthā-prajñāpāramitā-sūtras, as well as the Sūtra Requested by Akṣayamati, Maitreya’s Mahāyānasūtrālaṃkāra, and Dignāga’s Mañjughoṣastotra.241 A verse from Buddhajñānapāda’s Muktitilaka has been modeled on a wellknown verse from Nāgārjuna’s Mūlamadhyamakakārikās, and another verse in his Ātmasādhanāvatāra appears to be modeled on a verse from Dharmakīrti’s Pramāṇavarttika, so he was clearly familar with and inspired by those works, as well.242 Several of his writings include a number of Yogācāra- influenced passages indicating that all of the phenomenal world is mind. In the Muktitilaka, for example, Buddhajñānapāda writes: [As for] the outer yoga, the form aggregate and so forth, The conventional four elements, All moving things— They are pervaded by just one thing: All are nothing more than mind. Why? Because all sentient beings Are only the aggregate of consciousness. Stupid and confused beings Are unable to understand this truth at all: The mind, mental factors, wisdom, masters. As I discussed in Chapter One, Dīpaṃkarabhadra was a direct disciple of Buddhajñānapāda’s, and Vaidyapāda seems to have been a somewhat later disciple of both Buddhajñānapāda and Dīpaṃkarabhadra. Thagana is a slightly later figure who composed a commentary on the Guhyasamāja-tantra following the Jñānapāda School (Tōh. 1845), one on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana (Tōh. 1868), and one on Ratnākaraśānti’s Hevajra sādhana (Tōh. 1247). His having composed a commentary on Ratnākaraśānti’s sādhana places Thagana in the 11th century at the earliest. Isaacson (2002a, 459) records that the Blue Annals lists Thagana as a guru of Ratnākaraśānti’s in the Jñānapāda School. 240 Like many traditional Buddhist authors Buddhajñānapāda also incorporated verses and prose from other works directly without citation. This occurs both in his prose works (as in the verses incorporated from the Nāmamantrārthāvalokinī in his Ātmasādhanāvatāra mentioned in Chapter One and which I discuss further below), and his versified works (e.g. the verses from the Guhyasamāja-tantra incorporated into his Samantabhadrasādhana, and several lines from the Sarvabuddhasamāyoga-tantra incorporated into the Dvitīyakrama). But here I am referring to citations of the sort found in commentarial prose writings that are actually specified as textual citation by the author, whether or not he names their source. 241 While I had already identified several of the citations in this work, for other identifications I am indebted to Péter Szántó, who kindly shared with me his unpublished draft Sanskrit edition of the Ātmasādhanāvatāra, which includes his identifications of Buddhajñānapāda’s citations. 242 The verse from the Mūlamadhyamakakārikās is parallel with a well-known verse from the Samājottara setting forth the two stages of tantric practice, which Isaacson (2002a, 468-9) has pointed out as being based on Nāgārjuna’s verse. In an earlier conference paper (C. Dalton 2014) I demonstrated that the verse from the Muktitilaka is most likely earlier than that of the Samājottara, and the source on which the Samājottara’s is based. I revisit this argument, adding further evidence of the influence of Buddhajñānapāda’s thought and writings on the Samājottara, in Chapter Eight. I also discuss in that chapter the verse from the Pramāṇavarttika on which Buddhajñānapāda has modeled a verse in the Ātmasādhanāvatāra. 137 And mental objects—these are alike in being mind.243 While it is clear that Buddhajñānapāda here holds a very Yogācāra-like position, asserting all things to be mind, the context of the passage must also be considered. The preceding few verses describe the suchness of all things, or nondual bliss, as possessing a series of opposing qualities—“it is not existent nor non existent, [both] potent and non-potent, the essence of the elements and elementless, has form and is formless”—and state that “therefore it is seen through the path of yoga/union (sbyor ba).”244 But what is described in the passage I have cited here is the “outer yoga,” which Vaidyapāda indicates is the first step towards seeing or realizing suchness. What follows this passage in the Muktitilaka—after a short description of the fact that at the time of liberation, phenomena are experienced not as many, but as “just one”—is a passage on the “inner yoga” of the practice of directly cultivating suchness, which we have examined already above. The presentation in the Muktitilaka of phenomena as being “only mind,” then, is set forth as merely a step in the progression towards the realization of true suchness, not the final realization, nor the final view. In the Dvitīyakrama one of a set of four verses constituting a very short description of the generation stage practice preceding the perfection stage practice that is the Dvitīyakrama’s focus, likewise takes the practitioner through the stage of first viewing the world as mind alone, but then moves on to viewing the mind itself as empty: Looking at [it] as mind alone The outer world is seen to be empty of nature245 Seeing246 mind alone, as well, to be empty Remain in self-awareness alone. |158|247 Vaidypāda clarifies that this self-awareness is a state “beyond the two extremes.”248 Here again we see the viewing of the outer world as mind to be just a stage leading to a more profound realization. The step after cultivating the Yogācāra-like perspective of seeing the outer world as mind is the more Madhyamika-like perspective of seeing the mind itself to be empty. An identical progression is found in Buddhajñānapāda’s Ātmasādhanāvatāra. Here Buddhajñānapāda proceeds through a series of logical arguments to establish the fact that all phenomena are mind. First he shows the outer perceived world to be mind alone, and then he shows that the inner perceiver is likewise, just mind.249 But then he moves on to show that mind is not established either, because it cannot be established as being either one nor many.250 The 243 phyi yi sbyor ba gzugs phung sogs// kun rdzob ‘byung ba bzhi po dang// g.yo dang bcas pa’i chos rnams kun// de gcig pos ni khyab pa ste// thams cad sems tsam kho na’o// ci phyir sems can thams cad kun// rnam shes phung po kho na’o// blun po rmongs pa rnams kyis ni// de don cung zad shes mi ‘gyur// sems dang sems byung ye shes dang// shes bya (bya] em. based on Vaidyapāda’s Muktitilaka-vyākhyāna; pa D P) rnam shes gcig pa’o// (Muktitilaka, D 50a. 2-4; P 60a.3-5) 244 yod pa med cing med pa (pa] D, pa’i P) med// dbang po dang bcas dbang po med// ‘byung ba’i ngo bo ‘byung ba med// gzugs dang bcas shing gzugs med pas// sbyor ba’i lam gyis mthong pa’o// (Muktitilaka, D 50a.2; P 60a.2-3). I believe that here the word sbyor ba (yoga) is intended in terms of both of its meanings, “union” as well as “yoga” in the sense of a practice. 245 Tib. stong par bya; literally “made empty.” I have rendered the term less literally here, as the meaning is that the yogin is to see or perceives the world as empty, which is indeed its fundamental nature. Vaidyapāda comments that this means not to mentally engage with the appearance of the world as appearing separately (Sukusuma, D 113b.4). Khenchen Chodrak Tenphel explains that the term “made empty” is used here to indicate that the practitioner is to “make” his perception of the world accord with the way the world actually is (personal communication, February 2016). 246 Again, the same wording of “making empty” is used here. 247 sems tsam la ni blta bas te// phi rol rang bzhin stong bar bya// sems tsam de yang stong byas nas// rang rig tsamdu gnas par bya// (Dvitīyakrama, verse 158). 248 mtha gnyis dang ‘bral ba (Sukusuma, D 113b.5). 249 Ātmasādhanāvatāra, D 52b.3-53a.4. 250 Ātmasādhanāvatāra, D 53a.4-53a.7. 138 argument against the true establishment of an object because it is “neither one nor many” (gcig dang du ‘bral) is one among the logical arguments that became known—in Tibet, at least—as the four (or sometimes five) “Great Logical Reasonings of the Madhyamaka” (dbu ma’i gtan tshig chen po). Using this particular argument takes Buddhajñānapāda’s reasoning in the Ātmasādhanāvatāra in a more Madhyamaka-oriented direction. While philosophers who did not uphold a Madhyamaka position may also have used the “neither one nor many” argument, it is clear that Buddhajñānapāda, at least, did understand this argument to represent Madhyamaka thought. In the Dvitīyakrama, we find a doxography of non-Buddhist and Buddhist philosophical viewpoints in which the Madhyamaka view is given at the top of the philosophical hierarchy, above that of the Yogācāra.251 In that doxography Buddhajñānapāda gives a very short description of the philosophical position held by each system, and one of the descriptors of the Madhyamaka view that he uses is precisely the perspective that “[all things] are beyond the nature of being singular or multiple.” Immediately following a brief description of the Sautrāntika position, that which falls hierarchically just below that of Yogācāra, he writes, That which has parts is not the ultimate; This is the case even for subtle particles. One cannot observe them individually; They do not appear, but are just like a dream. |136| The wisdom252 that is free from subject and object Is the ultimate, pure like a crystal— This is what the Yogācārins understand. |137| All of these different traditions Are not the ultimate because [All things] are beyond the nature of being singular or multiple, Just like a lotus in the sky. |138| Peace [beyond] nonduality or non-nonduality Completely stainless like space— The intelligent Mādhyamikas understand [reality to be] thus.253 |139| It is clear from this, and another verse in the Dvitīyakrama in which Mādhyamikas are placed above Yogācārins,254 that among the philosophical positions that he addresses, Buddhajñānapāda places the Madhyamaka position on the top. Unlike with the proponents of other traditions listed in this doxography, the Mādhyamikas receive the qualifier of being “intelligent.” It is unclear whether Buddhajñānapāda meant to qualify Mādhyamikas on the whole as more intelligent than adherants of other philosophical positions, or whether he meant to specify a specific group of Mādhyamikas, as opposed to some other group (of, presumably, less intelligent Mādhyamikas). Vaidyapāda, however, does specify a specific group of Mādhyamikas in his comments on a verse in the Dvitīyakrama summarizing the relationship between the philosophical systems several lines later: after the Yogācāra position, Vaidyapāda mentions not that of the Mādhyamikas, but of the Yogācāra-Mādhyamikas (rnal sbyor spyod pa’i dbu ma pa).255 Indeed, the fact that 251 See Dvitīyakrama, verses 126-139. 252 Vaidyapāda’s commentary preserves a different reading of this line. Instead of reading ye shes te “that wisdom,” it reads rnam shes che, “that great consciousness” (Sukusuma, 111a.4). 253 yan lag can yang don dam min// rdul phran dag kyang de bzhin no// so sor snang ba mi dmigs pa// rmi lam lta bu mi snang ste// |136| gzung ‘dzin spangs pa’i ye shes te// don dam shel ltar dag pa ru// rnal ‘byor spyod pas rab tu rtogs// |137| so sor snang ba’i gzhung thams cad// don dam min te gcig pa dang// du ma’i rang bzhin bral ba’i phyir// rnam mkha’i chu skyes bzhin du ni// |138| gnyis med gnyis su med min zhi// shin tu dri med nam mkha’ ltar// blo ldan dbu ma pa yis rtogs// |139| (Dvitīyakrama, verses 136-39). 254 See Dvitīyakrama verse 149. 255 Sukusuma, D 111b.2. 139 Buddhajñānapāda’s guru Haribhadra is a well-known Yogācāra-Mādhyamika philosopher and that Buddhajñānapāda cites many Yogācāra works and includes Yogācāra influenced passages in his writings, but he ultimately chooses to place Madhyamaka on top of Yogācāra in the Dvitīyakrama’s doxography, would suggest Buddhajñānapāda’s own position to be a Yogācāra- Mādhyamika one. But, as we will see, that was clearly not the perspective he held to be the highest; the tantric perspective yet again takes precedence. Immediately following the philosophical doxography cited above, which places Madhyamaka above all other positions surveyed, Buddhajñānapāda makes clear that even the (Yogācāra-?)Madhyamaka position is not the view he holds as the highest. He writes, [Though] reality abides as suchness, [Beings] conceptualize it distinctly In these and countless other [ways]. Therefore all of these [perspectives] |140| Are not the genuine; they can be surpassed. The perspective of the higher yogins Is superior to that of the lower. The lower view is refuted |141| By the wisdom of the higher one. Therefore, by means of the higher stage The sahaja ācārya Performs the genuine blessing. |142| Luminous and perfectly joyful like the sky The self-arisen great *adhideva Is realized through spontaneously arisen wisdom In reliance on the words of the guru.256 143|257 According to Buddhajñānapāda, then, all of the philosophical perspectives, even the ones that he himself appears at times to advocate, are not genuine because they can all be surpassed. As for precisely what surpasses all of these views, it is the “self-arisen great *adhideva” that is “realized through spontaneously arisen wisdom/ in reliance on the words of the guru.” As we already learned above, the sahaja ācārya is the consort, and her blessing is union with her.258 In what appears to be an interesting use of word play, Buddhajñānapāda’s explanation as to why the perspective gained through tantric perfection stage practice is higher than the philosophical systems is expressed in the lines, “the lower view is refuted/ by the wisdom of the higher one.” The word “wisdom” (shes rab, prajñā), is used in tantric works to indicate the consort; indeed Vaidyapāda’s commentary identifies the sahaja ācārya precisely as “the prajñā” here.259 Buddhajñānapāda thus appears here to be playing on the double entendre of the word prajñā to 256 This passage is yet another instance in which Buddhajñānapāda states that the wisdom that the disciple gains in a tantric context comes from “the words of the guru.” Vaidyapāda makes this even more explicit. He writes, “From the words of the guru means, from what is transferred from the words of the great causal master, one directly experiences the bliss which is to be realized.” bla ma’i kha las zhes te/ de (de] P, de’i D) rgyu’i (rgyu’i ] P, rgyud D) slob dpon chen poi’i kha las rnam par ‘pho ba las mngon sum du bde ba rang la (la] D, las P) ‘byung ba rtogs par bya’o zhes so// (Sukusuma, D 111b.4-5; P 134a.8). As discussed above, the “causal master” is explained in the Sukusuma, as well as in the Muktitilaka-vyākhyāna, to refer to the guru who bestows initiation upon the disciple. 257 de la sogs te mtha’ yas pa// don de kho na la gnas nas// tha dad so sor rtog pa’o// de bas de dag thams cad kyang// |140| yang dag min pas bla dang bcas// ‘og ma’i ‘og ma’i rnal ‘byor blo// gong ma gong ma’i khyad par ‘phag// og ma’i blo ni gong ma yi// |141| shes rab kyis ni sun ‘byin no// de bas gong ma’i rim pa yis// lhan cig pa yi slob dpon gyis// byin gyis brlabs pa yang dag bya// |142| gsal zhing rab dga’ nam mkha ‘dra// rang byung lhag pa’i lha chen po// lhan cig skyes pa’i ye shes kyis// bla ma’i kha las rtogs par bya// |143| (Dvitīyakrama, verses 140-43). 258 Sukusuma, D 111b.3-4. 259 ibid. 140 mean both “wisdom” and “consort.” He is simultaneously asserting that “the lower views [of each of the lower philosophical systems mentioned] are refuted/ by the wisdom of the higher ones [i.e. the view of each of the philosophical systems higher than it],” and that “the lower views [of all of the philosophical systems described]/ are refuted by the [view attained through practice with the] consort of the higher [system of the perfection stage].” That is why, Buddhajñānapāda says, the sahaja ācārya performs the blessing by means of the higher stage (which Vaidyapāda indicates here means initiation—referring, of course, to the third initiation260) and thus in reliance upon the guru’s words the “self-arisen *adhideva” is realized. As I will explore below, Buddhajñānapāda (uniquely) uses the term *adhideva to refer to the tantric result. But what is abundantly clear from this passage is that Buddhajñānapāda is claiming that the view gained through perfection stage tantric initiation and practice is superior to that of any and all philosophical positions, even the Madhyamaka view that he places above the other philosophical perspectives. Vaidyapāda’s comments on the two lines from Dvitīyakrama verse 141 about the lower views being refuted by the wisdom of the higher views are worth citing here, as he uses them to give a short summary of the entire doxographical passage from the Dvitīyakrama (only part of which I cited above): Compared to the non-Buddhists, the Vaibhāṣikas [are higher because of asserting things to be] impermanent. Compared to them, the Sautrāntikas [are higher because of asserting those impermanent things to be] imputations. Compared to them, the Yogācārins [are higher] because [they assert that these imputations are] merely mind, and compared to them, the Yogācāra-Mādhyamikas [are even higher] because [they further assert the idea of things being merely the mind] to be just the relative level of things. Compared to them, those who uphold the *Niruttara-grantha(!) (bla na med pa’i gzhung pa) [improve further by asserting that even at the relative level, pheonomena] are nothing but wisdom.261 We see here both that Vaidyapāda takes the Mādhyamika view in question to be a Yogācāra- Mādhyamika one, as noted above, and also that he refers to the perspective that Buddhajñānapāda holds to be superior to all of the non-tantric philosophical positions as that of “those who uphold the *Niruttara-grantha”—whatever those texts might be! Vaidyapāda uses the terms Niruttara tantra (bla med rgyud) and Yoganiruttara tantra (rnal ‘byor bla na med pa’i rgyud) seemingly synonymously in the Sukusuma, and appears to understand the Yoganiruttara tantras, which he distinguishes from the Yoga tantras, to be the highest class of tantra.262 It is perhaps odd that he should here use the term grantha (gzhung) rather than tantra, but it is possible that he did so in order to include śāstras and other non-scriptural works—or quasiscriptural works like the Dvitīyakrama—that pertain to the Niruttara class, as well. In any case, given the clear description of perfection stage initiation and/or practice in the Dvitīyakrama passage on which he is commenting, I do believe that Vaidyapāda is here, with the phrase “upholders of *Niruttara-grantha” referring to the perspective of those who uphold the Yoganiruttara tantras and their associated literature. Indeed, the perspective that he describes— that even at the relative level all pheonomena are held to be nothing other than wisdom—is a perfect match for Buddhajñānapāda’s perspective on the all-pervasive nature of nondual wisdom that I discussed in the first section of this chapter. 260 ibid. 261 de yang mu stegs pa las bye brag tu smra ba ste mi rtag pa’o// de las mdo sde pa ste brtags pa’o// de las rnal ‘byor spyod pa ste sems tsam gyis so// de las rnal ‘byor spyod pa’i dbu ma pa ste kun rdzob tsam nyid kyis so// de las bla na med pa’i gzhung pa ste ye shes kho nas so// (Sukusuma, D 11b.1-2; P 134a.4-5). 262 See Sukusuma 89b.7 and 108a.6-108b.1. 141 This passage from the Dvitīyakrama shows us yet another way in which Buddhajñānapāda indicated the superiority of tantra: its perspective is, in his estimation, superior even to the highest among the philosophical perspectives that can be upheld. In another verse from the Dvitīyakrama discussing nondual wisdom, Buddhajñānapāda states that, It is not known by the śrāvakas Nor by the pratyekabuddhas The Yogācārins, Mādhyamikas, And bodhisattvas do not know it. |149|263 Here again we see clearly that even a [Yogācāra-?]Mādhyamika does not know the ultimate suchness of nondual wisdom. The reason for this lack of comprehension may be indicated in a passage from the Muktitilaka in which it is suggested that the very methodology of philosophical analysis is itself an obstacle to the realization of the ultimate nature. Buddhajñānapāda writes, This supreme nondual nature, The self that pervades all things And is beyond the purview of saṃsāra Is called the dharmadhātu. This type of perfect wisdom Is not known through direct perception and the rest Nor is it known by those who analyze.264 The intended referents in this passage of those who rely upon “direct perception and the rest,” or “those who analyze” are not immediately clear. “Direct perception and the rest” seems to be a reference to the two primary modes of knowing set forth in the writings on valid cognition (pramāṇa)—that is, direct perception (pratyakṣa, mngon sum) and inference (anumāna, rjes dpag)—that were routinely accepted and therefore addressed in nearly all subsequent Buddhist philosopical literature. “Those who analyze” could likewise easily be taken to refer to proponents of any philosophical system that approaches truth via the use of logical analysis, which of course encapsulates the practice of philosophers on the whole. Vaidyapāda explains that the verse is “taught in order to show that this type of truth is not know by the logicians (rtog ge ba, tārkika).”265 I therefore believe that we can understand this passage as a rejection of the modes of perception and analysis relied upon by proponents of any philosophical system as a means for accessing ultimate truth.266 A few verses later in the Muktitilaka Buddhajñānapāda writes, 263 nyan thos rnams kyis mi shes so// rang sangs rgyas kyis kyang mi shes// rnal ‘byor spyod dang dbu ma pa// byang chub sems dpas mi shes so// |149| (Dvitīyakrama, verse 149). 264 de gnyis med pa’i rang bzhin mchog// dngos po kun la khyab pa’i bdag// ‘khor bas rab tu ma zin pa// chos kyi dbyings zhes bshad (zhes bshad] D, kyi shes P) pa’o// de ‘dra rab mchog ye shes ni// mngon sum sogs (sogs] P V (D and P), tshogs D) pas mi shes so// de bzhin dpyad mkhan gyis mi shes// (Muktitilaka, D 47b.3-4; P 57a. 4-5). 265 de ‘dra ba’i don de rtog ge ba (ba] D, pa P) rnams kyis mi shes par bstan pa’i phyir/ (Muktitilaka-vyākhyāna, D 50a.1; P 335b.4). 266 This is not, however, precisely how Vaidyapāda has read the passage. He writes that these two pādas were included in order to show that the tārkikas (rtog ge ba rnams) do not realize nondual wisdom, but then suggests that the line on direct perception refers to the direct perception of the “outsiders” (i.e. non-Buddhists; phyi rol gyi) and the line about those who analyze refers to Brahmins (tshangs pa; the term must be understood to denote Brahmins here, not Brahmā (!)) (Muktitilaka-vyākhyāna, D 50a.1). The lines that follow explain that śrāvakas and pratyekabuddhas do not know nondual wisdom either, so Vaidyapāda is not out of line in considering these comments to refer to non-Buddhist thinkers, since Buddhajñānapāda does indeed often list different positions in a hierarchical order, in which case those positions listed before Buddhist views would logically be the non-Buddhist ones. However, given that Buddhajñānapāda in other contexts is very clear when making reference to non-Buddhist traditions (he identifies them by name), and given his comments in the Dvitīyakrama on the fact that even the (quite Buddhist!) Mādhyamikas still do not know nondual wisdom, I prefer to take his comments here more generally as referring to a style of engagement and analysis that could be applied equally by Buddhists or non-Buddhists, rather than to a particular philosophical position. 142 Why does buddhahood not come from concepts? Because it comes from the utterly pure nature.267 While these lines are found in the context of a refutation of conceptually-oriented ritual, rather than of philosophy, nonetheless the general tenor is similar to that of the above passage. For Buddhajñānapāda the tantric perspective—or at least that of the perfection stage—is superior to that of any philosophical position at least in part because it lies beyond the reach of the conceptual world, in the experience of nondual wisdom, where analysis and concepts find no purchase. The Suchness of the Second Stage: Distinguishing Suchness in a Tantric Context Yet another feature of Buddhajñānapāda’s writings through which he valorizes perfection stage tantric practice over other practices is the use of a particular term, “the suchness of the second stage,” with reference to suchness as experienced and cultivated through the perfection stage path. Buddhajñānapāda’s arguments in the Ātmasādhanāvatāra supporting the practice of deity yoga indicate his interlocutor there to be someone who accepts the possibility of meditation on suchness, but takes issue with the practice of deity yoga. Given that meditation on suchness is accepted by an interlocutor objecting to tantric practices, we can presume that what this interlocutor accepts is an exoteric non-tantric practice of the cultivation of suchness. This being the case, we might be inclined to think that the meditation on suchness advocated in Buddhajñānapāda’s oeuvre was not perceived as constituting suchness in a way that is distinct from the practices of the earlier tradition. However, the use in several places in his writings— including in the title of the Dvitīyakrama, the Oral Instructions on Training in the Suchness of the Second Stage268—of the more specific term, “the suchness of the second stage,” suggests otherwise. The term must be considered within the context of Buddhajñānapāda’s claims in both the Dvitīyakrama and Muktitilaka, examined above, that the cultivation of the suchness required for awakening is only possible through tantric sexual initiation and practices, and that the result of these practices is superior to that of the non-tantric Mahāyāna and of the lower stages of tantric practice. Given these facts, I believe that we must understand that Buddhajñānapāda held the “suchness” that he writes of cultivating by means of the perfection stage to be a genuine or full realization of suchness that was not accessible by means of non-tantric Mahāyāna methods, or indeed even by means of the lower stages of tantric practice; that is, it would follow logically from his statements on tantric superiority that what is cultivated through non-tantric Mahāyāna methods is not, in fact, the full genuine suchness that is the true nature of all phenomena. With respect to the difference between exoteric practices of meditation on emptiness and what may appear to be similar practices in the context of the perfection stage, Germano has noted that “the actual “content” and style of these meditations when isolated out from their context is near identical, and yet, when contextualized discursively and practically, the distinct semantic shapings of that similar “content” results in arguably quite different practices despite their formal similarties.”269 While it is doubtful that Buddhajñānapāda would have attributed such a difference just to “semantic shapings,” indeed it seems he did wish to draw a distinction between the discursive and practical aspects of training in “the suchness of the second stage” and that of the training in suchness advocated in earlier Buddhist traditions. It appears that with the use of the specific term “the suchness of the second stage,” Buddhajñānapāda is yet again asserting the superiority of the tantric path. Let us take a look at several of these references in his and Vaidyapāda’s works. 267 ji phyir rtog las sangs rgyas min// rang bzhin rnam dag las byung phyir// (Muktitilaka, D 47b.7; P 57b.1). 268 I discussed my translation of the Dvitīyakrama’s title in Chapter Two. 269 Germano, 1994 220-21. 143 The most prominent use of the phrase “the suchness of the second stage” is, of course, in the title of the Dvitīyakrama, which is an entire work devoted to the instructions on Training in the Suchness of the Second Stage. In that work Buddhajñānapāda uses the phrase several times apart from the work’s title. One such passage, following a list of the names, or synonyms, of suchness reads, Therefore, with the mind that has already [generated] faith, Genuinely maintain the nature of all phenomena, The profound, luminous, nondual great reality, The suchness of the second stage |283| Which has been taught by the guru.270 In this passage, the suchness of the second stage is equated with “profound, luminous, nondual great reality” itself. As we saw in the first section of this chapter, “profound,” “luminous,” and “nondual” are precisely the terms that Buddhajñānapāda frequently uses in his writings to describe suchness or the nature of reality, but here he specifies this “suchness” as specificaly pertaining or connected to “the second stage” of tantric practice. However, just prior to this passage Buddhajñanapāda writes: Non-meditation itself, The great pith instructions of the revered master, Transferred from ear271 to ear, Not known by the śrāvakas, |280| Not known by the pratyekabuddhas and others, The letterless itself, Wordless, inexpressible, and so on. In the sūtras and tantras |281| It has been expressed, and will be again, With these countless names and others. There is nothing at all taught there Besides this suchness. |282|272 Here he seems to take a more inclusive position, stating that this same suchness is precisely that which is expressed in all of Buddhist literature. And those who are specifically mentioned as left out from the knowledge are in this passage only the non-Mahāyāna śrāvakas and pratyekabuddhas, though the term “and so on” could easily include others such as the bodhisattvas or proponents of various philosophical views who, as we have seen, Buddhajñānapāda in other passages explicitly excludes from genuine knowledge of suchness. It seems, then, that while asserting the supremacy of the tantric view and at times explicitly stating that non-tantric paths cannot lead to the full result, Buddhajñanapāda is unwilling to deny that the truth of suchness itself is taught in the non-tantric and lower tantric scriptures. This would suggest that it is simply that the practices necessary for realizing and cultivating that suchness are not found in these lower paths, and that the second stage of tantric practice is thus unique in bringing the practitioner, through its practice of sexual initiation and sexual yogas, to the genuine realization of suchness, which is thus termed “the suchness of the second stage.” 270 de bas dad pa sngon ‘gro ba’i// sems kyis chos kun de bzhin nyid// zab gsal gnyis med don chen po/// rim pa gnyis pa’i de kho na// |283| bla ma’i gsung ni yang dag gzung// (Dvitīyakrama, verses 283a-284a). 271 rna] S P N V (D and P), sna D C 272 bsgom pa med pa nyid dang ni// rje btsun man ngag chen po dang// rna nas rna bar ‘pho byed dang// nyan thos kyis ni shes min dang// |280| rang sangs rgyas sogs mi shes dang// yi ge med pa nyid dang ni// tshig dang bral dang brjod med sogs// de la mdo dang rgyud rnams las// |281|de ‘dra rnam pa mtha’ yas pa// gsungs shing yang dag gsung ‘gyur ba// der ni de bzhin nyid ‘di las// |282| (Dvitīyakrama, verses 280-282). 144 In another passage from the Dvitīyakrama following the description of the ten bhūmis of sexual practice Buddhajñānapāda states: In this way, as for the suchness Of the second stage,273 Whichever yogin drinks this supreme nectar Together with the method |314| Certainly becomes a son of the buddhas, A companion of the bodhisattvas, A leader of the vidyādharas, The husband of the dākiṇīs. |315| The main guide, Leader of the śrāvakas and pratyekabuddhas, The revered master of ordinary beings.274 In this passage, likewise, it is specifically by means of the “suchness of the second stage” that the yogin attains the results of practice, and that suchness is here identified with the “supreme nectar.” Vaidyapāda seems to understand the “supreme nectar” that the yogin should drink to be wisdom, as he writes that it is to be received “from the guru’s mouth” and experienced again and again.275 I am, however, inclined to read the line more literally as a reference to the drop of bodhicitta consumed by the disciple both in the third initiation and in the context of postinitiatory sexual yogic practice. (Of course, one can also read the passage in both ways.) In any case, both of these passages, like the Dvitīyakrama’s title, specify that it is the “suchness of the second stage” that is to be cultivated in order to attain the supreme result. In his commentary, Vaidyapāda explains that this passage is meant as a “praise of remaining in the suchness of the second stage,”276 and defines “the suchness of the second stage” as the “perfection stage of the perfection stage.”277 This is presumably a reference to a passage from the Muktitilaka where the two stages of tantric practice are said to each be twofold themselves. In that passage, Buddhajñānapāda himself mentions the “perfection stage of the perfection stage,” which he there identifies both with the seven yogas, as well as with unsurpassed omniscience.278 I will discuss Buddhajñānapāda’s classifications of the two stages further in Chapter Five, but for now it is sufficient to note that, if we follow Vaidyapāda’s identification of this suchness with the “perfection stage of the perfection stage,” then the “suchness of the second stage” refers to suchness that is identical with the result of awakening, and it is precisely this which is cultivated or trained in through the practice of the perfection stage. 273 Here Vaidyapāda describes this as “the perfection stage of the perfection stage,” following Buddhajñānapāda’s four-fold classification of the creation and perfection stages in the Muktitilaka (Sukusuma, D 128b.6-7). 274 de ltar rim pa gnyis pa yi// de bzhin nyid ni thabs bcas pa’i// bdud rtsi mchog ‘di rnal ‘byor gang// ‘thung bar byed pa sangs rgyas kyi// |315| de yi sras su nges pa ste// byang chub sems dpa’ rnam kyi grogs// rig pa ‘dzin pa’i dpon po ste// mkha’ ‘gro ma yi khyo ru ‘gyur// |316| nyan thos rang ‘dren rnams kyi ni// ‘dren par byed pa’i gtso bo ste// sems can phal pa’i rje brtsun no// (Dvitīyakrama, verses 314a-316c). 275 rnal ‘byor gang ‘thung bar byed pa’i zhes pa ni bla ma’i zhal nas blangs nas de’i ro yag dang yang du myong byar byed pa’o// (Sukusuma, D 129a.1). 276 rim pa gnyis pa’i de kho na nyid la gnas pa’i sngags (Sukusuma, D 128b.6). 277 rim pa gnyis pa’i de bzhin nyid ni rdzogs pa’i rim pa’i rdzogs pa’i rim pa’o// (Sukusuma, D 128b.6-7). Vaidyapāda again identifies the “suchness of the second stage” with perfection stage practice in his Muktitilakavyākhyāna, when introducing the section of the Muktitilaka on the three bindu yogas of the perfection stage. There he writes, “Having [first] remained in the generation stage, [now] in order to teach the training in the suchness of the second stage, the text begins...” (da ni de ltar bskyed pa’i rim pa la gnas pas rim pa gnyis pa’i de kho na (kho na] D, P om.) nyid bsgom pa bstan pa’i phyir... (Muktitilaka-vyākhyāna, D 52b.3-4, P 339a.1-2). 278 Muktitilaka, D 52a.1-2. 145 Another passage from the Dvitīyakrama also equates suchness or reality with the second stage of tantric practice, and likewise identifies this with the final result, here called nirvāṇa: The sphere of the Buddha’s nirvāṇa The unborn vajra, manifest awakening, The supreme essence of all sugatas, This great nondual nonconceptual reality Is explained as the second stage.279 |34|280 The final line of the verse (rim pa gnyis par rab tu bshad) could equally well be translated so that the verse reads “this great nondual nonconceptual reality/ is explained in the second stage,” rather than as the second stage. Both readings of the Tibetan are perfectly suitable. But while it is certainly the case that nondual reality is explained in the teachings on the second stage, given the descriptions of the “suchness of the second stage” that we have seen above, and the equation of the suchness of the second stage with the final result of practice, I believe it is better to understand the verse to be saying that nondual reality “is explained as the second stage.” That is, the second stage is the “perfection” stage—utpannakrama could actually be translated more literally as the perfected stage—precisely because it is the stage of training in the reality that itself is already perfected, already manifest.281 Thus, again, in referencing the “suchness of the second stage,” Buddhajñānapāda raises the status of perfection stage practices, and their connection with the final, perfect tantric result, over those of other non-tantric and nonperfection stage practices. The use of the phrase “the suchness of the second stage” did not catch on among other tantric writers, however; within the translated Tibetan canon it is found only in Buddhajñānapāda’s writings and Vaidyapāda’s commentaries. This phrase thus appears to be a unique way in which Buddhajñānapāda chose to qualify and distinguish tantric—and specifically perfection stage—practice from that of the exoteric Buddhist traditions. Expressing the Result of Buddhist Practice in Tantric Terms Buddhajñānapāda writes of the result of Buddhist practice with an array of commonly used Buddhist terms: “nirvāṇa,” “buddhahood,” “dharmakāya,” “suchness,” “reality,” and others. However, there several places in his writings where he refers to the result of awakening in specifically tantric terms. Examining these passages gives us a better sense of what Buddhjñānapāda understood to constitute the result of tantric practice, which, as we saw above, he holds to be superior to the non-tantric result. At several points in the Dvitīyakrama Buddhajñānapāda employs an unusual term for the final result of tantric practice. The term that he uses there, the *adhideva (lhag pa’i lha),282 means superior or supreme deity, and can at times 279 Vaidyapāda gives three synonyms for the second stage: the spontaneously generated stage (lhan cig skyes pa’i rim pa), the perfection stage (rdzogs pa’i rim pa), and the stage of [things] just as they are (ji bzhin pa’i rim pa). He then gives a brief description of the generation stage: “as for the generation stage it is for the purpose of reversing the coarse delusions of the world and its contents. This yoga that involves engaging with the conceptual mind is the first [stage].” rim pa gnyis par rab tu bzhad zhes pa ni/ lhan cig skyes pa’i rim pa ‘am/ rdzogs pa’i rim pa ‘am/ ji bzhin pa’i rim pa (rim pa] P; om. D) rnam grangs so// bskyed pa’i rim pa ni snod bcud rags par ‘khrul pa bzlog pa’i phyir ro// blos rnam par gzhag pa’i rnal ‘byor pa ste dang po’o// (Sukusuma, D 96b.6-7). 280 sangs rgyas mya ngan ‘das pa’i khams// skye med rdo rje mngon byang chub// bder gshegs kun gyi snying po mchog// gnyis med rtog bral don chen te// rim pa gnyis par rab tu bshad// |34| (Dvitīyakrama, verse 34). 281 Moreover, as I explore in Chapter Six, Buddhajñānapāda most frequently uses the terms “perfection stage” (utpannakrama; rdzogs pa’i rim pa) and “second stage” (dvitīyakrama; rim pa gnyis pa) to refer to suchness itself, rather than the yogic methods that one uses to cultivate this experience. 282 Unfortunately, I cannot be certain about the reconstruction of the term lhag pa’i lha; there are a number of good candidates. The terms adhideva, adhidevatā, and adhidaiva, are all attested for lhag pa’i lha (Negi 1993, vol 16., 7546). The term adhidaivata is also closely linked with, and often used synonymously with these terms, though it 146 be used as synonymous with the term tutelary deity (iṣṭadeva(-tā), yi dam), the personal meditation deity of a practitioner. However, the way that Buddhajñānapāda employs the term in the Dvitīyakrama makes it clear that in his usage *adhideva does not refer (just?)283 to the tutelary deity, but rather to the final result of the Buddhist path—a result that is, according to his system, achieved only through tantric practice. At the conclusion of the Dvitīyakrama’s description of the types and qualities of tantric consort, and immediately prior to the description of the rituals for the second and third tantric initiations in which the consort plays a crucial role, Buddhajñānapāda states what seems to be the purpose of relying upon consort practices: In that way, by means of the illusory great mudrā Of that type of female, The so-called *adhideva, So difficult to encounter in the three realms, will be accomplished. |82|284 Here Buddhajñānapāda does not indicate clearly what he means by the *adhideva, but Vaidyapāda’s comments suggest that he understands the term to refer to Mahāvajradhara, whom he states to be superior to the gods, bodhisattvas, and even the buddhas. Vaidyapāda explains that Mahāvajradhara is present in all beings (presumably as their basic nature) but can only be accomplished using the “higher methods which seal by means of wisdom,” that is to say, practices involving a consort.285 Just a few verses later, in the first verse in the Dvitīyakrama dealing with the ritual for the third initiation, Buddhajñānapāda himself indicates that the

• adhideva refers to the final result of practice, and that it is for the purpose of attaining this result

that the disciple is to practice with his consort. He writes, And then, in order to bring about the realization Of the self-arisen dharmakāya, great joy That is equal to space, called the *adhideva, The girl is given to him [i.e. to the disciple].286 |86|287 Here the *adhideva is directly equated with “the dharmakāya, great joy that is equal to space,” a clear reference to one among the kāyas, or bodies, of awakening that are said to be the result of Buddhist practice. We should note that it is not either of the rūpakāyas, the form-based bodies of seems to be more commonly translate as rab tu che ba’i lha (see, for example, the Sarvabuddhasamāyayoga-tantra, I.2, which Buddhajñānapāda incorporates into verse 314 of the Dvitīyakrama, and which I discuss below). 283 As we saw earlier in this chapter, Buddhajñānapāda uses the term “profound and luminous” (zab gsal) to refer to suchness, and the way he describes the aspect of luminosity appears to connect this aspect to the mahāmudrā, the form of the deity. (The passage in the Muktitilaka in which we find this connection, which I already discussed above, reads: “Because it is the purification of the mind of oneself and others/ As illusory and rainbow-like/ In the form of the mahāmudrā/ It is called genuine luminosity,” phyag rgya chen po’i gzugs ‘chang ba// sgyu ma ‘ja’ tshon lta bu ru// rang dang gzhan gyi rgyud sbyong bas// yang dag gsal ba zhes bya’o// (Muktitilaka, D 47b.2-4).) Given this connection, it is possible that in using the term *adhideva to refer to the final result of tantric practice, Buddhajñānapāda is also bringing in a form-based aspect of the result of awakening—the practitioner actually manifesting in the form of her personal tutelary deity—though we will see below that in the Dvitīyakrama he explicitly links the *adhideva not with the form kāyas, but with the formless dharmakāya. 284 de ltar de sogs bud med kyi// sgyu ma’i phyag rgya chen pos ni// ‘jig rten gsum du rnyed dka’ ba’i// lhag pa’i lha shes bya ba bsgrub// |82| (Dvitīyakrama, verse 82). 285 ci’i phyir lhag pa’i lha zhes bya zhe na/ lha rnams las mchog tu gyur pa ni kha na ma tho ba med pa’o// de las mchog tu gyur pa ni byang chub sems dpa’o// de las mchog tu gyur pa ni sangs rgyas rnams so// de rnams kyi phul du gyur pa ni rdo rje ‘chang chen po sbyor ba bdun dang ldan pa’o// de lta bu sems can thams cad la mi slu ba’i tshul du gnas kyang ye shes kyis rgyas btab pa’i thabs gong ma dang bral na mi ‘grub pas/ de rnams dang lhan cig tu gyur pa’i lhag pa’i lha zhes bya ba sgrub ces bya’o// (Sukusuma, D 102b.3-5; P 123b.2-5). 286 de la. Vaidyapāda clarifies that in response to the disciple’s supplications the guru gives the girl “that he has blessed” into “the disciple’s right hand” and recites the subsequent verses. (Sukusuma, D 105b.6-7). 287 de nas de la rang ‘byung gi// chos sku rab dga’ mkha’ nmyam pa// lhag pa’i lha zhes bya ba ni// rtogs bya’i ched du bu mo byin// |86| (Dvitīyakrama, verse 86). 147 awakening, that Buddhajñānapāda equates with the *adhideva—which we might have expected given the usual use of the term to mean the tutelary deity, who indeed has a specific form—but rather the dharmakāya, the formless buddha kāya that is “equal to space.” And again, this passage clearly states that it is in order to bring about this final realization that the disciple is to take up practice with his consort. Buddhajñānapāda mentions the *adhideva once more, in a passage that we examined earlier, found at the conclusion of the Dvitīyakrama’s philosophical doxography in which he sets forth the highest—the tantric—position. There, just after a verse describing the sahaja ācārya (i.e. the consort) performing the genuine blessing (i.e. that of union), it is mentioned that the *adhideva is accomplished through the guru’s words: Luminous and perfectly joyful like the sky The self-arisen great *adhideva Is realized through innate wisdom288 In reliance on the words of the guru. |143|289 It is again clear here that the *adhideva is the result that is to be accomplished, and the context of the passage indicates that this is achieved specifically by means of sexual yogas performed with a consort. The first two lines of the final passage from the Dvitīyakrama in which we find the term

• adhideva are parallel with the first two pādas of Chapter One, verse 2 of the

Sarvabuddhasamāyoga-tantra.290 This unattributed citation from the Sarvabuddhasamāyoga employs the term “supreme deity” (adhidaivata, rab tu che ba’i lha), which Buddhajñanapāda’s verse equates, in the next line, with the *adhideva (lhag pa’i lha); indeed these two terms appear often to be used synonymously.291 This verse in the Dvitīyakrama gives more detail about the nature of the result that is to be accomplished, and in the final line frames that result in an even more specifically tantric context. The verse follows the passage homologizing the ten bhūmis with sexual practices and indicating that these sexualized bhūmis are superior to their exoteric Mahāyāna counterparts. As we saw above, that passage suggests that it is only through such sexual yogas that a practitioner can come to genuinely know reality. Buddhajñānapāda states that those who lack knowledge of such reality are not genuine buddhas, and then writes: This is the self-arisen bhagavan The sole supreme deity (adhidaivata), Called the *adhideva 288 lhan cig skyes pa’i ye shes, *sahajajñāna(?) 289 gsal zhing rab dga’ nam mkha ‘dra// rang byung lhag pa’i lha chen po// lhan cig skyes pa’i ye shes kyis// bla ma’i kha las rtogs par bya// |143| (Dvitīyakrama, verse 143). 290 Those two pādas read: asau svayaṃbhūr bhagavān eka evādhidaivataḥ. Thanks to Harunaga Isaacson for bringing this parallel to my attention and to Péter Szántó for sharing with me his draft Sanskrit edition of the Sarvabuddhasamāyoga-tantra. These same two pādas of the tantra are also incorporated into the Pañcakrama IV.2 (see Tomabechi 2006, 165). 291 The way the term adhidaivata as used in this particular passage of the Sarvabuddhasamāyoga-tantra (I.1-2) in fact appears somewhat related to the way Buddhajñānapāda uses the term *adhideva. The adhidaivata is identified there with Vajrasattva and with the bhagavan, terms that do, of course, represent the resultant state of awakening. However, the way that the term adhidaivata is used throughout the remainder of the tantra seems more consonant with the meaning of “tutelary deity,” or “personal deity,” which really does not correspond with the way that Buddhajñānapāda uses the term *adhideva. I suspect that the way that the term adhidaivata is used in the Sarvabuddhasamāyoga-tantra I.1-2 inspired Buddhajñānapāda’s use of the (easily synonymous) term *adhideva. Indeed, it seems likely that the Sarvabuddhasamāyoga-tantra exercised an even broader influence on Buddhajñānapāda’s thought; as we saw earlier in this chapter, that tantra’s comfort with the term “self” (ātman, bdag) is also reflected in Buddhajñānapāda’s writings, and I suspect that a line from the Sarvabuddhasamāyoga may possibly have acted as a scriptural basis for Buddhajñānapāda’s presentation of the three blisses occurring in tantric practice. I discuss the latter point in Chapter Six. 148 [And] explained to be the thirteenth bhūmi. |314|292 While the traditional Mahāyāna system has only ten bodhisattva bhūmis, later tantric systems sometimes added up to an eleventh, thirteenth, or even more bhūmis, thus setting the tantric result as distinct from and superior to that of the exoteric Mahāyāna. Here, in equating the

• adhideva with the “thirteenth bhūmi,” Buddhajñānapāda is yet again asserting the superiority of

the tantric result. It is noteworthy, as well, that both this and the previous passage use the term “self-arisen;” this term is used to refer directly to the *adhideva in the first passage, and to the bhagavan, who is equated with the *adhideva, in the second. Describing the ultimate result of practice in this way corresponds with Buddhajñānapāda’s assertion explored at the beginning of this chapter that nondual wisdom—which is equated with suchness and therefore with the final result of practice—is in fact the innate nature of all phenomena, including the nature of mind. In this passage the *adhideva is also called, or equated with, the “supreme deity” (adhidaivata, rab tu che ba’i lha), further emphasizing the “deity” (deva, daivata) aspect of the term. Indeed “supreme deity,” and “*adhideva” are somewhat curious terms to use for the result of the path of an ultimately non-theistic system like Buddhism. While the term *adhideva is certainly used in tantric Buddhist texts, in most such works—at least as far as I am aware—it is used in its aformentioned sense of the tutelary deity, as synonymous with iṣṭadeva, but not to indicate the result of tantric practice, as Buddhajñānapāda uses it.293 The term *adhideva is, however, used throughout non-Buddhist Indian religious literature to indicate any number of supreme deities in the many theistic religious systems that flourish there. The use of this term in Buddhajñānapāda’s writings to indicate the tantric result could be seen as a way of linking the practice of deity yoga as a crucial tantric method with the final result of practice (attained, of course, through the perfection stage), a “space-like” state that transcends even concepts like “deity;” but it could also be seen as a further indication of Buddhajñānapāda’s familiarity and interaction with non-Buddhist practice systems. In any case, like the phrase the “suchness of the second stage,” using the term *adhideva to qualify the final result of tantric Buddhist practice appears to be a unique feature of Buddhajñānapāda’s writings. In the Muktitilaka we find another presentation of the result of the path in specifically tantric terms. Here, Buddhajñānapāda makes reference to the different kāyas that arise from engaging in the two stages of tantric practice. After having first explained that the practice of the dharma is divided into two stages, the generation and perfection stages, he writes: Due to the distinction of the two stages There [arise] the mantrakāya and the jñānakāya. The jñānakāya has two [aspects]: The jñāna [kāya] and the kāya of complete purity (*viśuddhikāya?). This completely pure [kāya] is the seven yogas, The perfection stage of the perfection stage; It is unsurpassed omniscience Endowed with the supreme of all aspects.294 Vaidyapāda explains this passage as follows, 292 di ni rang byung bcom ldan ‘das// gcig pu rab tu che ba’i lha// lhag pa’i lha zhes bya ba ni// bcu gsum sa zhes bya bar bshad// |314| (Dvitīyakrama, verse 314). 293 Adhideva is used in the sense of iṣṭadeva, for example, in the Vajrāvalī (See Mori 2009 Vol. 2, 357). 294 bskyed pa’i rim pa rnam pa gnyis// de bzhin rdzogs pa’i rim pa’o// rim pa gnyis kyi bye brag gis// sngags kyi sku dang ye shes sku// ye shes sku la ye shes dang// rnam par dag pa’i sku gnyis so// rnam par dag pa sbyor ba bdun// rdzogs pa’i rim pa rdzogs rim ste// rnam pa kun gyi mchog ldan pa// thams cad mkhyen pa bla me ‘gyur// (Muktitilaka, D 52a.2-3; P 62b.2-4) 149 Then, in order to indicate the different types of bodies of the deity [that come about] by means of those two stages [he writes] Due to the distinction of... Here, the mantrakāya is the kāya that is generated. [It is so called] because it arises from [syllables like] hūṃ and so forth, [and] it is impure. The jñānakāya is the perfected kāya and is pure. That [perfected kāya] also has two divisions, the jñāna [kāya], or the illusory body, which is the pure kāyā [attained] by the yogin of the third [level] who has slightly [gained] the wisdom of realization; and the kāya of complete purity. This [kāya of complete purity] also has two [aspects]: that which remains in a state in which it displays characteristics, and is therefore [called] the unfailing kāya, and the resultant kāya, which is [the nature of that,] just as it is. These two are the primordially pure completely pure body, which is the seven yogas. That is also the perfection stage of the perfection stage. You should know that precisely this is also omniscience endowed with the supreme of all aspects.295 Since the distinction into the generation and perfection stages was made immediately prior to this passage, it is clear that the mantrakāya is meant as the kāya that comes about due to the practice of the generation stage, while the jñānakāya is that which results from the perfection stage. As we have already seen, Buddhajñānapāda makes it clear in his writings that the final result of practice comes about only through perfection stage practice, so the mantrakāya cannot be that final result. Indeed, as Vaidyapāda explains, the mantrakāya is an impure kāya that is socalled because it arises from syllables. This certainly refers to the visualized form of the deity that the practitioner generates from a seed-syllable during generation stage practice. Among the two aspects of the jñānakāya that arise due to the practice of the perfection stage, the second one, called the completely pure kāya, is identified with the final result of practice, described here as “unsurpassed omniscience,” a commonly used term for the state of awakening in the Mahāyāna tradition. The use of the common Mahāyāna term for the awakened state here in the Muktitilaka can be understood on the one hand to explicitly and clearly equate the more tantric terms, the “seven yogas,” and “the perfection stage of the perfection stage,” with the more common nontantric terms for awakening. But again, in explicitly connecting this common term for the result of awakening with the second stage of tantric practice, Buddhajñanapāda may also be indicating that the tantric path is the only way to bring about this final result. While the term jñānakāya itself is far from uncommon, Buddhajñānapāda’s association of the jñānakāya with perfection stage practice and his division of this jñānakāya into two aspects is worth examining a bit further here. Since Buddhajñānapāda himself does not elaborate on the first of these two aspects of the jñānakāya—he only notes that this first aspect is itself also called simply the jñānakāya—we must rely here on Vaidyapāda, who identifies this kāya as one that is produced by the perfection stage practices of “a yogin of the third [level] who has slightly [gained] the wisdom of realization.” The third level is one of a series of three (or four?) levels of progress in yogic practice mentioned in Buddhajñānapāda’s and Vaidyapāda’s writings, which appears to be the stage from which the yogin progresses on to the final result of awakening.296 This first aspect of the jñānakāya that arises in the practice of a third-level yogin, as it is 295 da ni rim pa gnyis kyis lha’i sku’i bye (bye] P., bya D) brag bstan pa’i phyir rim pa gnyis kyis zhes pa la sogs pa’o// de la sngags kyi sku zhes pa ni bskyed pa’i sku ste/ hūṃ la sogs pa las byung ba’i phyir te ma dag pa’o// ye shes sku zhes pa ni rdzogs pa’i sku ste dag pa’o// de la yang gnyis te ye shes dang zhes pa ni sgyu ma lta bu’i sku ste/ rnal ‘byor pa gsum pas rtogs pa’i ye shes cung zad dag pa’i sku’o// rnam par dag pa’i sku zhes pa dang gnyis so// de la yang gnyis so// mtshan nyid kyi tshul du gnas pas na mi (na mi] D, ni P) slu (slu] D, bslu P) ba’i sku dang/ de ji bzhin pa’i ‘bras bu’i sku ‘o// de gnyis ni ye nas shin tu dag pa’i sku ste sbyor ba bdun no// rdzogs pa’i rim pa’i rdzogs rim kyang de’o// rnam pa kun gyi mchog dang ldan pa’i thams cad mkhyen pa yang de nyid yin par shes par bya’o// (Muktitilaka-vyākhyāna, D 38b.2-6; P 366b.4-367a.2). 296 See note 139. 150 described by Vaidyapāda here, is a pure body called the illusory body (māyādeha), a term that is found more commonly in texts laying out the Ārya School perfection stage practices, where it describes the practitioner’s attainment of a type of “wisdom body” of the deity during the third among the so-called “five stages” (pañcakrama) of practice according to that tradition. The use of the term māyādeha in Vaidyapāda’s writings, in a way that appears rather similar to its usage in the Ārya School, is an interesting precursor to the use of that term in that tradition’s slightly later Guhyasamāja practice system.297 But, according to Vaidyapāda’s presentation here (and, perhaps not incidentally, also according to the Ārya School’s system) this is not yet the final result of practice. The aspect of the jñānakāya that is, according to the Muktitilaka, equivalent to the state of omniscience itself is the so-called kāya of complete purity (*viśuddhikāya?), which Buddhajñānapāda identifies with the seven yogas, and with the “perfection stage of the perfection stage.” The seven yogas, as I have mentioned above, are seven aspects of the state of awakening that are referred to in Buddhajñānapāda’s writings and presented in more detail in Vaidyapāda’s works,298 and the “perfection stage of the perfection stage” is likewise identified with suchness itself. Vaidyapāda, however, divides even this kāya of complete purity into two aspects: one aspect that displays characteristics, or form, which he refers to here as the “unfailing kāya” —presumably this corresponds with the rūpakaya, the “form body” of perfect awakening—and another aspect that he calls the “resultant kāya,” and describes as the very nature just as it is—presumably this would parallel the dharmakāya, free from any type of form or characteristics. These passages in the Muktitilaka and the Muktitilaka-vyākhyāna are important in providing us a sense of how both Buddhajñānapāda and Vaidyapāda understood the results of different types of tantric practice and their relationship with the state of final awakening. We see clearly again that it is only through the practice of the perfection stage that a yogin is able to achieve the kāyas that constitute the final state of awakening; the results that are achieved through lower levels of practice are inferior (and it seems that, at least according to Vaidyapāda’s presentation, even the perfection stage itself first brings about a pure kāya that is not yet the final result). However, it is worth noting that all of the kāyas mentioned here—including even those describing the manifestation of final awakening—seem to be in some way related to the deity. At the beginning of the passage in the Muktitilaka-vyākhyāna Vaidyapāda states that all of these different kāyas that are brought about through the generation and perfection stage practices are “different types of bodies of the deity” (lha’i sku’i bye drag), and the kāyas that are mentioned and described do indeed seem to be a series of progressively refined or purified forms of the deity. Thus even the seven yogas and the “perfection stage of the perfection stage,” which are terms or aspects of the final state of awakening, unsurpassed omniscience, are also somehow understood to constitute “types of bodies of the deity.” Describing awakening in this way certainly gives the presentation of the final result of Buddhist practice a distinctly tantric flavor. We have seen now quite a number of ways in which Buddhajñānapāda distinguished the tantric perspective, practice, and result from that of non-tantric paths in his writings, and both directly and indirectly indicated the superiority of tantric practice generally, and that of the perfection stage specifically. While the position of asserting the superiority of the tantric result was certainly held by some tantric Buddhist authors, as noted above, it seems that 297 As I discussed in Chatper One, the Ārya School appears to be slightly later than the Jñānapāda School. Vaidyapāda cites the Anuttarasandhi of Śākyamitra, who Toru Tomabechi has argued may well be an intermediary figure between the Jñānapāda and Ārya Schools, but otherwise does not seem to show awareness of any writings of the Ārya School. 298 The most detailed presentation of them in Vaidyapāda’s writings is found in his Yogasapta, which I discuss in Chapter Seven. 151 Buddhajñānapāda may perhaps have been in the minority in making this claim. The evidence found throughout his oeuvre, however, leaves little doubt that this was his position, and that he specifically associated the superior tantric result with the view and practices associated with the perfection stage. Let us now move on to look at another doctrinal feature of Buddhajñānapāda’s work, namely, the rhetoric of non-action, which we also find closely associated in his writings with the second stage of tantric pratice. V: Ritual and the Rhetoric of Non-action The yogin who holds actions to be the great path is like a wild animal chasing a mirage—the goal continually appears, but can never be grasped. -Mañjuśrī instructing Buddhajñānapāda, Dvitīyakrama Buddhajñānapāda’s writings illustrate a tension between the detailed ritual practices of the tantric path and a rhetoric of non-action, connected particularly with the second stage of tantric practice. It is clear from his writings that Buddhajñānapāda advocated the practice of the rituals of Mahāyoga tantra; he composed several sādhanas for generation stage practice, including his well-known Caturaṅga/Samantabhadra-sādhana, which set forth the elaborate ritual structures of that stage of tantric practice. He also composed a short work, the *Gativyūha, detailing the specifics of the physical postures to be assumed during certain ritual contexts. Moreover, even his more perfection stage-oriented works include ritual instructions, like the Dvitīyakrama’s ritual formula for the higher tantric initiations and perfection stage visualizations, and the Muktitilaka’s brief elaboration on the tantric samayas. In the latter passage Buddhajñānapāda declares flesh, marrow, blood, bone, and the subtle channels to be equivalent to the five buddhas, and gives an injunction to consume a number of such impure substances and also to offer them in the homa. However, both the Dvitīyakrama and the Muktitilaka also include passages which declare a variety of external ritual practices, including maṇḍala, homa, mantra recitation, and other practices, to be unnecessary and advise that the yogin who knows the true nature of reality should give them up. The Dvitīyakrama seems to advocate a middle-road perspective on the issue, negotiating the need for engaging in outward practices but also the need to abandon them, presumably in the context of perfection stage practice: Thus the maṇḍala, homa, Bali, recitation, the counting rosary, |151| Sitting cross-legged, maintaining postures, and so forth— Are in contradiction299 to the unelaborate, [Thus] they should not be [exclusively] taken up; but neither should they be [wholly] rejected Since they are emanated by the *adhideva. |152| The yogin who holds actions To be the great path Is like a wild animal chasing a mirage— [The goal] continually appears but can never be grasped. |153| When infected by the great sickness of actions, The one who heals [himself] with the great medicine Of unwavering wisdom is a sublime being. |154|300 299 Tib. rnam par slu ba, Skt. *visaṃvāda? 152 This passage emphasizes the problematic nature of action in general, and specificialy the various ritual practices that involve such outward actions, privileging “unwavering wisdom” as the antidote to such action-focused spiritual pursuits. This passage is preceded by one expressing the superiority of the nondual wisdom of the second stage of tantric practice, and followed by detailed instructions on cultivating suchness by means of the three bindu yogas of the perfection stage. Thus we can understand both “the unelaborate” and the “unwavering wisdom” that is mentioned here, to refer to the nondual wisdom of the perfection stage of tantric practice. When Buddhajñānapāda states that the various ritual practices including the maṇḍala, homa and the rest “should not be [exclusively] taken up” nor “[wholly] rejected,” he seems to be drawing a middle ground between advocating these practices in the context of the earlier stages of the tantric path—those that lead up to and frame the perfection stage—and cautioning that they must be left behind in the final training in the suchness of the second stage. But his statement that “they are emanated by the *adhideva” also seems to suggest that these more action-oriented practices are somehow expressions of the very wisdom that is the result of tantric practice. Vaidyapāda here explains further that, “neither should they be [wholly] rejected means that they should not be rejected since through remaining in these [activities] one perfects the accumulation of merit. The *adhideva is the characteristic-less deity, and [the text] says [those practices] are emanated by [the *adhideva] because it is the source of everything.”301 In this way, Vaidyapāda links the *adhideva—a term which, as we saw above, Buddhajñānapāda uses to refer to the result of practice—with the wisdom that is, in Buddhajñānapāda’s system the very source of the entire phenomenal world, here including the tantric practices that involve ritual action and elaboration. The Muktitilaka takes a somewhat stronger position on the point of avoiding action, but again the advice to abandon action-based ritual practices is specified as being “for the yogin who knows [the] reality” of suchness. The passage in question reads: For the yogin who knows this reality What is the point of fatiguing himself With the activity of the maṇḍala? With the activity of bali? With the activity of homa? With the activity of counting mantras? With the activity of sitting cross-legged and the rest? Since such things are [just] meant to fool302 beginners! Why does buddhahood not come from concepts? Because it comes from the utterly pure nature. All of those conceptual rituals, All of that weariness—give it up! Activity arises from this nature303 300 de la dkyil ‘khor sbyin sreg dang// gtor ma bzlas pa bgrang phreng dang// |151| skyil mo krung dang stang stabs sogs// spros bral rnam bar slu ba ste// bya ba ma yin dgag pa min// lhag pa’i lha yis sprul phyir ro// |152| bya ba rnams la rnal ‘byor pa// lam chen dag tu yongs ‘dzin pa// ri dwags smig rgyu snyeg pa ltar// rtag tu snang yang ma zin no// |153| bya ba’i nad chen gyis zin la// ye shes g.yo med sman chen gyis// gso byed skyes bu dam pa’o// |154| (Dvitīyakrama verses 151b-154). 301 dgag pa ma yin zhes te de la gnas pas bsod nams kyi tshogs rdzogs par bya ba’i phyir mi dgag ste/ lhag pa’i lha ni mtshan ma med pa’i lha ste/ de las thams cad ‘byung ba’i phyir na sprul phyir ro// zhes so// (Sukusuma, D 112b.2-3; P 135a.8-135b.1). 302 ‘drid. See note 306 below. 303 Vaidyapāda elaborates, “within the nature there is everything!” rang bzhin las ni thams cad yod do// (Muktitilaka-vyākhyāna, D 50b.3-4). 153 [And] action also arises from this nature.304 [But] having come to know that nature Activity is exhausted and there is no action.305 The verses immediately prior to this passage discuss the “supreme nondual nature,” and how it is beyond the purview of non-tantric practitioners. Thus, the Muktitilaka advocates that a practitioner who knows this reality abandon all sorts of action-based practices that are “meant to fool beginners!” This is certainly a provocative statement,306 but Vaidyapāda explains that it is because beginners are obscured by so many concepts that their minds are unable to remain in nonduality, such that even if that nonduality were taught, their minds would not be tamed. Thus in order to “fool” them—presumably into training in the practices that will eventually prepare them to approach nonduality directly—they are taught these action-based practices.307 Similar to the statement in the previous passage from the Dvitīyakrama that action-based practices are “emanated from the *adhideva,” here the Muktitilaka explains that action and activity do originate from the ultimate nature. However, when that nature is realized—which is precisely what happens via the second stage of tantric practice—action and activity are exhausted. What follows in the text is an example: it is just like when someone makes and uses a raft to cross a river, but once she has crossed the river she leaves the raft behind and goes along on her way. Again, the passage seems to indicate that action-based practice are necessary in the beginning stages of practice, but should be set aside when one has a direct experience of suchness. In this way, the rhetoric of non-action found in Buddhajñānapāda’s writings appears to be directed specifically towards the second stage of tantric practice, which is the focus of both the Dvitīyakrama and the Muktitilaka. One way of understanding these statements is to regard the rhetoric of non-action in Buddhajñānapāda’s writings as just that—a rhetoric, which does not necessarily serve to undermine the action-based practices that Buddhajñānapāda also clearly advocates. As we will see below, this is one way that scholars have understood the similar rhetoric found in the context of the writings of the Great Perfection tradition.308 Another possible way to understand the statements about non-action is to read them as denying just the external activies that were emphasized in earlier tantric systems that were more ritually focused, and suggesting that these are to be set aside in the context of performing perfection stage practices that were more directed towards inward meditative processes. Indeed, many of the examples given in these passages on what the yogin is to give up are precisely such outer rituals, like maṇḍala practices, bali offerings, homa, mantra recitation, and so forth. It does seem that Buddhajñānapāda is suggesting a rejection of these external ritual practices in the context of the second stage of tantric practice where there is more of a focus on the internal processes of visualization and the manipulation of the winds and energies of the subtle body. Indeed, Jacob 304 D om. this line 305 ‘di yi don shes rnal ‘byor pas// bya ba’i las ni dkyil ‘khor dang// bya ba’i las ni gtor ma dang// bya ba’i las ni sbyin sreg dang// bya ba’i las ni bgrang phreng dang// bya ba’i skyil mo krung la sogs// dal byed pas ni ci zhig bya// dang po pa rnams ‘drid phyir ro// ci phyir rtog las sangs rgyas min// rang bzhin rnam dag las byung phyir// rtog bcas cho ga de dag kun// ngal ba la sogs spangs byar bya// rang bzhin las ni byed pa’ang yod// rang bzhin las ni bya ba’ang yod// (rang bzhin las ni bya ba’ang yod// ] P, D om. this line) rang bzhin yongs su shes pa la// byed pa sed (sed] D, med P) cing bya ba’ang med// (Muktitilaka, D 47b.6-48a.1; P 57a.8-b.2). 306 The text reads ‘drid, and Vaidyapāda’s commentary supports this reading (Muktitilaka-vyākhyāna, D 50b.2). As this is a rather strong statement, I wonder if ‘drid could be a transmission error for ‘khrid, which would mean “guide” instead of “fool”? This would be a slightly less provocative statement that would also make sense, and is an easy slip. However, as tempted as I am to make this emendation, the text as stands clearly reads ‘drid, so I have no basis for doing so other than conjecture. 307 Muktitilaka-vyākhyāna, D 50b.1-2. 308 Bernard Faure has also made reference to such a “rhetoric of immediacy” in his work by that title on Chan/Zen Buddhist traditions (Faure 1991). 154 Dalton has written about a move towards the “internalization” of Buddhist tantric practice precisely in the 8th and 9th centuries withen Buddhajñanapāda was writing,309 and I believe that we can understand this shift in focus to be at least part of what these passages on non-action in Buddhajñanapāda’s writings suggest. However, what is not clear from his surviving writings is precisely how this advocacy of non-action is to be understood with regard to the ritualized and action-based practices that pertain to the perfection stage itself. That is, as we shall see in Chapter Six, many of the perfection stage practices described in Buddhajñānapāda’s works do involve action—sexual activity to be carried out with the tantric consort, manipulations of the wind and energies of the subtle body while in union with the consort, and the visualizations that accompany these practices. As we saw above, in addition to rejecting external ritual activity, the passages on nonaction also refer specifically to wisdom as a “remedy” for action, and describe the rituals that are to be given up as “conceptual.” We saw earlier in this chapter how Buddhajñānapāda’s writings emphasize freedom from conceptuality—that is, freedom even from conceptual “action”—as an important quality of nondual wisdom. Another possible way, then, of understanding the context and function of Buddhajñānapāda’s statements about non-action with regard to perfection stage practices is in relation to the passage in the Muktitilaka that divides the perfection stage itself into two stages. Vaidyapāda explains that the first of these is the “generation stage of the perfection stage,” which consists of the three bindu yogas—precisely the action-based and visualization-focused perfection stage practices I just noted—and the “perfection stage of the perfection stage,” which he explains to be the reality that is indicated by such practices, just as it is. Buddhajñānapāda’s injunctions on non-action can thus perhaps be understood to refer to the “perfection stage of the perfection stage,” that is the final result of suchness itself, which is what is trained in by means of the perfection stage practices. The internal and mental “activity” of the perfection stage practices are directed precisely to realizing this state of the “perfection stage,” which is itself nondual nonconceptual wisdom, free from any sort of action whatsoever. This “second part of the second stage” is also how Vaidyapāda glosses the term “great perfection,” which is used in the final verses of the Tibetan translation of the Dvitīyakrama, a point to which we will turn in the next chapter. VI. An Innovative Turn-of-the-9th-century Voice: Some Conclusions We have here surveyed a wide swath of doctrinal positions laid out in Buddhajñānapāda’s writings. But as I noted at the beginning of this chapter, this is just a first scratch along the surface of his his oeuvre. I have endeavored to show just some of the positions evidenced in Buddhajñānapāda’s works—those that stood out as particularly interesting or worthy of mention—and where possible to show how those ideas grew out of the roots of earlier traditions, and then became the roots for the further growth of Indian and Tibetan Buddhist tantric doctrine and practice. In Buddhajñānapāda we see a figure who wove the idea of nondual wisdom into a thoroughly tantric fabric, but also showed how a practitioner could, through the practice of that tantric path—and particularly through the sexual yogas of the perfection stage— come to perceive the truth that the thread of nondual wisdom has made up the basic fabric of all of appearance and existence from the very beginning. While framing his discourse in a clearly Buddhist context, he did not hesitate to draw upon non-Buddhist sources and ideas to communicate to his disciples and readers. Buddhajñānapāda adopted and adapted Śākyamuni’s own awakening narrative to emphasize the importance—indeed the necessity—of the higher tantric path as the means to awakening, and did not shy away from declaring that path to be the 309 J. Dalton 2004. 155 only way to attain the final result of full awakening. His works emphasize the importance of the oral transmission from a guru of the very nature of reality itself, allowing the disciple to experience her own nature directly, so that she knows experientially what she must train in as she completes the path. Buddhajñānapāda also indicated that once the practitioner thoroughly knows that nature, the effortful practices that have been instrumental in bringing her to that knowledge can then fall away. This emphasis on non-action is found specifically in the writings that focus on the perfection stage, in reference to which, as we saw above, the term “great perfection” (rdzogs pa chen po) appears twice in the Tibetan translations of Buddhajñānapāda’s works. Given the presence of this term in his writings, and the many references to its usage in his oeuvre in the work of later Tibetan Great Perfection authors, the relationship of Buddhajñānapāda’s writings and the early/proto- Great Perfection literature that was beginning to emerge in Tibet precisely around his time is a topic worth examining. It is to this inquiry that we will now briefly turn. 156 Chapter Four The Perfection Stage of the Perfection Stage: Buddhajñānapāda and the Great Perfection? Having come to fully understand this, the universal form of the wisdom of the great perfection, the perfectly pure body, Great Vajradhara, the essence of all the great glorious ones, is accomplished through this second stage. -Buddhajñānapāda, Dvitīyakrama Scholars familiar with early Great Perfection literature will have noticed in the previous chapter’s discussions that each one of Buddhajñānapāda’s positions that I have examined has at least some parallels in the early (or proto-) Great Perfection literature that began to surface in Tibet in the 8th and 9th centuries. It is therefore not just the use of the term rdzogs chen in Buddhajñānapāda’s writings—a point which has received note in both traditional and modern scholarship on the Great Perfection—that is deserving of our attention, but rather the ways in which the positions evidenced in his writings parallel the doctrinal positions found in early Great Perfection works. In this chapter I will briefly examine the views highlighted in Chapter Three in Buddhajñānapāda’s works and show similarities or parallels between these and some early Great Perfection writings, including the Guhyagarbha-tantra, Padmasambhava’s Garland of Views: A Pith Instruction (Man ngag lta ba’i ‘phreng ba), and several short works from Dunhuang. However, two points need to be mentioned before we get into this material. First, what follows is a very preliminary look into this topic, meant as a sort of gateway into the material and an invitation for further inquiry; and second, with regards to the early Great Perfection works whose positions I examine below, while I have familiarized myself with these writings, I am here relying primarily on the work of scholars who specialize in that literature since my own main focus in this study has been on the content of Buddhajñānapāda’s oeuvre. As I noted above, the mere use of the term rdzogs chen in Buddhajñānapāda’s works—the only point of relationship between his writings and the Great Perfection literature that has been heretofore examined in modern scholarship—is only part of what I would like to take a look at here, but the instances where this term is found in the Tibetan translations of Buddhajñānapāda writings are a good place to start. The Term “rdzogs chen” in Buddhajñānapāda’s Writings The term rdzogs chen, “great perfection,” is found in the Tibetan translations of two of Buddhajñānapāda’s surviving works, the Caturaṅga/Samantabhadra-sādhana and the Dvitīyakrama, though only the appearance in the Dvitīyakrama seems to have been noted in modern scholarship. As I mentioned in Chapter One, unfortunately the Sanskrit original for neither of these passages survives. However, we have two extant Tibetan translations of the Caturaṅga/Samantabhadra-sādhana (Tōh. 1855 and Tōh. 1856, respectively)—one translation under each of the two titles by which the text circulated—and one Tibetan translation of the Dvitīyakrama. I will first examine the term rdzogs chen as it is found in the Caturaṅga/Samantabhadra-sādhana, since we have more sources here—multiple translations of the sādhana, as well as multiple commentaries—and what we glean from this investigation will help us to better contextualize the term as it is found in the Dvitīyakrama, where we have fewer sources to work with. The transmission history of the sādhana is complicated (I will explore it further in Chapter Five), but for now suffice it to say that I believe the two recensions of the text 157 that circulated separately (for reasons that remain elusive) under these different titles did include some very minor variations, but otherwise constitute a single work. With regards to the passage in question, only one of the two translations—that of the Caturaṅga, translated by Smṛtijñānakīrti (an Indian paṇḍita who knew Tibetan well enough that he worked alone, without the assistance of a Tibetan collaborator), uses the term rdzogs chen, where it forms part of the phrase rdzogs chen ye shes, “the wisdom of the great perfection.” At this same place, the translation of the Samantabhadra-sādhana by Śraddhākaravarman and Rinchen Zangpo instead reads bsam yas ye shes, “inconceivable wisdom.” The verse is as follows: Having realized in this way, the one who is filled with compassion Instantly accomplishes the wisdom of the great perfection1/inconceivable wisdom,2 Glorious Samantabhadra, the innate state.3 Who, then, would not mediate upon this? |162|4 Among the four surviving commentaries on the sādhana that are available for study,5 two of them—by Thagana (Tōh. 1868) and Śrīphalavajra (Tōh. 1867)—comment on the Samantabhadra, and two—by Vaidyapāda (Tōh. 1872) and Samantabhadra (Tōh. 1869; yes, the commentator’s name is Samantabhadra!)—comment on the Caturaṅga. The commentaries generally follow the minor variants corresponding with the particular version of the sādhana (i.e. the Caturaṅga or the Samantabhadra) that they are commenting on. Therefore, paying particular attention to the terminology used in Vaidyapāda’s and Samantabhadra’s commentaries—since they both comment on the Caturaṅga—should give us a clue into whether the use of the term rdzogs chen in Smṛtijñānakīrti’s translation of the Caturaṅga-sādhana is an instance of an actual variant between the Caturaṅga and the Samantabhadra, or simply a difference in translation choice between Smṛtijñānakīrti and the team of Śraddhākaravarma and Rinchen Zangpo. With regard to the passage in question, the translation of only one of the four commentaries— Vaidyapāda’s Samantabhadrī-ṭīkā (yes, his commentary on the Caturaṇga-sādhana is titled Samantabhadrī!)—uses the word rdzogs chen ye shes in citing the root verse, which, not incidentally, Vaidyapāda glosses as dpag tu med pa’i ye shes, “limitless wisdom,” already coming closer to the “inconceivable wisdom” in Rinchen Zangpo and Śraddhākaravarman’s translation of the Samantabhadra-sādhana.6 Two of the other three commentaries, those by Samantabhadra and Thagana, both read bsam gyis mi khyab pa, “inconceivable” (though only in the case of Thagana’s work is this term clearly modifying “wisdom”; Samantabhadra glosses bsam gyis mi khyab pa alone). Śrīphalavajra’s commentary here reads rnam par mi rtog pa’i mtshan nyid can kyi ye shes, “wisdom characterized by nonconceptuality,” though it is unclear whether “nonconceptual” here is a gloss on a term from the root text or simply the author’s addition.7 1 rdzogs chen ye shes (reading from the translation of the Caturaṅga-sādhana). 2 bsam yas ye shes (reading from the translation of the Samantabhadra-sādhana). 3 gzhi yi dgnos po (reading from the translation of the Caturaṅga-sādhana); rang gi ngo bo (reading from the translation of the Caturaṅga-sādhana). I believe this minor difference to be simply due to variant translation choices. 4 de ltar shes nas snying rje ldan pas rdzogs chen ye shes rab tu ‘grub pa’i sems// dpal ldan kun tu bzang po gzhi yi dngos po gang gis bsgom par mi bya’o// (Caturaṅga-sādhana, D 42b.2-3; P 50b.8-51a.1). de ltar rtogs na snying rjer ldan pa dag// ye shes bsam yas skad cig gis sgrub pa// dpal ldan kun tu bzang po’i rang gi ni// ngo bo su zhig nges par sgom mi byed// (Samantabhadra-sādhana, D 36a. 1-2; P 42a.8) . 5 One additional Sanskrit commentary survives but is not available for study (see C. Dalton and Szántó, forthcoming). 6 Samantabhadrī-ṭīkā, D 177b.7. 7 Samantabhadra’s Sāramañjarī, D 45a.4, translated by Nyayanaśrī and Loden Sherab; Thagana’s Śrīsamantabhadrasādhana-vṛtti, D 230b.3, translated by Śraddhākaravarman and Rinchen Zangpo; Śrīphalavajra’s Samantabhadrasādhana-vṛtti, D 186b.1, translated by Vīryabhadra, Vibhākara(?), and Rinchen Zangpo. The 158 Due to the fact that we are in all cases working with texts in translation, it is difficult to say anything conclusive on the matter, but given the prevalence of the term bsam gyis mi khyab pa, “inconceivable,” in the commentaries as well as its use in Śraddhākaravarman and Rinchen Zangpo’s translation of the root text, it seems more likely that whatever the Sanskrit term in the sādhana was, it was something that was at the very least more easily translated into Tibetan as “inconceivable wisdom” than as the “wisdom of the great perfection.” However, the context of this passage, in which the innate state is equated with Samantabhadra, and also with wisdom, a translator working in context of 11th-century Tibet (which all of the translators of both the sādhana and its commentaries were), and particularly one such as Tsalana Yeshe Gyaltsen, who had some connection to the Nyingma traditions, might very well find it appropriate to translate the term qualifying wisdom here as “great perfection.” I cannot speculate on Smṛtijñānakīrti’s choice of the term rdzogs chen in his translation the sādhana. But, in the case of Kamalaguhya and Tsalana Yeshe Gyaltsen’s translation of Vaidyapāda’s commentary, Davidson has noted that Yeshe Gyaltsen’s translations were “influenced by Nyingma terminology and are therefore some of the few places where the term “great perfection” (rdzogs chen) is located in writings widely accepted as canonical.”8 In the end, it looks like the use of the term rdzogs chen in the translation of the Caturaṅga/Samantabhadra-sādhana and Vaidyapāda’s commentary may tell us more about the translators of those works than about Buddhajñānapāda’s use of a term that had the same semantic content as the Tibetan term rdzogs pa chen po. The Tibetan translation of the Dvitīyakrama also uses the term rdzogs pa chen po, though here we only have a single translation, and a single commentary—Vaidyapāda’s Sukusuma— which also survives only in its Tibetan translation. Both the root text of the Dvitīyakrama and Vaidyapāda’s commentary were translated by Kamalaguhya and Tsalana Yeshe Gyaltsen, and the term rdzogs pa chen po appears at this point in both. The verse in question reads: Having come to fully understand this, The universal form of the wisdom of the great perfection, The perfectly pure body, Great Vajradhara, The essence of all the great glorious ones, [is accomplished] through this second stage,9 |392|10 Given what we saw with respect to Kamalaguhya and Tsalana Yeshe Gyaltsen’s translation choice of term rdzogs chen in the Caturaṅga-sādhana, a bit of caution is warranted in our assessment of the passage here. Again we see a passage whose context—the universal form of wisdom, Great Vajradhara—is something that could clearly have been associated with the term rdzogs chen in 11th-century Tibet, especially with a translator who was familiar with Nyingma Sāramañjarī survives in part in multiple Sanskrit recensions, and the longer recension does include the commentary on this verse, but the whole line from the Tibetan recension of the Sāramañjarī that glosses the term bsam gyis mi khyab pa is absent in this recension of the work (see Szántó, unpublished edition, 150). Szántó, therefore, has not included any modifying adjective for jñāna in his Sanskrit reconstruction of the verse based on the Sāramañjarī. The presence of a line in the Tibetan translation that is absent in the Sanskrit recension of the Sāramañjarī is not, however, unusual. There are many discrepancies between the Tibetan translation and the multiple extant Sanskrit recensions of the work, which themselves are not uniform in length. The fact that all four commentaries extant in Tibetan translation give some kind of gloss here, though, suggests that there was likely an adjective used in the Sanskrit, whatever it may have been, and despite the fact that the Sanskrit Sāramañjarī does not comment on one. 8 Davidson 2005, 114. 9 Vaidyapāda clearly indicates that this is to be understood as the “second stage,” the perfection stage only, rather than the “two stages.” (Sukusuma, D 137b.7-138a.1). I have translated in accordance with his comments, somewhat (but not completely unfeasibly) against the grain of the Tibetan translation of the root text, which would be more easily translated as the “two stages.” 10 de ltar rab tu shes par byas nas su// rdzogs pa chen po ye shes spyi yi gzugs// yongs su dag sku rdo rje ‘chang chen po// dpal ldan kun gyi ngo bo rim gnyis ‘dis// |392| (Dvitīyakrama, verse 392). 159 traditions and terminology. It is also worth noting that the phrase in this verse, rdzogs pa chen po ye shes, “the wisdom of the great perfection” is the very same phrase in the Caturaṅgasādhana, which, as we saw, other translators chose to translate as bsam yas ye shes, “inconceivable wisdom.” But when we look at Vaidyapāda’s commentary on this passage in the Dvitīyakrama, it is clear that there is something different here. Whereas Vaidyapāda’s gloss on the term rdzogs chen ye shes in the Caturaṅga-sādhana was “immeasurable wisdom” (dpag med ye shes), here he glosses the term rdzogs pa chen po, “the great perfection” alone, as “the second stage of the second stage” (rim pa gnyis pa’i rim pa gnyis pa). This gloss appears to be drawing on Buddhajñānapāda’s division in the Muktitilaka of each of the two stages of tantric practice into two further sub-stages, where, as we saw before, he identifies the “perfection stage of the perfection stage,” with the seven yogas.11 Here in the Dvitīyakrama, then, given that Tsalana Yeshe Gyaltsen and Kamalaguhya translated both the root text and its commentary, it seems quite clear why they chose rdzogs pa chen po to translate whatever Sanskrit term they found in the Dvitīyakrama at this point, since Vaidyapāda glosses this term—again, whatever the Sanskrit may have been—as precisely what constituted the Tibetan term rdzogs chen in some of its earliest usages, as the culmination of the perfection stage (which I will address further below). What the Sanskrit term they translated actually was, and whether it held the semantic sense of “the great perfection” is still, however, in question. If, for example, the Sanskrit phrase were equivalent to that from the Samantabhadra/Caturaṅgasādhana— which is in fact a distinct possibility, since the Tibetan phrase in translation is indeed identical—what has been rendered here as “great perfection” may have been a term that could just as easily have been rendered “inconceivable.” In that hypothetical case, Vaidyapāda’s gloss on the term would still hold sense: “inconceivable refers to the second stage of the second stage.” Again, we are left without the possibility to make any definitive conclusions on terminological usage, since we are dealing entirely with works in translation. However, given what we saw regarding Yeshe Gyaltsen’s and Kamalaguhya’s translation choices in the Caturaṅga/Samantabhadra-sādhana, we have to acknowledge that it is certainly possible that a term with the semantic sense of “great perfection” was not found in Buddhajñānapāda’s work. But taking a look now at how the term rdzogs chen was used in other proto-/early great perfection from around the same period, I think we will be inclined to feel confident that the Tibetan term rdzogs pa chen po does indeed seem to correspond, in terms of content, to the way it has been used in the Tibetan translations of Buddhajñānapāda’s oeuvre. The term Rdzogs chen in Early Great Perfection Literature As a point of comparison with what we have seen in Buddhajñānapāda’s writings, let us look briefly at the use of the term rdzogs chen as it is found in some other works of Indic origin from around the same period. In these writings we see the term rdzogs chen used much in the same sense as whatever term(s) Smṛtijñānakīrti, Kamalaguhya, and Yeshe Gyaltsen translated as “great perfection” in Buddhajñānapāda’s works and Vaidyapāda’s commentaries—that is, as the culmination, or the outcome, of perfection stage practice. The closest parallels to the way the translators used the term rdzogs chen in Buddhajñānapāda’s works are found in the mid-8th century Guhyagarbha-tantra, a text that, while its origins have been the subject of some debate among traditional authors in Tibet as well as modern scholars, is ostensibly an Indian tantra, though it now survives only in its Tibetan translation. It is worth noting here, as well, that the only extant Indic commentary on the Guhyagarbha-tantra (again surviving only in Tibetan translation), the Spar khab, is attributed to Buddhajñānapāda’s guru Vilāsavajra. The term rdzogs 11 Muktitilaka, D 52a. 160 chen in the Guhyagarbha-tantra constitutes the earliest known appearance of the term in a way that resembles its usage in the later Great Perfection literature.12 It appears in the tantra just four times, and seems to indicate the ritual moment associated with the sexual climax of the perfection stage yogas immediately following the ritual consumption of the bodhicitta produced through those yogas, as well as in a more general sense to indicate the realization that is attained through the perfection stage.13 However, it is important to note that while the Guhyagarbhatantra’s references to the “great perfection” do indeed seem to indicate that it is meant as the culmination of the practices that characterize the perfection stage, neither the terms “generation stage” (bskyed rim) nor “perfection stage” (rdzogs rim) appear in the tantra itself. Indeed, the earliest usage of those specific terms (similar terms are used in Padmasambhava’s mid-to-late 8th-century Garland of Views, discussed below) of which I am aware is in Buddhajñānapāda’s Muktitilaka.14 Sam van Schaik has noted that the signification of the term rdzogs chen in the Guhyagarbha-tantra seems to differ little from later Great Perfection traditions: all qualities (yon tan) and enlightened activities (‘phrin las)—that is, the aims of the Buddhist practitioner—are complete (rdzogs) from the start (ye nas). That is to say, in another phrase that is used in the tantra far more often, everything is spontaneously present (lhun gyis grub). Furthermore, there is an emphasis on the transcendence of concepts in a state beyond the reach of thought (bsam gyis mi khyab).15 When looking at the ways the term rdzogs chen is used in the Guhyagarbha-tantra alongside the ways it is found in the Tibetan translations of Buddhajñānapāda’s writings and Vaidyapāda’s commentaries, we find of course the parallel in the use of the term rdzogs chen to indicate the culmination of the perfection stage. But we also find strong resonance with the identification in Buddhajñānapāda’s Muktitilaka of the “perfection stage of the perfection stage” with the seven yogas, which, as we saw above, are both seven aspects of the final result of awakening as well as being closely connected with perfection stage practices, through which they are meant to be experienced and realized. Moreover, the idea found in the Guhyagarbha-tantra of the result— “the aims of the Buddhist practitioner,” as van Schaik puts it—being already and spontaneously present from the beginning is echoed in Buddhajñānapāda’s identification of nondual wisdom with the very nature of the entirety of the phenomenal world, and of the mind itself. Finally, the Guhyagarbha-tantra is a Mahāyoga tantra, and the presence of what came to be called the “great perfection” in close association with a Mahāyoga tantric framework is also paralleled in Buddhajñānapāda’s writings. These commonalities include aspects connected with generation stage practices such as the basic view of the aggregates and elements as the male and female buddhas, as well as those connected with perfection stage practices, such as the use of the fourfold framework of sevā and the rest (the caturaṅga) which are applied to a perfection stage context in both the Guhyagarbha-tantra and the Dvitīyakrama. The fact that the Great Perfection tradition developed, at least in part, within Mahāyoga tantric context has been noted by a number of scholars, including Karmay, Germano, Jacob Dalton, and van Schaik.16 12 van Schaik 2004, 167. 13 Dalton 2004. 14 The terms generation stage (utpattikrama) and perfection stage (utpannakrama) are well known to appear in the Samājottara, which Isaacson (2002a, 468-9) identifies as their scriptural locus classicus. Nonetheless, the very verse in which the terms appear in the Samājottara is found in parallel in the Muktitilaka, and in an earlier paper (C. Dalton 2014) I have argued that Buddhajñānapāda’s verse is the earlier of the two. I examine this point more fully in Chapter Eight. 15 van Schaik 2004, 168. 16 Karmay 1988, especially Chapter 6; Germano 1988, 212-226; Dalton 2004, 8-21; van Schaik 2004, 167; Dalton 2016, 34-47. 161 Another work by an Indian author appearing around the time of Buddhajñānapāda’s writings that uses the term rdzogs chen is Padmasambhava’s mid-to-late-8th-century Garland Of Views: A Pith Instruction, a work on the thirteenth chapter of the Guhyagarbha-tantra.17 In this treatise rdzogs chen appears as part of the term “the mode of the great perfection” (rdzogs pa chen po’i tshul) following the “generation mode” (skyed pa’i tshul) and “perfection mode” (rdzogs pa’i tshul) as the third of the threefold “modes” of the vehicle of inner yoga.18 In fact, this classification occurs at the culmination of a doxography that is somewhat reminiscent of the Dvitīyakrama’s doxographical section discussed above.19 In the Garland of Views the “great perfection” is clearly connected with the generation and perfection stage practices of deity and sexual yogas, and also in some sense constitutes the culmination of the perfection stage, as the “mode” that supersedes it, the very highest perspective in Padmasambhava’s doxography. However, given that each of these practices, including generation and perfection, are referred to and discussed as “modes” (tshul) of practice, rather than the usual “stages” (rim pa), the way the term rdzogs chen is used in the Garland of Views may constitute a slight shift away from the more limited use that we saw in the Guhyagarbha-tantra itself—and in the Tibetan translations of Buddhajñānapāda’s works—as a ritual moment or a result that is the specific outcome of perfection stage practices. As van Schaik interprets its usage, rdzogs chen in the Garland of Views “primarily functions as an interpretive framework” for the “visualization of deities and the experience of bliss through sexual union.”20 And indeed, the term “mode” of practice does indeed suggest that rdzogs chen may serve as a framework of sorts, but the way the “mode of the great perfection” is defined in the Garland of Views, while it may perhaps shift slightly away from the other usages we have discussed, remains very much rooted in them. The Garland of Views defines the “mode of the great perfection” as “meditation [based in] having realized all mundane and supramundane phenomena as indivisibly and primordially possessing the nature of the maṇḍala of awakened body, speech, and mind.”21 As such, it still functions, as in the case of the Guhyagarbha-tantra and the usages of the term in translations of Buddhajñānapāda’s works, as a meditation on, or training in, the true state of things, the result that the yogin has come to realize. The method by means of which such realization is attained is, in the case of both Buddhajñānapāda’s works and in the Guhyagarbha-tantra, clearly indicated to be the perfection stage yogas. In the Garland of Views the presence of the generation and perfection “modes” as preliminary to the “great perfection mode” within the vehicle of “inner yoga” suggests that they are likely understood to form the basis for the perspective of the great perfection in Padmasambhava’s view, as well, thus indicating that the great perfection as portrayed in his Garland of Views is not just a framework for interpreting the earlier practices, but as a cultivation of the experience of the ultimate or resultant state, which is the outcome of those earlier practices, as well.22 Indeed, Karmay has noted that the term rdzogs chen in the Garland of 17 This is one of the few works traditionally attributed to Padmasambhava whose attribution is also generally accepted by modern scholars. 18 Man ngag lta ba’i ‘phreng ba, 4a.4. 19 Man ngag lta ba’i ‘phreng ba, 4b.2. I address this point further below. 20 van Schaik 2004, 171. 21 ‘jig rten dang ‘jig rten las ‘das pa’i chos thams cad dbyer med par sku gsung thugs kyi kdkyil ‘khor gyi rang bzhin ye nas yin par rtogs nas sgom pa. (Man ngag lta ba’i ‘phreng ba, 4b.2). 22 In terms of terminology, the Garland of Views is a particularly interesting case, since—assuming that the attribution is correct—it is the work of an Indian author that seems to have been composed directly in Tibetan; that is, there is no translator’s colophon to indicate that the work is a translation. This, then, may be the earliest known use of the term rdzogs chen in a work that was composed in the Tibetan language, rather than translated from Sanskrit. The method of this text’s composition is interesting to consider given that it is a Tibetan language composition by an Indian author. Considering Davidson’s (2005) model of “grey texts,” a category somewhere between original compositions and translations and which arose at the nexus of Indian authors and their Tibetan 162 Views, “appears only as an extension of the term rdzogs rim,”23 and van Schaik also holds that for Padmasambhava in this work rdzogs chen is “the culmination of the development and perfection stages.”24 What is more, the perspective on ultimate reality, the realization of which constitutes the great perfection in the Garland of Views—that all mundane and supramundane things have always had the nature of the maṇḍala of awakened body, speech, and mind—is very reminiscent of Buddhajñānapāda’s assertion that all phenomena have, from the very beginning, had nondual wisdom as their nature. Buddhajñānapāda’s writings also include a similar use of the term maṇḍala to describe the innate state. In a striking passage from the Muktitilaka Buddhajñānapāda writes Since everything is beyond conception So-called “saṃsāra” does not exist. Just like a rope “snake,” it appears But is held not to exist, Because when false perception is abandoned, Apart from the rope, no trace of anything else can be found. Likewise, when conceptual elaboration is abandoned, Within the maṇḍala no trace of saṃsāra can be found. Thus due to [experiencing] the profound and luminous maṇḍala I25 remain in unceasing nirvāṇa. Because ordinary beings don’t know this truth They are tormented by the suffering of existence.26 Setting aside for the moment Buddhajñānapāda’s explicit declaration of his own personal state of awakening (!) in this passage, his assertion that saṃsāra is merely a false perception while the true nature of all things is “profound and luminous maṇḍala” is quite similar to the perspectives on the ultimate state as constituting the maṇḍala of awakened body, speech, and mind found in the Garland of Views. In addition to sharing a similar perspective on the state of reality that is the culmination of the perfection stage, and which the Garland of Views clearly terms rdzogs pa chen po, there are a number of other commonalities between the Garland of Views and Buddhajñānapāda’s writings. The doxography in the Garland of Views, as mentioned briefly above, is quite reminiscent of the one in the Dvitīyakrama, and not only in terms of the simple fact that both texts contain such a doxographical presentation. Jacob Dalton has observed that most doxographical classifications of tantra from Indian sources were ritual-based rather than translators, it is possible that the Garland of Views’ content might have been dictated by its author (in an Indian language or in Tibetan) and then recorded in Tibetan by a Tibetan disciple. 23 Karmay 2008, 138. 24 van Schaik 2004, 179. J. Dalton (2016, 41-2) suggests that in the early great perfection literature the term rdzogs chen was used in quite a variety of different ways, including immersion in a nonconceptual state, as the interpretive “framework” for deity yoga, the nonconceptual state that arose from the ingesting of sexual fluids, and likely others, as well. 25 The Dpe sdur ma edition (962) here reads da, but the Derge (47b.1), Peking (56b.8) and Vaidyapāda (Muktitilakavyākhyāna, D 49a.7) clearly read nga. Vaidyapāda’s commentary makes the point even more clear, adding “I and others....” (bdag srogs) (Muktitilaka-vyākhyāna, D 49a.7). 26 thams cad rnam rtog dang bral bas// ‘khor ba zhes bya yod ma yin// ‘on kyang thag pa’i sprul lta bur// snang zhing med pa nyid du ‘dod// rab rib spangs pas thag pa las// gzhan ni cung zad rnyed mi ‘gyur// de bzhin spros spangs dkyil ‘khor la// ‘khor ba cung zad rnyed mi ‘gyur// des na zab gsal dkyil ‘khor bas// nga ni rtag tu mya ngan ‘das// de don skyes bos ma shes bas// srid pa’i sdug bsngal gyis gzir ro// (Muktitilaka, D 47a.7-47b.1; P 57b.5-58a.- 1) 163 philosophically-based.27 The Dvitīyakrama’s doxography, which begins with non-Buddhist views and then classifies Buddhist views in a hierarchical mode, thus stands out in being not ritually- but philosophically-based, while also including the perfection stage tantric perspective as the highest among the “views” surveyed. The Dvitīyakrama’s doxography does not, however, distinguish between the various classes of tantra. The Garland of Views’ doxography holds an interesting middle ground: in the section on exoteric traditions (both non-Buddhist and Buddhist, just as in the Dvitīyakrama) it focuses on view, but then the text transitions to an emphasis on ritual distinctions when distinguishing among the various classes of tantra. Their philosophical perspectives (“views”) still receive mention in Padmasambhava’s discussion of the different tantric systems, but the views he sets forth here are largely consistent among the classes of tantra, while their distinctions are expressed in ritual terms. The single exception here may be the “mode of the great perfection” where Padmasambhava identifies this mode as “meditation”— which we may take as a form of practice that could thus be identified as a “ritual”—but then describes that meditation not in terms of its ritual details (as he has done with the other tantric classes), but rather in terms of its view: the “mode of the great perfection is” “to meditate [on the basis of] having realized all mundane and supramundane phenomena as indivisibly and primordially possessing the nature of the maṇḍala of awakened body, speech, and mind.”28 Yet another point of commonality between Buddhajñānapāda’s writings and the Garland of Views is their reinterpretation of the four “branches” of sevā, upasevā, sādhana, and mahāsādhana. As we saw above, these four terms seem to have first been applied to the various practices of the generation stage, as in the Guhyasamāja-tantra (and indeed they continue to be so applied, even in the context of modern-day practice), but Buddhajñānapāda reinterprets them in the Dvitīyakrama to correspond to the stages of sexual union associated with the perfection stage practice. A similar, but more general, association of this schema with perfection stage practice is found in the Guhyagarbha-tantra.29 In the Garland of Views these same four categories are repurposed to span practices fitting into both the generation and perfection stages, beginning with the generation of oneself as the deity, up to the blissful moment at the culmination of the perfection stage. Though their way of interpreting the four branches is not identical, Buddhajñānapāda and Padmasambhava are clearly working within a common milieu, with similar material and terminology. Additionally, the Garland of Views is similar to Buddhajñānapāda’s works in making it clear that the highest path of tantra—in the Garland’s presentation, this is the path of the great perfection—is not suitable for everyone, and that it is to be kept secret because it is beyond the comprehension of ordinary individuals.30 One final curious connection between the Dvitīyakrama and the Garland of Views is a citation in Padmasambhava’s text that is identified in the commentarial literature as a quotation from the Guhyasamāja-tantra. Karmay notes that the passage in question is, however, not found in the Guhyasamāja-tantra, which is true, though I have identified a passage that contains a slightly extended form of basically same content in the Guhyasamāja-tantra, at the end of its fifteenth chapter.31 However, this particular passage cited in the Garland of Views is nearly identical to a 27 Dalton 2005, 119-120. 28 ‘jig rten dang ‘jig rten las ‘das pa’i chos thams cad dbyer med par sku gsung thugs kyi kdkyil ‘khor gyi rang bzhin ye nas yin par rtogs nas sgom pa (Man ngag lta ba’i ‘phreng ba, 4b.2). 29 Guhyagarbha-tantra, D 122a.5-6. 30 Man ngag lta ba’i ‘phreng ba, 8b.3-9a.1. 31 Karmay 2007, 158. The passage from Matsunaga’s edition reads: atha te sarvatathāgatāḥ sarvatathāgatakāyavākcittavajraṃ tathāgatam evam āhuḥ/ sarvatathāgatadharmā bhagavan kutra sthitāḥ kva vaā sambhūtāḥ/ vajrasattva āha/ svakāyavākcittasaṃsthitāḥ svakāyavākcittasambhūtāḥ/ bhagavantaḥ sarvatathāgatā āhuḥ/ svakāyavākcittavajraṃ kutra stitham/ ākāśasthitam/ ākāśaṃ kutra stitham/ na kvacit/ (Matsunaga, 85). Fremantle’s edition and translation give a version of the passage from the Guhyasamāja-tantra that is even closer to 164 passage from the Dvitīyakrama (which is not identified as a quotation).32 The passage as cited in the Garland reads: Phenomena abide in the mind; The mind abides in space; And space abides nowhere.33 The parallel passage from the Dvitīyakrama reads: ...Why is that? Because all phenomena Abide in the mind. This, as well, Abides in space. Space itself Abides nowhere; it is luminous.34 Whether both passages were separately adapted from the longer passage in the Guhyasamājatantra or are a quotation from some other shared source (perhaps the more likely scenario, given that Padmasambhava, at least, identifies the passage as a quotation) the presence of this parallel passage is further evidence that the two texts do very much seem to be coming out of a similar milieu, both in terms of the ideas found in the works, as well as their literary sources. Having noted these similarities between Buddhajñānapāda’s writings and some early Indic sources for the Great Perfection literature, let us now take a more topical approach, and look at instances in which some of the doctrinal features I have identified in Buddhajñānapāda’s writings appear in some early Great Perfection works. Nondual Wisdom as the Nature of Mind and all Phenomena As we saw in Chapter Three, Buddhajñānapāda clearly asserts at many points in his writings that nondual wisdom is not only the resultant state of awakening, but is itself the very identity of the mind and indeed the entire phenomenal world. Such an identification of the result of awakening with the nature of the mind and the phenomenal world is a common characteristic of many Great Perfection works, starting from a very early period. Van Schaik has noted that “all forms of the Great Perfection place a great emphasis on nonduality and assert often that the enlightened state is immanent in the everyday state.”35 The Guhyagarbha-tantra states that, “The mind itself is the perfect Buddha; Do not search for the Buddha anywhere else,”36 and a Mahāyoga treatise by the 9th-century Indic author *Madhusādhu that has been preserved at Dunhuang, and which shows strong affinities with early Great Perfection literature, likewise the quotation from the Dvitīyakrama and Garland of Views. Her edition reads: atha te sarvatathāgatāḥ sarvatathāgatakāyavākcittavajram evam āhuḥ/ sarvatathāgatadharmā bhagavan kutra sthitāḥ kva vā saṃbhūtāḥ/ vajrasattva āha/ svakāyavākcittasaṃsthitāḥ svakāyavakcittasaṃbhūtāḥ/ bhagavantaḥ sarvatathāgatā āhuḥ/ cittaṃ kutra sthitam/ ākāśasthitam/ ākāśaṃ kutra sthitam/ na kvacit/ (Fremantle 1970, 348; see also Fremantle 1970, 349 for the Tibetan edition of the passage. Fremantle’s English translation of the passage reads: “Then all the Tathāgatas said to the Tathāgata, Vajra Body, Speech and Mind of all Tathāgatas: O Blessed One, where do the dharmas of all the Tathāgatas exist and where do they come from? Vajrasattva said: they exist in your body, speech, and mind, and they come from your body, speech and mind. The Blessed Tathāgatas said: where does mind exist? He answered: it exists in space. They asked: where does space exist? He answered: nowhere.” (Fremantle 1970, 110). 32 As we have seen, Buddhajñānapāda seems to have freely adapted passages from other texts into his writings without the need to identify them as quotations, as in the passage parallel to the Sarvabuddhasamāyoga-tantra in verse 50 of the Dvitīyakrama (see my translation and notes for further details), and the numerous passages from the Nāmamantrārthāvalokinī adapted into his Ātmasādhanāvatāra that I addressed above (See also Appendix I). 33 chos rnams thams cad ni sems la gnas so// sems ni nam mkha’ la gnas so// nam mkha’ ni ci la yang mi gnas so// (Man ngag lta ba’i ‘phreng ba, 5b.2-3). 34 ci yi phyir na chos thams cad// |32| sems la gnas te de nyid kyang// nam mkha’ la gnas nam mkha’ ni// gang du min gnas ‘od gsal ba// (Dvitīyakrama, verse 32d-33c). 35 van Schaik 2004b, 51. 36 van Schaik 2008, 14. 165 echoes this idea: “It is sufficient to realize mind’s reality. It is not necessary to seek buddhahood anywhere other than in the mind.”37 In the lines for taking refuge at the beginning of his Caturaṅga/Samantabhadra-sādhana Buddhajñānapāda expresses this same identification quite clearly: “I constantly take refuge in the sugatas/ Who abide within my mind.”38 The Guhyagarbha-tantra, moreover, identifies the phenomenal world with the awakened state noting that, “all the tathāgatas and all phenomena are characterized by having the single essential identity of being primordially awakened.”39 As we will recall, Buddhajñānapāda shares a similar perspective in the Dvitīyakrama. He writes, The nature of phenomena, From form and the rest up to omniscience, Is the perfectly pure wisdom of the nonduality of the profound and the luminous, Which is like the center of space. |23| 40 In identifying phenomena with nondual wisdom, he essentially identifies them with the awakened state, though this idea is not made quite as explicit in Buddhajñānapāda’s works as it is in some of the early Great Perfection writings. The World Arises out of Nondual Wisdom: A Gnostic Cosmogony The cosmogonic narratives that identify nondual wisdom as the source of the phenomenal world explored above in Buddhajñānapāda’s writings also share parallels in some early Great Perfection works. As we will recall, Buddhajñānapāda wrote of the emergence of saṃsāra from the reality of nondual wisdom in the Dvitīyakrama: The reality which is like that Is present pervading all things. Yet,41 from beginningless time, even from this There was arising in the manner of the great thought. |35|42 And from that also the great elements [arose]: The maṇḍala of wind arose, And from that also, the great element of fire Arose and spread. |36| From that, the great element of water also Arose and spread, and from that also earth. From the essence of the gathering of the four [elements] Mountains, and so forth, and all sentient beings also [arose] |37| In all their variety: subtle and gross: Men, women, and hermaphrodites, The young and old, 37 ibid. 38 rang gi yid la gnas par gyur pa yi// bde gshegs rnams la rtag tu skyabs su chi// (Samantabhadra-sādhana, D 29a.6). 39 de bzhin gshegs pa thams cad dang/ chos thams cad ye nas sangs rgyas pa’i ngo bo nyid du gcig pa’i mtshan nyid yin pas... Guhyagarbha-tantra, D 112b.2. 40 chos rnams gzugs la sogs pa rnams// kun mkhyen bar gyi rang bzhin ni// nam mkha’ dkyil ltar rnam dag pa’i// zab gsal gnyis med ye shes te// |23| (Dvitīyakrama, verse 23). 41 Vaidyapāda explains further, “Although that kind of nonduality pervades and remains [as the nature of] all things, the reason that this is not apparent is explained with the lines beginning, “Yet, from beginningless time...” de yang gnyis su med pa de lta bus dngos po kun rnam par khyab ste gnas kyang de mi gsal ba’i rgyu ni thog med dus nas zhes te/ (Sukusuma, D 97a. 4). 42 de ‘dra’i don des dngos po kun// rnam par khyab ste rnam gnas kyang// thog med dus nas de las kyang// rnam rtog chen po tshul byung ste// |35| (Dvitīyakrama, verse 35). 166 Gods, nāgas, and yakṣas, |38| Evil spirits, planets, Yāma, The Lord of Water, Indra, hell beings, Pretas, animals, and those who abandon all of this,43 Beings who rely upon consciousness alone,44 Such beings remain, spread far and wide.45 |39| Therefore, the nondual nonconceptuality That is higher than that is completely obscured. Because of not realizing it, all beings Cycle around in saṃsāra. |40|46 The Guhyagarbha-tantra’s account of the emergence of saṃsāra is similar in identifying its source as “the essence of the sugatas” and the manifestation of confusion as “concepts,” from which all the various forms of saṃsāric existence unfold. Both works also identify conceptualization as that which obscures beings’ experience of their original nature or essence. The second chapter of the Guhyagarbha-tantra states: Amazing! [Even] from within the essence of the sugatas, [confusion occurs]. Emanated by an individual’s concepts and karma Diverse bodies and enjoyments, Places, suffering, and so forth, The self, [and things that are] mine are all separately perceived. No captor has bound us, bondage is nonexistent, Nor is there anything to be bound. Conceptualizing and grasping to a self We insist upon untying knots in space, [But] there is no bondage nor release.47 Moreover, the cosmogonic emergence of phenomena from nondual wisdom that is articulated in Buddhajñānapāda’s works can be seen as closely related to the idea of a single basis that manifests either as saṃsāra or nirvāṇa depending on the presence or absence of realization. This idea, which became very important in later Great Perfection works, is found in 43 Vaidyapāda identifies these as the śrāvakas and so forth. de kun spangs pa ni nyan thos la sogs pa’o// (Sukusuma, D 97b.1). 44 Vaidyapāda identifies these as those beings of the realm of Limitless Space, and so forth—inhabitants of the Formless Realm—since they have abandoned form. nam mkha’ tha’ yas la sogs pa ste/ gzugs spangs ba’i phyir ro// (Sukusuma, D 97b.1-2). 45 Vaidyapāda comments that the statement that these beings live far and wide means that, “having been produced by conceptuality, they appear in the ten directions.” de kun rgyas par gnas zhes pa ni rtog pas bzo byas nas phyogs bcu kun du snang ba’i phyir ro// (Sukusuma, D 97b.2). 46 de ‘dra’i don des dngos po kun// rnam par khyab ste rnam gnas kyang// thog med dus nas de las kyang// rnam rtog chen po tshul byung ste// |35| de las yang ni ‘byung ba che// rlung gi dkyil ‘khor nyid byung ste// de las kyang ni me yi khams// chen po byung nas khyab mdzad de// |36| de las chu khams chen po yang// byung ste khyab mdzad de las kyang// sa byung bzhi bsdus ngo bo las// ri sogs sems can thams cad kyang// |37| sna tshogs phra ba sbom po dang//skyes pa bud med ma ning dang// gzhon nu dang ni rgan po dang// lha dang klu dang gnod sbyin dang// |38| gdon dang skar ma gshin rje dang// chu bdag rgya byin dmyal ba dang// yi dags dud ‘gro dang de kun// spang dang shes tsam rab brten pa’i// ‘gro ba kun du rgyas par gnas// |39| de bas de yi gong ma yi// gnyis med rtog bral rab bsgribs te// ma rtogs pas na ‘gro ba kun// ‘khor bar rab tu ‘khor bar ‘gyur// |40| (Dvitīyakrama, verses 35-40). 47 e ma bde gshegs snying po las// rang gi rnam rtog las kyis sprul// sna tshogs lus dang longs spyod dang// gnas dang sdug bsngal la sogs pa// bdag dang bdag gir so sor ‘dzin// sus kyang ma bcings bcings med de// bcing bar bya ba yod ma yin// rnam rtog bdag tu ‘dzin pa yis// nan gyis mkha’ la mdud pa dor// bcings med rnam par grol med pa’i// (Guhyagarbha-tantra, D 112b.2-4). 167 the eleventh chapter of the Guhyagarbha-tantra,48 as well as in the 9th-century treatise of

• Madhusādhu preserved at Dunhuang, which reads:

Because the phenomena of nirvāṇa (mya ngan las ‘das pa) and saṃsāra (‘khor ba) manifest depending on whether there is realization or non-realization, they are nondual. Therefore they are called the single basis (gzhi gcig) or the single truth.49 While Buddhajñānapāda’s works do not use the terminology of a “single basis,” the same perspective is not only implied by the cosmogonic narratives in his writings in which nondual wisdom acts as the source for saṃsāric phenomena, which, when their nature is realized, are none other than nirvāṇic, but is found more overtly in statements such as this one, from the Muktitilaka, already cited above: Since everything is beyond conception So-called “saṃsāra” does not exist. Just as a rope “snake,” it appears But is held not to exist, Because when false perception is abandoned, Apart from the rope, no trace of anything else can be found. Likewise, when conceptual elaboration is abandoned, Within the maṇḍala no trace of saṃsāra can be found. Thus due to [experiencing] the profound and luminous maṇḍala I50 remain in unceasing nirvāṇa. Because ordinary beings don’t know this truth They are tormented by the suffering of existence.51 This passage clearly indicates that the only distinction between saṃsāra and nirvāṇa is the presence or absence of conceptual elaboration, which also is what makes the distinction as to whether or not an individual is experiencing the “profound and luminous maṇḍala.” To be sure, this idea of the distinction between saṃsāra and nirvāṇa coming down to just concepts is a much older one in Buddhism. In a verse from his Mañjughoṣastotra, cited by Buddhajñānapāda in the Ātmasādhanāvatāra (and which, incidentally, may have been the inspiration for the line in the Muktitilaka in which Buddhajñānapāda states that he himself “remain[s] in unceasing nirvāṇa”), Dīgnāga wrote: There is nothing called saṃsāra Apart from concepts Therefore when free from concepts You remain in unceasing nirvāṇa.52 48 Dorje n.d., 61. The term in the Guhyagarbha-tantra here reads rgyu gcig, “single cause” rather than gzhi gcig “single basis,” but the idea is the same (Guhyagarbha-tantra, D 122a.2). 49 van Schaik 2008, 17. 50 The Pedurma edition (962) here reads da, but the Derge (47b.1), Peking (56b.8) and Vaidyapāda (Muktitilakavyākhyāna, D 49a.7) clearly read nga. Vaidyapāda’s commentary makes the point even more clear, adding “I and others....” (bdag srogs) (D 49a.7). 51 thams cad rnam rtog dang bral bas// ‘khor ba zhes bya yod ma yin// ‘onkyang thag pa’i sprul lta bur// snang zhing med pa nyid du ‘dod// rab rib spangs pas thag pa las// gzhan ni cung zad rnyed mi ‘gyur// de bzhin spros spangs dkyil ‘khor la// ‘khor ba cung zad rnyed mi ‘gyur// des na zab gsal dkyil ‘khor bas// nga ni rtag tu mya nga ‘das// de don skyes bos ma shes bas// srid pa’i sdug bsngal gyis gzir ro// (Muktitilaka, D 47a.7-47b.1; P 57b.5- 58a.1) 52 The passage as it is cited in the Ātmasādhanāvatāra reads: rnam rtog las gzhan ‘khor ba zhes// bya ba ‘ga’ yang yod min te// de phyir rnam rtog bral gyur pas// rtag tu khyod ni mya ngan ‘das//. I am grateful to Péter Szántó who shared with me his draft Sanskrit edition of the Ātmasādhanāvatāra in which he identifies the source of the passage as Dignāga’s Mañjughoṣastotra. 168 But the connection of this idea with a cosmogony in which the reality of nondual wisdom or the “essence of the sugatas” acts as the source of the phenomenal world, that the awakened nature itself is the single basis of both saṃsāra and nirvāṇa does seem to be a unique expression of this doctrine that is common to Buddhajñānapāda’s writings and early Great Perfection works, and which finds much further expression in the later Great Perfection tradition.53 The Transmission of Nondual Wisdom: Pointing It Out We saw in Chapter Three how Buddhajñānapāda’s writings describe the direct transmission of the reality of nondual wisdom from a guru to his disciple, usually specified as taking place through verbal means, in the context of the third tantric initiation or immediately afterward—precisely what, in later traditions, came to be known as a “fourth initiation.” This fourth initiation became extremely important in later Great Perfection traditions, where it is also called the “precious word initiation” (tshig dbang rin po che). In these later Great Perfection traditions, as well as in the later tradition of Mahāmudrā, the technique of bestowing a transmission of the nature of reality directly to the student was also abstracted from the context and sequence of tantric initiation, and took the form of a symbolic conveyance of the nature of reality from guru to disciple called an “introduction” (ngo sprod). As David Germano has noted with reference to early Great Perfection meditation practices, “subsequent Great Perfection traditions indicate that such styles of meditation begin with a symbolic indication of the mind’s nature in an encounter with one’s teacher referred to as a “pointing to” or “introduction to” the mind’s nature (sems khrid; ngo sprod).”54 The function of this (usually) verbal conveyance of the nature of reality in these later Great Perfection traditions—as the initial method by means of which the disciple is brought to a direct experience of reality that serves as the basis for her later training in it—is indeed identical to the function of the transmission of suchness that we find in Buddhajñānapāda’s writings. But despite the emphasis in the later Great Perfection traditions on the importance of such a transmission or “introduction,” the earliest Great Perfection writings— at least those that have so far received attention from modern scholars—do not seem to discuss this procedure. This important feature of Buddhajñānapāda’s writings thus is in little evidence in early Great Perfection works. Given the importance of such a transmission in the later tradition, the lack of its mention in the early writings is notable. The lacuna might well be attributed to the fact that the early Great Perfection tradition was still so enmeshed in the tradition of Mahāyoga tantra that a “pointing out” instruction outside the structures of tantric initiation had not yet developed, and since the early works (at least those that have been studied by modern scholars) tend not to focus on tantric ritual structures, while they certainly do mention tantric initiation, they do not generally discuss its details. Given that, like we have seen in Buddhajñānapāda’s writings, this kind of “transmission” formed, from the late 8th/ early 9th century, part of the initiatory sequence, the lack of a discussion of the details of ritual in general, and initiation in particular, in early Great Perfection texts would mean that such a transmission might not have been mentioned in the texts, though it may have been taking place in practice. In any case, there appears to have been little research done on the origins and development of the practice of “introduction” or “pointing out” within Great Perfection traditions. Besides Germano’s brief reference, cited above, Yamamoto mentions the term’s occurrence in the works of the 12th- 53 This idea continued to be developed and expressed in the Great Perfection traditions. For example, the idea of a “genuine foundation of unification (sbyor ba don gyi kun gzhi)” that unifies all saṃsāra and nirvāṇa because it is “both the cause for all thoroughly afflicted phenomena to appear and the basis for their purification” is described in Nubchen Sangye Yeshe’s 10th century Armor Against Darkness (Mun pa’i go cha) (J. Dalton 2016, 42). For a 12thcentury Tibetan presentation of the idea of the single basis see Yeshi and J. Dalton 2018, esp. 263-272. 54 Germano 1994, 228. 169 century Tibetan master Lama Zhang in conjunction with the Mahāmudrā tradition, and Hatchell, in reference to 14th-century developments in the Great Perfection describes such an introduction as a “fundamental feature of the Great Perfection tradition.”55 But while these scholars define and briefly discuss the term, none discuss its provenance or its development. As far as I am aware, such a practice receives no mention at all in the modern scholarship on the early Great Perfection. What does receive brief mention in some early works of the Great Perfection, however, is the related issue of sudden versus gradual realization, and here we again find a point of connection with Buddhajñānapāda’s perspectives. As we saw, Buddhajñānapāda seems to have accepted aspects of both sudden and gradual realization; that is, the path of higher tantric practice as he describes it includes both sudden experiences of realization at the outset of the path when receiving the transmission of reality from the guru, as well as and during or at the conclusion of the path. But Buddhajñānapāda also acknowledges a gradual process of training in that reality as the disciple progresses along the path. Jacob Dalton writes of the works of Pelyang, that he likewise takes “a diplomatic position on the sudden-gradual debate...as he allows for gradual progress along a path that culminates in a moment of sudden enlightenment.”56 The 10th-century Great Perfection author Nubchen Sangye Yeshe is perhaps even more explicit on this point.57 This feature of a balanced approach to the issue of gradual versus sudden realization, then, seems to be another feature shared between Buddhajñānapāda’s thought and at least some writings of the early Great Perfection tradition. Privileging Tantric Practice: The Superiority of the Tantric Path and its Result Buddhajñānapāda’s position that the tantric path and result are superior to that of the nontantric path is an important aspect of his thought. As we saw above, this view is evidenced throughout his writings, both directly and indirectly. A similar position seems to be found in the treatise on the four yogas by Madhusādhu, which likewise asserts the superiority of the Vajrayāna, including the superiority of its result.58 However, Madhusādhu’s position does differ somewhat from Buddhajñānapāda’s in claiming that the Mahāyāna path can still lead to full buddhahood.59 While his view on the superiority of tantra may not have been precisely shared with early writings from the Great Perfection tradition, Buddhajñānapāda’s method for showing the superiority of tantric practice—homologizing tantric practices with non-tantric ones, or generation stage practices with perfection stage ones, and always privileging the “higher” practices—does appear in several early Great Perfection works. As we saw above, like Buddhajñānapāda’s writings, the Guhyagarbha-tantra applied the fourfold structure of sevā, sādhana, and the rest to perfection stage practices (though without actually mentioning the term “perfection stage”), including its culmination in the “great perfection,” and Padmasambhava’s Garland of Views likewise applied this framework to span practices from the generation stage up through the great perfection. Germano has noted that the reinterpretation of this four-fold schema continued up through the works of the 14th-century Tibetan scholar Longchenpa, who interpreted all four stages in terms of the practices of the great perfection.60 The Questions and Answers of Vajrasattva, a short treatise found at Dunhuang that lies at the intersection between Mahāyoga 55 Yamamoto 2009, 300; Hatchell 2009, 160. 56 J. Dalton 2012, 188. 57 J. Dalton 2016, 42. Thanks to Jacob Dalton for drawing my attention to Nubchen’s position on this issue. 58 van Schaik 2008, 13. 59 ibid., 27. Van Schaik does not elaborate on the apparent tension between Madhusadhu’s holding the Vajrayāna result to be superior and at the same time asserting that the bodhisattva path can lead to full awakening. 60 Germano 1994, 223-4. 170 and the early Great Perfection, likewise recasts the practices of sevā and sādhana in terms of freedom from an actor or action and freedom from effort, important aspects of the Great Perfection tradition that are also emphasized in Buddhajñānapāda’s writings, as we shall discuss below.61 Thus while the works of the early Great Perfection tradition did not necessarily clearly articulate the position that the result of the Vajrayāna was superior, they did use the same technique of reinterpreting the terms of Mahāyoga to suit and characterize the practice traditions that they held to be the pinnacle of Buddhist practice. Another point of intersection between Buddhajñānapāda’s works and some early Great Perfection writings in this regard is the clear preference for the tantric view above that of the various philosophical schools current and popular in their time. We saw that Buddhajñānapāda sometimes upheld a Yogācāra-Madhyamaka position in his writings, and clearly asserted the (Yogācāra-)Madhyamaka position to be the superior one among philosophical views, but at least in his later writings he always indicated this to be lower than the tantric view. Likewise, Padmasambhava’s doxography in the Garland of Views, though it deals only with the larger category of the “vehicles” of Buddhist practice rather than the details of specific philosophical systems within those vehicles, places the tantric view, and especially that of the great perfection, higher than that of all other systems. As noted above, Padmasambhava also discusses the great perfection specifically in terms of its “view,” or philosophical perspective, rather than presenting it in terms of its ritual aspects, like he does with the lower tantric traditions. The Questions and Answers of Vajrasattva, is, like Buddhajñānapāda, more specific in placing the perspective of the great perfection above that of the highest philosophical positions. The Questions and Answers notes that the yogin must discard as delusion all philosophical distinctions “even those of the two highest philosophical schools in early Tibet, the Yogācāra-Madhyamaka and the Sautrāntika- Madhyamaka.”62 Ritual and the Rhetoric of Non-Action Yet another point of contact between Buddhajñānapāda’s writings and those of the early Great Perfection tradition is their shared emphasis on the rhetoric of non-action. This type of rhetoric is an important feature of the writings of the Great Perfection tradition, starting from the earliest period. A number of early or proto- Great Perfection works like Mañjuśrīmitra’s Meditation on the Awakened Mind (Byang chub sems bsgom pa), Buddhagupta’s The Secret Handful (Sbas pa’i rgum chung, IOL Tib J 594), an unascribed commentary on Cuckoo of Awareness (Rig pa’i khu byug IOL Tib J 647), and Palyang’s Six Lamps (Sgon me rnam drug) feature the rhetoric of nonaction.63 However, as van Schaik points out, other early treatises like the Questions and Answers of Vajrasattva, in which the great perfection approach is “firmly embedded in [a] Mahāyoga treatise,” suggest that the rhetoric of nonaction found in these other works should not be taken to imply an actual rejection of all practice.64 Germano has likewise noted that, “in the history of Buddhism we often find the rhetorical negation of a practice serves a variety of functions without necessarily entailing the literal rejection of the practice in question.”65 He suggests that such a rhetoric can serve the function of discouraging a rigid fixation on techniques as producing experiences.66 In Buddhajñānapāda’s writings, just like those of the early Great Perfection tradition, we see the rhetoric of nonaction alongside a Mahāyoga 61 Dalton 2012, 194. 62 J. Dalton 2012, 188; 197. 63 van Schaik 2008, 4-5. 64 ibid. 65 Germano 1994, 227-8. 66 ibid. 171 ritual context. Indeed, Buddhajñānapāda’s works seem to simultaneously advocate ritual practices and reject them. As we saw in the Dvitīyakrama, Buddhajñānapāda wrote that Thus the maṇḍala, homa, Bali, recitation, the counting rosary, |151| Sitting cross-legged, maintaining postures, and so forth— Are in contradiction67 to the unelaborate, [Thus] they should not be [exclusively] taken up; but neither should they be [wholly] rejected Since they are emanated by the *adhideva. |152|68 Likewise, the author of the Questions and Answers of Vajrasattva “is keen to get the message across that the practice of deity yoga is emphatically not to be abandoned, but any concept of the practices as a cause for enlightenment, or of the deities as separate from one’s own primordially enlightenment (sic) mind, is to be abandoned.”69 Some of these early texts even share a similar metaphorical language on nonaction with Buddhajñānapāda. The passage cited above from the Dvitīyakrama continues: The yogin who holds actions To be the great path Is like a wild animal chasing a mirage— [The goal] continually appears, but can never be grasped. |153| When infected by the great sickness of actions, The one who heals [himself] with the great medicine Of unwavering wisdom is a sublime being. |154|70 The Cuckoo of Awareness (Rig pa’i khu byug) likewise describes effort as a “sickness” which is to be “abandoned.”71 As discussed above, the rhetoric of non-action in Buddhajñānapāda’s writings seems to be primarily found in the context of the practices of the perfection stage, and specifically in terms of the yogin’s having come to know the wisdom that is gained through perfection stage practice. In that way, outward ritual can be understood in his system to constitute a foundation for practices that become increasingly unelaborate. In a passage from the Muktitilaka cited above Buddhajñānapāda writes, Why does buddhahood not come from concepts? Because it comes from the utterly pure nature. All of those conceptual rituals, All of that weariness—give it up! Activity arises from this nature [And] action also arises from this nature.72 [But] having come to know that nature 67 Tib. rnam par slu ba, Skt. *visaṃvāda? 68 de la dkyil ‘khor sbyin sreg dang// gtor ma bzlas pa bgrang phreng dang// |151| skyil mo krung dang stang stabs sogs// spros bral rnam bar sluba ste// bya ba ma yin dgag pa min// lhag pa’i lha yis sprul phyir ro// |152| (Dvitīyakrama, verses 151c-152). 69 van Schaik 2004a, 172-3. 70 dbya ba rnams la rnal ‘byor pa// lam chen dag tu yongs ‘dzin pa// ri dwags smig rgyu snyeg pa ltar// rtag tu snang yang ma zin no// |153|bya ba’i nad chen gyis zin la// ye shes g.yo med sman chen gyis// gso byed skyes bu dam pa’o// |154| (Dvitīyakrama verses 153-154). 71 van Schaik 2004b, 72. 72 D om. this line 172 Activity is exhausted and there is no action.73 Here, as was already discussed above, while acknowledging that action and activity have their source in the “utterly nature pure nature” of nondual wisdom, Buddhajñānapāda notes that once the yogin has come to know that nature, activity and action are no longer needed. This perspective on the rhetoric of non-action, that it is is to be understood within the larger framework of action-based practices, but as their culmination or fruition when the ultimate nature has already been known—known precisely by means of having engaged in such actionbased practices—may also be a helpful framework for addressing the rhetoric of non-action in the early Great Perfection tradition, as well. In both the cases there is likely some combination of the two perspectives: on the one hand the rejection of ritual is rhetorical, meant to prevent the practitioner from getting too fixated on ritual methods as producing an outcome, especially since that outcome is, in these systems, already present as the very nature of the practitioner and the world from the very beginning; but such statements against action and effort may also be meant to indicate the actual abandonment of such elaborations as the yogin’s practice comes to fruition. The Nexus of Mahāyoga Tantra and Poetic Pith Instructions on the Nondual Nature One final point of intersection between Buddhajñānapāda’s works and those of the early Great Perfection tradition concerns the presence of both traditional Mahāyoga tantra-based content as well as more poetic statements on nonduality and primordial purity. Both Germano and Karmay have suggested that the early Great Perfection developed at the nexus of the literature of Mahāyoga ritual traditions and literature in the style of the “mind section” of the Great Perfection, that is a “pristine, ritual-free discourse” of “siddha-style yogic practitioners.”74 Van Schaik has, more recently, argued that these were not two separate traditions, but rather that “the early or proto- Great Perfection texts were written by the same people who were producing more conventional direct exegesis on the tantras. There were clearly two different kinds of texts being written, yet the tradition of placing those two kinds of discourse into two different textual categories, Mahāyoga and Atiyoga, had yet to be developed.”75 The 9th-century author Palyang, he says, “wrote treatises on Mahāyoga ritual with a particular emphasis on nonduality and spontaneous presence, as well as short treatises that took the ideas of nonduality and spontaneous presence and expounded them without reference to the ritual universe of the Mahāyoga tantras.”76 While Buddhajñānapāda’s works seem, like Palyang’s writings, to fit more into the category of Mahāyoga-based works that have particular emphasis on nonduality, we do find in his writings short poetic passages that sound very much like mind series literature, and seem as if they could be abstracted and indeed stand alone from the rest of the work as short aphorisms, or pith instructions. That is, it is almost as if in Buddhajñānapāda’s writings we find both strains that are described by scholars as constituting the roots of the Great Perfection tradition present together in the same text. One such pithy passage from the Dvitīyakrama reads: Just as someone possessed of mantras and medicines |145| Enacts the slaying of snakes, 73 ‘di yi don shes rnal ‘byor pas// bya ba’i las ni dkyil ‘khor dang// bya ba’i las ni gtor ma dang// bya ba’i las ni sbyin sreg dang// bya ba’i las ni bgrang phreng dang// bya ba’i skyil mo krung la sogs// dal byed pas ni ci zhig bya// dang po pa rnams ‘drid phyir ro// ci phyir rtog las sangs rgyas min// rang bzhin rnam dag las byung phyir// rtog bcas cho ga de dag kun// ngal ba la sogs spangs byar bya// rang bzhin las ni byed pa’ang yod// rang bzhin las ni bya ba’ang yod// (rang bzhin las ni bya ba’ang yod// ] P, D om. this line) rang bzhin yongs su shes pa la// byed pa sed (sed] D, med P) cing bya ba’ang med// (Muktitilaka, D 47b.6-48a.1; P 57a.8-57b.2). 74 van Schaik 2004a, 166-67; Germano 1994a, 215. 75 van Schaik 2004a, 195. 76 ibid, 201. 173 Likewise, when the great lord of yogins, Seals [them] with the medicine of wisdom What can the afflictive emotions do? |146| What can the rain do To someone with an umbrella in his hand? Likewise, when carrying the umbrella Of nondual wisdom |147| Even if a rain of concepts should fall How could they do any harm?77 Thus while Buddhajñānapāda’s works certainly frequently refer to nondual wisdom in the context of Mahāyoga doctrines like the visualization of the deity, recitation of mantra, and especially the practices of the perfection stage that involve visualizations and sexual yogas, we do also see glimpses like this of statements about nondual wisdom that are abstracted from that kind of framework. Such passages are indeed reminiscent of works of the early Great Perfection. The presence of passages like this in his works may not have been noted in modern scholarship, but the doctrinal similarity between Buddhajñānapāda’s writings and those of the mind series did receive mention by the 14th-century Tibetan historian Gö Lotsāwa, who even goes so far as to cite a number of parallel passages between Buddhajñānapāda’s works and mind series texts.78 Buddhajñānapāda and the Great Perfection? What, then, shall we make of the later Tibetan claims that Buddhajñānapāda was an Indian upholder of the tradition of the Great Perfection? Was some Sanskrit term with the semantic equivalent of the term “great perfection”—whatever that term may have been—used in Buddhajñānapāda’s own writings? Without the Sanskrit originals of the Caturaṅga/Samantabhadra-sādhana and the Dvitīyakrama we will never know, but as we saw above, at least in the case of the Caturaṅga/Samantabhadra-sādhana, and probably in the Dvitīyakrama, as well, the answer is most likely no. However, was the idea of what came to be identified as “the great perfection,” especially according to the earliest literature of that tradition, found in his writings—“the perfection stage of the perfection stage,” “Great Vajradhara,” “Glorious Samantabhadra, the innate state,” a state of wisdom that the yogin could recognize and train in, which at the same time constituted the final result of awakening as well as the foundational ground of phenomenal existence? Yes, by all means, that certainly does appear to be the case. And were quite a number of the doctrines espoused in his writings very much in line with those found in the early or proto- Great Perfection writings appearing in Tibet at precisely the time of Buddhajñānapāda’s life? Again, here the answer is a definitive yes. As van Schaik has noted, “though there is little doubt that most of the texts in the canons of Great Perfection scriptures originated in Tibet, Indic models may well have existed at one time.”79 Van Schaik may perhaps be referencing specifically the type of Great Perfection texts in which the great perfection is abstracted from the Mahāyoga context out of which it seems to have arisen, and here we do not find such a textual model in Buddhajñānapāda’s works. But Buddhajñānapāda’s writings do appear to be a set of Indic texts that share in much the same context and flavor of the very earliest Tibetan Great Perfection works in which the tradition was still enmeshed within, 77 ji ltar snags dang sman ldan pas// |145| sbrul dag gsod par byed pa bzhin// de ‘dra’i rnal ‘byor dbang phyug che// ye shes sman gyis rgyas btab pas// nyon mongs gyis ni ci byar yod// |146| gang zhig lag na gdugs thogs la// de la char pas ci byar yod// de bzhin gnyis med ye shes kyi// gdugs thogs la ni rtog pa yi// |147| char pa rab tu ‘bab ‘gyur yang// de la de yis ji ltar gnod// (Dvitīyakrama, verses 145d-148b). 78 Roerich, 168-9. 79 van Schaik 2004a, 201. 174 and perhaps just beginning to emerge out, of its Mahāyoga chrysalis. Here, indeed, the parallels are very strong, and it is in that sense that I believe we can acknowledge Buddhajñānapāda’s writings as Indic works that are very much in accord with the earliest writings of the Great Perfection tradition. Having familiarized ourselves in Chapter Three with the broader doctrinal ideas espoused in Buddhajñānapāda’s work, and here in this chapter with their relationship to the early Great Perfection traditions that were developing at this time, in the next section we will shift our focus to take a look at the more practical details of the ritual systems that provided the means for Buddhajñānapāda’s disciples and the practitioners of his tradition to immerse themselves in these doctrines in a direct and experiential way. 175 Ritual 176 Chapter Five The Two Stages of Tantric Practice: Generating Self as Deity in Buddhajñānapāda’s Generation Stage Sādhana In an isolated place or on the edge of town, having appropriately completed all the required tasks, sit down on a comfortable seat. Then bring to mind all sentient beings by means of the four great brahmacaryas. With these and the rest, purify the karmic obscurations in one’s mind-stream. Looking at it as being mind alone, the outer world is seen to be empty of nature. Seeing mind alone, as well, to be empty, remain in self-awareness alone. Then that awareness is imagined as a moon, and so forth, upon a seat, which when struck with the pen of the syllable becomes a characteristic implement. From that, generate oneself as the deity. And while possessing divine pride, seal with the four mudrās and emanate the deities of the maṇḍala-cakra, then please them. Accustom oneself to this through training. -Mañjuśrī’s brief generation stage instructions to Buddhajñānapāda, Dvitīyakrama The division of tantric Buddhist practice into two stages is an important feature of the tradition that continues, up to the present day, to provide a fundamental framework for the tantric path. This division into the generation stage (utpattikrama, bskyed rim) and perfection stage (niṣpannakrama, utpannakrama, rdzogs rim)1 developed in the 8th century, as newer practice techniques involving sexual yogas and the manipulation of the inner winds and energies of the practitioner’s subtle body began to be integrated with the earlier practices of self-generation in the form of the deity. Harunaga Isaacson has identified the scriptural locus classicus of the two stages in the Samājottara,2 which first circulated separately but was later integrated into the Guhyasamāja-tantra as its eighteenth chapter. A parallel verse on the distinction into two stages is, however, found in Buddhajñānapāda’s Muktitilaka. A comparison between the two verses suggests that Buddhajñānapāda’s is likely the earlier of the two, and indeed his writings show no clear knowledge of the Samājottara, but rather seem likely to have influenced that work in a number of places.3 Tanemura has even suggested that Buddhajñānapāda “was probably the first person who integrated the two systems of meditation (i.e. the generation and perfection stages) into Buddhist tantric practice.”4 Buddhajñānapāda was certainly a forerunner in the integration of the generation and perfection stages into a system of tantric practice, and his writings may well be the earliest still extant to present a complete and integrated system of both stages, but reference to the two stages is also made in other late 8th-century works like Padmasambhava’s Garland of Views (assuming that the attribution of that work is correct). The division of tantric practice into the generation and perfection stages thus seems to be one that was coming into currency, in both authored works—composed by multiple authors—and Buddhist scriptures, right around the turn of the 9th century. At this far of a historical remove, we may not be able to say much more than that about the emergence of these important categories. 1 Both the terms niṣpannakrama and utpannakrama are used in Sanskrit Buddhist literature to refer to what is translated into Tibetan as rdzogs rim, and what I refer to here as the perfection stage. We do not know for certain which of the Sanskrit terms Buddhajñānapāda used for this stage of practice (or if he perhaps used both?), as none of his writings using the term survive in Sanskrit. However, a verse from the Muktitilaka setting forth the two stages of practice is parallel with a verse in the Samājottara, which does survive in Sanskrit, and the Samājottara verse uses the term utpannakrama (to be more precise, it reads kramam autpannakaṃ), perhaps suggesting that this was the term Buddhajñānapāda himself used. 2 Isaacson 2002a, 468-9. 3 C. Dalton 2014. I explore these influences further in Chapter Eight. 4 Tanemura 2015, 329. 177 The generation stage of tantric practice focuses on the processes by means of which the practitioner gradually envisions herself in the form of a tantric deity, including visualization of the deity’s consort (if applicable), complete retinue, and entire celestial abode. Beginning with a recollection of the state of emptiness, it involves a complete reimagining of the entire cosmos in the purified form of the deity’s pure abode, while all of the living beings within the cosmos— with the practitioner-as-deity5 central among them—are reimagined as deities. In the generation stage, then, the practitioner “generates,” or gradually mentally creates, a purer—and according to the tantras, therefore a more accurate—vision of herself and the cosmos in which the world and beings, headed by the practitioner-as-deity, participate in a shared awakened reality. This purified vision of reality that is imagined in the generation stage is taught to align more closely with the nature of reality in its apparent aspect (the empty aspect of reality being undifferentiated in terms of purity or impurity). The procedures by means of which this pure re-envisioning takes place are homologized with the ordinary processes by means of which beings are born into saṃsāric existence, thus making explicit the fact that in the generation stage the practitioner is recreating this same process, but in a pure form. Having generated herself in the form of the deity, the practitioner then stabilizes this vision by remaining within this pure reality while carrying out the performance of a variety of tantric activities. As for the perfection stage, this term is generally used to describe two further aspects of tantric practice. The first of the ways that the term perfection stage is used—described in the later Tibetan tradition as the “perfection stage with attributes” (mtshan bcas rdzogs rim)—is to refer to practices that involve the manipulation of the internal winds and energies of the practitioner’s subtle body using yogic techniques—generally while the practitioner is visualized in the form of the deity. Such manipulations of the subtle body are meant to allow the practitioner to cultivate, and thus bring about a full actualization of, the awakened reality that was envisioned in its apparent form in the generation stage, but the focus in the perfection stage is more on the cultivation of the empty or non-apparent aspect of reality, the direct and nonconceptual experience of suchness itself. These practices sometimes involve applying yogic techniques while in sexual union with a partner (who is sometimes specified as visualized, and other times specified as flesh-and-blood, though often the texts do not make either specification). Perfection stage practices, particularly those associated with yogic techniques involving the manipulation of the winds and energies in the subtle body, are often said to bring about a series of processes and signs that are homologized, not with birth like in the generation stage, but instead with the process of death. The perfection stage practices thus bring about these experiences in an intentional and controlled manner via yogic techniques, thus (re-)producing the death process, as well, but like in the generation stage, in a purified matter that is intended to result in awakening rather than in the continued experience of saṃsāra. A second aspect of the perfection stage—in the later Tibetan tradition termed the “perfection stage without attributes” (mtshan med rdzogs rim)—refers to the practice of cultivating the direct experience of suchness within the context of traditions like that of the Great Perfection and some Mahāmudrā traditions, in which suchness is cultivated without reliance on the previously mentioned yogic techniques. In these traditions less (or no) emphasis seems to be placed on bringing about the signs of the death process, but rather on the practitioner’s simply remaining for more and more prolonged 5 While I use this particular neologism to make clear that the deity here means the practitioner self-visualized as the deity, it is not so different from the way the practitioner-as-deity is referred to in the traditional literature. The Sanskrit recension of the Sāramañjarī, a commentary on Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana uses a similar phrase to describe the practitioner during the practice of the sādhana in which he is self-visualized as the main deity, Mañjuvajra. He is referred to in the Sāramañjarī as the “yogin whose identity is Mañjuvajra” (Mañjuvajrātmayogī) (see, for example, Tanaka 2017, 66). 178 periods—eventually constantly—within the state of suchness itself, which is identified directly with the state of awakening. In contrast to these common uses of the term “perfection stage” to refer to the various methods and techniques by means of which suchness is directly cultivated, in Buddhajñānapāda’s writings we will see that this term, and its synonym the “second stage,” is used much more frequently to refer not to those various methods for the cultivation of suchness (which in his system are a series of yogas performed with a partner in which the winds and energies of the subtle body are manipulated), but rather directly to the experience suchness itself. I will explore Buddhajñānapāda’s use of the term “perfection stage” in more detail in the next chapter, but for now let us first take a look at the way that his works make a distinction between the two stages of tantric practice. Distinguishing Generation and Perfection in Buddhajñānapāda’s Writings As noted above, the classical verse distinguishing the two stages of tantric practice found in the Samājottara has an earlier parallel in Buddhajñānapāda’s Muktitilaka. The verse, in Buddhajñānapāda’s rendering, reads: The dharma taught by the buddhas Abides authentically in two stages: The generation stage And the perfection stage.6 As Isaacson has already noted, the Samājottara’s verse is clearly modeled on Nāgārjuna’s wellknown verse from the Mūlamadhyāmakakārikās that sets forth the distinction between the two truths, the relative truth and the ultimate truth, and the parallels between the two stages and the two truths that the verse’s structure invokes was undoubtedly intentional.7 Unlike the Samājottara, the Muktitilaka goes on to divide both the generation stage and the perfection stage into two subsidiary stages. Immediately following the verse cited above, Buddhajñānapāda writes: The generation stage has two aspects And the perfection stage likewise. Due to the distinction of the two stages There [arise] the mantrakāya and the jñānakāya. The jñānakāya has two [aspects]: The jñāna [kāya] and the kāya of complete purity. This completely pure [kāya] is the seven yogas, The perfection stage of the perfection stage; It is unsurpassed omniscience Endowed with the supreme of all aspects.8 Vaidyapāda explains this passage as follows, The generation stage has two aspects means the generation stage [aspect of the generation stage], the ādiyoga-[samādhi] and so forth, and the perfection stage [aspect of the generation stage], the [practices involving the] bindu and so forth. And the perfection stage likewise means the generation stage [aspect of the perfection stage], 6 sangs rgyas rnams kyis chos bstan pa// rim pa gnyis la yang dag gnas// bskyed pa yi ni rim nyid dang// rdzogs pa’i rim pa kho na’o// (Muktitilaka, D 52.1-2 P 62b.1-2). 7 Isaacson 2002a, 468-9. 8 bskyed pa’i rim pa rnam pa gnyis// de bzhin rdzogs pa’i rim pa’o// rim pa gnyis kyi bye brag gis// sngags kyi sku dang ye shes sku// ye shes sku la ye shes dang// rnam par dag pa’i sku gnyis so// rnam par dag pa sbyor ba bdun// rdzogs pa’i rim pa rdzogs rim ste// rnam pa kun gyi mchog ldan pa// thams cad mkhen pa bla me ‘gyur// (Muktitilaka, D 52a.2-3; P 62b.2-4) 179 meditating on the three bindus; and the perfection stage [aspect of the perfection stage], the reality, just as it is, which is indicated by that. Then, in order to indicate the distinction in the bodies of the deity [that come about] by means of those two stages [he writes] Due to the distinction of... Here, the mantrakāya is the kāya that is generated. [It is so called] because it arises from [syllables like] hūṃ and so forth, [and] it is impure. The jñānakāya is the perfected kāya and is pure. That [perfected kāya] also has two divisions, the wisdom [kāya], or the illusory body, which is the pure kāyā [attained] by the yogin of the third [level] who has slightly [gained] the wisdom of realization; and the kāya of complete purity. This [kāya of complete purity] also has two [aspects]: that which remains in a state in which it displays characteristics, and is therefore [called] the unfailing kāya, and the resultant kāya, which is [the nature of that,] just as it is. These two are the primordially pure completely pure body, which is the seven yogas. That is also the perfection stage of the perfection stage. You should know that precisely this is also omniscience endowed with the supreme of all aspects.9 Though Buddhajñānapāda’s verses only directly identify the fourth among the four subdivisions of the two stages, “the perfection stage of the perfection stage,” Vaidyapāda’s comments fortunately provide more details with respect to these subdivisions. The first of the four, “the generation stage of the generation stage,” he describes as “the ādiyoga[-samādhi] etc.” Ādiyoga is the name of the first of the three samādhis that structure the practice of deity yoga in the Yoga tantras and continue to structure the generation stage in Indic tantric texts all the way up through the late Yoginī tantras, so this is clearly a reference to what are commonly known as generation stage practices. However, Vaidyapāda’s statement “the ādiyoga[-samādhi] etc.” here leaves the precise line of demarcation between these and the next set of practices, the “perfection stage of the generation stage” unspecified. We might presume that his “etc.” is meant to encompass all three of the three samādhis that are often associated with the generation stage, but it seems that this may not be the case; although it is not made clear in this passage, we will see below that the third among the three samādhis, the karmarājāgrī-samādhi, appears to be, in Buddhajñānapāda’s system, the place where perfection stage practices began to be integrated into the already established system of the Yoga tantra’s three samādhis of deity yoga. But what we do see described here in Vaidyapāda’s commentary on this passage from the Muktitilaka is an indication of precisely this same blurring of the line, or overlap, between the two stages that we find in Buddhajñānapāda’s ritual writings, which, presumably, serve as Vaidyapāda’s reference for these statements. That is, the distinction between generation and perfection stage practices that I described above—which is what came to be the standard distinction between the two stages—keeps generation stage practice limited to visualization practice of oneself as the deity and its maṇḍala, usually accompanied by the recitation of mantra, and other associated activities. Practices involving the manipulation of the subtle energies in the body—such as the bindu yogas that Vaidyapāda mentions—would normally be counted as part of the yogas of the perfection 9 bskyed pa’i rim pa rnam pa gnyis/ zhes pa ni bskyed pa’i rim pa ni dang po’i (po’i] D, po P) rnal ‘byor la sogs pa dang/ rdzogs pa’i rim pa ni thig le la sogs pa’o// de bzhin rdzogs pa’i rim pa’o// zhes pa ni de bzhin du bskyed pa’i rim pa thig le gsum bsgom pa dang/ rdzogs pa’i rim pa des mtshon pa’i ji bzhin pa’i don to (to] D, no P)// da ni rim pa gnyis kyis lha’i sku’i bye (bye] P., bya D) brag bstan pa’i phyir rim pa gnyis kyis zhes pa la sogs pa’o// de la sngags kyi sku zhes pa ni bskyed pa’i sku ste/ hūṃ la sogs pa las byung ba’i phyir te ma dag pa’o// ye shes sku zhes pa ni rdzogs pa’i sku ste dag pa’o// de la yang gnyis te ye shes dang zhes pa ni sgyu ma lta bu’i sku ste/ rnal ‘byor pa gsum pas rtogs pa’i ye shes cung zad dag pa’i sku’o// rnam par dag pa’i sku zhes pa dang gnyis so// de la yang gnyis so// mtshan nyid kyi tshul du gnas pas na mi (na mi] D, ni P) slu (slu] D, bslu P) ba’i sku dang/ de ji bzhin pa’i ‘bras bu’i sku ‘o// de gnyis ni ye nas shin tu dag pa’i sku ste sbyor ba bdun no// rdzogs pa’i rim pa’i rdzogs rim kyang de’o// rnam pa kun gyi mchog dang ldan pa’i thams cad mkyen pa yang de nyid yin par shes par bya’o// (Muktitilaka-vyākhyāna, D 38b.2-6; P 366b.4-367a.2). 180 stage.10 But here in Vaidyapāda’s description of the subdivisions of the two stages we see some overlap in the categories of the “perfection stage of the generation stage,” which he describes as “the [practices involving the] bindu and so forth,” and the “generation stage of the perfection stage,” which he describes as “meditating on the three bindus.” It is not clear from these statements precisely what the difference between the bindu-related practices that comprise these two categories might be, or indeed what the difference between something called the “perfection stage of the generation stage” and the “generation stage of the perfection stage” might be. As we shall see, the matter is indeed not completely clear.11 When we look to Buddhajñānapāda’s ritual works, particularly the Caturaṅga/Samantabhadra-sādhana and the section of the Dvitīyakrama that outlines the perfection stage yogas, we see precisely this same slight overlap or blurring of the lines between the two stages reflected in Vaidyapāda’s comments above. In the Caturaṅga/Samantabhadrasādhana, after the main generation stage practices of generating the deity, the consort, and the maṇḍala deities, consecrating each of them and making offerings, praises, and the “tasting of nectar”—that is, after the completion of the first two samādhis, the ādiyoga- and the maṇḍalarājāgrī-samādhis—we find, at the start of the karmarājāgrī-samādhi, a series of two bindu yogas prior to the mantra recitation practice. While all of the other practices listed above are quite expected in a generation stage sādhana, the presence of these bindu yogas at this point is a bit unexpected. The first of the bindu yogas involves the contemplation of the bindu at the yogin’s heart center, but the second, the sūkṣma yoga, performed at the “nose tip”—which is specified in multiple commentaries as being the “tip of the lotus’ nose”—indicates that the visualization takes place at the point of the yogic partners’ conjoined sexual organs.12 While the Caturaṅga/Samantabhadra-sādhana is generally known as the generation stage sādhana of Buddhajñānapāda’s system, these yogas, both in their description in the Caturaṅga/Samantabhadra-sādhana itself, as well as in terms of the further details that are provided in multiple commentaries on the sādhana, are very similar—though not precisely identical (they are less elaborate)—to the first two of the three bindu yogas of the perfection stage described in the Dvitīyakrama and Muktitilaka. We may guess, then, that the two simpler practices involving the contemplation of the bindu at the yogin’s heart center and the sūkṣma yoga at the “nose tip” described in the Caturaṅga/Samantabhadra-sādhana are what Vaidyapāda refers to in his commentary on the Muktitilaka as “the [practices involving the] bindu and so forth” that constitute the “perfection stage of the generation stage,” while the three more elaborate bindu yogas described in the Dvitīyakrama are what Vaidyapāda refers to as “meditating on the three bindus,” which constitutes the “generation stage of the perfection 10 The practice of the sūkṣma yoga (one of the types of bindu yogas described in Buddhajñanapāda’s writings) has precursors in the Sarvatathāgatatattvasaṃgraha-tantra, where the distinction between the two stages of tantric practice is not yet made. We can thus see Buddhajñānapāda’s writings as a sort of intermediary between that earlier system, and later systems of tantric practice in which the stages of generation and perfection are more clearly divided in terms of the presence or absence of yogic techniques involving the manipulation of energies in the subtle body. 11 In addition to serving as an important resource for clarifying the distinction between the two stages as articulated by Buddhajñānapāda and Vaidyapāda, the passages from the Muktitilaka and the Muktitilaka-vyākhyāna just cited are also very important for what they tell us about the result of tantric practice as it was understood by both of these authors. I examined some of these aspects already in Chapter Three. 12 Whether the consort for these practices is a person or simply a visualized partner is not specified in the sādhana or the commentaries. At this point in the sādhana the yogin is continuing to visualize himself as Mañjuvajra, who is in any case in union with his consort throughout the whole of the sādhana. However, some of the instructions found in the commentaries on the second bindu yoga do suggest that a flesh-and-blood partner is most likely intended here. See, for example Samantabhadrī-ṭīkā, D 168a.2-4, Sāramañjarī, D 37b.6-38a.3, and Szántó unpublished 125. 181 stage.” While it does seem likely that Vaidyapāda’s classifications could be applied to the practices described in Buddhajñānapāda’s writings in this way, this remains speculative, as none of our texts states this directly. In any case, all of this taken together suggests that in Buddhajñānapāda’s system some aspects of what came later to be relegated to the perfection stage were integrated with the practices pertaining to the generation stage. Later traditional authors seem to have noticed this idiosyncrasy of Buddhajñānapāda’s system, as well. In a commentary on the Sampuṭa-tantra Gambhīravajra writes that, “regarding ‘that which has the form of the chickpea’13 the pith instruction texts of Jñānapāda indicate this as the bindu of bodhicitta that pertains to the generation stage. But other masters’ writings say [it pertains to] the perfection [stage].”14 Such a blurring of the lines—or overlap—between what came to be the standard distinction between the two stages is, however, not unexpected in the work of one of the earliest writers to describe tantric practice within the framework of these categories. Generation stage and perfection stage practices are similarly found together in sādhanas at Dunhuang, which seem to reflect more or less the same period of tantric development as Buddhajñanapāda’s writings.15 In fact, some later systems of Indian Buddhist tantra likewise appear to uphold a form of practice that combines generation and perfection stage practices, so the more common later paradigm that separates the two stages into two separate ritual practices does not appear to be the only one.16 Nonetheless, despite the slight blurring, or overlap, of the two stages in Buddhajñānapāda’s system, the basic character of the distinction of the generation and perfection stages as it came to characterize tantric Buddhist practice already holds from the time of Buddhajñānapāda’s writings. Indeed, even Buddhajñānapāda’s division of the perfection stage into two aspects or stages, which Vaidyapāda has termed the “generation stage of the perfection stage” and the “perfection stage of the perfection stage,” corresponds quite well with the later Tibetan traditions’ categories of the “perfection stage with attributes” and the “perfection stage without attributes.” It therefore seems that such a dual function of the term “perfection stage” was already present from the time of the early writings on these practices. Having briefly explored the distinction between the two stages as it is articulated in Buddhajñānapāda’s writings I will, in the remainder of this chapter, focus on the generation stage writings found in Buddhajñānapāda’s oeuvre, before taking up the perfection stage practices in the next chapter. Buddhajñānapāda’s Generation Stage Sādhanas Five generation stage sādhanas that can be confidently attributed to Buddhajñānapāda— the Caturaṅga/Samantabhādhra-sādhana (Tōh. 1855 and 1856), Bhaṭṭārakāryajambhalajalendra-sādhana (Tōh. 1861), Guhyajambhala-sādhana (Tōh. 1862), Vistarajambhala-sādhana (Tōh. 1863), and Śrīheruka-sādhana (Tōh. 1857)—survive in Tibetan 13 The bindu is in some practices described as being the size of a chickpea, and in others the size of a mustard seed. 14 rtsa na ka yi zur gzugs can/ zhes bya ba ni ye shes zhabs kyi man ngag gi gzhung gis/ bskyed pa’i rim pa’i phyogs su byang chub sems kyi thig le’o// slob dpon gzhan ma’i gzhung gis de ni rdzogs par gleng/ (Śrīguhyārthaprakāśamahādbhūta, D 127b.6-7). 15 J. Dalton 2004, 8. Also see, for example IOL Tib J 464/1 as described in J. Dalton and van Schaik 2006, 60-1. The Guhyasamāja sādhanas at Dunhuang show no knowledge of a distinction between the Jñānapāda School or the Ārya School, so they are thought to represent an early stage of Guhyasamāja practice. However, as I noted in Chapter One, the Mahāyoga sādhanas from Dunhuang seem to hew to a different ritual paradigm than do those from much of the Indic tradition (see Chapter One note 126). 16 Ratnākaraśānti’s Mahāmāyā-sādhana (Tōh. 1643) for instance, seems to combine both stages of practice. Thanks to Harunaga Isaacson for pointing out thisx feature of the sādhana to me. 182 translation (one of them in two different Tibetan translations), and portions of three of these sādhanas—the Caturaṅga/Samantabhadra-sādhana, Vistarajambhala-sādhana, and Śrīherukasādhana— survive in their original Sanskrit. Four of the five works are minor, very short sādhanas, while the Caturaṅga/Samantabhadra-sādhana is a major work that inspired the composition of a number of Indic commentaries and related writings, and provides the basis for the nineteen-deity generation stage maṇḍala that characterizes the Jñānapāda School of Guhyasamāja practice. As I noted in Chapter One, while the Mukhāgama (again, not to be confused with Buddhajñānapāda’s Dvitīyakramatattvabhāvanā-mukhāgama, which is referred to as the Mukhāgama in some modern scholarship!), a generation stage sādhana based closely on the Caturaṅga/Samantabhadra-sādhana, is attributed both traditionally and in modern scholarship to Buddhajñānapāda, its concluding verses clearly indicate that it was compiled/composed by Śākyamitra on the basis of Buddhajñānapāda’s oral instructions, and an earlier part of the text suggests that Śākyamitra wished to distinguish his own style of composition from that of Buddhajñānapāda’s. Thus, I do not consider this work to be among Buddhajñānapāda’s generation stage writings. However, it does provide some clarifications on certain points of the generation stage practice that are not included in the Caturaṅga/Samantabhadra-sādhana itself. The fact that the Mukhāgama shares a number of details in common with Vaidyapāda’s commentary on the Caturaṅga suggests that it very well may have been based on Buddhajñānapāda’s oral instructions on the practice, and it therefore remains an important resource in the study of the generation stage according to his system. I will focus my comments in this chapter on the Caturaṅga/Samantabhadra-sādhana, as this is clearly the most important of Buddhajñānapāda’s generation stage works. However, before getting into that more elaborate work, a few observations on his shorter sādhanas are in order. Three of the short sādhanas, the Bhaṭṭārakāryajambhalajalendra-sādhana, Guhyajambhala-sādhana, and Vistarajambhala-sādhana, are practices of the wealth deity Jambhala, who, as we may recall, played a rather important role in Buddhajñānapāda’s life story; Jambhala is credited in the later autobiographical section of the Dvitīyakrama with having provided Buddhajñānapāda and his disciples their daily provisions when they lived together in the Parvata cave near Vajrāsana in the later part of Buddhajñānapāda’s life. Of these three sādhanas, the Vistarajambhala-sādhana also survives in Sanskrit, as Sādhanamāla No. 285,17 though it is not precisely identical with the Tibetan translation in several places. The three Jambhala sādhanas are grouped together in the Tengyur with a common colophon. Like the Caturaṅga/Samantabhadra-sādhana, these Jambhala sādhanas give a short description of the standard preparations for the practice, such as generating bodhicitta, making offerings, confessions and the like, and then describe the generation of the practitioner-as-deity in a very simple way—he emerges from his seed syllable jaṃ. Two of the three Jambhala sādhanas involve the process of nyāsa, the installation of syllables or deities on the body of the deity, which also features in the Samantabhadra-sādhana. The Jambhala practices also describe a number of ritual activities that the practitioner is to engage in, including making water offerings to Jambhala and the wealth goddess Vasudhārā, creating drawn images, and a wrathful means of accomplishing Jambhala involving Vajra Hūṃkara, which is given as a failsafe if the peaceful methods of accomplishment do not prove effective. Like the Caturaṅga/Samantabhadrasādhana, one among the three, the Bhaṭṭārakāryajambhalajalendra-sādhana, gives details on how to practice during post-meditative activities like eating. The fourth short generation stage sādhana attributed to Buddhajñānapāda is the Śrīheruka-sādhana for the practice of an otherwise un-named Heruka. This work is clearly 17 Bajracharya 1928, No. 285. 183 associated in its commentary with the Guhyasamāja-tantra and shares several features with the Caturaṅga/Samantabhadra-sādhana, including the generation of the main deity by means of a causal or progenitor deity who is identical in terms of form, color, and symbolic implements to the causal deity described in the Caturaṅga/Samantabhadra. Both are likewise referred to by the common epithet for a causal deity, the “vajrasattva.” While the cremation ground aesthetic with which the Heruka in the Śrīheruka-sādhana is depicted is not reflected in Buddhajñānapāda’s other works, much of the rest of the structure of this brief sādhana aligns with the structure of the Caturaṅga/Samantabhadra-sādhana, albeit in a very condensed way. This is the only one among the short sādhanas attributed to Buddhajñānapāda that is overtly connected with the Guhyasamāja-tantra. His most important Guhyasamāja generation stage work is, however, the Caturaṅga/Samantabhadra-sādhana, to which we now turn. The Caturaṅga/Samantabhadra-sādhana and Associated Works Judging only by the number of extant Indic commentaries on this sādhana—there are five!—the Caturaṅga/Samantabhadra-sādhana may have been Buddhajñānapāda’s most popular work. The sādhana is preserved in two different Tibetan translations which each bear one of the two different titles under which the work seems to have circulated: the Samantabhadra-sādhana (Tōh. 1855), which was translated into Tibetan by Śraddhākaravarman and Rinchen Zangpo, and the Caturaṅga-sādhana (Tōh. 1856), which was translated by Smṛtijñānakīrti, working alone. Despite the overwhelming parallels between the two works (which are, in fact, nearly identical in content), the existence of two separate Tibetan translations with different titles and the additional fact that the translation of the Caturaṅga-sādhana in the Tibetan canon preserves a mixed-up page order (it appears that the pages of the text were quite literally shuffled, and the translation was somehow preserved in the Tengyur with this shuffled page order), there has been some question in earlier scholarship as to whether these two translations constitute two separate works, or are separate translations of a single work.18 The two translations are quite different in style: Śraddhākaravarman and Rinchen Zangpo’s Samantabhadra-sādhana breaks the verses of the Sanskrit text into four nine-syllable lines, while Smṛtijñānakīrti has rendered the verses of the Caturaṅga into two lines ranging from thirteen to nineteen syllables. The former has resulted in a translation that is more readable in Tibetan, but the overall content of the two translations is in most places identical. That is, a careful reading of the two translations side-by-side shows that most of the differences between the two can be attributed to differences in translation word choice, but there do remain some minor variants that indicate that these may represent two separate recensions of the same work.19 The Tibetan translation of the Caturaṅga also shows more evidence of textual corruption than that of the Samantabhadra, and there are a number of instances where recourse to the Samantabhadra is necessary to make sense of the Caturaṅga’s corruptions. This includes the use of the Samantabhadra to restore the correct order of the Caturaṅga.20 18 See Kikuya 2012, 1265 and Tanaka 2017, 29. 19 Here I concur with Tanaka (1996, 264) who has come to the same conclusion. Tanaka has compared a few phrases from the two translations (1996, 263-64). 20 See Kikuya 2012 for a table of the correspondences. I did not have access to the editions Kikuya references in this article and was thus unable to make sense of his correspondences. I therefore restored the verse order myself when reading the two texts side-by-side, so I am able to separately confirm that the verse order of the complete Caturaṅga (with the exception of three pādas and three half-pādas that seem to have been lost in the page-shuffling process) can be fully restored with recourse to the Samantabhadra. 184 A comparison of the available commentaries on the sādhana adds both clarity and complexity to our picture of the transmission of the work. Four commentaries survive in Tibetan translation, two of which indicate in their titles that they are commentaries on the Caturaṅga— Vaidyapāda’s Caturaṅgasādhanopāyikā-samantabhadrī-nāma-ṭīkā (Tōh. 1872) and Samantabhadra’s Caturaṅgasādhana-ṭīkā-sāramañjarī-nāma (Tōh. 1869)—and two of which indicate in their titles that they comment on the Samantabhadra—Śrīphalavajra’s Samantabhadrasādhana-vṛtti (Tōh. 1867) and Thagana’s Śrīsamantabhadrasādhana-vṛtti (Tōh. 1868). Two Sanskrit manuscripts of Samantabhadra’s Sāramañjarī survive: one fragmentary manuscript discovered by Kimiaki Tanaka in the Nepal National Archives, which has been edited, translated, and studied by him, and another nearly complete manuscript of a significantly more extensive recension of the work, photographs of which were brought to Europe by Giuseppi Tucci, and which is currently under study by Péter Szántó.21 However, neither of these recensions is precisely identical with the Tibetan translation of the Sāramañjarī, and there are passages where even the two Sanskrit texts do not correspond.22 (To make the situation even more complicated, the longer Sanskrit recension also incorporates at least some passages from Vaidyapāda’s commentary—which survives only in Tibetan—that are not found in the Tibetan translation of the Sāramañjarī!23) The Tucci manuscript of the Sāramañjarī, however, does include a colophon confirming that the Sāramañjarī styles itself a commentary on the Caturaṅga-sādhana, rather than on the Samantabhadra.24 A fifth Sanskrit commentary survives, but is unavailable for study.25 The fact that two of the surviving Indic commentaries indicate by their titles that they are commenting on the Caturaṅga, while two self-identify as commentaries on the Samantabhadra, and yet all four of these are, in fact, commenting on a sādhana with the same content, suggests that the work was preserved and circulated under two different names in India. Among the Indic commentators on the work, Vaidyapāda and Samantabhadra are the earlier of the four, and thus the closest to Buddhajñānapāda. Given that both use the title Caturaṅga for the sādhana, we can surmise that this is the earlier of the two titles under which the work circulated. Regarding their respective chronologies, Vaidyapāda and Samantabhadra were likely contemporaries, some time in the mid-9th century. As was discussed in Chapter One, Vaidyapāda was probably a direct disciple of Buddhajñānapāda himself, as well as a disciple of some of Buddhajñānapāda’s senior disciples, like Dīpaṃkarabhadra; Samantabhadra seems to have been a disciple of Buddhajñānapāda’s co-disciple (under their guru Pālitapāda) Śrīkīrtipāda, 21 See Tanaka 2017 for the most recent edition and English translation of the shorter Sanskrit recension of the Sāramañjarī from the Nepal National Archives. Tanaka had already published the Sanskrit edition of the manuscript in a series of earlier articles and book chapters (Tanaka 1991; 1995; 1996; 2007b; 2010). It is not unexpected that we find a manuscript of the Sāramañjarī preserved in Kathmandu given that, according to the colophon of its Tibetan, the work was translated into Tibetan “in the palm of Nepal” (bal yul gyi thil du bsgyur), which I assume to be the Kathmandu Valley (Sāramañjarī, D 45b.4). See Szántó 2015 for a description of the longer Sanskrit recension of the Tucci manuscript. I am grateful to Péter Szántó for sharing with me his unpublished draft Sanskrit edition of the longer Sanskrit recension (and hope he will publish it soon!). 22 In fact, it is possible that the differences among the various “recensions” are significant enough to make it difficult to identify them as a single work, though there are certainly enough parallels that it is tempting to do so, and I tentatively do. Further comparative study of the texts is necessary to determine their relationship more clearly. 23 See, for example Samantabhadrī-ṭīkā D 145b.2-4, corresponding with the Sanskrit in Szántó, unpublished, p. 48, 39.2. 24 The colophon reads Sāramañjarī nāma Caturaṅgasādhanasya ṭīkā samāptā. kṛtir iyam ācāryaśrīSamantabhadrapādānām iti. (Szántó, unpublished draft Sanskrit edition of the Sāramañjarī, 152.) This is helpful as the Sanskrit titles given in the Tibetan translations of works in the Kangyur and Tengyur are sometimes reconstructions and are thus not always reliable (as in the case of the Dvitīyakrama!). 25 The existence of this commentary is reported in Kawasaki 2004. 185 who he notes in the colophon of the Sāramañjarī commanded him to compose that work.26 The other two commentators, Śrīphalavajra and Thagana, who both comment on the Samantabhadra, are from a later period, with Thagana probably living in the eleventh century.27 Vaidyapāda’s and Samantabhadra’s commentaries are strikingly similar. Significant elements are nearly or actually identical between the two works, even down to the exact wording in a number of passages. Given that they seem to have lived around the same time, we are left with only textual evidence from the two commentaries to determine their relationship. A close reading of the two commentaries side-by-side has led me to the conclusion that Vaidyapāda’s work was likely earlier, and that Samantabhadra probably relied on Vaidyapāda’s commentary when composing his own.28 Another interesting point on the relationship between these two commentaries is that while Samantabhadra’s Sāramañjarī identifies itself as a commentary on the Caturaṅga, there are a number of passages where the Sāramañjarī diverges from Vaidyapāda’s Caturaṅga-ṭīkā in which the Sāramañjarī is closer to the text of the Samantabhadra-sādhana as it is preserved in Tibetan translation, while the Caturaṅga-ṭīkā hews more closely to the Caturaṅga-sādhana as preserved in the Tibetan.29 While it is certainly possible that some of these instances result from differences arising out of the translation process, there are enough such instances that seem to actually represent minor variants in the sādhana itself, that it appears as if the Sāramañjarī may be commenting on a version of the sādhana that has already undergone a small amount of change from the version that Vaidyapāda commented on, and that those minor changes in the sādhana were later associated with its alternative title, the Samantabhadra-sādhana. This is just a tentative suggestion, however, and warrants further research. But the fact that the two different titles, Caturaṅga-sādhana and Samantabhadra-sādhana were understood by Indian commentators to refer to the same work is 26 On the latter point see Szántó 2015, 554. 27 See my Chapter Three note 239. 28 There are, for example, instances where both cite the same passage of a different work to support their comments on the Caturaṅga, and often in those cases Samantabhadra’s citation is a longer version of the one that Vaidyapāda provides, adding further context. This is the case even for the Tibetan recension of the Sāramañjarī, which is not nearly extensive as the longer Sanskrit recension, which includes even more extensive citations not present in either of the other two recensions. See for example Caturaṅga-ṭīkā, D 136b.7 and Sāramañjarī, D 8a.1-2. In the occasional instances where Vaidyapāda provides a longer version of a citation that is common to the two commentaries, Samantabhadra follows his abbreviated citation with a prose summary of the rest of the full citation that Vaidyapāda has provided (See for example Caturaṅga-ṭīkā, D 138b.1 and Sāramañjarī, D 9a.7-9). There are yet other sections where Vaidyapāda explains that a certain topic ought to be learned from the oral instructions, but Samantabhadra gives more detail on that very topic in writing in his commentary (See, for example, Caturaṅga-ṭīkā, D 139a.3 and Sāramañjarī, D 10b.3). Of course, even given this evidence, my conclusions on the relationship of Vaidyapāda’s and Samantabhadra’s works remain speculative, especially given that we know, from comparing its various recensions, that Samantabhadra’s commentary itself was augmented over time. A study of the relationship between the various recensions of the Sāramañjarī would certainly provide a fruitful way to study the development of texts from this period. As something of an aside, I will add that I have seen a similar situation in the case of a much later (18th-century) Tibetan generation stage commentary that was similarly augmented—often with additional citations, which is precisely what we see in the case of the Sāramañjarī. Certainly texts grew, but it may be that this genre of practical sādhana instructions was particularly susceptible to development and augmentation. 29 See, for example Caturaṅga-ṭīkā, D 141b.2 which reads ‘dul ba, following the Caturaṅga-sādhana, and the Sāramañjarī D 13a.3, which reads ‘dul bya following the Samantabhadra-sādhana; See also Caturaṅga-ṭīkā, D 142b.3 which reads zas lnga following the Caturaṅga-sādhana and Sāramañjarī D 14a.5 which reads mchod pa lnga following the Samantabhadra-sādhana. See also Caturaṅga-ṭīkā, D 143b.3 which reads tshad med ‘od gzhi following the Caturaṅga-sādhana and Sāramañjarī D 14b.2, which reads ‘od zer dpag du med pa (omitting bzhi) following the Samantabhadra-sādhana. See also Caturaṅga-ṭīkā D 143b.5 which reads mchog gi rdo rje’i ye shes gzugs sogs pa following the Caturaṅga-sādhana’s rdo rje dam pa ye shes gzugs, and Sāramañjarī D 15a.2, which reads nor bu dmar ‘dra zhes pa ni / padma rāga dang ‘dra ba’o// following the Samantabhadra-sādhana’s rdo rje bzang ‘dir nor bu dmar ‘dra la//. These are just a few of the many instances of this pattern. 186 strongly suggested by the fact that Thagana, commenting in the 11th century on the Samantabhadra-sādhana appears to have been reading Samantabhadra’s 9th-century commentary on the Caturaṅga-sādhana. Thagana follows Samantabhadra’s commentary on quite a number of points, many of which are places where Samantabhadra’s comments diverge from Vaidyapāda’s.30 There is certainly more to learn about the transmission history of the Caturaṅga/Samantabhadra-sādhana, but even just this limited comparison of the extant commentaries shows us that the sādhana seems to have undergone some minor changes in India, and that it was preserved there under two different titles that were nonetheless recognized by exegetes as referring to the same work.31 In addition to the several commentaries that it inspired, the Caturaṅga/Samantabhadrasādhana also served as the basis for at least two other Indic works: Śākyamitra’s Mukhāgama and Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi. Both of these texts paraphrase or repurpose large sections of the sādhana. Śākyamitra’s text, which he claims to be a record of Buddhajñānapāda’s oral instructions on the Samantabhadra-sādhana, is an unusual work in that it seems to be something of a cross between a sādhana and a sādhana commentary, but almost entirely in verse. The Mukhāgama basically follows the structure of the Caturaṅga/Samantabhadra-sādhana, describing the same maṇḍala of deities from the Caturaṅga/Samantabhadra, but with added details on certain points that are not elaborated in Buddhajñānapāda’s work, including an extensive elaboration on the rakṣacakra of wrathful deities that form the protective boundary for the practice of the sādhana. Some of the details added in the Mukhāgama also appear in Vaidyapāda’s commentary on the Caturaṅga-sādhana, which does enhance the possibility that both works were indeed composed on the basis of oral instructions on the practice of the sādhana. The Mukhāgama also superimposes a structure of sorts on the practice, including a brief preliminary discussion of the initiatory prerequisites, and a structuring of the generation stage process in terms of the twelve links of dependent origination (pratītyasamutpāda) and the four boundless attitudes (apramāṇa).32 The Mukhāgama, however, departs from the order of the Caturaṅga/Samantabhadra-sādhana in several places, and also seems to repeat certain processes of the sādhana more than once with alternate descriptions. Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi directly paraphrases most of the Caturaṅga/Samantabhadra-sādhana, which it integrates as the framework for an initiation manual in Buddhajñanapāda’s Guhyasamāja tradition.33 While Dīpaṃkarabhadra’s work follows the Caturaṅga/Samantabhadra much more closely than does the Mukhāgama, Dīpaṃkarabhadra, like Śākyamitra, also rearranges the verse order of the Caturaṅga/Samantabhadra in several places, and changes the content of some verses, such as, 30 For evidence that Thagana is reading Samantabhadra, see, for example, Sāramañjarī D 6a.2-3 and Śrīsamantabhadrasādhana-vṛtti, D 192a.5-6 on the offerings to the buddhas of the three times which are discussed in the later two pādas of verse 8 of the sādhana. 31 My observations here on the relationship among the commentaries were made without reference to Szántó’s draft edition of the long recension of the Sāramañjarī, which I received after I had already read the sādhana and the extant Tibetan commentaries alongside Tanaka’s edition of the shorter recension of the Sāramañjarī in Sanskrit. A closer look at this longer recension of the work will certainly add more to our picture of the relationship between the commentaries, as well as providing a fascinating look at way a single commentarial work developed. 32 Several scholars have attributed this structuring of the generation stage practice in terms of the twelve links of dependent origination to Buddhajñānapāda himself, but as I have discussed above, the Mukhāgama is clearly attributed to Śākyamitra. While such categories may very well have been part of Buddhajñānapāda’s oral instructions on the Caturaṅga/Samantabhadra-sādhana, it is difficult to assert categorically that they are his contributions when they are nowhere represented in his own writings. 33 Daisy Cheung, a doctoral student at the University of Hamburg, is currently preparing a doctoral dissertation on Dīpaṃkarabhadra’s text, which will be a very welcome addition to the research on Buddhajñānapāda’s Guhyasamāja tradition. 187 for example, the verses on taking refuge in the three jewels, which have a content very different from Buddhajñānapāda’s refuge verses. Dīpaṃkarabhadra’s popular work is the subject of two extant Indic commentaries, by Vaidyapāda and Ratnākaraśānti, and appears to have been quite influential. But not more influential, it seems, than the Caturaṅga/Samantabhadra-sādhana itself, which, including its five commentaries and the two works based on it, served as the direct inspiration for at least seven different extant Indic works.34 Structure of the Generation Stage in the Caturaṅga/Samantabhadra-sādhana As described above, the practice of the generation stage involves the practitioner visualizing herself in the form of a tantric deity and the deity’s retinue. This “generation” of oneself as the deity is generally preceded by some preliminary practices including taking refuge in the Buddha, Dharma, and Sangha; generating bodhicitta, the aspiration to achieve awakening for the benefit of all sentient beings; and generating a protective circle within which the practice will unfold, along with the deity’s pure abode. Once the full maṇḍala of the deity, consort (if there is a consort), and retinue have been completely generated out of the state of emptiness, the so-called wisdom deities (jñānasattva) are invited and descend into the practitioner-as-deity after which offerings and praises are presented to the deity and maṇḍala. This is usually followed by some mantra recitation and ultimately with the dissolution of the visualization back into emptiness and the closing of the practice with dedications and aspirations. Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana follows this general model, with some elaborations at various points. The commentaries on the Caturaṅga/Samantabhadra-sādhana place its generation stage processes within the structure of three samādhis, which were already used to structure the practice of self-generation as the deity in the Yoga tantras. According to the commentarial tradition these three samādhis—the ādiyoga-samādhi, maṇḍalarājāgrī-samādhi, and the karmarājāgrī-samādhi—constitute the larger framework within which the sādhana is to be practiced. Indeed, these same three samādhis are employed as the framework for generation stage sādhana in most Indian tantric traditions from the Yoga tantras onwards, all the way up through works on the later Yoginī tantras like the Cakrasaṃvara, Hevajra, and Saṃputa-tantra. Generally, the three samādhis encompass the generation stage process as follows: the ādiyoga- 34 In addition to being Buddhajñānapāda’s work that has received the most attention from Indic commentators, the Caturaṅga/Samantabhadra-sādhana is also the work that has received the most attention from modern scholars. Kimiaki Tanaka has done the most extensive research on the sādhana, having edited and published the Sanskrit fragment of the Sāramañjarī, as well as translating it in full and providing a very brief study of the generation stage process of the Caturaṅga/Samantabhadra (Tanaka 1996 (in Japanese); 2010 (in Japanese with English Chapter Summaries); and 2017). Yukei Matsunaga wrote (in Japanese) about the Caturaṅga/Samantabhadra-sādhana in his work on the history on the development of tantric Buddhism (Matsunaga 1980 A history of the formation of esoteric Buddhist scriptures, cited in Tanaka 2017). Kazuo Kano has published an edition of verses 20-54 of the sādhana preserved in a manuscript in Tibet (Kano 2014). The full text is presumably preserved in the manuscript from which Kano derived the verses published in this article. However, it is kept on display at the Tibet Museum in Lhasa and scholars are not given access to the manuscript. Kano published his edition of the verses that happened to be on display at the museum on a given day and which he was able to photograph. Péter Szántó has produced an as-yetunpublished edition of the long Sanskrit recension of the Sāramañjarī. Ryuta Kikuya has reconstructed the verse order of the Caturaṅga on the basis of the Samantabhadra (Kikuya 2012; this article was published in Japanese, but Kikuya gave a 2014 conference presentation in English the content of the article). Tsutomu Sato has also written in Japanese about the maṇḍala of the Samantabhadra-sādhana (Sato 1995, The Composition of Maṇḍala in the Jñānapāda-school, cited in Tanaka 2017) and Chizuko Yoshimizu studied the four branches of the Caturaṅga/Samantabhadra-sādhana in her unpublished master’s thesis, also in Japanese (Tanaka 2017, 36). 188 samādhi involves the preliminaries as well as the practitioner’s self-generation as the deity (and, if there is one, the generation of the deity’s consort), the maṇḍalarājāgrī-samādhi includes the generation of the remaining deities that constitute the retinue of the main deity, and the karmarājāgrī-samādhi consists of the ensuing activities performed by the yogin while selfvisualized as the deity, along with the dissolution of the maṇḍala and the concluding activities of dedication and aspiration. The Caturaṅga/Samantabhadra-sādhana adheres to this paradigm quite closely. The ādiyoga-samādhi of the Caturaṅga/Samantabhadra-sādhana consists of the preliminary practices and the self-generation of the practitioner as the main deity Mañjuvajra along with his consort, as well as the consecration of both; the maṇḍalarājāgrī-samādhi encapsulates the generation of all of the other eighteen deities of the maṇḍala and their consecration, along with the offerings and praises that are made to Mañjuvajra and the surrounding deities of the maṇḍala; and the karmarājāgrī-samādhi consists of a number of activities including a philosophical analysis of reality, a contemplation of the symbolism of the maṇḍala deities, the practice of bindu yogas, the dissolution of the maṇḍala, and concluding prayers and aspirations. The Indic commentaries are uniform in their division of the Caturaṅga/Samantabhadra-sādhana into these three samādhis, though they differ on the minor point of whether to also include in the third samādhi a number of instructions on post-meditative conduct, as well as some short rites to be practiced in conjunction with the sādhana, as needed.35 The next level of structure of the sādhana involves dividing the practice into the four branches—caturaṅga—the division that has given the sādhana one of its names. While the structuring of the text into these four branches is also set forth only in the commentarial tradition—the sādhana itself makes no mention of any of the four branches—the fact that the Caturaṅga-sādhana seems to be the earlier of the two titles by which the work was known suggests that these divisions are likely not a later addition to the tradition but were part of the way Buddhajñānapāda himself structured the practice of the generation stage. (Presumably this is also the case with the three samādhis discussed above, since that division was already used to structure Yoga tantra sādhana.) The four branches themselves—sevā, upasādhana, sādhana, and mahāsādhana—are described in Chapter Twelve of the root Guhyasamāja-tantra. That passage reads (in Fremantle’s translation): Absorption in the sacred law of sevā, the arising of upasādhana, the sacred law whose object is sādhana, and mahāsādhana the fourth, —having understood their distinctions, then perform the Acts. Absorption in the samādhi of sevā is to meditate on ultimate enlightenment, in the great siddhi of upsādhana examine the vajra senses, in sādhana visualize the Mantra Lord—this is called arousing, and at the time of mahāsādhana the vajra wisdom will succeed by visualizing the image of the Vajra of his mantra with the Lord in his crown.36 These verses are followed by a short section of the tantra describing the practice of sevā and the vows of upasādhana, sādhana, and mahāsādhana. It seems that the four branches as described here in the Guhyasamāja-tantra provide just a very basic guideline for the way in which they are employed in the Caturaṅga/Samantabhadrasādhana. That is, the branches as they are described as pertaining to the sādhana correspond 35 Vaidyapāda indicates that the conclusion of the karmarājagrī-samādhi takes place with the dissolution of the maṇḍala deities in verse 139. Samantabhadra includes the following section of additional practices within the third samādhi and holds that the karmarājagrī-samādhi concludes only after verse 157. 36 Fremantle, 1971, 70. sevāsamayasaṃyogam upasāadhanasambhavam/ sādhanārthasamayaṃ ca mahāsādhanacaturthakam/ vijñāya vajrabhedenga tataḥ karmāṇi sādhayet/ 60 sevāsamādhisaṃyogaṃ bhāvayed bodhim uttamam/ upasādhanasiddhyagre vajrāyatanavicāraṇam/ 61 sādhane codanaṃ proktaṃ mantrādhipatibhāvanam/ mahāsādhanakāleṣu bimbaṃ svamantravajrinaḥ/ 62 makuṭe ‘dhipatiṃ dhyātva sidhyate jñānavajriṇaḥ/ 63 (Matsunaga 1980, 42-3). 189 only loosely with the way they are described in the tantra. In accordance with the commentators on the Caturaṅga/Samantabhadra-sādhana—who are uniform in dividing the sādhana into these four branches—sevā consists of generating the deity; upasādhana consists of the practice of nyāsa, filling the body of the deity with syllables that represent deities; sādhana involves the consecration of the deity’s body, speech, and mind with the syllables oṃ āḥ and hūṃ, representing awakened body, speech, and mind; and mahāsādhana consists of the deities receiving consecration and having their crown ornamented by the presiding buddha of their particular buddha-family. These four branches are described in the Indic commentaries as occurring three times in the Caturaṅga/Samantabhadra-sādhana in three different “grades:” lesser (mṛdu), middling (madhya), and greater (adhimātra).37 The lesser four branches involve the four processes described above as applied to the main deity of the sādhana, Mañjuvajra. The middling four branches involve the same four processes as applied to Mañjuvajra’s consort. Both the lesser and middling four branches pertain to the ādiyoga-samādhi. The greater four branches involve applying the same four processes to the other deities of the maṇḍala, and thus pertain to the maṇḍalarājāgrī-sāmadhi. The first of these three sequences of the four branches—the lesser four branches as applied to Mañjuvajra himself—is presented in the sādhana in the most detail, constituting sixteen verses of the sādhana, whereas the middling set is much more abbreviated, in just four verses. In the case of the third set of four branches, that relating to the maṇḍala deities, only the first branch of sevā—the generation of the maṇḍala deities—is described in detail in the sādhana itself, while the other three branches—the practices of nyāsa; consecrating body, speech, and mind; and ornamenting the crown of the maṇḍala deities—are merely alluded to in a single line of the sādhana, but are elaborated in the commentarial tradition. To give a further sense of the details of the generation stage practice as described in the Caturaṅga/Samantabhadra-sādhana, I include below a detailed topical outline of the contents of the work.38 Many of the major topical headlines in the outline are derived from the commentarial tradition, primarily Vaidyapāda’s Samantabhadrī-ṭīkā, while the minor divisions are my own and are meant to give a sense of the practical details of the generation stage processes that the sādhana sets forth. Following the topical outline, I summarize the sādhana’s contents, with recourse to the commentaries for some additional details—thus elaborating further on the topical outline—and finally present a brief discussion some of the notable features of the sādhana. Topical Outline of the Caturaṅga/Samantabhadra-sādhana I. Textual Preliminaries 1. Author’s Homage and Pledge to Compose the Sādhana [1-2] II. Main Part of Text A. The Main Part of the Generation Stage Practice I. Ādiyoga-samādhi 1. Injunction to Practice [3-6] 2. The Preliminaries to the Session a. Sitting Down to Practice, Visualizing the Buddhas [7-8] b. Offering and Confession [9-10] 37 Tanaka (2017, 34-36) discusses the three grades of the four branches as they are presented in the short Sanskrit recension of the Sāramañjarī. 38 Tanaka has included what appears to be a topical outline of the Samantabhadra-sādhana in his 2009 publication on maṇḍalas (501-503). This outline is, however, in Japanese, which I am unfortunately unable to read, so I have not relied upon it in preparing my own. 190 c. Rejoicing and Dedicating [11] d. Taking Refuge in the Three Jewels [12-14] e. Arousing Bodhicitta [15-17] 3. The Preliminaries to Generating the Deity a. Contemplating Three Gates of Liberation [18] b. Meditating on the Protection Circle [19a] c. Generating the Dharmodaya [19b-20] d. Generating the Celestial Palace [21-27] e. Generating the Seats of the Deities [28] 4. Generating the Main Deity: The Four Branches—Lesser Stage a. Lesser Sevā: Self Generation as Mañjuvajra with Consort i. Generating the Causal Deity 1. Generating the Causal Deity and Consort [29-34] 2. Unique Preliminaries: Purifying and Generating Embodied Beings and the Inanimate World as Deity a. Buddhas Enter into Causal Deity [35] b. Goddesses Emerge from Causal Deity and Dissolve into Consort [36] c. Maṇḍala emerges into Consort’s Lotus, All Beings Placed in Maṇḍala [37] d. Buddhas Enter Maṇḍala and Confer Initiation on Beings [38] e. Beings are Purified and Emanated as Maṇḍala Deities from Syllables [39-43] 3. Practitioner’s Mind as oṃ āḥ hūṃ Enters into the Consort’s Lotus and Melts [44] 4. Causal Deity and Consort Melt [45] 5. Goddesses Sing for Deity to Emerge [46-49] ii. Generating Resulting Deity: Mañjuvajra [50-54] 1. Light from Syllable maṃ Summons Buddhas who Dissolve into Oneself who Becomes Mañjuvajra [50-52] 2. Description of Deity and Blessing [53-54] b. Lesser Upasādhana: Bodhisattvas Fill Mañjuvajra’s Sense Faculties [55a-b] c. Lesser Sādhana: Consecration of Mañjuvajra’s Body, Speech, Mind [55c-63] d. Lesser Mahāsādhana: Consecration and Crowning of Mañjuvajra with Family Lord [64-65] 5. Generating the Consort: Four Branches—Middle Stage [66] a. Middling Sevā: Generate the Consort [66a-b] b. Middling Upasādhana: Bodhisattvas Fill Consort’s Sense Faculties [66c] c. Middling Sādhana: Consecration of Consort’s Body, Speech, Mind [66d] d. Middling Mahāsādhana: Consecration and Crowning of Consort with Family Lord [66d] 6. Consecrating the Consort’s Body with The Five Families [67-68] 7. Joining in Union with the Consort to Please the Tathāgatas [69] 191 II. Maṇḍalarājāgrī-samādhi [70-108] 1. Generating the Maṇḍala Deities: Four Branches—Greater Stage [70- a. Greater Sevā: Generate the Maṇḍala Deities [70- i. Summon Buddhas and Emit Maṇḍala into Consort’s Lotus [70- 71] i. The Five Buddhas 1. Cittavajra (=Akṣobhya) Who Dissolves into Mañjuvajra [72-73] 2. Kāyavajra (=Vairocana) in East [74] 3. Ratneśa (=Ratnasambhava) in South [75] 4. Amitabha in West [76] 5. Amoghasiddhi in North [77] 6. A Description of Their Common Features [78] ii. The Four Consorts 1. Locanā in Southeast, resembling Vairocana [79] 2. Mamakī in Southwest, resembling Akṣobhya [80] 3. Pandara in Northwest, resembling Amitabha [81] 4. Tārā in Northeast, resembling Ratnasambhava [82] 5. A Description of their Common Features [83] iii. The Six Sense Goddesses, Rūpavajrā etc. at Four Corners and Two Sides of Main Door [84] iv. Wrathful Gatekeepers 1. Yamantaka, resembling Akṣobhya in Eastern Gate [85- 86] 2. Aparājita, resembling Vairocana in Southern Gate [87] 3. Hayagrīva, resembling Amitabha in Western Gate [88- 89] 4. Amṛtakuṇḍalin, resembling Akṣobhya, northern gate [90] 5. Crowning Gatekeepers with Family Lords [91] b. Inviting the Wisdom Maṇḍala [92] c. Making Yamāntaka etc. Protect the Maṇḍala [93] d. Greater Upasādhana, Sādhana, and Mahāsādhana [94 c] 2. Making Offerings [94a-b, d; 95-100] 3. Praises [101-106] 4. Tasting the Nectar [107-108] III. Karmarājāgrī-samādhi [109-139] 1. Bindu Yoga: Emanate out Buddhas who Make Beings into Buddhas and Bring them Back into Seed-Syllable [109] 2. The Philosophical Investigation: Establishing All Phenomena to be Mind, Which is Free From Perceiver and Perceived [110-126] a. Pure Equivalences of the Maṇḍala Deities [121-124] 3. Bindu Yoga at Heart [127-129] 4. Sūkṣma Yoga at “Nose Tip” [130-131] 5. Mantra Recitation [132-138] 6. Request to Depart and Dissolving the Maṇḍala [139] B. Branch Practices I. Dedication and Aspiration [140] 192 II. Instructions in Post-Meditation Practices 1. Maintain Identity as Main Deity, World as Cittavajra etc. [141-143] 2. How to Eat [144-146] 3. When Enjoying the Sense Pleasures Make Offerings [147] 4. Perform All Actions Within Equipoise of Being the Deity [148] 5. All Acts of Body and Speech are Forms of Mudrā and Mantra and [149-151] 6. Rite for Mending Samaya [152-153] 7. How to Sleep [154] 8. How to Wake Up [155] 9. How to Receive Accomplishment [156] 10. How to Avert Obstacles during the Rite for Receiving Accomplishment [157] III. Conclusion 1. Presentation of Nature of the Saṃsāric Predicament and its Remedy, According with Reality [158-161] 2. Injunction to Practice Deity Yoga [162] 3. Author’s Dispensing with Pride [163-164] 4. Dedication with Signature Line [165] Summary of Contents of the Caturaṅga/Samantabhadra-sādhana Preliminaries The Caturaṅga/Samantabhadra-sādhana begins with Buddhajñānapāda’s homage to Mañjuśrī (verse 1), and his commitment to compose the sādhana on the basis of a request from his spiritual teacher (verse 2). Vaidyapāda and Samantabhadra identify the spiritual teacher who requested the sādhana as Buddhajñānapāda’s guru Pālitapāda, while Vaidyapāda also provides the alternative interpretation that this could also refer to Mañjuśrī himself, who in the visionary encounter described in the Dvitīyakrama commanded Buddhajñānapāda to compose a generation stage sādhana of the Guhyasamāja-tantra.39 These textual preliminaries are followed by several verses in which Buddhajñānapāda gives the injunction that a qualified disciple—one who is compassionate, has aroused bodhicitta, who practices generosity, and who has received the proper initiations—should take up the practice of the sādhana of Mañjuśrī (verses 3-6). Our commentators do not indicate whether this short section is considered part of the Ādiyogasamādhi, or whether the samādhi proper begins with the next verse that begins to describe the actual meditation. The meditation proper begins with the practitioner visualizing a radiant syllable maṃ, Mañjuvajra’s seed syllable, in his heart center and bringing to mind the “three gates of liberation” (rnam thar sgo gsum), a series of contemplations on emptiness (verse 7). He then visualizes the buddhas of the three times (verse 8) in front of whom he will make offering and confess his past negative actions (verses 9-10). He rejoices in the merit that has been accumulated by others and dedicates that merit to the awakening of all beings (verse 11). The practitioner then takes refuge in the Three Jewels—the Buddha, Dharma, and Sangha (verses 12- 14)—and arouses the altruistic attitude of bodhicitta, the intention to establish all sentient beings in the state of awakening (verses 15-17). This group of verses, beginning with that on confession, seems to have become a popular liturgical set, as they are cited (without attribution) 39 Samantabhadrī-ṭīkā, D 131b.5; Sāramañjarī, D 2b.7; Szántó unpublished, 4. In verse 367 of the Dvitīyakrama Mañjuśrī does, indeed, command Buddhajñānapāda to compose a generation stage sādhana for the practice of the Guhyasamāja-tantra, and in verse 377, Buddhajñānapāda himself mentions composing some sādhanas at the instigation of his guru Pālitapāda. 193 in a number of Nepalese Buddhist Sanskrit manuals, as well as transliterated and translated, as a set, into Chinese in the Miaojixiang pingdeng bimi zuishang guanmen dajiaowang jing (Taishō 1192).40 Generation of the Rakṣacakra and Celestial Palace What follows are some preliminaries to the actual generation of the deity, which again begin with a contemplation of emptiness by means of the three gates of liberation, this time along with the recitation of the mantra oṃ śūnyatājñānavajrasvabhāvātmako’ham (verse 18). Then the practitioner is to contemplate the protective space in which the practice will unfold. In the sādhana itself this is described simply as a ground made of vajras. The commentaries, however, outline a much more detailed process that involves visualizing a ten-spoked wheel on which ten wrathful protective deities are visualized (the so-called rakṣacakra), though Vaidyapāda notes that the details of this ritual are to be learned, not from a textual source, but from the oral instructions.41 This process of generating the rakṣacakra is also elaborated in detail in Śākyamitra’s Mukhāgama, which he indicates is based on Buddhajñānapāda’s oral instructions. This contemplation is followed by the visualization of the dharmodaya—the “source of phenomena”— in the form of a white triangle that has emerged from Samantabhadra. At the center of this triangle is a lotus and a crossed vajra (verses 19-20). It is here that a “maṇḍala”—understood by the commentators to refer to the deity’s celestial palace—is visualized. This palace emerges from Vairocana and consort, who have themselves emerged from a cakra, which was produced from the syllable bhrūṃ (verse 21). The celestial palace is described in the sādhana some detail, including the pure correspondences (viśuddhi) of its architecture with the positive qualities of the Buddhist path and of the state of awakening (verses 22-28). Generation of the Causal Deities The generation of the principal deity proper, and thus the process of “lesser sevā” the first branch among the four branches of the “lesser” grade, begins at this point in the sādhana.42 The generation of the practitioner-as-deity in the Caturaṅga/Samantabhadra-sādhana involves a two-fold generation process, first of the “causal” deity—referred to by the commentators both as Vajrasattva, a common epithet for the causal deity, and as Vajradhara. This causal deity is actually a pair of deities in union who function as the progenitors of the main deity of the sādhana, Mañjuvajra, and his consort. This causal pair is generated, according to the commentators, by means of the process of the five manifestations of awakening (pañcākarābhisambodhi), a procedure for deity generation already known from the Yoga tantras. The process of these five stages is, however, not fully laid out in the text of the Caturaṅga/Samantabhadra-sādhana itself; one must have recourse to the commentaries for the full details. The causal deity is described as “arisen from the seed-syllable of Paramādya” (paramādyahṛdayasambhavam),43 who is identified in the commentaries as Vajrasattva or Vajradhara—terms which again seem to simply be used as epithets of the causal or progenitor 40 Tanaka 2017, 30-31. 41 See Samantabhadrī-ṭīkā, D 139b; Sāramañjarī, D 10b; Mukhāgama, D 19a.1-20b.4. 42 The Indic commentaries actually do not identify the point at which this sevā starts, but they do point out where it ends. In a much later (18th-century) Tibetan commentary by Jamgön Kongtrül it is mentioned that this first part of the lesser grade of sevā includes both the generation of the causal deities as well as the resultant deities, so I have followed that presentation here (Kongtrül 2008, 77). 43 For the Sanskrit edition of this verse see Kano 2014, 66. 194 deity (verses 29-31).44 This causal deity is white in color with three faces—the right black, and the left red—and six arms, two of which embrace his consort and the other four of which hold a vajra, sword, lotus, and jewel (verses 32-33). He is described as the “body who produces all of the Victors” (jinajanakatanum) and the yogin is instructed to visualize himself as this deity (verse 34). This factor of self-visualization as the progenitor deity is an unusual feature of Buddhajñānapāda’s sādhana, but one which is preserved in the later works of the Jñānapāda School; in most sādhanas pertaining to other traditions that employ the causal deity method of generation, the causal couple is visualized in front of the practitioner.45 I discuss some of the implications of this unusual feature of the sādhana below. Unique Preliminaries: Purifying Embodied Beings and the Inanimate World and Generating them as the Deity What follows are a series of preliminaries to the process by which the main deities of the sādhana, Mañjuvajra and consort, are “born” from the causal deities, and which seem to constitute yet another unique feature of the Caturaṅga/Samantabhadra-sādhana. This begins with the buddhas entering into the body of the practitioner-as-causal-Vajradhara (verse 35). Then, the nine goddesses of the maṇḍala—Locanā and the rest of the four buddha consorts, along with Rūpavajrā and the rest of the six sense goddesses—are emanated out from his body and dissolve into the body of his consort (verse 36). A maṇḍala of rays of “stainless moonlight” emerges into her lotus, and the practitioner then mentally places all sentient beings in that maṇḍala (verse 37). Then, Vairocana and the other buddhas enter into the maṇḍala “in the form of the fluid moon” (indudravarūpa)—that is, in the form of bodhicitta—and confer initiation upon beings there (verse 38). After this initiation, the sādhana simply describes the emanation, by means of their seed syllables, of a number of deities (verses 39-42). First are the six bodhisattvas, Kṣitigarbha, etc., who are the pure embodiment of the sense faculties—the eyes, and so forth—but who do not form part of the main set of the sādhana’s maṇḍala deities; in the generation of the main maṇḍala later in the sādhana, these bodhisattvas are visualized on/as the sense faculties of the main maṇḍala deities (verse 39). Next are the six goddesses, Rūpavajrā and the rest, who are the purification of the sense objects, and who do form part of the 19-deity maṇḍala generated later in the sādhana (verse 40a-b). Then Locanā and the rest of the four buddha consorts who are the purified forms of earth and so forth, the four elements (verse 40cd), are emanated out. They, too, constitute part of the main 19-deity maṇḍala. Finally come the buddhas, Vairocana and the rest of the five buddhas, who are the purified form of form and other five aggregates (verse 41). The five buddhas also constitute members of the main maṇḍala-cakra that is generated later in the sādhana (but, as we shall see, when they are generated as part of the main maṇḍala Akṣobhya merges with Mañjuvajra, as he does not have a separate seat in the maṇḍala). These deities—that is, all of the main deities of the 19-deity maṇḍala with the exception of the main deity Mañjuvajra (who is here represented by Akṣobhya), his consort, and the four gate guardians, and with the addition of the six bodhisattvas who mark the sense faculties of all the maṇḍala deities—are emanated by means of their individual seed syllables (verses 42-43). 44 Vaidyapāda (Samantabhadrī-ṭīkā, D 143b.4) identifies him as Vajrasattva, as does Samantabhadra in the long Sanskrit recension of the Sāramañjarī (Szántó unpublished, 43). The Tibetan translation of the Sāramañjarī here reads rdo rje ‘chang, Vajradhara (Sāramañjarī, D 15a.1). 45 Schwind 2012, 76. This feature of self-identification with the causal deity is also seen in several sādhanas preserved at Dunhuang. However in those sādhanas the practitioner then somehow “steps back out of that form to be absorbed back into it and become the resultant deity” (Jacob Dalton, personal communication). 195 While the Caturaṅga/Samantabhadra-sādhana itself states clearly that all sentient beings are ushered into the maṇḍala at the point of union where they are initiated or consecrated by the buddhas, the sādhana does not elaborate on the nature of the deities that are subsequently emanated out from this maṇḍala. The commentaries, however, elaborate on this point rather significantly. Samantabhadra’s commentary explains that after the initiation within the maṇḍala located in the casual consort’s lotus, it is sentient beings themselves who, having been purified by means of the process of transforming them into the deities like Kṣitigarhba and so forth (by means of their seed syllables), are emanated out from the celestial palace in the lotus of the causal consort.46 Vaidyapāda similarly explains that during the initiation that takes place in the maṇḍala at the point of union, the dharmas of all sentient beings “melt as the moon” (zla bar zhu) and then transform into the various syllables, which transform into symbolic implements, which transform into the deities, who are then emanated out from this maṇḍala.47 That is, the commentaries describe a process by means of which the entirety of the constituent parts of beings as well as the inanimate world—the sense organs of beings, the sensory objects, the four great elements, and the five aggregates of beings—are dissolved into fluid through the initiation and re-emerge in their pure forms as the maṇḍala deities. What is not made clear in either the sādhana or the commentaries is the location to which these sentient beings-purified-as-deities are emanated. Indeed, these very same maṇḍala deities are generated again in the main generation stage of the sādhana where they are “born” from the union of Mañjuvajra and his consort and emanated out into the maṇḍala. This issue of their “double emanation”—first being emanated out as deities who are purified sentient beings from the union of the causal deities here, and later being emanated out as the maṇḍala deities who are buddhas born from the resultant deiies in the main generation stage of the sādhana—is, to my knowledge, not addressed in the commentarial tradition.48 However, the entire process of the drawing in, purification, and emanation out in purified form of the beings and the inanimate world at this point in the sādhana is, to my knowledge, a unique feature of the Caturaṅga/Samantabhadra-sādhana. I discuss the function and doctrinal implications of this unusual feature of the sādhana in more detail below. Generating the “Resultant” Principal Deity: The “Lesser” Four Branches The next step in the practice is the generation of the practitioner as the main deity of the sādhana, the “resultant deity” Mañjuvajra, together with his consort. This involves the practitioner’s mind—specified in the commentaries as the antarabhavacitta, the mind of the socalled “intermediate state,” the “bardo” state between death and rebirth—being visualized in the form of the syllables oṃ, āḥ, and hūṃ, concealed between two syllables hōḥ.49 The practitioner’s mind, in the form of these syllables, enters into the causal deity—the commentaries specify that it enters through his mouth and travels out via the vajra path—and emerges into the lotus of the causal consort where, due to her passion, it melts into liquid (verse 44). The light that emerges from the melting of the syllables thus causes the causal deity and consort themselves to “become the liquid moon,” that is, to melt into the form of bodhicitta, to which the goddesses Locanā and 46 Sāramāñjarī (see Szántó unpublished, 48). The important details included in this section of the Sāramañjarī are not present in the Tibetan translation of the text. 47 Samantabhadrī-ṭīkā, D 145a.7-145b.6. 48 It is definitely not addressed in either Vaidyapāda’s or Samantabhadra’s commentary. 49 As I mentioned above and discuss further below, the generation stage process re-envisions the ordinary processes by means of which rebirth into saṃsāric existence takes place in a purified form. The reference to the practitioner’s mind as being the antarabhavacitta, the mind of the being in the intermediate state prior to taking rebirth makes this point quite explicit. 196 the rest sing in order to incite this bindu of bodhicitta to emerge as the deity (verse 45).50 The passionate songs that they sing are taken directly from Chapter 17 of the Guhyasamāja-tantra (verses 46-49).51 In response to the goddesses’ inspiring songs, light rays from the seed-syllable maṃ turn all of the buddhas into forms of Mañjuvajra, which are then dissolved into the practitioner, who thus instantly becomes Mañjuvajra (verses 50-51).52 He is red in color like saffron, with three faces—the right face is black and the left, white—and has six arms, two of which embrace his consort and the rest of which hold a sword, an arrow, a blue utpala, and a bow (52-53). The practitioner is instructed to recognize that phenomena are free from subject and object and to perform the blessing of reciting the mantra oṃ dharmadhātusvabhāvāthao’haṃ (verse 54). Next comes the branch of lesser upasādhana, through which seed syllables for the bodhisattvas Kṣitibarbha and the others—which the commentators tell us emerge from the maṃ at Mañjuvajra’s heart center—transform into those bodhisattvas who then fill the eyes and other sense organs of Mañjuvajra (verse 55a-b). This is followed by the branch of lesser sādhana, through which Mañjuvajra’s mind, speech, and body (in that order) are consecrated with the syllables hūṃ, āḥ, and oṃ, respectively, which become Cittavajra, Vācvajra, and Kāyavajra53 at his heart, throat and crown, respectively (verses 55c-56). Rays of light from their hearts and from the point of union of Mañjuvajra and consort make offerings to the Victors, who then bestow blessings of vajra mind, speech, and body (verse 57-63). The verses of supplication for the bestowal of these blessings are drawn directly from Chapter 12 of the Guhyasamāja-tantra, and are punctuated in the sādhana with mantras through which the practitioner self-identifies with vajra mind, speech, and body.54 This is followed by the branch of lesser mahāsādhana in which Mañjuvajra is consecrated by a gathering of goddesses arisen from light rays that have emerged from the buddhas, who themselves were invoked with light rays from Mañjuvajra’s heart center. The consecration takes place with water from vases poured by the goddesses, resulting in Mañjuvajra’s becoming ornamented by his “family lord”—the presiding buddha of his “family”—who the commentators identify as Akṣobhya (verses 64-65). Generating the Consort: The “Middling” Four Branches The next step in the generation process is the “middling” four branches, applied to Mañjuvajra’s consort, who is not named in the Caturaṅga/Samantabhadra-sādhana, nor in Vaidyapāda’s nor Samantabhadra’s commentaries.55 The sādhana simply states that the 50 Neither the sādhana nor the commentaries explain where these four goddesses came from, but Vaidyapāda, in his commentary on a section of the Dvitīyakrama that gives a very condensed version of the generation stage, explains that the four goddesses are transformations of the four boundless attitudes. He writes, “By means of the three syllables entering, and so forth, [the causal deities] melt into light and then the previously mentioned four boundless attitudes transform into goddesses. In response to their inciting, this [light] transforms into the seed syllable of one’s deity.” yi ge gsum ‘jug pa la sogs pas ‘od du zhu ba las gong gi tshad med pa bzhi lha mor gyur/ de bskul bas rang gi lha’I sa bon du gyur/ (Sukusuma, D 114a.3; P 137a.6-7). 51 = Guhyasamāja-tantra, 17.72-75. 52 The commentators describe this second-generation process as again taking place via the five manifestations of awakening, but the sādhana itself does not indicate this. 53 That is, Akṣobhya, Amitabha, and Vairocana, respectively. 54 The verses are Guhyasamāja-tantra 12.74-75 for vajra mind, Guhyasamāja-tantra 12.72-73 for vajra speech, and Guhyasamāja-tantra 12.70-71 for vajra body. 55 Tanaka (1996, 260) identifies Mañjuvajra’s consort as Māmakī but does not note the source of this information. Tanaka 2009 again identifies Mañjuvajra’s consort as Māmakī, explaining this with the statement that the Mukhāgama (which Tanaka regards as Buddhajñānapāda’s work; as I have discussed already, I hold it to be Śākyamitra’s composition) “clearly explains that Akṣobhya has Māmakī as his consort” (Tanaka unpublished English translation of Chapter Five of his 2009 work, 289; 289n38). Indeed, the Mukhāgama does identify 197 practitioner is to visualize that from his own seed-syllable emerges a consort who resembles the main deity (verse 66 a-b). Vaidyapāda, however, elaborates that this process takes place through the same procedure as the generation of Mañjuvajra—she is generated first through the process of the pañcākarābhisambodhi, then the three syllables enter her mouth, and she dissolves into light and is incited by the songs of the goddesses upon which she emerges via the syllable maṃ, then a symbolic implement, and finally in her form as Mañjuvajra’s consort.56 The branches of upasādhana, sādhana, and mahāsādhana—through which her eyes and other sense organs are filled with Kṣitigarbha and the other bodhisattvas, her body, speech, and mind are consecrated, and she is crowned with the lord of her family—are expressed succinctly in just two lines (verse 66 c-d). The consort is then consecrated, through placing syllables symbolizing the five buddha families at five points on her body, and the transformation of the syllables āḥ and hūṃ, respectively, into the pericarp and petals of her lotus (verse 67-68). The practitioner-as- Mañjuvajra then engages in sexual union with his consort, with the thought that doing so is pleasing to the tathāgatas (verse 69) and recites the mantra oṃ sarvatathāgatānurāganavajrasvabhāvātmako ‘ham. This completes the first of the three samādhis, the ādiyoga-samādhi. Generating the Maṇḍala Deities: The “Greater” Four Branches The second samādhi, the maṇḍalarājāgrī-samādhi consists of the “greater” stage of the four branches, those pertaining to the maṇḍala deities. First is the greater sevā, the generation of the maṇḍala deities, which takes place by means of their “birth” from Mañjuvajra and his consort, who, as we saw above, have just begun to undertake the act of union. In this sequence light rays from the seed-syllable in Mañjuvajra’s heart summon the buddhas, who enter into Mañjuvajra and emerge from his vajra and into the consort’s lotus in the form of bodhicitta (verse 70). These buddhas-as-bodhicitta are then imagined taking on the form of the maṇḍala deities and are thus emanated out in order to benefit the world (verse 71). The emergence of the individual deities follows, starting with the five buddhas. Cittavajra—that is to say Akṣobhya—is summoned with the mantra vajradhṛk (verse 72). He is drawn in and made to enter into oneself as Mañjuvajra (verse 73). All of the other deities who will be emanated take seats at various places in the maṇḍala. The dissolving of Akṣobhya into Mañjuvajra indicates that Mañjuvajra here functions as an embodiment of Akṣobhya, the representative of the vajra family at the center of the maṇḍala. Kāyavajra—that is, Vairocana—is summoned with the mantra jinajik and placed in the east (verse 74). Ratneśa—another name for Ratnasambhava—is summoned with ratnadhṛk and placed in the south (verse 75). Amitābha is summoned with the mantra arolik and placed in the west (verse 76). Amoghasiddhi is summoned with prajñādhṛk and placed in the north (verse 77). All of the five buddhas are adorned with beautiful locks, a crown, and jewels, and are to be visualized seated in union with their consorts on sun discs (verse 78). The four buddha consorts are the second group of deities to be summoned by their mantras. First is Locanā, who resembles Vairocana, summoned with the mantra moharatī and placed in the southeast (verse 79). Māmakī, resembling Akṣobhya, is summoned with dveṣaratī and placed in the southwest (verse 80). Pandaravaṣinī, resembling Amitabha, is summoned by rāgaratī and placed in the northwest (verse 81). Tārā, resembling Ratnasambhava, is summoned by vajraratī and placed in the northeast (verse 82). These consorts hold a wheel, red utpala, Akṣobhya’s consort as Māmakī, and in the Caturaṅga/Samantabhadra-sādhana Akṣobhya merges with Mañjuvajra after he is generated; this latter point is, however, less clear in the Mukhāgama. I am grateful to Dr. Tanaka for sharing with me his unpublished English translation of Chapter Five of his 2009 publication in Japanese. 56 Samantabhadrī-ṭīkā, D 151a.6-151b.2. Samantabhadra also summarizes this process (Sāramañjarī, D 21a.6- 21b.2; Szántó unpublished, 63). 198 lotus, and a yellow utpala, and otherwise appear like the lords of their families (verse 83).57 Next, the six goddesses representing the sense objects, Rupāvajrā and the others, are generated by means of the three syllables oṃ, āḥ, and hūṃ and placed in the four corners of the maṇḍala and at the two sides of the main door. They also are similar in appearance to Akṣobhya and the other buddhas, in accordance with their particular buddha families (verse 84). Finally, the wrathful gatekeepers who occupy the four gates of the maṇḍala are summoned. Yamāntaka, who resembles Akṣobhya is summoned with yamāntakṛt and placed at the eastern gate (verses 85-86). Aparājita (or Prajñāntaka), resembling Vairocana, is summoned with prajñāntakṛt and placed in the southern gate (87). Hayagrīva, resembling Amitabha, is summoned with padmāntakṛt and placed in the western gate (88-89). Amṛtakuṇḍalin, also resembling Akṣobhya, is summoned with vighnāntakṛt and placed at the northern gate (verse 90). These wrathful deities hold a club, staff, lotus, vajra, and other implements and their crowns are adorned with Akṣobhya, Vairocana, Amitabha, and Amoghasiddhi, respectively (verse 91). Having thus generated all of the deities of what is generally called the visualized “samaya maṇḍala,” the deities of the “wisdom maṇḍala” are then invited in order to merge with and consecrate this visualized maṇḍala. This procedure involves summoning the sugatas with hooklike rays of light from the seed-syllable at Mañjuvajra’s heart, which, according to the commentaries, incites the four gate guardians to carry out the work of inviting, binding, and pleasing the wisdom maṇḍala, which they are then made to protect (verses 92-94b). The commentaries add the branch of greater upasādhana in which the deities’ sense organs are filled with the bodhisattvas, as above, and the branch of greater sādhana in which the deities’ body, speech, and mind are consecrated. Then initiation is conferred, and the deities are crowned by their respective family lords, which the commentaries note constitutes greater mahāsādhana (verse 94c). What follows are a series of verses in which offerings are made to the deities of the maṇḍala. Here, the ten goddesses of the maṇḍala are drawn into Mañjuvajra, emitted as bodhicitta in the consort’s lotus, and emanated out through the consort’s pores as light rays with goddesses at their tips who are beautifully adorned and who make offerings to the buddhas of the maṇḍala. The practitioner is instructed to engage in this practice of offering while remaining free from the concepts that there is someone who is offering, an offering being made, and the like (94d-100). This sequence is concluded with the mantra oṃ sarvatathāgatapūjāvajrasvabhāvātmako ‘ham. Then, verses of praise, taken from the Guhyasamāja-tantra, are made to five buddhas of the maṇḍala (verses 101-106).58 The final two verses of the maṇḍalarājāgrī-samādhi are the verses for “tasting the nectar,” in which the practitioner visualizes making an offering of the five nectars to the deities of the maṇḍala who consume it through vajra-tongues marked with the syllable hūṃ and are satisfied by the offering (verses 107-108). Activities Performed While Visualized as the Deity: The Karmarājāgrī-Samadhi The third and final samādhi is the karmarājāgrī-sāmadhi, which involves a series of practices that the practitioner performs while maintaining his self-visualization as the deity and maṇḍala. The karmarājāgrī-samādhi of the Caturaṅga/Samantabhadra-sādhana consists primarily of the practice of several bindu yogas, along with the practice of mantra recitation, but it also contains a somewhat elaborated philosophical excursus of the sort not usually found in the middle of a sādhana. It is in this samādhi that we see the peculiarity of what appear to be 57 The fact that the four buddhas surrounding Mañjuvajra are said to be in union with their consorts at the cardinal points of the maṇḍala, and that those very same consorts are also said to be seated in the intermediate directions of the maṇḍala is not addressed in either the sādhana nor the commentaries. 58 Verses 101-106 of the sādhana are drawn from the Guhyasamāja-tantra 17.1-5. 199 perfection stage practices included in what is generally regarded as a generation stage sādhana. The philosophical section is preceded by a short bindu yoga practice in which the practitioner emanates buddhas out from the seed-syllable in his/Mañjuvajra’s heart center. Those buddhas clear away sentient beings’ ignorance, transform them into buddhas, and usher them back into the seed-syllable at Mañjuvajra’s heart center (verse 109). The philosophical section immediately follows this verse. Vaidyapāda tells us that the arguments therein are meant to refute the concern that it would not be possible for all that appears and exists to enter into the bindu if the outer world were materially existent,59 and Samantabhadra gives a similar reason.60 The verses that follow closely parallel several of the arguments set out in Buddhajñānapāda’s Ātmasādhanāvatāra, and state that neither outer phenomena nor an inner apprehending subject are truly established. This section also includes a set of verses that set out the pure correspondences (viśuddhi) of many of the maṇḍala deities. Some of these correspondences are the same as those given in the section of Buddhajñānapāda’s Ātmasādhanāvatāra that contains passages taken (without attribution) from Vilāsavajra’s Nāmamantrārthāvalokinī.61 At the conclusion of this section of the sādhana Buddhajñānapāda explains that the arguments were set forth here in order to prevent the practitioner from fixating on any type of distinctions (verses 110-126). Following this excursus, the sādhana returns to the practice of the bindu yogas. The text instructs that, having considered all things to exist in a way that accords with the investigations just performed, the practitioner should proceed with the meditation of “his own mind” in the form of a “mantra bindu” in the center of the symbolic implement in his/Mañjuvajra’s heart center (verse 127). This bindu emanates out light that illuminates the body, speech, and mind of the yogin visualized as Mañjuvajra, and then dissolves back into the bindu, causing a downward flow of nectar from the bindu (verse 128). Subsequently, the light from the bindu again fills the interior of the body and then radiates out from each of Mañjuvajra’s pores, first filling the area of the maṇḍala and then extending even further (verse 129). The commentaries explain that this light is to be gathered back in again and made to dissolve into the bindu, again inspiring the production of nectar, and that this entire process is to be engaged in repeatedly. The next yoga described in the sādhana is the practice called the “subtle yoga” (sūkṣma yoga), which involves meditating on the entire maṇḍala of deities inside of the deity’s symbolic implement, this time at the tip of the “[lower] nose,” that is to say at the tip of the yogin’s vajra (i.e. the penis) (verse 130). The commentaries explain that this is effected by first emanating out light rays from the seed-syllable at Mañjuvajra’s heart center which draw in the tathāgatas, and melting them into bodhicitta, which descends into the center of the consort’s lotus where it transforms into a maṇḍala in the center of the deity’s symbolic implement; in the instance of Mañjuvajra, the symbolic implement is a sword. The practitioner then visualizes the tip of his own vajra as hooked light rays which draw the maṇḍala from the center of the consort’s lotus to the “nose tip” of the lotus.62 The sādhana indicates that when the practitioner sees the “signs of stability” with respect to this practice he should repeatedly emanate out forms of the buddhas together with their symbolic implements (verse 131). The commentaries identify these signs of stability as a series five experiences that indicate the dissolution of the elements—seeing something like a mirage, 59 Samantabhadrī-ṭīkā, D 161a.3-4. 60 Sāramañjarī, D 27b.7-28a.2; Szántó unpublished, 97. 61 Not all of the correspondences are identical, of course, given that the maṇḍala described in the Nāmamantrārthāvalokinī (and thus in the Ātmasādhanāvatāra) is that of the Mañjuśrīnāmasaṃgīti, not the Guhyasamāja maṇḍala present here in the Caturaṅga/Samantabhadra-sādhana. 62 Vaidyapāda specifies that these are the practical instructions (man ngag) for the practice (Samantabhadrī-ṭīkā, D 168a.2-4.) See also Sāramañjarī, D 37b.6-38a.3 and Szántó unpublished 125, where Samantabhadra gives the same instructions. 200 smoke, a bright sky (or fireflies)63, a flickering lamp, and a cloudless sky—that are mentioned in the Dvitīyakrama (verses 196-99) as unfolding in the context of perfection stage practice.64 The mention of the appearance of these signs in relation to a practice from the Caturaṅga/Samantabhadra-sādhana is again indicative of the overlap between the generation and perfection stages in Buddhajñānapāda’s works. The next section of the sādhana describes the ritual for mantra recitation, in which the practitioner is to acknowledge all phenomena as having the identity of his own mind. Then he meditates on his mind in the form of his mantra—which the commentators indicate to mean his seed-syllable—radiant with light. He emanates this light outwards and uses it to draw back in the forms of the maṇḍala deities while reciting the mantra, in connection with the inhalation and exhalation of his breath (132-134b). Then offerings are made “as before,” which the commentators specify as following the same procedure from the maṇḍalarājāgrī-sāmadhi, in which the goddesses are drawn into the consort’s lotus and then emanated out of her pores, make offerings to the tathāgatas, and then dissolve back into the consort (verse 134c-d). These tathāgatas who have received offerings are then gathered back in, and dissolve into the body of the yogin-as-Mañjuvajra, while he experiences the bliss of being joined in union with his consort (verse 135). The sādhana specifies that if the practitioner’s body or mind becomes weak or tired through the practice, the consorts, Locanā and the rest, sing songs as before to rouse the practitioner-as-Mañjuvajra, and offerings are again made just as before during the maṇḍalarājāgrī-samādhi (verse 136). Then, bodhicitta is drawn in by means of the breath, which takes the form of hooked light rays, and made to descend into Vairocana and the rest “according to the ritual” in order to dispel weariness or passion (kheda). The commentators describe this ritual of drawing in bodhicitta with the breath as involving the visualization of a moon disc the size of a handspan, located a handspan’s height above the practitioner’s head, where the syllable oṃ rests above the syllables of the five nectars. The light from these hooks draws in the tathāgatas of the five families who dissolve into syllables and melt, and this liquid is then brought down into Vairocana and the other buddhas by means of the hooked light rays of the inhalation and exhalation of the breath, filling their nāḍīs and satisfying them (verse 137). The yogin should remain in this practice for a moment, while continually reciting the mantra (verse 138). Having completed these yogas, the practitioner should then make the request for the deities of the wisdom maṇḍala to depart by first repeating the offerings and praises (and the tasting of the nectar, adds Vaidyapāda), and then dissolving the maṇḍala deities into the syllables oṃ, āḥ, and hūṃ (Vaidyapāda gives details on which deities dissolve into which syllables), which syllables then themselves dissolve. According to Vaidyapāda, this dissolution of the maṇḍala deities marks the completion of the third of the three samādhis, the karmarājāgrī-samādhi, and therefore the formal completion of the practice of the sādhana as such (verse 139).65 Samantabhadra includes the following section of additional practices within the third samādhi and holds that the karmarājāgrī-samādhi concludes only after verse 157.66 “Branch Practices” Performed Outside of Formal Sessions 63 The Dvitīyakrama itself (verse 198) reads “like a bright sky” but Vaidyapāda’s Sukusuma specifies that this is “like fireflies,” the more common description of this particular sign (Sukusuma, D 117b.1-2). 64 These same signs are also mentioned in the Samājottara (vv. 150cd-151d). Samantabhadra cites the passage on these signs from the Samājottara in his commentary (Sāramañjarī, D 28a.3-4 Szántó unpublished,126). 65 Samantabhadrī-ṭīkā, D 170b.3. 66 Sāramañjarī, D 43b.5-6. Samantabhadra notes that because these various procedures are connected to the karmarājāgrī-sāmadhi they are also included therein. This passage absent in the longer Sanskrit recension of the Sāramañjarī. 201 What follow are, according to Vaidyapāda, several “branch practices” that are to be pursued at a number of different occasions outside of the formal practice session. The first of these post-meditative instructions, presumably to be carried out immediately following the final samādhi, instructs the practitioner, who has by means of the sādhana “seen suchness,”67 to make dedications and aspirations toward the liberation of all beings (verse 140). Having thus emerged from the samādhi, the practitioner should continue to mentally regard himself in the form of the deity, and to regard his surroundings, likewise, as deities, rather than just as ordinary forms. He should additionally regard all phenomena as being undifferentiated, sharing in the single nature of Vajradhara (verse 141-142). Freed from mistaken conceptuality, he carries out whatever activities ought to be done and constantly makes offerings to the maṇḍala-cakra (verse 143). The next instruction is about how to eat, which involves the yogin’s visualizing the seed-syllable at his heart center, incanting food with the syllables oṃ, āḥ, and hūṃ, and imagining that while eating he is satisfying the deities with the food. Consuming food is likened to the outer homa (the practice of making burnt offerings) but performed now with the mind—a mental homa. The sādhana then notes that the supreme homa is the process by means of which the fire of wisdom consumes the fuel of the aggregates (144-146). Likewise, the text indicates, enjoyment of any of the sense pleasures is to be regarded as making an offering to the deity, and engaging with the sense pleasures in this way is to be understood as a samaya (verse 147). The practitioner should engage in all physical and verbal acts while remaining in a state of equipoise (verse 148). In this way, all of his physical and verbal actions become forms of mudrā and mantra (verse 149). Since different perceptions of a given object arise due to different conceptual states, because the practitioner’s mind is pure, all of his physical and verbal acts are likewise pure (verses 150-151). Next, a series of additional rituals and practices are given. The first of these is the ritual for mending samaya if it has been broken (verse 152-153). Then follow the practices for going to sleep (verse 154) and waking in the morning (verse 155). There are instructions on the ritual for receiving accomplishment (siddhi) at the conclusion of the practice (156), and the way to avert obstacles during the rite for receiving accomplishment (verse 157). Conclusion of the Sādhana Finally, the conclusion to the composition begins by identifying conceptuality as the central cause of saṃsāric suffering. The sādhana then states that there is no conceptuality within “that which is by nature profound and exalted,” which the commentators identify as the maṇḍala-cakra (verses 158-159). Buddhajñānapāda goes on to describe the process of using a remedy—which the commentators explain to be meditation on the maṇḍala-cakra—to counteract its opposing factor, which the commentators identify as saṃsāric suffering. He then advocates meditating upon one’s own nature as Samantabhadra in order to accomplish wisdom instantly (verses 160-162). Finally, Buddhajñānapāda writes two verses dispelling his own pride with regards to the act of composing the sādhana (verse 163-164), and a single dedicatory verse, into which—like in many of his works—he has inscribed his own name, Buddhajñāna (verse 165). Some Notable Features of the Caturaṅga/Samantabhadra-sādhana Two-Stage Generation and Self-Generation as Causal Deities 67 Again this reference to the practitioner “seeing suchness” in the context of the sādhana uses language that tends, in Buddhajñānapāda’s writings, to be connected more to the perfection stage than the generation stage. And, indeed, the bindu yogas that have just been performed do seem to constitute perfection stage practices. 202 The Caturaṅga/Samantabhadra-sādhana, along with its associated commentarial literature, deserves a full study; my outline and summary if the practice above, and my comments here, remain just a preliminary investigation into this material. However, I will address here in brief several features of the sādhana that I find notable. The first point of particular interest is the process by means of which the deities are generated in Buddhajñānapāda’s sādhana. His is a relatively early example of sādhana literature in general, and much work still needs to be done on the process of development of the structure of deityyoga sādhana as a genre. As noted in Chapter Three, the process by which the practitioner is to generate herself in the form of the deity following the tradition of the Yoga tantras is through the pañcākarābhisambodhi, the “five manifestations of awakening,” the canonical source for which is the Sarvatathāgatatattvasaṃgraha-tantra68 (though there is some evidence that the sequence in the tantra is a narrativized version of a previously existing ritual procedure). In the later Yoganiruttara tantras, this generation process was expanded to encompass a two-fold process involving what have come to be called the “causal” (hetu, rgyu) and “resultant” (phala, ‘bras bu) deities, a process followed in many such later generation stage sādhanas. Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana is an early example of a sādhana that uses this formula of two-fold generation. In the Caturaṅga/Samantabhadra-sādhana the causal deity, together with his consort, is generated by means of the procedure of the pañcākarābhisambodhi, just as in earlier Yoga tantra practice, though the procedure is only partially spelled out in the sādhana, and one must rely on the commentaries for further details. It is this causal deity and consort who then produce the main deity of the sādhana Mañjuvajra, who appears together with his own consort. As already mentioned above, Buddhajñānapāda’s sādhana is notable here in that the causal deity is also a self-generation, whereas in the later works of other practice systems the practitioner is instructed to identify only with the resultant, or main, deity of the sādhana. This slightly unusual feature, which is, however, preserved in later Jñānapāda School works,69 may be a further indication that the Caturaṅga/Samantabhadra-sādhana is one of the earlier sādhanas to use this method of causal and resultant deity generation, since the idea of the resultant deity as being somehow self-produced—produced by a causal deity with whom the yogin already identifies—is slightly awkward compared to the later and more streamlined version of the process where the progenitor deities are not visualized as oneself, but oneself-as-deity is generated or “born” from them. However, this self-identification with both the progenitor deities as well as the resultant deities does fit well with Buddhajñānapāda’s assertion that the entirety of the phenomenal world is nothing other than nondual wisdom, which is itself the nature of the practitioner’s own mind. From this doctrinal perspective, which emphasizes the singular awakened nature of the outer world and the practitioner’s own mind, the practitioner’s identification with both the progenitor deities as well as the resultant deities is unproblematic, and even fitting. Purifying Sentient Beings and the Inanimate World as Deity What follows the self-generation of the practitioner as the causal deities is a sequence in which sentient beings are drawn into the visualized maṇḍala at the causal deities’ point of union, purified, and then emanated back out in a purified form. This is another process that appears to 68 As noted in Chapter Three, these five processes are clearly present in the tantra, but the term pañcākarābhisambodhi itself is not found there. Thanks to Harunaga Isaacson for pointing out to me the fact that the term itself does not appear in the tantra. 69 See Schwind 2012, 76. As I also noted above, this feature of self-identification with the causal deity is also seen in several sādhanas preserved at Dunhuang. However in those sādhanas the practitioner then somehow “steps back out of that form to be absorbed back into it and become the resultant deity” (Jacob Dalton, personal communication). 203 be unique to Buddhajñānapāda’s system.70 As I have already briefly described in the summary of the sādhana above, in this part of the visualization, first all of the buddhas are first drawn into the causal deity, and all of the female deities of the maṇḍala are emanated out and absorbed back into the causal deity’s consort. Presumably this serves to consecrate, in some sense, the causal deities, identifying them with all buddhas and buddha consorts, as well as expressing the ultimate identity of the buddhas and the buddha consorts, themselves. A maṇḍala of bodhicitta then emerges from the causal deity’s vajra into the consort’s lotus, and all sentient beings are mentally ushered into that maṇḍala. The tathāgatas then enter the same maṇḍala in the form of bodhicitta, via the causal deity’s vajra, and confer initiation upon sentient beings there. As I described above, the sādhana itself states only that the deities of the maṇḍala—with the exception of Mañjuvajra, his consort, and the four gate guardians, and with the addition of the six bodhisattvas who are not deities of the maṇḍala proper, but instead are installed on the sense organs of all the maṇḍala deities—are to be emitted from this maṇḍala by means of their seed syllables. The commentaries, however, explain that the deities being emitted from this maṇḍala are none other than the very same sentient beings who entered the maṇḍala, were purified through their consecration by the tathāgatas, and thereby transformed into deities. According to the commentaries, the sentient beings who have entered into the maṇḍala first dissolve into bodhicitta, then transform into syllables, and then into symbolic implements, and finally into deities, in which form they are emanated out from this maṇḍala, presumably back into the world from whence they came (though our commentators are silent on this latter point of their destination). The groups of deities emanated out from the maṇḍala are described in the sādhana as the pure forms (viśuddhi) of a series of sets of phenomena—the sense organs, sensory objects, elements, and aggregates—that amount to the entirety of the constituent parts of beings, and indeed of the entire world. Vaidyapāda’s commentary gives more details about the process of this transformation: beings’ sense organs dissolve into bodhicitta and emerge as Kṣitigarbha and the others of the six bodhisattvas (who, in the sādhana itself are installed in those sense organs of the maṇḍala deities); form (as a sensory object) and the rest of the sense objects dissolve into bodhicitta and emerge as Rupavajrā and the rest of the six sense-object goddesses; the earth and the rest of the four great elements dissolve into bodhicitta and emerge as Locanā and the others among the four buddha consorts; the aggregate of form (or matter, rūpa) and the rest of the five aggregates dissolve into bodhicitta and emerge as Vairocana and the others of the five tathāgatas.71 This process involves a purification of each of the constituent parts that make up beings—their aggregates and sense organs—along with the entire inanimate world—the elements and sensory objects—such that all of those constituent parts, and thereby those beings, and that world, re-emerge in the form of deities. This is, in effect, a ritual enactment of the doctrine, articulated in Chapter 17 (vv. 51-52) of the Guhyasamāja-tantra, that makes precisely the same equation of the buddhas with the aggregates, the consorts with the elements, and the bodhisattvas with the sense faculties that is made in the Caturaṅga/Samantabhadra-sādhana, and on which the sādhana’s statements are certainly based.72 70 At least I, myself, have not seen this feature elsewhere and its being unusual was confirmed to me by both Jacob Dalton (personal communication) and Harunaga Isaacson (personal communication), neither of whom are familiar with its being used in other sādhanas. 71 I have, in this section on the viśuddhi of the deities, closely paraphrased Vaidyapāda’s Samantabhadrī-ṭīkā, D 145a.7-145b.2. See also Samantabhadra’s Sāramañjarī (D 16b.5-17a.5; Szántó unpublished 48-50), which describes essentially the same process. Vaidyapāda is clearer in some places, Samantabhadra in others. 72 The six goddesses and their associate with the sense objects are not mentioned in the tantra, but their very names—Rupavajrā (“Vajra Form”), etc.—make their association with the sense objects clear. The absence of Mañjuvajra and consort among the deities that are emitted in this process is consonant with the fact that the yogin himself will emerge as Mañjuvajra in the next process of the sādhana, and the omission of the wrathful gate 204 Linking Deity Generation to Human Reproduction While the generation of the main deity, Mañjuvajra, proceeds by means of the usual method for sādhanas that employ a two-step generation process, as I noted above the Caturaṅga/Samantabhadra-sādhana is important as an early representative that employs the method of double-generation. The procedure involves the practitioner visualizing his own mind—described in the commentaries as the mind in the intermediate-state (antarabhava) between death and rebirth—in the form of syllables (in this case oṃ āḥ hūṃ) enters into the causal deity through his mouth, is emitted into the consort’s lotus and melts into bodhicitta. Due to their state of passion, the two causal deities themselves melt into a bindu of bodhicitta. They are roused back into form with songs of passionate (yet soteriologically-oriented) longing, sung by each of the four buddha consorts, in turn.73 The process by means of which the resultant deities, Mañjuvajra and consort, emerge from this bindu is explained much more simply in the sādhana itself than in the commentaries. The sādhana states that light rays from the seedsyllable maṃ turn all of the buddhas into forms of Mañjuvajra, which are dissolved into the practitioner, who thus instantly becomes Mañjuvajra. The commentarial tradition explains the emergence of the practitioner-as-deity as taking place through the five-stage process of the pañcākarābhisambodhi, through which the causal deities were previously generated.74 In any case, the generation of the practitioner-as-deity from the pair of causal deities in the two-step generation process, as well as the generation of the maṇḍala deities from the point of union of Mañjuvajra and consort, very clearly augments the non-sexualized generation process of the pañcākarābhisambodhi from the Yoga tantras with a sexualized generation process that mimics the process of human reproduction. As such, this two-fold generation procedure, which we find in sādhanas starting from the Guhyasamāja tradition onwards, marks an important development in the structure of tantric sādhana. The fact that deities in these newer tantric systems were commonly depicted in sexual union allowed for this direct homologizing of the process of self-visualization as the deity with the process of human reproduction. In fact, Śākyamitra’s Mukhāgama—which is clearly based on, but not identical to, the Caturaṅga/Samantabhadra-sādhana, and which, according to Śākyamitra, follows Buddhajñānapāda’s oral instructions—structures the process of sādhana within the framework of the unfolding of the twelve links of dependent origination (pratītyasaṃutpāda) that characterize the stages of the process of rebirth within saṃsāra, making the connection of the generation stage with reproduction and birth even more explicit.75 Earlier Buddhist traditions explained that the process of the twelve links needed to be reversed or severed in order to free the practitioner of the bonds of saṃsāra; Śākyamitra’s relating the twelve links—in their ordinary forward guardians may perhaps be due to their identification as wrathful forms of compassionate wisdom; it might be unseemly for ordinary sentient beings—even purified ones—to appear in the forms of compassionate wrath. 73 As noted above, these are GST 17.46-49. 74 It is unclear whether this represents a development in the ritual tradition between the composition of the sādhana and the composition of its commentaries—the usual process in later sādhanas does indeed repeat the sequence of the pañcākarābhisambodhi here—or whether the use of the pañcākarābhisambodhi generation process for the resultant deities, as well, is part of the oral tradition dating back to Buddhajñānapāda himself. Both of these possibilities are tenable; sādhanas as a genre—especially in when composed in verse—frequently abbreviate or omit details that are assumed knowledge, which is usually passed down through an oral or textual commentarial tradition. 75 Tanaka 2009 states that this feature of dividing the sādhana in to the twelve links of dependent origination is also shared with the Caturaṅga/Samantabhadra-sādhana, but I have seen no evidence of this either in the sādhana nor in the commentarial tradition. Tanaka gives, in this work, a summary of the process of deity generation as connected to the twelve links based on the Mukhāgama, with additional reference to these twelve processes according to the Abhidharmakośa. I am grateful to Dr. Tanaka for sharing with me his as-yet-unpublished English translation of Chapter Five of his 2009 book, which was originally published in Japanese.) 205 progression, not reversed—to the process of generating the deity and the maṇḍala indicates that the generation stage was seen as a method for freeing the practitioner from the bondage of saṃsāra precisely by re-enacting the very same stages by which he is normally bound in sāṃsāra in a pure manner, resulting in the generation of a pure wisdom deity rather than an impure sentient being.76 The Philosophy Section Another notable feature of the Caturaṅga/Samantabhadra-sādhana is the philosophical excursus that occurs in the context of the bindu yogas towards the end of the sādhana. As I mentioned above, such an extensive philosophical investigation is not common in a sādhana. Moreover, when more abbreviated philosophical reflections do occur in sādhana literature— which is not terribly uncommon—they are usually found at the beginning of a sādhana at the point where the practitioner is to bring to mind emptiness prior to generating herself as the deity.77 In that more common case, the philosophical reflections function as a sort of analytical meditation meant to assist the practitioner in bringing about an experiential sense of emptiness out of which the deity and entire maṇḍala are to be generated. At that earlier point in the Caturaṅga/Samantabhadra-sādhana, we may recall, Buddhajñānapāda included a single verse (verse 18) about the three gates of liberation, presumably meant to assist the practitioner in bringing to mind the empty nature of phenomena prior to reciting the mantra oṃ śūnyatājñānavajrasvabhāvātmako ‘ham and visualizing the emergence of the maṇḍala out of the state of emptiness. Here, however, in the context of the bindu yogas, he has included eleven verses (verses 110-120) refuting the existence of materially existent outer phenomena, as well as an inner apprehending consciousness or mind that could perceive them. The verses also state that the mere reversal of the concept of self constitutes the purification of the saṃsāric state and emphasize the identity of saṃsāra and nirvāṇa from the perspective of their ultimate nature. As noted above, Vaidyapāda explains that the function of these verses is to help the practitioner overcome the concern that, were the world materially existent, it would be impossible for all appearance and existence to enter into the bindu;78 Samantabhadra’s explanation of the verses’ function is very similar.79 It seems, then, that these verses in the Caturaṅga/Samantabhadrasādhana were meant to serve as a kind of philosophical justification, underpinning, or support for the practices of the bindu yogas of the “perfection stage of the generation stage” found in the sādhana (and perhaps also for the practices of the “generation stage of the perfection stage” in the Dvitīyakrama and Muktitilaka, as well). In the Ātmasādhanāvatāra Buddhajñānapāda sets forth a philosophically-based defense of the practice of deity yoga against an unidentified interlocutor who suggests that deity yoga amounts to nothing more than the superimposition of the idea of the form of a buddha on the body of the practitioner, a position that Buddhajñānapāda rejects.80 Here in the Caturaṅga/Samantabhadra-sādhana, then, it seems that Buddhajñānapāda 76 Tanaka 2009 suggests that it was Buddhajñānapāda who “introduced sexology into the visualization of the maṇḍala.” The Caturaṅga/Samantabhadra-sādhana is the earliest sādhana of which I am aware to use this twostage generation process that employs the causal and resultant deities, but I am hesitant to attribute such a significant development in the process of sādhana to a single individual. There are, for example, sādhanas at Dunhuang that also use this model, and which seem to represent a pre-Buddhajñānapāda stage of development, or at least to show no knowledge of a distinction between the Jñānapāda and Ārya Schools (Jacob Dalton, personal communication). 77 See Isaacson 2007, 292. 78 Samantabhadrī-ṭīkā, D 161a.3-4. 79 Sāramañjarī, D 27b.7-28a.2; Szántó unpublished, 97. 80 The body maṇḍala debate between Khedrub Je and Ngorchen Kunga Zangpo in 15th-century Tibet studied by Yael Bentor (2015) and Rae Dachille (2015) appears to be a debate over a very similar question some six centuries later. Thanks to Jacob Dalton for drawing my attention to this. 206 is providing a similar set of philosophically-based arguments (some of them, in fact, parallel to those found in a different section of the Ātmasādhanāvatāra) for the practices of the bindu yogas in his practice system. A key aspect of these arguments, which is again mentioned later in the sādhana in the section on mantra recitation (verse 132), is the fact that all phenomena must be understood as being nothing other than the practitioner’s mind. Perfection Stage Yogas in “Generation Stage” Sādhana Yet another noteworthy aspect of the Caturaṅga/Samantabhadra-sādhana that was already mentioned above is the inclusion in the sādhana of bindu yogas that seem clearly to be related, in some manner, to the perfection stage. It is possible that Vaidyapāda may be referring to precisely these bindu yogas from the Caturaṅga/Samantabhadra-sādhana when he identifies the second of the two aspects of the generation stage mentioned by Buddhajñānapāda in the Muktitilaka as the “perfection stage of the generation stage.” The section of the Caturaṅga/Samantabhadra-sādhana in which the bindu yogas appear is the third of the three samādhis, the karmarājāgrī-samādhi. There are, in fact, several indications that the practice of the karmarājāgrī-samādhi, or at least part of that samādhi, may have been considered in Buddhajñānapāda’s system to relate or pertain to the perfection stage. One of these indications is found, not in the Caturaṅga/Samantabhadra-sādhana, but in the Dvitīyakrama. In that text, just before the instructions on the three bindu yogas, we find five verses (vv. 156-160) that summarize the practice of the generation stage, but only up through the practices included in the second samādhi, the maṇḍalarājāgrī-samādhi. Vaidyapāda’s commentary confirms this reading of these verses, identifying which of the practices mentioned in the verses pertain to each of the first two samādhis.81 The verses culminate in the injunction for the practitioner to “accustom oneself to this through training”—which Vaidyapāda explains as meaning “to apply oneself assiduously to the two yogas,”82 still referring, presumably, to the first two samādhis. The Dvitīyakrama goes immediately on to explain that having so trained, this the practitioner should then “train in the ultimate suchness/ That is the buddhas’ supreme sphere of experience,”83 and proceeds with the instructions for the three bindu yogas of the perfections stage. Given that the generation stage instructions given in these verses correspond with the practices of only the first two among the three samādhis, the samādhi that would logically follow to encompass the bindu yoga practices would be the karmarājāgrī-samādhi. And, indeed, that third samādhi is precisely where we see seen the bindu yogas in the Caturaṅga/Samantabhadra-sādhana, albeit in significantly more abbreviated form than those found in the Dvitīyakrama. Vaidyapāda introduces the bindu yogas of the Caturaṅga/Samantabhadra-sādhana’s karmarājāgrī-samādhi writing, “In order for the yogin, who has in the [preceding] manner perfected the maṇḍala-cakra, to bring his mind under control by means of training in suchness, the greatest of [all] great things, he [i.e. Buddhajñānapāda] teaches the bindu yoga.”84 As we saw in Chapter Three, “training in suchness” is a term that in Buddhajñānapāda’s practice system, is generally used to describe the second stage of tantric practice. Taken together, these passages suggest that in Buddhajñānapāda’s system the karmarājāgrī-samādhi served as the place where perfection stage practices—whether the “perfection stage of the generation stage” or “the generation stage of the perfection stage”—were 81 Sukusuma, 114a.1-7; P 137a-137b.4. 82 De la goms pa bslabs pas zhes pa ni sbyor ba gnyis la nan tan du byas pas zhes so// (Sukusuma, D 114a.7-b.1; P 137b.5) 83 sangs rgyas rnam kyi spyod yul mchog// mtha’ yi de nyid bsgom par bya// (Dvitīyakrama, verse 161b-c). 84 de ltar dkyil ‘khor gyi ‘khor lo rdzogs pa’i rnal ‘byor bas/ che ba’i che ba de kho na nyid bsgoms pas sems dbang du bya ba’i phyir thig le’i sbyor ba gsungs pa/ (Samantabhadrī-ṭīkā, D 160b-7-161a.1). 207 integrated into the already established system of deity yoga. This resulted in a system in which the two modes of practice appear not to have been practiced in isolation, but in an integrated way where the first two samādhis of the generation stage served as the framework for perfection stage practices that were integrated as part of the third samādhi. Since the bindu yoga practices in the karmarājāgrī-samādhi of the Caturaṅga/Samantabhadra-sādhana are simpler, while the practices of the first two samādhis in that sādhana are more elaborate, it is possible that this style of combining the two stages was emphasized for a practitioner towards the beginning of his practice. The more elaborate bindu yogas in the Dvitīyakrama and the Muktitilaka may have been combined with a simplified version of the first two samādhis, similar to their brief summary in verses 156-160 of the Dvitīyakrama, for a more advanced practitioner whose practice then emphasized the more elaborated yogas of the perfection stage. These comments are merely speculative, but they are based in part on later perfection stage practice manuals that often involve a more simplified form of generation; none of the early works of the Jñānapāda School are, to my knowledge, explicit about how a yogin should employ the variety of practices available in that tradition at various stages in his development as a practitioner. As I discussed above, the slight overlap of perfection stage practices into the generation stage may be reflective of the fact that the two stages were being newly distinguished at the time when Buddhajñānapāda’s system was developed. However, the inclusion of bindu yogas in the Caturaṅga/Samantabhadra-sādhana also brings up the issue of whether or not these bindu yogas were meant to be practiced while in sexual union with a partner—and the description of at least the sūkṣma yoga suggests that it was—which also raises the question of which initiations would have been necessary for the practitioner to engage in this sādhana. As I will discuss further in Chapter Seven, the initiation rituals as described in the Dvitīyakrama suggest that the kalaśābhiṣeka may have been bestowed in a separate ritual context from the guhya- and prajñājñānābhiṣekhas, but the presence of these bindu yogas in the Caturaṅga/Samantabhadrasādhana might suggest otherwise. Indeed, given the bindu yogas that form what appears to be an integral part of the sādhana, it seems unlikely that the kalaśābhiṣekha alone would have sufficed for the practice of the Caturaṅga/Samantabhadra-sādhana. The Development of Generation Stage Sādhana: Some Conclusions This brief look at the generation stage in Buddhajñānapāda’s writings, with a focus on the Caturaṅga/Samantabhadra-sādhana, serves as an indication that we still have a lot to learn about the development of sādhana as a genre, and even about a topic as basic as the division of tantric practices into two stages, and precisely what constitutes those stages according to different authors. As one of the earliest surviving authored sādhanas, Buddhajñanapāda’s work is an early example of many features of generation stage sādhana, and a more detailed study of his system—especially in relation to the sādhana-like sections found in the writings of earlier authors, like the fifth chapter of Vilāsavajra’s Nāmamantrārthalokāviṇī, and to the sādhanas of later authors—would further facilitate our understanding of the development of the genre. More study on the relationship between individual sādhanas and the tantras that provide their bases and on which they clearly draw would also be illuminating. For Buddhajñānapāda himself it seems that the generation stage served as the ritual framework for the practice of the perfection stage, which was, in his perspective, the real vehicle leading to awakening. But his attention to the details of the generation stage and his composition of multiple generation stage sādhanas indicates that it was a very important framework, and one that could also be used to accomplish other aims. Indeed, while the Caturaṅga/Samantabhadra208 sādhana contains perfection-stage-style bindu yoga and sūkṣma yoga practices, Buddhajñānapāda’s other shorter generation stage sādhanas do not. These shorter sādhanas, especially the three sādhanas of the wealth deity Jambhala, seem more focused on the use of the sādhana as a framework for the application of tantric rites for accomplishing more worldly aims, like the production of wealth. But even the short Heruka sādhana, which lacks such rituals for the accomplishment of worldly aims, is also free from any overtly perfection-stage-oriented practices. It thus appears that Buddhajñānapāda may have regarded the generation stage as having multiple functions, both the more ordinary and worldly function of achieving a variety of desired aims, as well as serving as a framework for the liberative practices of the perfection stage. Let us now turn our focus back to the Dvitīyakrama and the Muktitilaka and take a look at some of the details of the perfection stage practices that Buddhajñānapāda sets forth in those works. 209 Chapter Six The Ultimate Path to Awakening: The Perfection Stage in Buddhajñānapāda’s Writings In the space of the lotus, the jewel of the vajra and the heart of the lotus join, and in vajra posture the mind is observed within the jewel. The bliss that arises is ascertained—that itself is wisdom. This is explained by all of the genuine supreme gurus to be the perfection stage. -Mañjuśrī’s instructions to Buddhajñānapāda, Dvitīyakrama The second stage of tantric practice, the perfection stage, was extremely important for Buddhajñānapāda; he indicates at multiple places in his writings that it is only by relying on this “second stage” that a practitioner is able to reach full awakening. As I have already discussed in Chapter Five, Buddhajñānapāda was among the early authors to make a distinction between the two stages of tantric practice and thus to use the term “perfection stage.” His writings therefore give us a window into how this second stage was understood and practiced in the earliest period of its development within the tantric Buddhist tradition. In this chapter I will first examine a bit more closely the way that the term “perfection stage” and its synonym, the “second stage,” are used in Buddhajñānapāda’s writings. I will then take a look at some of the perfection stage practices that he outlines, primarily in the Dvitīyakrama and the Muktitilaka, as well as the way in which Buddhajñānapāda relates these practices to the schema of three blisses (his system lists only three, rather than the four blisses of later tantric traditions) that are said to arise in the context of these practices, and to three among the six branches of the classic “six branch yoga” (ṣaḍaṅgayoga) that are mentioned in his writings. Then, because the perfection stage practices in his system primarily consist of sexual yogas that involve practice with a partner, I will also examine what Buddhajñānapāda has to say on the topic of the tantric consort and reflect about the way that the tradition of kāmaśāstra appears to have influenced his writings (and perhaps vice versa!). Finally, I will briefly discuss the practice of utkrānti, the yogic “ejection of consciousness” at the moment of death, for which the Dvitīyakrama is an early Buddhist source, and which that text specifies as a practice appropriate for yogins who have already been introduced to the suchness of the perfection stage. But before we get into the details of the perfection stage practices outlined in Buddhajñānapāda’s writings, let us first look carefully at how he uses the terms “perfection stage” and “second stage” in his oeuvre, so that we begin with a clearer understanding of what these terms meant for him. The “Perfection Stage,” or the “Second Stage,” in Buddhajñānapāda’s Writings As I noted in Chapter Five, the term “perfection stage” in tantric Buddhist traditions has come to be used in two different ways. One is to refer to yogic practices that involve the manipulation of the internal winds and bindus within the channels that are held to constitute part of the practitioner’s subtle body, as a means for cultivating the direct experience of suchness itself. The second is to refer to the cultivation of the direct experience of suchness which, in some later traditions of Mahāmudrā and the Great Perfection is said to be done without reliance on such yogic techniques, or even without any technique at all. In the later Tibetan tradition these two different aspects are referred to by the terms “perfection stage with characteristics” (rdzogs rim mtshan bcas) and “perfection stage without characteristics” (rdzogs rim mtshan med). In his writings, however, Buddhajñānapāda uses the term “perfection stage” synonymously with the 210 term “second stage,” primarily to refer not to the method by means of which suchness is cultivated—which in the case of his practice system is the bindu yogas performed with a partner—but rather to refer directly to the experience of suchness itself.1 The term “perfection stage” appears four times in the Muktitilaka and three times in the Dvitīyakrama, while the “second stage” is used eight times in the Dvitīyakrama, and not at all in the Muktitilaka. The first appearance of the term “perfection stage” in the Dvitīyakrama is in reference to wisdom that is experienced by the disciple in reliance upon the sexual bliss of the third initiation. The text stages: In the space of the lotus, the jewel of the vajra and the heart of the lotus join, and in vajra posture The mind is observed within the jewel. The bliss that arises is ascertained, and that itself is wisdom. |124| This is explained by all of the genuine supreme gurus to be the perfection stage.2 Later in the Dvitīyakrama, “the perfection stage” is given as one among a list of the names of suchness, immediately after “the ultimate truth.”3 The third use of the term in the Dvitīyakrama makes reference to the “yogin of the perfection stage,” who is equated with “the supreme Vajradhara.”4 As was discussed in the previous chapter, it is in the Muktitilaka that Buddhajñānapāda sets forth the distinction between the two stages of tantric practice. In his presentation of the perfection stage in that text, he notes that it has two aspects, one of which he identifies further as the “perfection stage of the perfection stage.” Buddhajñānapāda equates this “perfection stage of the perfection stage” with the seven yogas, which, as we have seen in Chapter Three, are seven aspects of the final result of awakening, and therefore correspond to suchness.5 All of the references to “the perfection stage” in Buddhajñānapāda’s writings thus 1 One might be tempted to equate this usage with the later Tibetan term “perfection stage without characteristics” (rdzogs rim mtshan med), but the latter term, I believe, is used in contrast to the “perfection stage with characteristics” (rdzogs rim mtshan bcas) specifically to distinguish the methods (or lack thereof) by means of which suchness is cultivated—in the “perfection stage without characteristics” type of practices of the Great Perfection the “method” is considered to be precisely the absence of any method or act of cultivating suchness. 2 padma’i mkha’ la rdo rje nor bu pad snying gnyis la ‘byor dang rdo rje skyil krung sems// nor bu’i bar du mthong byas gang de bde ba ‘byung ba nges par de nyid ye shes te// |124| ‘di ni rdzogs pa’i rim pa yin par bla ma mchog rnams kun gyis yang dag bshad// (Dvitīyakrama, 124c-125a). While I believe that the text is quite clear here in identifying the wisdom of suchness itself with the perfection stage, it may be worth noting that Tsongkhapa likewise seems to have read this passage in a similar way. In his Lamp to Illuminate the Five Stages, in what I believe to be a reference to precisely this passage in the Dvitīyakrama he writes, “This is like Buddhaśrījñāna’s Oral Teachings of Mañjuśrī explaining that the innate wisdom that arises in a similar way during initiations, before meditation on the path has begun, is the completion stage” (Kilty 2013, 532). The first two lines of the passage I have cited here are themselves cited—with some variant readings—in a number of other tantric texts, in some of which the verse is attributed to the Paramādya-tantra, though Isaacson and Sferra (2014, 297 n 239) note that it is not found in any of the surviving recensions of that tantra. C.f. Abhayapaddhati (MS A fol. 15v2), the Sekanirdeśapañjikā (Isaacson and Sferra 2014, 185), the Kriyāsaṅgrahapañjikā (Sakurai 1996, 514) and the Yamāritantramaṇḍalopāyikā (fol. 24r.3), and the Āmnāyamañjarī (D 68b.1-2). This list of sources is given by Isaacson and Sferra (2014, 297 n 239) in reference to the verse’s citation in the Sekanirdeśapañjikā; the page numbers I have given here are those provided in Isaacson and Sferra’s citation, and include sources that I, myself, have not looked at. The interested reader is therefore directed to Isaacson and Sferra’s bibliography for further details). I address this verse further below in terms of its relevance to the three blisses in Buddhajñānapāda’s perfection stage system. 3 don dam pa yi bden pa dang// rdzogs pa yi ni rim pa dang// (Dvitīyakrama, v 276c-d). 4 bdag po rdo rje ‘chang ba mchog// rdzogs pa’i rim pa’i rnal ‘byor pa// (Dvitīyakrama, v 293c-d.) 5 rnam par dag pa sbyor ba bdun// rdzogs pa’i rim pa rdzogs rim ste// (Muktitilaka, D 52a.3; P 62b.2). Vaidyapāda’s commentary on this passage names the other aspect of the perfection stages as the “generation stage of the perfection stage,” which he identifies with the three bindu yogas. This “generation stage of the perfection stage” amounts to a use of the term “perfection stage” in a way that corresponds more closely with what came to be called the “perfection stage with aspects” (mtshan bcas rdzogs rim) in the later Tibetan tradition; that is, it refer to the 211 seem to use the term to indicate not the practices by means of which suchness is experienced, but rather the actual experience of suchness itself. The term “the second stage” is used more frequently in the Dvitīyakrama than the term “the perfection stage,” most prominently, of course, in the title of that text: The Oral Instructions on Training in the Suchness of the Second Stage. This term, too, seems most often to be used to reference the direct experience of suchness itself, though there is a single instance where Buddhajñānapāda uses the term the “second stage” to refer to the techniques of perfection stage practice, as well. As I already discussed in Chapter Three, the phrase “the suchness of the second stage” is equated in the Dvitīyakrama with reality itself, but several references to “the second stage” in the Dvitīykarama indicate that Buddhajñānapāda likewise understands the term “the second stage” alone to also be synonymous with suchness. For example, The method for training in the second stage Is the meditation upon the indestructible bindu.6 |271| In this instance it is clear that “the second stage” refers to suchness itself, since the bindu meditation is explicitly described as the “method for training” in that. Later in the Dvitīyakrama, after listing all of the synonyms of suchness (one of which, as noted above, is “the perfection stage”) Buddhajñānapāda advises: Therefore, with a mind that has already [generated] faith, Genuinely maintain the nature of all phenomena, The profound, luminous nondual great reality, The suchness of the second stage, |283| Which has been taught by the guru. Maintaining this, by means of the previously-described procedures, The individual who constantly habituates himself to it Based on this [practice] will give rise to the signs |284| As if leaping from bhūmi to bhūmi!7 Here the instruction to “maintain” the “suchness of the second stage” by means of the “previously-described procedures,” which in this case refers to the procedures of the three bindu yogas that were just described in the text, again suggests that “the second stage” refers directly to suchness itself. This suchness, according to the passage, is what is to be maintained by means of the bindu yogas, which here would (like in the above passage) constitute the methods for maintaining or cultivating that suchness. As I suggested already in Chapter Three, I believe we should understand the genitive relationship in the phrase the “suchness of the second stage” (and thus the relationship of the two terms in the Sanskrit compound) (rim pa gnyis pa’i de kho na nyid, *dvitīyakramatattva) to be appositional, and thus to mean “the suchness which is the second stage.” Another verse from the Dvitīyakrama likewise equates “the second stage” with wisdom itself: Having come to fully understand this, [One knows] the universal form of the wisdom of the great perfection, yogas that are meant to bring about the experience of suchness itself, rather than directly to that experience of suchness. It is possible—perhaps even likely—that Buddhajñānapāda himself intended this “generation stage of the perfection stage” as the other aspect of the perfection stage, and thus may also have used the term “perfection stage” in this sense, as referring to the yogic practices that help to bring about the experience of suchness. However, he does not indicate this directly in his own writings. 6 rim pa gnyis pa bsgom pa’i thabs// mi shigs thig le bsgom pa’o// |271| (Dvitīyakrama, 271e-f.) 7 de bas dad pa sngon ‘gro ba’i// sems kyis chos kun de bzhin nyid// zab gsal gnyis med don chen po// rim pa gnyis pa’i de kho na// |283| bla ma’i gsung ni yang dag gzung// de gzung gong ma’i rim pa yis// rtag tu goms byed skyes bu gang// de la brten pa’i rtags skyes nas// |285| sa nas sar ni ‘phar ba ltar// (Dvitīyakrama, 283a-285a). 212 The perfectly pure body, Great Vajradhara, The essence of all the great glorious ones, this second stage.8 |391|9 Vaidyapāda’s commentary is explicit that “the second stage,” which he identifies as “the perfection stage alone,” is synonymous with all of the other terms listed in this verse.10 One further verse also seems to identify the “second stage” with suchness: The sphere of the buddhas’ nirvāṇa The unborn vajra, manifest awakening, The supreme essence of all sugatas, This great nondual nonconceptual reality Is explained as11 the second stage. |34| As I discussed in Chapter Three, the last lines of this verse could equally well be translated as “This great nondual nonconceptual reality/ Is explained in the second stage.” But while nondual reality is certainly explained in the teachings on the second stage, I think that given the other references in the Dvitīyakrama that directly equate suchness with “the second stage” it is better to understand this verse to likewise state that nondual reality “is explained as the second stage.” Vaidyapāda’s comments on this verse also give three synonyms for “the second stage:” he says it is also called the “spontaneously arisen stage” (lhan cig skyes pa’i rim pa), the “perfection stage” (rdzogs pa’i rim pa), and the “stage of [things] just as they are” (ji bzhin pa’i rim pa).12 There is a single use of the term “the second stage” in the Dvitīyakrama that suggests Buddhajñānapāda may have also employed this term to sometimes refer to the techniques of practice that are meant to bring about the experience of suchness, rather than as referring directly to the experience of suchness itself.13 That verse reads as follows: Here in the second stage, the practitioner Practices one-pointed retention.14 The fact that this verse makes reference to practices done “in the second stage” suggests that the “stage” here is meant to refer to the procedures of perfection stage practice, as opposed to the experience of suchness, as the term otherwise seems to be used in Buddhajñānapāda’s writings. 8 rim gnyis ‘di (‘di] sugg. em. based on V (D and P), ‘dis D C P N S). Vaidyapāda clearly indicates that rim gnyis ‘di is to be understood as the “second stage,” the perfection stage (or, according to P, the “perfection stage of the perfection stage”) only, rather than to the “two stages.” (Sukusuma, D 137b.7-138a.1; P 166a.5). I have translated in accordance with his comments, somewhat (but not completely unfeasibly; rim gnyis could very easily be an abbreviation of rim pa gnyis pa made for metrical reasons) against the grain of the grammar of the Tibetan translation of the verse, which would otherwise be more easily read as the “two stages.” Also, given that the topic of the verse is wisdom, “the second stage” is really the only reading that makes sense here. 9 de ltar rab tu shes par byas nas su// rdzogs pa chen po ye shes spyi yi gzugs// yongs su dag sku rdo rje ‘chang chen po// dpal ldan kun gyi ngo bo rim gnyis ‘di// |391| (Dvitīyakrama, vese 391). 10 de nyid la rim pa gnyis ‘di zhes te rdzogs pa’i rim pa (pa] D, pa’i rdzogs pa’i rim pa P) de kho na zhes so// (Sukusuma, D 137b.7-138a.1; P 136a.5). Note that P reads “the perfection stage of the perfection stage alone.” 11 rim pa gnyis par. 12 Sukusuma, D 96b.6-7; P 116a.5. 13 Vaidyapāda definitely uses the term in this way at times. For instance, in a passage of his Muktitilaka-vyākhyāna cited in Chapter Three in which Vaidyapāda comments on the verses on the Muktitilaka where Buddhajñānapāda discusses the two stages, Vaidyapāda identifies the two aspects of the perfection stage as the “generation stage of the perfection stage” and the “perfection stage of the perfection stage.” The former, which he defines as the three bindu yogas, clearly uses the term perfection stage to refer to the yogic methods used for cultivating suchness. The “perfection stage of the perfection stage,” however, Vaidyapāda equates with suchness itself (Muktitilakavyākhyāna, D 38b.2-3; P 366b.5-6). 14 rim pa gnyis ‘dir sgrub pa po// rtse gcig pa yis rnam par gzung// (Dvitīyakrama, 324a-b). In fact, this line in the Tibetan more literally reads “Here in these two stages…” rather than “in this second stage.” However, the context of the passage makes it much more likely that what is meant here is the “second stage.” I suspect that the Tibetan translation reads gnyis rather than gnyis pa here simply for metrical reasons. 213 However, it is also possible to understand the locative particle in the verse as topical—i.e. “With regard to the second stage/ The practitioner practices one-pointed retention”—in which case the second stage could still be understood as referring not to practice techniques, but again to suchness itself. In either case, this brief survey of the ways in which Buddhajñānapāda uses the terms “the perfection stage” and “the second stage” clearly shows that, for him, their primary referent is not the method by means of which a yogin trains in cultivating suchness, but rather that very suchness itself.15 Perfection Stage Practices in Buddhajñānapāda’s Oeuvre As should be evident from the sources cited in the above discussion on terminology, it is only in Buddhajñānapāda’s Dvitīyakrama and Muktitilaka that we find direct mention of the perfection stage or the second stage. Likewise, it is in these two texts that we find details of the perfection stage practices according to his system. Indeed, these two works are identified in the later Tibetan tradition and in modern scholarship as presenting the perfection stage practices of Buddhajñānapāda’s system, while the Caturaṅga/Samantabhadra-sādhana is understood to set forth the generation stage procedures according to his system. However, as discussed in Chapter Five, some of the practices outlined in the Caturaṅga/Samantabhadra-sādhana appear to be less complex versions of several of the very same yogas delineated in Dvitīyakrama and the Muktitilaka, and the Indic commentaries on the Caturaṅga/Samantabhadra-sādhana discuss these practices in ways that make reference to signs of accomplishment ordinarily connected with perfection stage practice. A study of the perfection stage practices of Buddhajñānapāda’s system must therefore also take the Caturaṅga/Samantabhadra-sādhana into account. Generally speaking, the Dvitīyakrama provides the most detailed presentation of the perfection stage practices described in Buddhajñānapāda’s oeuvre. In fact, there is not a single perfection stage practice referenced in any of his other writings, including the Muktitilaka, that is not presented in significantly more detail in the Dvitīyakrama. Yet even the Dvitīyakrama’s presentation of some of these practices is difficult to follow without reference to the commentarial tradition. The perfection stage yogas in Buddhajñānapāda’s system—or, to more closely mirror his own use of the term, the yogas by means of which one trains in the perfection stage—consist primarily of a set of three “bindu yogas,” outlined in both the Dvitīyakrama and the Muktitilaka. As I noted above, a simplified version of two of these yogas are also found in the Caturaṅga/Samantabhadra-sādhana. In the Dvitīyakrama, these practices are described in verses 156-271, amounting to more than a quarter of that work’s total content. It is worth considering the context of the verses describing the perfection stage yogas within this text because, while the Dvitīyakrama’s content does include quite a diversity of topics, there is a coherent progression in the text’s overall narrative into which the perfection stage practices fit. I described the contents of the Dvitīyakrama in more detail in Chapter Two (and, of course, the entire Dvitīyakrama is translated in Part II of this dissertation). What follows here is just a very brief summary, intended to indicate the way in which the perfection stage practices fit within the overall structure of the text, because this also indicates how they fit within the context of the path of tantric practice in Buddhajñānapāda’s system. The Dvitīyakrama begins with Buddhajñānapāda’s autobiographical narrative leading up to his vision of Mañjuśrī, followed by Mañjuśrī’s instructions, which, as we will recall, constitute the majority of the text. These instructions begin with a general description of reality, or 15 There are two further uses of the term “the second stage” that I have not addressed in this discussion, in verses 155 and 314 of the Dvitīyakrama, but neither of these provides any further information on what Buddhajñānapāda intends by this term beyond what we have already examined in the verses above. 214 suchness, followed by a discussion of the appropriate characteristics of both the student and the teacher of the path that leads to realizing suchness. Then, since according to this work it is not possible to come to this realization without relying on a consort, the characteristics of four different types of consort, as well as the appropriate and inappropriate behaviors for such a partner, are set forth. This general information is followed by a description of the initiatory rituals that allow a practitioner to take up the practice of suchness, including the various practices by means of which he first comes to a direct recognition of the suchness of the perfection stage during that initiation. Following these initiatory rituals and practices, meant to bring the student to a direct experience of what he should then train in, is a doxography of views that places the tantric perspective recognized through those just-described rituals at the very top. It is then that the methods for training in suchness—the three bindu yogas of the perfection stage and a final perfection stage instruction involving the practice of dissolution—are described in some detail. These instructions are followed by a list of the names of suchness and the benefits and results of training in it. Here there is an excursus that departs somewhat from the narrative progression of the text, in which the ten bhūmis are homologized with sexual practices. This is followed by a praise of the yogin who trains in these procedures, and a condemnation of those who would deprecate him, concluding with the assurance that a practitioner who perfects this training will achieve the final result of awakening within this very life. Next, Mañjuśrī gives instructions in the yoga of utkrānti—the yogic ejection of consciousness at the moment of death—which is prescribed for individuals who have already received the suchness of the second stage but were unable to perfect its realization through training in the perfection stage yogas during this lifetime. Mañjuśrī elaborates on the importance of the instructions that he has given and commands Buddhajñānapāda to compile them and to compose supplementary texts relating to Guhyasamāja practice. The Dvitīyakrama concludes with Buddhajñānapāda’s closing autobiographical narrative, starting from the dissolution of his vision of Mañjuśrī, followed by a final injunction to practice, and the dedications and aspirations that conclude the text. We can see through this brief description of its overall content that the Dvitīyakrama focuses entirely on the perfection stage, describing the reality of suchness itself, the requisites that are necessary to come to this realization (a qualified guru, student, and consort), the way that the practitioner comes to directly experience suchness through initiation, and the practices through which he then trains in this suchness that he has obtained from his guru, either during this life via the three bindu yogas, or at the moment of death via the practice of utkrānti. Thus, while the title of the work—Oral Instructions on Training in the Suchness of the Second Stage— only promises instruction on how to train in suchness, the text in fact provides the full structure of the perfection stage path from the first moment of its recognition during initiation, through the process of training in it, up to its full realization resulting in awakening. The bindu yogas through which that training takes place are a central part of the process, and thus constitute a major focus of the text. Indeed, if we also include the instructions on the yoga of utkrānti along with the other perfection stage yogas (which I believe makes sense, given that the utkrānti instructions are specified as appropriate for a practitioner who has already received suchness from his guru), these practical instructions make up more than one-third of the Dvitīyakrama’s verses. Meditating on the Indestructible Bindu: The First Bindu Yoga The first bindu yoga described in the Dvitīyakrama is called meditation on the “indestructible bindu” (mi shig pa’i thig le), which Vaidyapāda specifies as having “the nature of 215 the wisdom of bliss (dga’ ba’i ye shes).”16 This is a reference to the wisdom connected with the first of the three blisses, called simply “bliss” (dga’ ba, ānanda), that are said to arise during sexual yogic practices in Buddhajñānapāda’s system.17 Indeed, later in the Dvitīyakrama Buddhajñanapāda explicitly states that the meditations on the three bindus correspond with the three blisses.18 Instructions on this first bindu yoga practice are also found in the Muktitilaka, and there are parallels in the Caturaṅga/Samantabhadra-sādhana, as well.19 The description of this yoga in the Dvitīyakrama indicates that the framework for its practice is the generation stage. As I noted in Chapter Five, the generation stage practices summarized at this point in the Dvitīyakrama are practices included in the first two among the three samādhis, the ādiyogasamādhi and the maṇḍalarājāgrī-samādhi, concluding with the emanation and pleasing of the deities of the maṇḍala-cakra that completes the maṇḍalarājāgrī-samādhi.20 Then the practitioner is instructed, to “cast away the outer body” and “train in the ultimate suchness that is the buddhas’ supreme sphere of experience.”21 Vaidyapāda explains the casting away of the outer body as the practice of purifying the four processes of birth (from heat and moisture, miraculously, from a womb, and from an egg22) by means of training in the generation stage, and notes that the perfection stage instructions begin with the injunction to train in suchness.23 What follows is a series of visualizations in which the practitioner-as-deity meditates on a bindu at his heart center, from which he emanates out and reabsorbs light rays that illuminate a series of syllables and maṇḍalas, and then draw these back into the bindu. The procedure begins with the generation of sentient beings in the form of buddhas by means of emanating light out from the seed syllable visualized in the center of the samayamudrā (i.e. the jñānasattva) 24 of the yogin-as-deity’s heart center, which emerges from the right nostril, and from which tathāgatas and the deities of the maṇḍala-cakra emerge and fill the world.25 All beings are then purified and generated in the form of buddhas,26 who then melt into nectar, “the 16 dga’ ba’i ye shes kyi rang bzhin mi shigs pa’i thig le bsgom pa (Sukusuma, D 114b.2-3; P 137b.8). 17 Buddhajñānapāda’s works are among the early writings to give a typology of blisses in the context of tantric practice. I address this topic in more detail below. 18 This was the authentic teaching of the ritual/ Of meditating on the three bindus/ That correspond with the three blisses. / |241| de ltar dga’ gsum bye brag gi// thig le rnam gsum bsgom pa yi// cho ga yang dag bstan pa’o// |241| (Dvitīyakrama, verse 241). I discuss this point further below. 19 See Muktitilaka D 49a.2-5, and Caturaṅga/Samantabhadra-sādhana, verses 109 and 127-129. 20 See Dvitīyakrama verses 156-160. As I noted in Chapter Five, the first of the bindu yogas included in the Caturaṅga/Samantabhadra-sādhana likewise appears at precisely this point in the sādhana’s structure— immediately following the completion of the maṇḍalarājāgrī-samādhi. 21 phyi rol lus ni log byed de// sangs rgyas rnam kyi spyod yul mchog// mtha’ yi de nyid bsgom par bya// (Dvitīyakrama, verse 161a-c). 22 These are the four modes by which sentient beings take birth according to Buddhist Abhidharmic systems. 23 Sukusuma, D.114b.1-3; P 137b.5-8. 24 The term used in the Dvitīyakrama here, the samayamudrā, refers to a smaller image of the deity visualized at the practitioner’s heart center. This is more commonly termed the jñānasattva and in fact Vaidyapāda’s commentary glosses the term samayamudrā as used in this passage as the jñānasattva (Sukusuma, D 114b.4; P 138a.1-2). Buddhajñānapāda himself uses the term jñānasattva instead of samayamudrā in a later passage in the Dvitīyakrama to reference this smaller deity visualized in the heart (see Dvitīyakrama verses 248, 252, 261, and 262). 25 Dvitīyakrama verses 161-164. 26 Vaidyapāda elaborates on this process: “They are generated in the form of buddhas. How is this done? They are made to melt as the moon, dissolve, and are purified, which means that the emanated maṇḍala, which has melted like the moon, dissolves into sentient beings and they become as above. That itself, as well, enters into oneself as the essence of wisdom means that they are gathered as the essence of wisdom which is pure like water and ushered into one’s left nostril.” de rnams sangs gyas kyi skur bskyed pa’o// de gang gis she na/ zla bar zhu byas te/ thim pas rnam par dag byas nas/ zhes (zhes] D, shes P)\ pa ni ‘phros pa’i dkyil ‘khor zla ba lta bur zhu ba sems can thams cad la thim pas de rnams gong ma lta bur gyur pa’o// de nyid kyang ye shes ngo bo ru (bo ru] P, bor D) rang la zhes te chu ltar dang (dang] P, dangs D) ba’i ye shes kyi ngo bor ‘dus zhing rang gi g.yon p’ai ha sar 216 essence of wisdom,” and are then drawn back in that form into the practitioner through his left nostril. This nectar is ushered into the seed syllable at his heart center, which is then visualized as the indestructible bindu, the size of a chickpea and blazing with five-colored light.27 The yogin is to visualize all phenomena as gathered within his mind, within that bindu.28 Light from the bindu gradually radiates outward, first illuminating the area around itself, then the interior of the samayamudrā visualized at the practitioner’s heart, and then gradually out into the practitioner’s body, where it illuminates sixteen syllables representing bindus of bodhicitta29 zhugs par bya zhes so// (Sukusuma, D 114b.7-115a.2; P 138a.6-8). A similar (but not identical) process involving the purification of sentient beings is found in the process of the generation of the causal deities in Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana. 27 Vaidyapāda clarifies that the seed syllable becomes the indestructible bindu. This also makes sense, since first there is a syllable there, and then the practice is done with a bindu, rather than a syllable, so some sort of transformation from syllable to bindu must take place at some point. Vaidyapāda writes: “Once it has been made to enter inside in that way, it is brought into the seed syllable that abides in the center of the symbolic implement mentioned above; this sets forth the source and locus of the practice. By means of that [process], what does it become? This is expressed in the verse beginning, This itself…” nang gi la yang de bzhin du zhugs nas gong gi mtshan ma’i dbus na gnas pa’i sa bon la zhugs par bya ste ‘grub pa’i rgyu (rgyu] D, rgyud P) dang gnas bstan (bstan] P, brtan D) pa’o// des cir ‘gyur zhes na/ (des cir ‘gyur zhes na/] D, P om.) de nyid ces pa la sogs pa’o// (Sukusuma, D 115a.2; P 138a.8-b.1). 28 Dvitīyakrama verses 165-168. The procedure described in verses 161-166 is also set forth (in a less elaborated form) in the Caturaṅga/Samantabhadra-sādhana in a single verse—verse 109—at the outset of the karmarājāgrīsamādhi. 29 The verses identifying these sixteen syllables and their location on the body (verses 171-174 of the Dvitīyakrama) are attested in Sanskrit in a citation (?) given in the single surviving manuscript of Kalyāṇavarman’s Catuṣpīṭhapañjikā (see note 325 in my translation of the Dvitīyakrama for the Sanskrit and for further details). The verses appear as part of a larger citation, in which they are preceded by part of verse 111 from the Dvitīyakrama and followed by some verses that are not found in the Dvitīyakrama. This set of verses—which essentially consist of a combination of verses from two different sections of the Dvitīyakrama along with verses that are not from the Dvitīyakrama—are attributed in the Catuṣpīṭhapañjikā to the “Aṣṭāṣṭaka,” which seems to be the title of a work, though this is not entirely clear. The verses in the Dvitīyakrama (171-174) that describe the sixteen syllables located at the sixteen places in the body conclude with a single somewhat cryptic line: “These are completed at the time of the sixteenth” (bcu drug dus su rdzogs ‘gyur ba//), which Vaidyapāda explains to be a reference to sixteenth day of the lunar calendar—the fullest moment of the full moon (Sukusuma, D 115a.6). Vaidyapāda elaborates—unfortunately still somewhat cryptically—on this point: “Also, one should know that this is with regard to the stages of the first day [of the month] and so forth. They are completed at the time of the sixteenth, means that at the time when the outer moon comes to fullness, these are also perfected. One must understand that this is then reversed. Regarding being perfected at the time of the sixteenth, the sixteen places that are stirred up through practice also become “the sixteen.” These then [become] the bindu and this becomes like the moon, which produces the consciousness of joy. The previous light rays hook, means that they hook the sixteen syllables and draw them into the bindu. By slightly holding one’s mind, like the first wisdom, there for a moment, what happens? [The text then says] Meditate with determination/ On the great [maṇḍala]-cakra of deities together with its support.” de yang tshes gcig la sogs pa’i rim par shes par bya’o// bcu drug dus su rdzogs gyur pa zhes pa ni phyi’i zla ba rdzogs par’i dus su de yang rdzogs pa’o// de ne bzlog ste shes par bya’o// bcu drug dus su rdzogs par ‘gyur ba ni sgrub (sgrub] P, bsgrub D) pas dkrugs pa bcu drug/ de (de] D, ste P) yang bcu drug par ‘gyur/ de yang thig le/ de yang zla ba lta bur song nas/ dga’i ba’i shes pa ‘byung ba’o// gong gi ‘od kyis rnam pa bkug ste zhes pa ni yi ge bcu drug po rnam par bkug nas thig le’i nang du bcug la der rang gi sems dang po’i ye shes ltar bag zhad bzung bas cir ‘gyur zhe na/ lha’i ‘khor lo che/ rten dang bcas pa mos pas bsgom/ (Sukusuma, 115a.6-115b.1; P 138b.6-139a.2). These statements are very similar to Vaidyapāda’s comments at an earlier point in the Sukusuma: “Moreover, through practicing, by means of the agitation of the locations, the sixteen syllables appear, and these, then, become the sun and moon. Having transformed into a bindu like that, they go to the tip of the vajra. This itself, in a form which blazes with thousands of light rays, is meditated upon by the yogin in accordance with the ritual that will come below. When this happens, the suchness that has been spoken of will be realized, [and that is the] purpose [of this practice.]” / (de yang bsgrub pas gnas rnams dkrugs pa las yi ge rnams bcu drug par gyur/ de yang nyi zlar gyur/ de lta bu’i thig ler gyur nas rdo 217 located at different points in the body, and draws these syllables into the bindu at the heart center.30 Then the practitioner should contemplate the entire maṇḍala within a tiny bindu at the center of the lord of the maṇḍala who is himself visualized at the center of a slightly larger—but still small!—maṇḍala within the center of the chickpea-sized bindu. The procedure of visualizing light rays emanating out from that smaller bindu is repeated in the same way as above, until finally they emerge out of the outer body of the yogin-visualized-as-deity, illuminating the maṇḍala-cakra of which he himself is a part, and touch all of the tathāgatas.31 These rays then enter their mouths, collect nectar from the bindu in their hearts, and emerge from the vajra path before returning and entering into the practitioner’s right and left nostrils,32 and finally dissolving into the bindu at his heart.33 He is to hold his mind with concentration on this bindu, which is white tinged with red, blazing with five-colored light, and “has the nature of rje rtse mor ‘gro ba ste/ de nyid ‘od zer stong du ‘bar ba’i gzugs su rnal ‘byor pa rnams kyis ‘og nas ‘byung ba’i cho gas bsgoms nas/ ji skad du gsung pa’i de kho na nyid rtogs par ‘gyur pa’i phyir ro/) (Sukusuma, D 88a.4-5). In his much later (15th-century) instruction manual on the perfection stage rituals of the Jñānapāda School, Tārānātha, who reports having received initiation into and teachings on the Jñānapāda lineage from his master Buddhaguptanātha, gives an ever-so-slightly clearer presentation of this practice that does not seem substantially different from what is already here in Buddhajñānapāda’s text, with Vaidyapāda’s clarifications. He writes, “For the second part, the light from the bindu illuminates the jñānasattva, and from that light radiates forth and illuminates the interior of the foundational body. Like holding up a lamp in darkness, one sees clearly the sixteen bindus, which are the white substance.... [He lists here the syllables at all of the locations on the body, exactly as they are described in the Dvitīyakrama]...All of these are white and radiate white light. Think of them as being of the nature of bliss. The light from the heart center, either in stages or all at once, as one prefers, dissolves those syllables into the indestructible bindu at the heart center, and [it] then blazes with light and causes a strong increase in the essence of bliss. Contemplate thus.” (gnyis pa ni thig le’i ‘od kyis ye shes sems dpa’i sku gang / de las ‘od ‘phros gzhi lus kyi nang gsal zhing gang bar byas/ mun khung du sgron me bteg pa ltar dkar cha thig le bcu drug po rnams gsal bar mthong ba ni/ ...... thams cad kyang kha dog dkar po ‘od zer dkar po ‘phro ba/ bde ba’i rang bzhin can du bsam/ snying ga’i ‘od kyi yi ge de rnams rim pas sam cig car gang mos kyis snying ga’i mi shigs pa’i thig ler bstims pas/ ‘od zer ‘bar zhing/ bde ba’i ngo bo lhag par rgyas par bsam mo// (Dpal grol ba’i thig la’i khrid yig, 243-244). Tāranātha does not include any reference to the line “They become complete at the time of the sixteenth,” which is unfortunate because Vaidyapāda’s commentary on this point is rather cryptic. 30 Dvitīyakrama verses 169-175. This same procedure is described in a single verse in the Muktitilaka (D 49a.2); Vaidyapāda’s commentary there adds many details from the Dvitīyakrama to fill in the parts of the procedure that are not elaborated in the Muktitilaka’s version of the instructions (Muktitilaka-vyākhyāna, D 52b.5-52b.7) 31 Dvitīyakrama verses 176-179. The same procedure, with slight variation (the light from the bindu at the heart is described as emerging from the nose, rather than radiating outward through the body), is described in the Muktitilaka (D 49a-3-4). 32 Vaidyapāda explains that what enters into the right nostril is the nectar emerging from Akṣobhya, and so forth, the male deities, while what enters into the left nostril is what emerges from Locanā and so forth, the female deities (Sukusuma, D 115b.7-116a.1; P 139b.1-2). Tāranātha likewise describes this process in the same way, but is more explicit, explaining that the substance emerges from the vajra of the male deities and the lotus of the female deities. (Dpal grol ba’i thig la’i khrid yig, 244). 33 These lines in the Dvitīyakrama are slightly perplexing. The text reads, “Then it dissolves into the wisdom bindu/ Via the apparent bindu.” de nas ye shes thig le la// snang ba’i thigs par thim par ‘gyur// Vaidyapāda explains: “The wisdom bindu is the first bindu. The apparent bindu is the one that appeared from that, which is suchness.” ye shes kyi thig le zhes pa ni dang po’i thig le’o// snang ba’i thigs pa (thigs pa] P, thig le D) zhes pa ni de las snang ba’i thig le ste chos nyid do// (Sukusuma, D 116a.1; P 139b.2-3). The wording of these two lines is a bit strange, suggesting that the transmission may be corrupted. I am tempted to emend as follows: de nas ye shes thig le las/ snang ba’i’ thigs par (or thig ler) thim par ‘gyur. Following Vaidyapāda’s commentary, we could suggest emending in this way, but it’s not clear if he is actually glossing the root text there or merely explaining. The translation, following both emendations, would be: “Then, it dissolves into the bindu that has appeared from the wisdom bindu.” For now, I will simply leave this here as an alternative way to read the line. Tāranātha, for what it’s worth, does not mention anything about two bindus here, but he also left out the nesting sequence above in which the second bindu was visualized. He simply states that the two flows of nectar dissolve into the “root bindu” (rtsa ba’i thigs las) (Dpal grol ba’i thig la’i khrid yig, 244). 218 dripping.” He should then repeat this process of emanating light, absorbing nectar from the buddhas, and holding that nectar and his concentration within the bindu.34 This practice is said to bring about an encounter with the “mind as the vajra of cessation” and the signs of “the glorious wish-fulfilling gem that is the great receptacle of all the buddhas.”35 These attainments, which are mentioned in Chapter Eleven, verse 41 of the Guhyasamāja-tantra, in the context of a different yoga,36 are here specifically identified as the results of this first bindu yoga; a different result—the arresting of the breath—is mentioned below in the context of the second bindu yoga performed with the “secret bindu” at the tip of the vajra. The signs that occur with the attainment of “the vajra of cessation” are, according to Vaidyapāda, the same signs mentioned by Buddhajñānapāda himself later in the Dvitīyakrama as indications of the effectiveness of perfection stage practices: laughter, yawning, and trembling.37 Meditating on the Secret Bindu: The Second Bindu Yoga The second bindu yoga described in the Dvitīyakrama is called the yoga of the secret bindu and, according to Vaidyapāda, has the nature of the wisdom of the “intermediate bliss” (dga’ ba bar ma’i ye shes), the second of the three blisses according to Buddhajñānapāda’s system.38 This practice involves moving the bindu that was previously meditated upon at the heart center down to the tip of the penis, where it is held and meditated upon following a visualization similar to the one described above for the indestructible bindu at the heart center. The process by which this moving of the bindu takes place involves emanating light rays in the form of hooks from the bindu at the heart center that hook the sugatas and their maṇḍala-cakras and draw them—in the form of wisdom nectar—into the bindu at the heart center, which causes 34 Dvitīyakrama verses 180-183. 35 Dvitīyakrama verse 184. These lines are part of a verse adapted from the Guhyasamāja-tantra (11.41) and are also included in the Muktitilaka (D 49a.4-5). The verse from the Guhyasamāja-tantra reads: nirodhavajragataṃcittaṃ yadā tasya prajāyate/ sa bhavec cintāmaṇiḥ śrīmān sarvabuddhāgradhārakaḥ/ (I have emended Matsunaga’s edition of the line from the tantra to follow the variant -dhārakaḥ found in two of his manuscripts, since that is the reading found here in the Dvitīyakrama (it is also, incidentally, the reading found in the Tibetan translation of the tantra). Vaidyapāda explains: “The mind will become the vajra of the cessation of all entities, and the signs of stability [in that] arise. That is to say, Eventually, one will come to encounter means at some time [one] will suddenly encounter the goddess within the unchanging bindu, and at that time.... [the signs will authentically arise] of having become the glorious wish-fulfilling gem/ That contains all the great buddhas... ” ” sems dgnos po kun las ‘gog pa’i rdo rjer gyur nas brtan pa’i rtags skyes pa ste/ nam zhig de la reg gyur pa/ zhes pa ni dus nam zhig na mi ‘gyur ba’i thig le la lha mo (mo] P, mo’i D) lam gyis reg par gyur pa de’i tshe/ sangs rgyas kun gyi mchog ‘dzin pa // yid bzhin (bzhin] sugg. em based on Dvitīyakrama D C, and Guhyasamāja-tantra, 11.41), sbyin V (D and C)) dpal dang ‘dra bar ‘gyur ba’ ‘o// (Sukusuma, D 116a. 4; P 139b.6-7). Vaidyapāda’s commentary on the parallel lines from the Muktitilaka reads: “The vajra of cessation within the mind itself indicates that [this takes place] by means of the mind [which engages in] the action of inhalation into(?) that bindu. Whoever comes to eventually encounter that, means one should not have any doubts that by means of encountering [it?] with that mind the signs will arise.” ‘gog pa’i rdo rje sems nyid du/ zhes pa ni thig le de la(?) (la?] sugg. em., las D P) dbugs ‘jug pa’i spyod pa’i sems kyis so// nam zhig de la sus reg pa/ zhes pa ni sems des reg pas rtags rnams skye bar ‘gyur ba la som nyi mi bya’o// (Muktitilaka-vyākhyāna, D 53a.4-5). 36 The yoga described in the Guhyasamāja-tantra here reads, in Fremantle’s translation: “By joining the great fivepointed vajra, adorned with five flames, to the five places, you will attain vajra perception. Visualize your mantra as a wheel densely filled with sparks of fire, and joining it to the five vajras you will attain vajra perception. Visualize a wheel shining with the flames of the Buddhas at the centre of vajra space, and with the entrance of the Buddhas become their dwelling. Place Vairocana in your body at the centre of the maṇḍala of the Buddha, and visualizing OṂ in his heart meditate on your consciousness in the mantra; when your mind enters the vajra state of suppression you will become the glorious Wishing-gem which contains all the great Buddhas.” (Fremantle 1971, 63-4). 37 Sukusuma, D 116a.5-6; P 140a.3-4. 38 dga’ ba bar ma’i ye shes kyi rang bzhin gsang ba’i thig le… (Sukusuma, D 116a.7; P 140a.3-4). 219 it to descend “in the form of rajas, tamas, and sattva” to the center of the jewel of the vajra (the head of the penis). Vaidyapāda specifies that this process is performed by means of the “moving wind that has the form of hūṃ phaṭ,”39 and that the aspects of rajas, tamas, and sattva descend via the right, left, and central channels respectively.40 Then the practitioner is to visualize the subtle symbolic implement of his deity at the tip of the vajra, and within it the full support and supported maṇḍala together with the bindu in an area the size of a mustard seed. The Dvitīyakrama then notes that if while regarding the maṇḍala within the bindu, the practitioner’s mind becomes weary, the bindu may emerge from the center of the vajra, and instructs that if this takes place the yogin should “make it remain at the tip of the nose41 and examine by means of the bliss of cessation,” which is the third of the three blisses in Buddhajñānapāda’s system.42 This short passage in the Dvitīyakrama that gives instructions for what to do if the bindu emerges from the vajra during the practice appears to be something of an aside, providing instruction for how to practice in the case of emission; the explanation of the yoga—presumably the practice that the yogin is otherwise meant to continue prior to emission—continues below. What is more, the practice as described in the Muktitilaka lacks this ancillary instruction on what to do in the case of emission, and simply continues to present the practice of the bindu yoga.43 The practice follows the same stages of visualization as the yoga of the indestructible bindu: the emanation of light out of the bindu, gradually filling the space around itself and eventually emerging out through the nostril, touching the buddhas, drawing in their wisdom as nectar, and ushering it back into the bindu.44 This is described as the “branch of emptying,” (stong pa’i yan lag) and is said to “stop the breath.” This branch is the first among three of the yogas pertaining to the “six branch yoga” (ṣaḍaṅgayoga) that receive mention in the Dvitīyakrama.45 The yogin is instructed to hold his mind gently within the bindu, which brings about a state of “entitylessness” (dngos po 39 hūṃ phaṭ rnam pa dang ldan par gyo ba’i rlung gis so// (Sukusuma, D 116b.1; P 140a.4-5). 40 Sukusuma, D 116b.3-4; P 140a-8-140b.1. This use of the three guṇas from the Saṃkhya system to describe the constituent aspects of the bindu is unusual, and is an example of Buddhajñānapāda’s use of non-Buddhist terminology, suggesting that he was indeed operating in an eclectic milieu. 41 sna rtse rnam par gnas byas nas// dga’ bral dga’ bas brtag par bya// (Dvitīyakrama, verse 192. c-d). While looking only at the root verse, one might presume that the “nose” referred to here is (as is often the case) the tip of the yogin’s penis. Vaidyapāda, however, specifies that one is to “make it remain at the tip of the nose of the goddess’ lotus.” lha mo’i padma’i sna rtser rnam par gnas par byas nas/ (Sukusuma, D 116b.6-7; P 140b.4). This might lead us to speculate that the yogin may withdraw from his partner at this point, leaving his penis resting against, but not within, her vagina, presumably in order to assist with the arrest of the bindu at the tip of the penis. However, given the fact that we find an association of the bliss of cessation precisely with the process, not of retention, but rather of emission (or at least in the context of emission following retention) at several places in Buddhajñānapada’s writings, it seems unlikely that the process of making the bindu remain at the nose-tip of the lotus is related to the process of retention. Moreover, a similar passage in the Caturaṅga/Samantabhadra-sādhana is described in Vaidyapāda’s and Samantabhadra’s commentaries as indicating a process by means of which the yogin appears to be instructed to draw the bindu that was previously emitted into the lotus of the consort out onto the “nose tip” of her lotus by means of transforming the “prong” of his vajra into hook-like light rays that hook the bindu and draw it out to this location. (See Caturaṅga/Samantabhadra-sādhana verse 130 (Samantabhadra-sādhana, D 34b.3), Vaidyapāda’s Samantabhadrī-ṭīkā (D 168a.2-5), and Samantabhadra’s Sāramañjarī (D 38a.1-3; Szántó unpublished 125). This again suggests that the practice of making the bindu remain at the nose-tip of the lotus is not connected to the practice of seminal rentention, or follows retention, because it clearly follows emission of the bindu into the lotus. 42 This statement in the Dvitīyakrama is one of several references in Buddhajñānapāda’s works to the observation of the third of the three blisses—the “bliss of cessation” (dga’ bral gyi dga’ ba) in the context not of seminal retention, but rather of emission, a point that I will discuss further below. 43 Muktitilaka, D 49b.5-6. 44 Dvitīyakrama, verse 193-4. Sukusuma, D 116b.7-117a.3; P 140b.5-141a.1. 45 This particular branch is more normally termed prāṇāyāma. I discuss the three among the six branches that receive mention in the Dvitīyakrama below. 220 med pa’i gnas).”46 Then, through training in “retention,” another yoga from among the six branch yogas, five signs indicating the dissolution of the elements—earth, water, fire, wind, and space—into one another occur; these signs, along with the element whose dissolution or withdrawal they indicate, are each described in the Dvitīyakrama.47 The process of the dissolution of the elements is generally said to accompany the death process. Bringing about this process intentionally within the controlled context of the practice of the perfection stage, which is accompanied and thus identified by the appearance of the five signs is, in the Dvitīyakrama, explained to bring about the result of non-abiding nirvāṇa.48 Next, in order “to make this reality pervasive,” the yogin is instructed to “emanate it from the vajra path into the realm of space,” a practice which is described as the practice of “recollection” (anusmṛti, rjes ‘dren), which is also the next among the practices of the six branch yoga.49 Here in the Dvitīyakrama this involves a series of “recollections” of various aspects of the dharma, in what amounts to a tantric version of the traditional Buddhist practice of anusmṛti (recollection). The yogin visualizes the emanation and absorption of the maṇḍala-cakras of the various maṇḍala deities, and engages in the practice of several other rituals from the generation stage practice, like impassioning, tasting of nectar, and so on. Vaidyapāda repeatedly introduces the various practices of recollection with the phrase, “Regarding the recollection of … (any given factor: the maṇḍala; samaya; body, speech, and mind, etc.), what should one recall with respect to the reality that was previously seen?”50 This suggests that in the practice of recollection here is understood to bring to forth different aspects or qualities of the reality of suchness that the practitioner has previously experienced through the practice of the bindu yoga. Vaidyapāda clarifies further that “it is called the branch of recollection because of recalling [something] with respect to the reality that one has previously seen.”51 The Dvitīyakrama concludes the section on the secret bindu by noting that it is by means of the three branches mentioned in the text—i.e. the branches of emptying, retention, and recollection—that the practitioner trains in the practice of the secret bindu. Meditating on the Emanated Bindu, or Vajrajapa: The Third Bindu Yoga The third bindu yoga described in the Dvitīyakrama, which Vaidyapāda specifies as having the nature of the wisdom of the bliss of cessation (dga’ bral gyi ye shes)—the third among the three blisses in his system—is that of the “emanated bindu” (sprul pa’i thig le).52 This same practice is also quite commonly—including in the Dvitīyakrama— called vajra recitation (vajrajapa, rdo rje zlas pa), and involves visualizing the emanation of elemental maṇḍalas from various centers in the body along with the exhalation, inhalation, and retention of the breath. In the practice as described in the Dvitīyakrama the yogin should visualize a smoke- 46 Vaidyapāda explains that this does not mean that entities are empty because they are destroyed or overcome. Rather, by means of the yoga of lacking nature (one of the seven yogas mentioned by Vaidyapāda in his Yogasapta) one turns away from other mental states, and since one therefore remains only in suchness, the state of entitylessness ensues. (Sukusuma, D 117a.4; P 141a.2-3) 47 Dvitīyakrama verses 196-199. These signs are also described in the Samājottara (vv. 150-151) and referenced in the Muktitilaka. I discuss more about the relationship between the Samājottara and Buddhajñānapāda’s works in Chapter Eight. 48 Dvitīyakrama, verse 200. 49 Dvitīyakrama, verse 201. 50 For example, dam tshig rjes su dran pa ni smgon mthong ba’i don la gang dran zhe na/ (Sukusuma, D 118a.4; P 146ba.7). 51 de rnams ni sngon mthong ba’i don la rjes su dran pas na rjes su dran pa’i yan lag go// (Sukusuma, D 118b.6; P 143a.3-4). 52 da ni dga’ bral gyi ye shes kyi rang bzhin sprul pa’i thig le gsungs pa/ (Sukusuma, D 118b.6-7; P 143a.4). 221 colored wind maṇḍala marked by a white syllable oṃ within a symbolic implement in the heart center of Vairocana who is located at the crown of the yogin’s head. At his throat he should visualize a white water maṇḍala marked by a red syllable āḥ within a symbolic implement in the heart center of Amitābha. At his heart center he should visualize a red fire maṇḍala marked by the black syllable hūṃ in the heart center of Akṣobhya. In heart center of the samayamudrā between his two breasts the yogin should visualize the seed syllable of his own deity in the center of a yellow earth maṇḍala. The syllable oṃ represents coming, āḥ represents abiding, and hūṃ represents going, while the seed syllable of his personal deity represents freedom from these three. The practitioner is to visualize emanating these elemental maṇḍalas out of his body through his nostrils (different nostrils and different amounts of force are specified for each maṇḍala), drawing them back in, and then holding, along with the exhalation, inhalation, and holding of the breath and the syllables hūṃ, oṃ, and āḥ, respectively. This recitation practice is said to bring about “a wisdom that is free from drawing in, abiding, and letting go.”53 Following a slightly unclear (to me) series of mathematical explanations, it is stated that for a great yogin, this recitation is performed constantly, and thus takes place 21,600 times each day. This number is standard in later tantric literature as the number of movements of the internal winds that occur in the body on a daily basis.54 Performing such recitation is said to bring the yogin to the realization of all phenomena as illusory and to “share the fortune of the lords of the tenth bhūmi,” thus linking the abilities gained through this practice with the accomplishment of the highest bodhisattva bhūmi accomplished through exoteric Mahāyāna practice.55 This “natural recitation” is said to have been taking place since beginningless time, but without reliance upon a guru the practitioner is unable to realize it.56 The Dvitīyakrama distinguishes the “natural recitation” of vajrajapa from “external recitation,” and states that it is unnecessary for a yogin who abides within the reality of the natural recitation to practice external recitation, which is described as an “obstacle to meditation.57 Vaidyapāda explains further that external recitation, involving focusing upon making the sounds of mantras and counting them, is an obstacle to his practice because it is a distraction for the yogin.58 However, the Dvitīyakrama later notes that once he “abides within that reality” it is not a contradiction for a yogin to also engage in external recitation.59 Taken together, these statements suggest that “external” (i.e. ordinary!) mantra recitation may be considered an obstacle to the yogin’s full realization of suchness, because it distracts him from the focus on the more internal practices through which he cultivates that suchness, but once he has realized it to the degree that he “abides in that reality,” external mantra recitation is no longer an obstacle to his practice.60 Remaining in the practice of this ritual is said to “transfer great omniscience” into the mind of the practitioner, and Vaidyapāda adds that this extraordinary result happens in a single lifetime because of traversing an extraordinary path.61 53 ‘jug dang gnas dang ldang ba slas// grol ba’i ye shes… (Dvitīyakrama, verse 229c-d). 54 See, for example, Dorje and Kongtrül 2014, 254. 55 Sa bcu pa’i// dbang phyug rnams dang skal mnyam ste// (Dvitīyakrama, verse 233, c-d). 56 While this point is clear in the Dvitīyakrama itself (vv. 234-235) Vaidyapāda makes it even more explicit. He writes: “Because since beginningless time all sentient beings have arisen together with wind, they remain in the vajra recitation. But without being accepted by a teacher one will not realize this.” thog ma med pa’i dus nas sems can thams cad kyang rlung dang lhan cig tu byung bas na rdo rje bzlas pa la gnas kyang bla mas ma zin pas rtogs par mi ‘gyur ro// (Sukusuma, D 120b.5-6; P 145b.2-3) 57 Dvitīyakrama, verse 235. 58 Sukusuma, D 120b.6-7; P 145b.3-4. 59 Dvitīyakrama, verse 238ab. 60 See also my discussion in Chapter Three on the rhetoric of non-action in Buddhajñanapāda’s writings. 61 Sukusuma, D 121a.5-6; P 146a.3-4. 222 The Function of the Indestructible Bindu, the “Relative Form” of Nondual Wisdom Following the Dvitīyakrama’s presentation of the bindu yogas, all three of which, we remember, are also described (though in less detail) in the Muktitilaka, and the first two of which appear in an abbreviated form in the Caturaṅga/Samantabhadra-sādhana, we find a short excursus on the nature of indestructible bindu as the source of the entire relative phenomenal world. As I discussed already in Chapter Three, the Dvitīyakrama contends that this indestructible bindu alone remains when all else is destroyed at the end of an aeon, and then functions as the source for the phenomenal world’s re-emergence. In this passage the Dvitīyakrama refers to this indestructible bindu as the “relative form” of nondual wisdom. As we saw in Chapter Three, Buddhajñānapāda frequently describes nondual wisdom more generally (i.e. not specifically in its “relative form” as the indestructible bindu) as both identical to awakening and as the source of the phenomenal world. While the Dvitīyakrama does not address this point directly, identifying the indestructible bindu with nondual wisdom, even just as its relative form, may serve to support the theoretical framework behind the perfection stage yogas. That is, if the bindu that is manipulated through these yogic sequences is itself the relative form of nondual wisdom, this would serve to support the contention that such manipulations are effective in bringing about a stabilization of the yogin’s experience of that wisdom, which was first indicated to him during initiation, thus leading to its full realization and thus to awakening. Following this short presentation of this important function of the indestructible bindu, the Dvitīyakrama asserts that for as long as the bindu remains embodied, it brings about engagement in all sorts of actions. It is for this reason, the text states, that the meditation upon this indestructible bindu is taught.62 We can perhaps understand this to mean that since while the bindu—nondual wisdom in its “relative form”—remains in the body, an embodied being will engage in karmic activity, and will therefore not escape from saṃsāric existence. The practices of the bindu yogas are therefore taught in order that through their practice a yogin can transcend the state of an embodied sentient being and thus arrive at the state of awakening. Repetition of the Bindu Yoga with Maṇḍala Dissolution The instructions found in the next section of the Dvitīyakrama are not found in Buddhajñānapāda’s other writings on the perfection stage. What is presented here is precisely the same as the earlier description of the indestructible bindu practice (the first bindu yoga), up until the point when the nectar from the indestructible bindu in the hearts of the tathāgatas is taken up and emerges from the vajra path. At this point, in the previous version of the practice the nectar was drawn into the practitioner through his nostrils and dissolved into the bindu. Here in this second version of the practice, we find instead a dissolution process, similar to those found at the conclusion of many generation stage sādhanas. As the nectar from the tathāgatas is drawn inwards, it “draws all sentient beings and buddhas along with the inanimate together,” bringing them first into the vajrapañjara—the outer protection circle that was visualized earlier in the generation stage practice—which itself gradually dissolves inwardly, until nothing remains but the indestructible bindu, within which the yogin should gently hold his mind. Finally, even the maṇḍala appearing within that bindu is gathered in until the yogin holds the mind—Vaidyapāda specifies that this is done “according to the guru’s instructions”—within the “self-appearing bindu,” the bindu that is itself “blessed by one’s innate nature.”63 The practitioner is encouraged to meditate on this for as long as possible, bringing the mind repeatedly into the bindu. When 62 Dvitīyakrama, verse 245. 63 Dvitīyakrama, verses 259-263. Sukusuma, D 123b.1-123b.6; P148b.6-149a.4. 223 the mind emerges from there, the yogin should repeat the process of re-illuminating the whole of the “nested” maṇḍalas of the previous visualization, which again gradually dissolve into one another.64 As the mind is held within the bindu, and the life-force is also held there, the elements gradually dissolve—in the same way as was described above in the meditation on the secret bindu—and the five signs that accompany this dissolution occur “because one has entered into Vajrasattva.”65 Vaidyapāda specifies that these signs only occur when the practice is done with the secret bindu, suggesting that perhaps that these instructions on the dissolution phase can be applied to the practice of both of the bindu yogas, but that it is only when applied to the practice of the secret bindu that the signs actually unfold.66 He explains further that “Vajrasattva”—into which the Dvitīyakrama says the yogin “enters” by means of this process is the—“fundamental wisdom without reference point.”67 When the practitioner who holds his mind in the bindu in this way begins to experience signs such as yawning, laughing, trembling, and so forth, the bindu should be “emanated, by means of the higher stage, making it pervade everything” which Vaidyapāda explains to mean directing one’s focus from the bindu to the symbolic implement, jñānasattva, and so forth, emanating light from the bindu outwards and making it pervade everywhere, just as before. Vaidyapāda explains that this procedure is done in order to reverse the occurrence of the five signs.68 The result that the Dvitīyakrama says is attained through this practice—great non-abiding nirvāṇa—is obtained, according to Vaidyapāda, specifically through repeatedly bringing about and then reversing these signs.69 As noted above, under ordinary saṃsāric circumstances these signs accompany the death process, so bringing them about in a controlled way in the yogic context, but also reversing them (such that actual physical death does not take place) is the process by means of which the yogin attains the final result of awakening. The Blisses and the Branches of Yoga: Precursors of Later Yogic Systems in Buddhajñānapāda’s Writings While the system of three bindu yogas appears to be unique to Buddhajñānapāda’s Guhyasamāja practice system, a number of its individual elements, including the sūkṣma yoga and the practice of vajrajapa with the visualization of elemental maṇḍalas, are very much part of later systems of perfection stage yogic practice. There are also several systems found within Buddhajñānapāda’s perfection stage practices that appear to be precursors to what we find in later yogic systems. In particular, the typology of three blisses mentioned in the Dvitīyakrama and elaborated in Vaidyapāda’s Sukusuma and his Muktitilaka-vyākhyāna is one of the earliest such categorizations of the blisses (dga’ ba, ānanda) associated with sexual yoga, which in later systems were usually typified as four-fold. Additionally, three among the six practices of the “six-branch yoga” (sbyor ba yan lag drug, ṣaḍaṅgayoga) that became important aspects of many later perfection stage practice systems are mentioned in the Dvitīyakrama, though not always by the names that are typically used for them in the later literature. Thus, Buddhajñānapāda’s perfection stage system provides an important window into the early development of these later, more widely known and studied systems of perfection stage practice. The Three Blisses 64 Dvitīyakrama, verses 264-266. 65 Dvitīyakrama, verses 267-269a. 66 Sukusuma, D 124a.2; P 149a.8-b.1. 67rtsa ba’i ye shes dmigs pa med pa (med pa] P, D om.) ni rdo rje sems dpa ste/ (Sukusuma, D 124a.3; P 149b.2). 68 Sukusuma, D 124a.4; P 149b.3. 69 Dvitīyakrama, verse 270. Sukusuma, D 124a.6; P 149b.5-6. 224 The typology of “blisses” (ānanda) associated with sexual yogic practice is an important feature of most later systems of tantric Buddhist sexual yogic practice. Ronald Davidson has identified Buddhajñānapāda’s typology of blisses set forth in the Dvitīyakrama as the first emergence of such a system in the tantric Buddhist textual record.70 As noted in the discussion above, the three bindu yogas described in the Dvitīyakrama are connected in Buddhajñānapāda’s writings with a system of just three blisses (dga’ ba gsum), rather than the four of the later tantric traditions. At the end of the section on the three bindu yogas in the Dvitīyakrama Buddhajñānapāda makes this connection between the three bindu yogas and the three blisses explicit: This was the authentic teaching of the ritual Of meditating on the three bindus That correspond with the three blisses. |241|71 The three blisses themselves are listed in verses 290-91 of the Dvitīyakrama as bliss (dga’ ba, ānanda), middling bliss (dga’ ba bar ma, *madhyamānanda72), and the bliss of cessation (dga’ ba dang bral ba, viramānanda).73 While the names of the blisses here are not precisely identical with those in later systems, the three-fold set represented in Buddhajñānapāda’s works does appear to serve as the basis for the later four-fold systems. The later systems of four blisses differentiate between bliss (ānanda) and supreme bliss (paramānanda), a distinction that is not found in Buddhajñānapāda’s system; lack the middling bliss (*madhyamānanda) of his system, but instead include something called innate bliss (sahajānanda); and correspond with Buddhajñanapāda’s system in their inclusion of the bliss of cessation (viramānanda). There is, with regard to these later systems of four blisses, quite a bit of debate in traditional sources on 70 Davidson suggests that “the process of ecstatic differentiation and its eventual association with sahaja first emerges in the later writing of Buddhajñānapāda, whose Dvikramatattvabhāvanā-mukhāgama shows an evolution in this direction,” (Davidson 2002b, 60). 71 de ltar dga’ gsum bye brag gi// thig le rnam gsum bsgom pa yi// cho ga yang dag bstan pa’o// (Dvitīyakrama, verse 241). 72 To my knowledge the name of this second bliss in Buddhajñānapāda’s system is not attested in any Sanskrit sources. The first and third blisses from his system, however, have the same name as the first, and the third or fourth (depending on the system) blisses of later tantric systems, which are attested in Sanskrit sources. 73 Dvitīyakrama, verses 290-91. I speculate (perhaps wildly) that the names for these three blisses and the fact of their being three in number may be related to a line from the Sarvabuddhasamāyoga-tantra that was incorporated into the Dvitiīyakrama in the verse immediately after the one where the three blisses are, on my reading, first mentioned; both verses describe the culmination of the third initiation. The line from the Sarvabuddhasamāyogatantra reads, “Neither passion, nor dispassion, nor something in between is perceived.” (‘dod chags chags bral bar ma mi dmigs (Dvitīyakrama verse125b), c.f. Sarvabuddhasamāyoga, 1.3ab. na rāgo na virāgaś ca madhyamā nopalabhyate|). While the context of this line in the first chapter of the Sarvabuddhasamāyoga-tantra, from which it is drawn, is not one of sexual yoga, given the strong parallels between this verse and the names of the three blisses that are given in Buddhajñānapāda’s system, along with the incorporation of this line into the Dvitīyakrama precisely in the context of sexual yogic initiatory practice in which the three blisses are said to be experienced, I wonder if the line from the Sarvabuddhasamāyoga may have served as a scriptural source or inspiration for the classification of the blisses as three-fold, as well as for the names ascribed to them in Buddhajñānapāda’s system. Certainly, Vaidyapāda does not take the line that way, however; he understands it to be a three-fold description of suchness itself, which he says is free from a conceptualization of any of these three aspects (Sukusuma, D 109b.7- 110a.1; P 132a.5-6). Nonetheless, my suspicion remains. Ronald Davidson has suggested that the source of the three blisses referenced in the Dvitīyakrama may have been the oral tradition, specifically the teachings of Buddhajñānapāda’s guru Pālitapāda (Davidson erroneously refers to this guru as *Bālipāda as he did not, at the time of writing his article, have access to Sanskrit sources that we now have confirming the name of this guru) (Davidson 2002, 62). My speculation here with regards to a scriptural inspiration for the three-fold system of blisses does not necessarily contradict Davidson’s suggestion that this system may have been passed down to Buddhajñānapāda by means of an oral tradition. 225 the precise sequence in which the four blisses arise,74 and while the sequence of the blisses in Buddhajñānapāda’s system is clear—the so-called “middling bliss” is obviously the second of the three—the precise nature and function of the second and third blisses in his system is not entirely clear. As we shall see, this point may be clarified to some degree with reference to the debates that took place on the sequence of the four blisses in later sources. There are some indications in both Buddhajñānapāda’s and Vaidyapāda’s writings that the second bliss in Buddhajñānapāda’s system, the middling bliss, corresponds with the innate bliss of the later systems. While this latter term does not occur in Buddhajñānapāda’s own writings, Vaidyapāda—who also seems to uphold a system of only three blisses in accordance with Buddhajñānapāda’s presentation—does on at least two occasions use the term “innate bliss” though it is not made entirely clear how (or, indeed, even if) he connects this innate bliss with the three blisses of Buddhajñānapāda’s system. In his commentary on a passage of the Dvitīyakrama that instructs the yogin, Then, with great passion Engage in physical practice with her; Practicing this play in an isolated place You should examine bliss. |105|75 Vaidyapāda identifies the “bliss” that the yogin should examine as “innate bliss” (lhan cig skyes pa’i dga’ ba) which he says “is composed of the three [aspects].”76 The term “innate bliss” appears at least once more in Vaidyapāda’s writings, this time in a citation that he provides (in both the Sukusuma and the Muktitilaka-vyākhyāna) and ascribes to a work called The Precious Garland (Rin chen phreng ba).77 That citation mentions “innate bliss” (dga’ ba lhan cig skyes) as something that arises through practice with a tantric consort. In the passage in the Sukusuma where Vaidyapāda provides this citation he indicates that Buddhajñānapāda himself cited this particular passage, though the citation is not found in any of his extant writings.78 It is possible that Vaidyapāda is referencing Buddhajñānapāda’s use of the citation in the context of oral instructions, but in any case, none of Buddhajñānapāda’s surviving writings employ the term “innate bliss,” at all. Although Buddhajñānapāda himself clearly states, as we saw above, that the three blisses are associated with the three bindu yogas, we must rely on Vaidyapāda’s Sukusuma to specify the details of this association, though it proceeds in a rather straightforward manner. That is, according to Vaidyapāda’s commentary, the first bindu yoga of the indestructible bindu is associated with bliss, the second bindu yoga of the secret bindu is associated with the middling bliss, and the third bindu yoga of the emanated bindu (otherwise known as vajra recitation) is 74 See Isaacson and Sferra 2014 (esp. 96-100) for an overview of the different positions on the sequence of the four blisses. 75 de nas rab tu chags ldan pas// de dang lhan cig lus kyi ni// spyod pas dben pa’i gnas su spyad// rol pas dga’ ba brtag par bya// |105| (Dvitīyakrama, verse 105). 76 lhan cig skyes pa’i dga’ ba ste gsum gyis bsdus pa’o. (Sukusuma, D 107b.2; P 129a.7). It seems that these three aspects, for Vaidyapāda, are actually the three kāyas. In his comments on verse 397 of the Dvitīyakrama Vaidyapāda describes the bliss that is experienced during the prajñājñānābhiṣeka as “comprised of the three kāyas” (sku gsum gyis bsdus pa’i dga’ ba) (Sukusuma, D 139a.2; P 167b.3). 77 This is the verse that describes the three types of ācārya that are mentioned several times in Buddhajñānapāda’s and Vaidyapāda’s works. The verse is cited in both the Muktitilaka-vyākhyāna and the Sukusuma, but the citation is only identified as a verse from the Rin chen phreng ba (phreng ba] P, phrod pa D) in the Muktitilaka-vyākhyāna (Muktitilaka-vyākhyāna, D 47b.5-7; Sukusuma, D 88a.6-7). I have been unable to identify this work beyond the title that Vaidyapāda himself provides for it. See note 10 of my translation of the Dvitīyakrama for more details on this citation. 78 Sukusuma, D 88a.6. 226 associated with the bliss of cessation.79 In this way, the practice of the three bindu yogas might appear to represent a progression through the blisses, with the “bliss of cessation” of the third bindu yoga as the culminating experience. However, it is not entirely clear that this is the case, and there is some indication that the second of the three blisses in Buddhajñānapāda’s system, the middling bliss, is equated with the experience of a glimpse of suchness during the prajñajñānābhiṣeka. Precisely how this equation relates to the association of that second bliss with the second yoga is also not completely clear. And, as we shall see below, the third yoga, and therefore the third among the three blisses, is associated in several places in Buddhajñānapāda’s writings with seminal emission, not only in the context of the third initiation (in which emission is, as we will see in Chapter Seven, a standard feature), but also in the context of yogic practice. Moreover, both the second and the third yogas are described in the Dvitīyakrama in ways that might suggest their supremacy—the five signs that come about through practicing the secret bindu yoga are said to result in non-abiding nirvāṇa,80 and the practice of the third bindu yoga of the emanated bindu is said to transfer great omniscience into a practitioner’s mindstream.81 Thus, despite the fact that, in terms of their order, the bliss of cessation clearly comes later, it is not immediately obvious that one of these two blisses is somehow superior to the other. However, there are some further indications in Buddhajñānapāda’s own writings, and in Vaidyapāda’s, which, when taken in the context of the later debate on the sequence of blisses, and particularly in terms of some helpful observations about that debate that have been made by Harunaga Isaacson and Francesco Sferra, can perhaps bring us a bit closer to an understanding of the three blisses in Buddhajñānapāda’s system. Let us first take a look at one passage in his writings that suggests a connection between the middling bliss of Buddhajñanapāda’s system and the innate bliss of later systems, and that also suggests, as just noted, that this middling bliss may be equated with the experience of a glimpse of “an absence” that is identified with the suchness seen directly during the prajñajñānābhiṣeka. That passage—part of which is cited in quite a number of later works precisely in reference to the debate on the sequence of blisses (but with a slightly different reading)82—is found in the Dvitīyakrama in a set of verses on the culmination of the third initiation, where the yogin is instructed to observe and stabilize the blissful experience that comes about through sexual union with a partner: From the uniting of the realm of space and the vajra, great bliss that has genuine vision arises, which brings about genuine bliss. 79 dga’ ba’i ye shes kyi rang bzhin mi shigs pa’i thig le bsgom pa… (Sukusuma, D 114b.2-3; P 137b.8); dga’ ba bar ma’i ye shes kyi rang bzhin gsang ba’i thig le… (Sukusuma, D 116a.7; P 140a.3-4); dga’ bral gyi ye shes kyi rang bzhin sprul pa’i thig le….(Sukusuma, D 118b.6-7; P 143a.4). 80 Dvitīyakrama, verse 200. 81 Dvitīyakrama, verse 240. 82 Apart from its inclusion, along with most of the rest of verses 124-125 from the Dvitīyakrama in Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi , the second line (with some variants) is cited alone in a number of other later works: the Caturmudrānvaya (attributed, by some authors at least, to Nāgārjuna; p 32), the Abhiṣekanirukti (fol. 43 r) Kumāracandra’s Ratnāvalī (p. 102), Rāmapāla’s Sekanirdeśapañjikā (see Isaacson and Sferra 2014, 275), and the Kriyāsaṅgrahapañjikā (chapter 6, prajñājñānābhiṣekavidhiḥ, st. 13ab). This list of citations is provided in Isaacson and Sferra (2014, 275 n 120) in the notes to their translation of the Sekanirdeśapañjikā, which cites the passage twice. The page numbers that I give here are those provided in their citation, and include sources I, myself, have not looked at. The interested reader is therefore directed to Isaacson and Sferra’s bibliography for further details. I discuss the implication of two of the variant readings below. 227 Between the [bliss of] cessation and bliss an absence83 is seen and should be stabilized.84 Here, in the second line, we find a mention of “bliss” (dga’ [ba]), “the bliss of cessation” (chag bral85), and something that lies “in between the two blisses” (dga’ gnyis bar du). I take these three to refer precisely to the three blisses—bliss, the bliss of cessation, and middling bliss—of 83 dben] sugg. em. based on V (D and P), bden D C P N S) nyid (nyid] sugg. em. based on parallel verse in Vaidyapāda’s Yogasapta (see Yogasapta, D 71a.3; P84b.4), gnyis D C P N S). Here Vaidyapāda states, “The absence of the two blisses should be seen by means of the oral instructions, and [the text] is stating that one should stabilize that.” dga’ ba gnyis gyis dben pa de man ngag gis mthong ba de la blo brtan par gyis shig ces gdams pa’o// (Sukusuma, D 109b.5; P 132a.2). I have emended the Dvitīyakrama here in accordance with Vaidyapāda’s commentary and the Yogasapta to read “an absence” (dben nyid), rather than the implausible “the two truths,” (dben gnyis), found in all recensions of the Dvitīyakrama itself. However, Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi (as well as all of the other citations of this line in later Sanskrit sources of which I am aware, on which see note 82 above), instead of “an absence,” here reads “the target” (lakṣya). Although after much deliberation I have not chosen to do so, I will note here that it was very tempting to make an even more serious emendation of the Dvitīyakrama from bden gnyis not just to dben nyid following Vaidyapāda, but to ‘ben nyid, “the goal,” to match the Sanskrit (lakṣya) of all of the later Sanskrit sources of which I am aware that cite this passage. However, I have resisted doing this because, in addition to the fact that for this more metaphorical meaning of lakṣya (i.e. as the “target” of awakening rather than a physical archery “target”) one would prefer mtshon bya rather than ‘ben (though ‘ben is attested as a translation of lakṣya, at least in the archery target sense), while making the emendation to ‘ben might work in the Dvitīyakrama itself, it is much more difficult (though not absolutely impossible) to coherently make the emendation to ‘ben in Vaidyapāda’s commentary on the relevant passages. In the passage from the Sukusuma cited above in this note it would not be terribly difficult to make that emendation, but a later passage from the Sukusuma—“How is it that there is the absence of two blisses? Neither passion, dispassion, nor something in between are observed means that there is no conceptualization in terms of these three.” (dga’ ba gnyis kyis ji ltar dben zhe na/ ‘dod chags chags bral bar mi dmigs zhes te ‘di gsum gyis rtog pa med pa’o// (Sukusuma, D 109b.7-110a.1)—would be rendered significantly less coherent were one to emend dben to ‘ben. Thus, the Sukusuma reads much more smoothly and naturally without this emendation, and Vaidyapāda’s commentary is several centuries earlier than any of the Sanskrit sources that include this line with the reading of lakṣya, so my guess for the moment is that the Dvitīyakrama and the Sukusuma represent an earlier recension of the line that, in fact, read “an absence” (dben nyid; I unfortunately cannot guess what the Sanskrit may have been), and which later underwent some change to the reading of “the target” (lakṣya). As further evidence that Buddhajñānapāda’s own text probably did not read ‘ben/lakṣya, the line as cited in Abhayākaragupta’s Āmnāyamañjarī, which reflects the later reading, translates lakṣya in a way that is more expected, not as ‘ben but as mtshon bya (Āmnāyamañjarī, D 67a.1). In either case, however, I understand the verses to be saying more or less the same thing with either reading. That is, in the Dvitīyakrama the “absence” is identified several verses later with the perfection stage, which, as we have seen, is for Buddhajñānapāda identified with suchness itself. This is nothing other than the “target” that is referenced in the later recension of the verse. 84 This verse has been transmitted (and perhaps also translated) problematically in the Dvitīyakrama, but fortunately a parallel verse survives in Sanskrit in Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, which, along with the second line as cited in a number of other sources (see note 82 above for a list of these) is helpful in clarifying some, but not all, parts of the verse as it appears in the Dvitīyakrama. The parallel verse (actually 1.5 verses) in the Saṃkṣiptābhiṣekavidhi reads: khadhātuvajrasaṃyogāt saṃsparśāc ca mahādbhutaṃ/ sukham utpadyate yat tat paramānandadāyakaṃ// 10// viramānandayor madhye lakṣyam vīkṣya dṛḍhīkuru/ (ed. Sakurai 1996, 418. Thanks to Harunaga Isaacson for bringing this parallel to my attention). I have relied on the Sanskrit parallel verses in providing a clearer reading of the Tibetan translation of these verses from the Dvitīyakrama. For example, mahādbhutaṃ/ sukham utpadyate yat tat paramānandadāyakaṃ// viramānandayor madhye is extremely helpful in clarifying the confusing bde chen ‘byung byar ‘gyur// gang gang yang dag dga’ byed chags bral dga’ gnyis bar du, with which it appears to be precisely parallel, but which would otherwise not be naturally read that way just on the basis of the Tibetan. See notes 263-268 in my translation of the Dvitīyakrama for more details. 85 The Sanskrit from Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi reads viramānandayor, and it seems that this is indeed probably what the Tibeatan translators were reading here. They seem to have understood virama in the compound to mean viramānanda (thus they the two members of the compound would be viram[ānanda] and ānanda), and thus chags bral is here a translation of viramānanda. The more common translation of viramānanda, however, would be dga’ bral rather than chags bral. 228 Buddhajñānapāda’s system.86 If indeed the reference to “in between” (which could equally well be translated as “in the middle”) does refer to the “middling bliss” in Buddhajñānapāda’s system, this would mean that this middling bliss is also identified with the “absence,” that the verse says is to be seen and stablilized. This absence is identified several lines later in the text with the perfection stage. This seems, then, to refer to the glimpe of suchness that is experienced directly by a disciple in the context of the prajñājñānābhiṣeka. It would also therefore suggest that Buddhajñānapāda’s “middling bliss” corresponds to the “innate bliss” of later systems, since, as Isaacson and Sferra have shown, this line is frequently cited by proponents of the view that the innate bliss—which they point to as the referent of the term “in between” here, and which is, for them, the culminating moment of the prajñājñānābhiṣeka—occurs between the supreme bliss and the bliss of cessation.87 The recension of the verse cited by later authors in support of the position that innate bliss holds the third, rather than the fourth, place in the sequence of blisses varies slightly from the verse as it reads here in the Dvitīyakrama, and though the variant is small it may help us to understand Buddhajñānapāda’s verse, and the role of the middling bliss here more effectively. While in the Dvitīyakrama88 and Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi89 the line reads: viramānandayor madhye…, the line in Rāmapāla’s Sekanirdeśapañjikā (and apparently also in the Caturmudrānvaya, from which Rāmapāla appears to cite the line) reads: paramaviramayor madhye lakṣyaṃ vīkśya dṛdhīkuru.90 In both cases the first compound refers to two among the series of blisses that arise in the prajñājñānābhiṣeka. In Buddhajñānapāda’s system, in which there are just three blisses, the compound seems to refer to the bliss of cessation (viramānanda) and bliss (ānanda), the third and first of the blisses, respectively. In the later systems under discussion in the Sekanirdeśapañjikā, there are four blisses, among which the compound appears to reference supreme bliss (paramānanda; the second) and the bliss of cessation (viramānanda; the fourth according to the system upheld by Rāmapāla, who cites the passage). On the one hand this is not a significant difference, given that what is at stake is what lies between the two, and in Buddhajñānapāda’s system there is no division into the first two blisses (ānanda and paramānanda, respectively) of the later system, so bliss in his system could correspond with 86 Following the parallel Sanskrit verse from Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi (see note 84), the three blisses are mentioned in the first pāda of the second line, which reads viramānandayor madhye. As did the Tibetan translators, I take the two members of the compound to be viramānanda (dga’ bral), “the bliss of cessation” (see previous note), and ānanda (dga’ ba), “bliss.” Madhye, “in the middle” here, I suggest, refers to *madhyamānanda (dga’ ba bar ma), the “middling bliss,” in Buddhajñānapāda’s system. 87 Isaacson and Sferra 2014, 98. As I noted in note 83 above, the later texts that cite this passage have the reading lakṣya (mtshon bya), “the target,” rather than dben nyid, “an absence.” However, again, given the fact that this “absence” is identified in the Dvitīyakrama with the perfection stage, it seems to me that the difference in terminology here does reflect a difference in the ultimate referent of the two terms. 88 This citation of the line in Sanskrit is of course from the Saṃkṣiptābhiṣekavidhi rather than the Dvitīyakrama, given that the latter does not survive in Sanskrit. However, the Saṃkṣiptābhiṣekavidhi does appear to precisely parallel the Dvitīyakrama here at the beginning of the line, and it is easier to use the Sanskrit verse to show parallels with the later verse from the Sekanirdeśapañjikā, which is also extant in Sanskrit. 89 While Vāgiśvarakīrti’s Saṃkṣiptābhiṣekavidhi dates to around the 11th century, and he therefore certainly knew systems of four blisses, rather than just three, his initiation manual is devoted to initiation in the Guhyasamāja tradition, and he incorporates not just this single line, but most of verses 124 and 125 from the Dvitīyakrama (with a few variants). It therefore seems very likely that the Dvitīyakrama itself was his source for the verses, including this line, and therefore the fact that his recension of this line corresponds with Buddhajñānapāda’s, in whose system there were only three blisses, with no distinction made between ānanda and paramānanda suggests that the verse may have been modified by later authors to read paramaviramayor, in order to more clearly uphold their position on the sequence of blisses. The modification to lakṣyam from whatever Sanskrit is behind dben nyid (a modification that has made its way into the Samkṣiptābhiṣekavidhi) may also be related to this same concern. 90 For the line as cited in the Sekanirdeśapañjikā and in the Caturamudrānvaya see Isaacson and Sferra 2014, 98. 229 either (or both) of the first two blisses (bliss and supreme bliss) in the later systems. But on the other hand, the comparison between the two verses makes it even more clear that what occupies the middle place between the two, which corresponds in Buddhajñānapāda’s system to the middling bliss (*madhyamānanda), and in Rāmapāla’s system to the innate bliss (sahajānanda), should indeed be understood as parallel: both function as the “target” (lakṣyam) that is to be marked by the practitioner during the initiation. Although the term used in the Dvitīyakrama here is not “the target,” but rather “an absence” (dben nyid), that “absence” is identified several lines later in the Dvitīyakrama with the perfection stage, which is indeed precisely the “goal” that is glimpsed during the third initiation, so I do no think that the difference in terms in the verses is here reflective of a significant difference in the ultimate referent of the terms. If the middling bliss of Buddhajñanapāda’s system is indeed effectively parallel to the innate bliss of the later systems, then we might also profitably look to the relationship of the innate bliss to the bliss of cessation in those systems in order to better understand the relationship between the middling bliss and the bliss of cessation in Buddhajñanapāda’s system. Here the issue becomes more complicated, since, as I noted above, there are two different positions on the succession of the four blisses in these later systems, and, according to Isaacson and Sferra’s compelling analysis of the topic, these differences seem to be related precisely to different understandings of what constitutes the bliss of cessation, and therefore how it relates to the innate bliss. In brief, Isaacson and Sferra propose that the difference between the two positions on the sequence of blisses arose on the basis of a distinction in the understanding of the term (and presumably also the function) of the bliss of cessation, viramānanda. For the proponents of the sequence that places innate bliss in the third place in the four-fold sequence, before the bliss of cessation (i.e. those who cite precisely the line that parallels the one from the Dvitīyakrama: “Between the [bliss of] cessation and bliss an absence (or, in later versions, “the goal”) itself is seen and should be stabilized” ), they argue that the term virama in viramānanda is understood as “Cessation, as a post orgasmic experience of the descent to a lower state, re-entering the world of conceptual constructions, vikalpa,” thus necessitating that the crucial moment of seeing reality directly—the innate bliss—had to come before this descent. 91 Proponents of the other sequence, in which the bliss of cessation is the culmination of the sequence, they suggest, “understood the prefix vi- in virama or virama as having either intensifying sense…or alternatively as expressing diversity.”92 Isaacson and Sferra, moreover, suggest that prior to the fully developed debate on the sequence of four blisses, which seems to have arisen in relation to, and perhaps even out of the Hevajra-tantra, there were precursors to the two divergent positions on the sequence of blisses. They suggest that the first proto-position, corresponding to the later position in which the bliss of cessation follows innate bliss, is exemplified by “the (probably scriptural) passage…quoted in the Caturmudrānvaya,” i.e. precisely the line referencing the three blisses 91 Isaacson and Sferra 2014, 100. 92 ibid. 230 that we find in the Dvitīyakrama (with small variations as noted above).93 (The alternative protoposition, they note, is attested in the Guhyasiddhi.94) Following this proposal, the presence of the line in the Dvitīyakrama, then, might suggest that for Buddhajñānapāda, the middling bliss is the supreme one, and that the bliss of cessation involves a “post orgasmic…descent to a lower state” involving “the world of conceptual constructions, vikalpa.”95 However, in the Dvitīyakrama the association of the bliss of cessation with the aspect of a “descent” into conceptuality does not seem to apply, in particular with reference to the explicit link of the bliss of cessation with the third bindu yoga, the yoga of the “emanated bindu.” While the Dvitīyakrama corresponds with the first position described by Isaacson and Sferra in identifying the moment of the middling bliss (equivalent here to the innate bliss of later systems), the absence/goal that is to be glimpsed—the perfection stage or the dharmakāya itself96—as something that precedes the bliss of cessation, the bliss of cessation itself appears to be associated in the Dvitīyakrama not so much with a descent from the dharmakāya as with an emanation from or emergence out of that state.97 The bliss of cessation is explicitly associated in several places in Buddhajñānapāda’s writings with seminal emission, which is itself associated with emanation—of the maṇḍalacakra, or of other forms. The verse in the Dvitīyakrama introducing the practice of the third bindu yoga of the “emanated bindu” (sprul pa’i thig le)—which, as noted above, is linked specifically with the bliss of cessation reads: Even when it emerges from the jewel It is made to pervade the three realms— This is called meditation upon the emanated bindu.98 Vaidyapāda’s writes: “Then, having performed the meditation on the secret bindu, now, in order to indicate the purpose of the meditation on the emanated bindu [the text] states, Even when it emerges from the jewel/ It is made to pervade the three realms. Thus, after the conclusion of the initiation, due to abiding in the branch of increase,99 [there is] the emanated [maṇḍala]-cakra, the meditation upon the nirmāṇakāya; or, alternatively, because of emanating the four bindus 93 Isaacson and Sferra (who, it seems, were not aware of this line’s presence in the Dvitīyakrama) suggest that it may derive from a lost tantra, since the sources that cite it tend to give it the reverence normally attributed to scripture (ibid., 98-99). They note, however, that Abhayākaragupta (who, it should be noted, holds a position on the sequence of blisses that seems to be contradicted by this statement) in his Āmnāyamañjarī, casts doubt on its scriptural authority (ibid.). Although Abhayākaragupta does, indeed, as Isaacson and Sferra have noted, cast doubt on the scriptural authority of the line, he does still provide (“in the case that it is scriptural….” lung yin na de’i cha/….) a way of interpreting the line that does not undermine his position on the sequence of blisses (Āmnāyamañjarī, D 67a.1). While Isaacson and Sferra may be correct that this line is a from a lost tantra, it is also possible that the issue of its scriptural authority or lack thereof may be with reference to its presence here in the Dvitīyakrama, in a mukhāgama, a work that lies precisely on the borderline of scripture and authored commentary. The fact that thus far we are unaware of any earlier attestations of this line, makes this possibility worth considering. It should be noted, however, that while Abhayākaragupta questions the scriptural authority of this particular line, he cites as scriptural—and attributes to the Paramādya-tantra—a verse that is parallel with the very next two lines (124 cd) of the Dvitīyakrama (Āmyāyamañjarī, D 68a.1-2)! 94 Isaacson and Sferra 2014, 99. 95 Ibid, 100. 96 I address the issue of whether the dharmakāya itself is glimpsed in the prajñājñānābhiṣeka according to Buddhajñānapāda’s system (I argue that it is!) in Chapter Seven. 97 One might say that in this sense it corresponds with the aspect of the second position described by Isaacson and Sferra in construing “the prefix vi- in virama or virama…as expressing diversity” (Isaacson and Sferra 2014, 100). 98 nor bu las ni byung nas kyang// khams gsum khyab par byed pas ni// sprul ba’i thig le bsgom par bshad// (Dvitīyakrama, verse 214a-c). 99 This may be a reference to the fifth among the seven yogas, the branch of unfolding/increasing compassion (thugs rjes rgyas pa). 231 and the rest [it is called the practice of] the emanated [bindu]. This meditation is [now] explained.”100 Here the idea of emission—the emergence of the bindu from the jewel—is linked to the idea of emanation: the emanation of the maṇḍala-cakra (similar, perhaps, to the generation stage practice in which the maṇḍala-cakra is emanated precisely from the deities’ point of union as the result of the emission of bodhicitta there), or the emanation of the “four bindus and so forth,” which may perhaps here refer to the emanation of the four elemental maṇḍalas in the practice of vajra recitation, described above, which constitutes the yoga of the emanated bindu. Despite the explicit linking of the yoga of the emanated bindu with emission in this verse, it seems unlikely that the yoga itself, the practice of vajra recitation, was meant to be performed following seminal emission. Here the association seems to be more in the second sense that Vaidyapāda mentions—the emanation of the four bindus, probably the four elemental maṇḍalas that characterize the practice of vajra recitation in this tradition. However, another verse linking the bliss of cessation with emission, found in the middle of the explanation of the secret bindu yoga, does appear to be an instruction on what the yogin is to do if, while practing yoga, he (accidentally?) emits semen. Having described a practice in which the yogin is to focus on the maṇḍala in the center of a bindu at the tip of the vajra, the Dvitīyakrama makes what appears to be something of an aside: If while [regarding] that |191| One’s mind becomes dull or weary And it emerges from the vajra Make it remain at the tip of the nose101— And examine by means of the bliss of cessation. |192|102 The description of the secret bindu yoga, the practice of which seems be contingent upon the yogin’s not having emitted the bindu of bodhicitta, continues in the subsequent verses. In fact, as I mentioned above, the less detailed instructions on the practice of the secret bindu yoga in the Muktitilaka lack this instruction on what to do if the mind becomes tired and the bindu emerges from the vajra, and simply proceed with the description of the procedures for the secret bindu yoga. This statement then, appears to be an additional instruction specific to the case in which the yogin is unable to continue retaining the bindu and emits it from the vajra, and the injunction to “examine by means of the bliss of cessation,” thus seems to associate the third of the three blisses, the “bliss of cessation” specifically with the moment of emission in the context of yogic practice. In this case, examining “by means of the bliss of cessation” seems to be a method for the yogin to also employ the circumstance of emission as part of his practice. 100 De gsang ba’i thig le bsgom par byas nas/ sprul pa’i thig le bsgom pa’i dgos pa gsungs pa/ nor bu las ni byung nas kyang// khyams gun khyab par byed pas ni// zhes te dbang gi mthar thugs rjes rgyas pa’i yan lag tu gnas pas sprul pa’i ‘khor lo ste (‘khor lo ste] D, P om.) sprul pa’i sku sgom par byed pa’am yang na thig le bzhi la sogs par spros pa’i phyir na sprul pa ste de bsgom pa (bsgom pa] P, bsgoms D) bshad do zhes so// (Sukusuma, D 118b.7- 119a.2; P 143a.8-143b.2). 101 While looking only at the root verse, one might presume that the “nose” referred to here is (as is often the case) the tip of the yogin’s penis. Vaidyapāda, however, specifies that one is to “make it remain at the tip of the nose of the goddess’ lotus.” lha mo’i padma’i sna rtser rnam par gnas par byas nas/ (Sukusuma, D 116b.6-7; P 140b.4). A similar passage in the Caturaṅga/Samantabhadra-sādhana is described in Vaidyapāda’s and Samantabhadra’s commentaries as indicating a process by means of which the yogin appears to be instructed to draw the bindu which was previously emitted into the lotus of the consort out onto the “nose tip” of her lotus by means of transforming the “prong” of his vajra into hook-like light rays that hook the bindu and draw it out to this location. (See Caturaṅga/Samantabhadra-sādhana verse 130 (D 34b.3), Vaidyapāda’s Samantabhadrī-ṭīkā (D 168a.2-5), and Samantabhadra’s Sāramañjarī (D 38a.1-3; Szántó unpublished 125). 102 gal te de la rang gi sems// |191| bying bar song ngam skyo ba na// rdo rje las ni phyir byung ste// sna rtse rnam par gnas byas nas// dga’ bral dga’ bas brtag par bya// |192| (Dvitīyakrama, verses 191d-192d). 232 In any case, there is much that still remains unclear about the three blisses in Buddhajñānapāda’s system and their relationship to the three bindu yogas. While it seems that the second bliss is linked to the direct experience of suchness in the initiatory context and the third to seminal emission in a yogic context (and perhaps also in the initiatory context?), the idea of emission here seems to be understood here not as a type of “descent” out of the state of suchness but rather as an emanation from it, like the emanation of the maṇḍala deities in the generation stage. What I have set forth here is only a preliminary look at the topic of the three blisses in Buddhajñānapāda’s system, a more detailed examination of which will certainly reward further inquiry. Perhaps a study of the bindu yogas of Buddhajñānapāda’s system in the later Jñānapāda School literature, or as they continue to be practiced in the living Tibetan tradition, will shed further light on the three blisses, and in particular on the question of the relationship between the second and third blisses, in Buddhajñānapāda’s system. The “Branches” of the Six-Branch Yoga In addition to providing an early example of a typology of blisses in the context of sexual yogic practice, Buddhajñānapāda’s writings also offer early evidence of the inclusion of several “branches” among what came to be the standard set of a “six-branch yoga” (ṣaḍaṅgayoga) in tantric Buddhist perfection stage practice.103 Systems of ṣaḍaṅgayoga (in addition to another important system of eight-branch yogas (aṣṭāṅgayoga)) are found in many different Indic yogic traditions, certainly not limited to Buddhism, and the individual branches that are included in these systems can vary rather widely.104 The earliest Buddhist system of a six-branch yoga appears to be that found in the Samājottara (vv. 141-154).105 As I have noted before, Buddhajñānapāda’s works show no evidence of familiarity with the Samājottara, while instead the Samājottara appears to show evidence of the influence of Buddhajñānapāda’s writings and practice system. Part of the evidence for this relationship can be found in the fact that the Dvitīyakrama makes reference to the practice of just three among the six yogas of the six-branch yoga as listed in the Samājottara, and one of these three yogas is mentioned in the Dvitīyakrama with a different name than the standard one by which it came to be known from the time of the Samājottara onwards. Francesco Sferra has shown the importance of the six-branch yoga as it is presented in the Samājottara, by means of the fact that even when the descriptions of the practices differ in later Buddhist systems, those works still hew to the names of the yogas from the list given in the Samājottara and frequently cite passages from that scripture even when explaining the yogas in ways that are at odds with their presentation therein.106 Given the obvious popularity of the Samājottara as a source for these yogas, the fact that Buddhajñānapāda’s works only reference three among the six practices and use a different name for one of them serves a further indication that his writings likely preceded the Samājottara. The three among the six branches that receive mention in the Dvitīyakrama are the branch of emptying (gtong pa’i yan lag), the branch of retention (bzung pa’i yan lag), and the branch of recollection (dran pa’i yan lag). With reference to the Samājottara’s classical list of six yogas, these three constitute the third, fourth, and fifth of the six yogas, respectively. Despite the fact that a full set of six yogas is not mentioned anywhere in Buddhajñānapāda’s writings, these three practices are nonetheless referred to in the Dvitīyakrama as “branches.” At the 103 Ryūta Kikuya (2000) has written on the ṣaḍaṅgayoga system in the Jñānapāda School, but the article is in Japanese, so I unfortunately have not been able to take Kikuya’s work into consideration in my discussion here. 104 See Sferra 1990 (esp. 11-15) for a comparison of a number of Buddhist and non-Buddhist ṣaḍaṅgayoga systems and a discussion of their relationship with various aṣṭāṅgayoga systems. 105 Sferra 1990, 15. 106 ibid. 233 conclusion of the section on the practice of the secret bindu yoga, the section of the text in which all three of these practices are presented, Buddhajñānapāda writes, “In this way, by means of these three branches,/ Meditate upon the secret bindu.”107 The first of these practices, termed in Buddhajñānapāda’s system the “branch of emptying,” is more commonly referred to in the Samājottara and later sources as the branch of prāṇāyāma.108 Several pieces of evidence allow us to make this identification. First, the description of the practice of the “branch of emptying” in the Dvitīyakrama (vv. 186-194) corresponds with the branch of prāṇāyāma as described in the Samājottara (vv. 147-148). The identification of the branch of emptying with prāṇāyāma is also suggested by Vaidyapāda’s Samantabhadrī-ṭīkā, his commentary on Buddhajñānapāda’s Caturaṅga-sādhana, where he mentions that the practice of the sūkṣma yoga according to that sādhana—which corresponds with the secret bindu yoga—constitutes “the branch of emptying, or cessation, the third.”109 In the traditional list of the six-branch yoga as given in the Samājottara—a work with which, Vaidyapāda, unlike Buddhajñānapāda, was definitely familiar (he wrote a commentary on it!)— prāṇāyāma is the third. A third piece of (admittedly much later) evidence is found in the writings of Tāranātha, who, in his commentary on the perfection stage practices of Buddhajñānapāda’s system, Guidance Manual on the Glorious Muktitilaka (Dpal grol ba’i thig le’i khrid yig), retains the name “branch of emptying” used in Buddhajñānapāda’s text, but also directly correlates this branch with the branch of prāṇāyāma.110 The branches of retention (dhāraṇā)111 and of recollection (anusṃṛti) are more easily identifiable as members of the six-branch yoga, as they are referred to in the Dvitīyakrama by their commonly used names. While the Dvitīyakrama mentions only half of the classic set of the six-branch yoga given in the Samājottara, it remains an important early source for these practices in a Buddhist text.112 107 de ltar yan lag gsum gyis ni// gsang ba’i thig le bsgom byas nas// |213| (Dvitīyakrama, verse 213c-d). 108 There appears to be some confusion with regard to the name of this branch in Buddhajñānapāda’s system. While all of the available recensions of the Dvitīyakrama itself identify the practice as “the branch of emptying” (stong pa’i yan lag), Vaidyapāda Sukusuma reads “the branch of casting out” (gtong pa’i yan lag) (as does Vaidyapāda’s Samantabhadrī-ṭīkā, which also mentions the practice). Tāranātha’s later commentary on the perfection stage practices of Buddhajñānapāda’s system follows the Dvitīyakrama in reading stong pa’i yan lag for this practice— Tāranātha calls it the “branch of emptying which stops the breath” (dbugs dgag stong pa’i yan lag) (Dpal grol ba’i thig le’i khrid yig, 247). 109 dgag pa dang gtong ba’i yan lag gsum pa’o (Samantabhadrī-ṭīkā, D 168a.6). 110 Dpal grol ba’i thig le’i khrid yig, 247. Tāranātha’s work generally remains quite faithful to Buddhajñānapāda’s writings in its presentation, including maintaining the unique vocabulary of this particular system. For example, in the case of the branch of dhāraṇā, Tāranātha retains the idiosyncratic spelling of gzung ba’i yan lag found in the Tibetan translations of Buddhajñānapāda’s and Vaidyapāda’s writings—he calls this branch “the branch of retention where the signs appear” (rtags snang gzung ba’i yan lag)—when listing the branches as found in the Jñānapāda School practices, but later refers to the same practice using the more common Tibetan translation ‘dzin pa’i yan lag (Dpal grol ba’i thig le’i khrid yig, 247-48). 111 This branch is more commonly rendered in Tibetan as ‘dzin pa, though this does not present much of a problem given that gzung ba and ‘dzin pa are simply different tenses—future and present, respectively—of the same verb. 112 If I am correct in suggesting that Buddhajñānapāda did not know the Samājottara, the source from which he may have drawn these yogas (apart, of course, from the fact that the entire practical content of the Dvitīyakrama is asserted to have been revealed to him by Mañjuśrī) remains unknown. However, the first branch that he mentions— the “branch of emptying”—is also described in the Dvitīyakrama as “stopping the breath” (dbugs dgag) (Dvitīyakrama, verse 194.b). One version the lists of the six-branch yogas from non-Buddhist sources listed by Sferra, includes prāṇasaṃrodha, “stopping the breath” as the name of a branch that in other versions of the list corresponds with prāṇāyāma, although the sources that Sferra lists that use this term seem to be later than Buddhajñānapāda, (Sferra 1990, 13). Also worth noting is the fact that Vaidyapāda, though he clearly is familiar with the full set of six branches, having composed a commentary on the Samājottara, (the Saṃyagvidyakara), makes no attempt to discuss the other three yogas of the six-fold set in relation to the Dvitīyakrama. 234 A Typology of the Tantric Consort and Sexual Practices: The Relationship between Buddhajñānapāda’s Writings and Kāmaśāstra The passages from the Dvitīyakrama that deal with sexual practices performed with a tantric consort are consistently focused on the soteriological function and aim of the bliss generated through sexual union. However, several passages nonetheless bear evidence of a relationship with the Indian tradition of kāmaśāstra, which was generally focused on a more worldly type of pleasure. It seems as if the Buddhist tradition of sexual yogic practice adopted some techniques from the kāmaśāstric tradition and adapted these to the aims of Buddhist practice. However, as we will see, it appears that the influence of these two literary genres was mutual, or at the very least that Buddhist tantric and kāmaśāstric authors were participating in overlapping textual communities. Daud Ali has explored some of the ways in which Buddhist tantric literature and practice seems to have influenced kāmaśāstra,113 and Buddhajñānapāda’s writings seem to demonstrate further evidence of a relationship of mutual influence. I will briefly explore here just a few of the instances where we can see evidence of the interplay between these traditions in Buddhajñānapāda’s writings: there are instances where his writings appear to rely upon kāmaśāstric sources, others where his writings may have themselves influenced the kāmaśāstric tradition, and still other instances where we can turn to kāmaśāstric sources to clarify otherwise obscure passages in Buddhajñānapāda’s and Vaidyapāda’s writings. Buddhajñānapāda’s Four-fold Typology of Tantric Consorts and The Four-fold Kāmaśāstric Typology of Women As I showed in Chapter Three, Buddhajñānapāda’s writings make it clear that he holds the practice of the sexual yogas of the perfection stage performed with a consort to be an essential component of the path to awakening. Given the importance of a partner for the tantric practices that his writings espouse, it is perhaps not surprising that the Dvitīyakrama includes a passage describing the characteristics and qualities of such a tantric consort. Since, like most (if not all?) Indic Buddhist tantric works, the Dvitīyakrama is written from the perspective of a male practitioner, it is a female partner who is described. However, the passage in the Dvitīyakrama that discusses the tantric consort not only describes the general qualities of an ideal female partner for tantric practice, it goes on to outline four different types of consort that correspond with the four buddha consorts from the Guhyasamāja-tantra’s maṇḍala. The Dvitīyakrama’s presentation of the four-fold typology of the tantric consort begins with a verse that emphasizes the crucial importance of the yogin’s relying upon a female partner in order to come to a direct realization of suchness. Because this realization is not possible without practicing with a partner, the text, incorporating a line from the Sarvabuddhasamāyoga-tantra, indicates the superiority of women among all of the myriad types of illusions that make up the practitioner’s experience of the phenomenal world: That which is luminous and joyful, equal to space—114 One will not know115 it any other way. Thus, a woman, the illusory mudrā, 113 Ali 2011, esp. pp 54-55. 114 Vaidyapāda specifies that this refers to nondual wisdom (gnyis su med pa’i ye shes) (Sukusuma, D 99a.7; P 119a.7). 115 rig] S P V(P), rigs D C N V(D). 235 Is superior among all illusions.116 |50|117 Following this statement, Buddhajñānapāda lists and describes the four consort types, which have as their “pure forms” (viśuddhi) the four Buddha consorts from the Guhyasamāja-tantra’s maṇḍala: Māmakī, Pāṇḍaravāsinī, Tārā, and Locanā.118 The four types, called kamalī,119 śāṅkhinī, citriṇī, and hastinī, are associated in the text not just with the four buddha consorts, but also with types of animals: nāgas, tigers and lions, the wild black antelope, and elephants, respectively. Each of the four types is then described in terms of mostly physical, but in some cases also personality-related, characteristics. I will not cite the full passage here (see verses 50- 67 of my Dvitīyakrama translation) but just to give a sense of the way the consort types are presented, the first of the four types, kamalī, is described as follows (and the description of the other three types follows essentially the same structure): Here, as for kamalī I will explain her shape and characteristics: |53| She is a girl who is redolent with the scent of lotus Her face is round, the tip of her nose like a mustard seed, Her nails are red and her back is bent [out of respect].120 The soles of her feet rest flat upon the earth. |54| Her body hairs coil and she is golden.121 Her breasts122 are like the fruit of the mustard plant. 116 bud med sgyu ma’i phyag rgya ni/ sgyu ma kun las khyad par ‘phags//. These two lines have strong parallels with the first two lines of the Sarvabuddhasamāyoga-tantra 1.4, which read, in Sanskrit, sarvāsām eva māyānāṃ strīmāyā praviṣiṣyate |, and in Tibetan translation, sgyu ma dag ni thams cad pas// bud med sgyu ma khyed par che// (D 151a.3). The Sarvabuddhasamāyoga also mentions the woman as a mudrā in the last two lines of the immediately preceding verse: sarvastrīmāya mudreyam advayaṃ yānam uttamam |; bud med kun gyi sgyu ma’i rgya// ‘di ni gnyis med theg pa’i mchog// (D 151a.2). Thanks to Ryan Damron for bringing these Sarvabuddhasamāyoga parallels to my attention and to Péter Szántó for sharing with me his draft edition of the Sarvabuddhasamāyoga-tantra. The two lines from the Dvitīyakrama are also paralleled in Śākyamitra’s Anuttarasandhi, included as the second stage in Nāgārjuna’s Pañcakrama, which reads: sarvāsām eva māyānāṃ strī māyaiva viśiṣyate/ (Mimaki and Tomabechi 20); sgyu ma dag ni thams cad las/ bud med sgyu ma khyad par ‘phags (Pañcakrama, D 49a.7; Mimaki and Tomabechi 20). Tomabechi (2006, 132n128) has already noticed all of these parallels and additionally notes that a passage identical to that in the Pañcakrama is found in the Vajramaṇḍālaṃkāra. 117 gsal shing rab dga’ mkha’ mnyam pa// gzhan du rig par mi ‘gyur bas// bud med sgyu ma’i phyag rgya ni// sgyu ma kun las khyad par ‘phags// |50| (Dvitīyakrama, verse 50). 118 The four consort types are described in the Dvitīyakrama verses 50a-67b. 119 All versions of the Dvitīyakrama read kamalī, which is an unusual and unexpected form. Kamalinī would be the expected feminine form that would correspond with the names for the other types. It is possible that that Tibetan translators may have simply shortened the form for metrical reasons. Vaidyapāda’s Sukusuma, however, also reads kamalī, but this may again simply be because the translators of the commentary were referencing the Tibetan translation of the root text. I have not taken the liberty of changing the text in my edition, however, as this would render unmetrical all verses in which it occurs. In any case this is an unusual term for this particular type. The type, when it is mentioned in later texts, is instead called padminī (see below). I am grateful to Mattia Salvini for a helpful conversation on this topic. 120 rgyab sgur (Skt. *kubjā?). The term normally means hunchbacked, but Khenchen Chodrak Tenphel explains that here it is meant to indicate a respectful body posture. (Khenchen Chodrak Tenphel, personal communication, January 2016). 121 ser] sugg. em. based on V(D and P), sen D C, se P N S. Vaidyapāda makes it clear that this refers to her coloring. Below Buddhajñānapāda states that her skin color is reddish. In the Cakrasaṃvara-tantra the type of woman who corresponds to kamalī is described as “reddish-golden” (Gray 2007, 236). 122 dkar ‘chang. Vaidyapāda makes it clear that this term refers to breasts: “White means milk. That which holds this are breasts.” dkar ba ni ‘o ma’o// de ‘chang ba ni nu ma ste/ (Sukusuma, D 100a.7; P 120b.6). I believe that this is likely a translation of one of the Sanskrit terms for breasts payodhara—literally “that which holds milk.” The 236 She has three wrinkles at her waist.123 Her chest is lovely and she has the [leisurely] gait of an elephant. |55| The taste of her blood is sour124 Her skin is reddish. The pure form of this goddess is Māmakī.125 The four-fold typology of tantric consorts articulated in the Dvitīyakrama, is found in some later Buddhist sources, including in Chapter 18 of the Samvarodaya-tantra,126 but it also corresponds precisely with the classic four-fold typology of women found throughout late Indian kāmaśāstra literature. The presentation of this typology in the Dvtīyakrama is an early one in Buddhist literature, and indeed in Indian literature on the whole. In terms of Buddhist sources, the Dvitīyakrama’s is the earliest mention of such a four-fold classification—either in the scriptural or commentarial literature—with which I am familiar. With regards to non-Buddhist Indian literature, Vātsyāyana’s famed 3rd-4th century Kāmasūtra does not include a four-fold categorization of women; Vātsyāyana instead has a six-fold schema. Following the composition of this foundational work, there is a significant historical gap from which period, it seems, no kāmaśāstric sources remain extant. Kokkoka’s Ratirahasya, and the Buddhist author Padmaśrī’s Nāgarasarvasva are considered to be the earliest of the “later” kāmaśāstra works that followed the Kāmasūtra,127 and the Ratirahasya is reported to be the first text to describe women in these classical four types.128 That text is difficult to date: its dates are given by some scholars as 9-10th century, by others as the 10-12th century, and by others as late as the 13th century.129 The 9th-10th century dates, however, seem to be based on erroneous and/or ambiguous references to the Ratirahasya in the works of the 10th century author Somadevasūri, and all that may be said with certainty is that the Ratirahasya is cited by commentators beginning only in the 13th century.130 In any case, even with the earliest dates posited, it seems that the Ratirahasya is certainly later than Buddhajñānapāda’s late 8th/early 9th-century Dvitīyakrama. The Ratirahasya lists precisely the same four types of women as the types of consorts in the Dvitīyakrama (though as in the Samvarodaya, the first type according to Kokkoka is called the more commonly used padminī rather than Buddhajñānapāda’s unusual term kamalī131). There is one other possibly early but difficult-to-date source for this four-fold categorization: *Surūpa’s Kāmaśāstra (Tōh. 2500), Tibetan dkar ‘chang can also be understood to mean this, given that the term dkar is often used for milk products, in general. However, I have not been able to find any other uses of the term dkar ‘chang in Tibetan. 123Vaidyapāda comments: “Below her navel [she has three wrinkles] that look like a triśūla.” (lte ba’i ‘og tu tri shū la (shū la] D shu la P) lta bu zhes pa’o// (Sukusuma, D 100b.1; P 120b.7). This is a classical mark of beauty in Indian literature. 124 skyur] C P N S V (D and P), skar D 125 de la ka ma lī yi ni// dbyibs dang mtshan nyid bstan par bya// |53| bu mo padma’i dri bro zhing// ngo zlum sna rtse til ‘dra ste// sen mo dmar zhing rgyab sgur dang// rkang mthil kun gyis sa la reg |54| pa spu ‘khyil ldan ser mo can// dkar ‘chang til gyi ‘bras bu ‘dra// gsus pa’i gnyer ma gsum ldan zhing // brang mdzas glang chen ltar ‘gros dang// |55| rākta’i ro ni skyur ba’o// sha mdog dmar te lha mo ‘di// mā ma kī yis rnam par dag// (Dvitīyakrama, verses 52c-56c). 126 In the Samvarodaya the first type, instead of kamalī, is called padminī, which is in fact the much more common name for this particular type; the two categories obviously correspond, though, given that both are derived for words meaning “lotus.” The Samvarodaya passage on the four types appears likely to have been influenced by the Dvitīyakrama, as there are a number of parallels. For the Samvarodaya passage see Tsuda 1994, 155-57 and 324- 35. The Samvarodaya is quite a bit later than the Dvitīyakrama; Harunaga Isaacson has suggested it may date to as late as the 12th century (English 2002, xxi, 384n2). 127 Ali 2011, 43. 128 Datta 1988, 1203; Ali 2011, 45. 129 See Ali 2011, 44 and 44n14; Datta 1988, 1203; and Hopkins 1992, 35 and 35n4. 130 Ali 2011, 44n14 and 44. 131 See note 119. 237 which, like the Dvitīyakrama, is not extant in Sanskrit but is preserved in Tibetan translation (and may therefore not have been considered by Indologists discussing early kāmaśāstric literature). We know nothing of the author of this treatise, though the homage and one of the concluding verses of the work suggest he was a Buddhist, and he tells us that his work was composed on the basis of Nāgārjuna’s treatise on erotics. Vogel, who has edited and translated the Kāmaśāstra, suggests that this must be the “tantric” Nāgārjuna, who he dates to the 6th century.132 However, if it is indeed a work by this Ārya School author Nāgārjuna on which

• Surūpa’s work is based (and this point is itself not entirely certain)133 this author, we now know,

was likely writing slightly later than Buddhajñānapāda, in the 9th, not the 6th, century. (A Nāgārjuna, by the way, is also listed as a source in Kokkoka’s Ratirahasya.134) *Surūpa’s work, then, also appears to be later than Buddhajñānapāda’s, making the Dvitīyakrama the earliest known locus, Buddhist or non-Buddhist, of this important four-fold kāmaśāstric categorization of women.135 Sexual Acts in a Yogic Context: Interrelation Between Tantric Buddhist and Kāmaśāstric Writings While the fact that the Dvitīyakrama appears to be the earliest extant text to include a four-fold typology of women may indicate the influence of Buddhajñānapāda’s works on later kāmaśāstra, we cannot be sure that this is the case; it is certainly possible that Buddhajñānapāda himself was relying upon a no-longer-extant kāmaśāstric source in making a four-fold typology of tantric consorts.136 And indeed there are other passages in his writings that suggest he may at times have been relying on kāmaśāstric sources. A passage from the section of the Dvitīyakrama that describes the third initiation, labeled by Vaidyapāda as describing the processes by means of which the initiate couple “physically cultivate passion,” includes some detail on the sexual acts by means of which the yogic partners are meant to arouse one another prior to coitus.137 The description of these acts includes a series of five postures, which Vaidyapāda explains as different postures that the partners are to assume in order to gaze upon one another to incite 132 Vogel 1965, 4. 133 Vogel, though, notes that “the work admittedly used by Surūpa as his source does not apera to be identical with any of the several known treaties entitled Ratiśāstra that go by the name of Nāgārjuna and differ widely from each other.” (Vogel 1965, 5). 134 Ali 2011, 60. 135 The (much) later Tibetan commentaries on the Guhyagarbha-tantra such as Longchenpa’s 14th-century Phyogs bcu mun gsel commentary also have a four-fold typology, but it is different from the classical kāmaśāstra typology (See Dorje 1987, 902). The Guhyagarbha-tantra itself, which we can date to the 8th century, only says “discriminating between devīs, nāginīs, and female mudrās of inferior species, or else without discrimination,” at the point where Longchenpa gives his extensive commentary on the four types of consorts (Dorje 1987: 883). Other Buddhist tantras have different schemas of classification of consorts, or of women in general, like the tantras pertaining to the Cakrasaṃvara tradition, including the Cakrasaṃvara-tantra itself and the Abhidhānottara-tantra, which have a seven-fold classification (See Gray 2007, 227-29 and Kalff 1979, 237-38, respectively); and the Sampuṭa-tantra and Caṇḍamahāroṣaṇa-tantra, which each have a five-fold one corresponding to the five buddha families (See Sampuṭa-tantra, 1.1.42-1.1.45 and Caṇḍamahāroṣaṇa-tantra, 8.15-8.17). 136 Buddhajñānapāda explains that the typology is four-fold precisely due to the fact of the correspondence of the types with the four buddha consorts, writing “ This illusion here in this world,/ Because of having Locanā and so forth as its pure forms, / Is of four types” | sgyu ma de yang ‘jig rten ‘dir// spyan la sogs par rnam dag pas// rigs ni rnam pa bzhir ‘gyur te// (Dvitīyakrama, verse 51a-c). This could be taken as evidence that the classification does indeed spring from the Buddhist tradition, even if not necessarily from Buddhajñānapāda himself, but of course it is also possible that Buddhajñanapāda adopted an already existing four-fold typology because it fit well with the system of the buddha consorts. It seems unlikely that we will be able to determine this question with any certainty. 137 Sukusuma, D 107b.1; P 129a.5-6. These acts are described in the Dvitīyakrama, verses 105-113. 238 passion (this section of the text occurs before the actual sexual union that constitutes the main part of the third initiation). The verse reads: First, coming together Then the [posture] characterized by the elbows Additionally, the one [characterized by] extending [the legs] And likewise, the [posture] characterized by lifting up And then the complete extending [of the legs]—these are the five. |106|138 In his commentary on this verse Vaidyapāda writes, “How should one look? Demonstrating the five principal [ways] as taught in the *Sarvasaṃcalaśāstra(???), (skyod byed thams cad kyi gtsug lag, the “Treatise on All Kinds of Movement”???)139 the text says First…”140 Vaidyapāda goes on to provide what appear to be citations, in verse, presumably from this text—if indeed the

• Sarvasaṃcalaśāstra is meant to be the name of a text—describing each of the postures

mentioned in the verse. The first of these citations reads: The woman firmly embraces [him] around neck And the man’s forearms Are placed against her elbows141 It is also said that her calves should be brought together;142 This is [how] to perform the position.143 The term “position” (bsdam pa) is likely a translation of the Sanskrit term bandha, used in kāmaśāstra to refer to sexual positions. Indeed, it is possible that Vaidyapāda may be citing a kāmaśāstric source (the aforementioned Skyod byed thams cad kyi gtsug lag?) here and in his subsequent comments on the positions described in verse 106 of the Dvitīyakrama. As I noted above, there seem not to be many (any?) extant kāmaśāstric works from the period between the 3rd-4th-century Kāmāsūtra and the later kāmaśāstra texts starting with the work of Kokkoka and Padmaśrī, both dated to no earlier than the 9th century, and quite possibly as late as the 12th. Vaidyapāda’s citations here may thus provide a window into a kāmāśāstric source from this intermediary period, which may well have served as an inspiration for the erotological aspects of Buddhajñānapāda’s work. However, there is a reference in one of the passages that Vaidyāpāda cites here to “stages of bliss,”144 which is either a noteworthy reference to stages of bliss in a 138 dang por ‘dus pa byas nas su// de nas gru mo mtshan nyid dang// yang ni brkyang par bya ba dang// de bzhin yar bteg mtshan nyid dang// shin tu brkyang pas lnga ru ‘gyur// |106| (Dvitīyakrama, verse 106). 139 skyod (skyod] P, skyed D) byed (byed] D, phyed? P) thams cad kyi gtsug lag. Harunaga Isaacson (personal communication) suggests the (admittedly rather speculative) possibility that skyod byed thams cad could be translating something like sarvaparvartana, sarvasaṃcālana, or sarvaprakampaka, with the sense of “one who makes all (beings in Kāmadhātu) move/act,” as an epithet for Kāma, thus skyod byed thams cad kyi gtsug lag would actually be referring to Kāmaśāstra, but if this were indeed the intended meaning, this would assume a rather poor (or at least overly literal) translation on the part of the Tibetan translator. As I note below, it is also not completely clear whether this work is a kāmaśāstric source. 140 ji ltar blta bar bya zhe na/ skyod (skyod] P, skyed D) byed (byed] D, phyed? P) thams cad kyi gtsug lag las bshad pa’i gtso bo lnga gsungs pas/ dang por zhes pa la sogs pa’o// (Sukusuma, D 107b.2-3; P 129a.7-9). 141 While holding her head in his hands, as described in the subsequent verse cited by Vaidyapāda? 142 Presumably with her legs around his body. 143 de’i mtshan nyid kyang ji skad du/ bud med mgul par (par] D, pa P) dam ‘khyud de// pho yi dung pa gnyi ga yis// gru mo gnyis la bzhag pa yag// bud med rje ngar ‘dus par bshad// ces te/ bsdam pa’i bya ba’o// (Sukusuma, D 107b. 3-4; P 129a.8-129b-1). 144 “The woman’s bent knees/ are to be placed on the man’s elbows/ This [posture] is called “knees on elbows”/ These [postures] are asserted to be (to produce?) the stages of bliss./ This is the act of looking closely.” de nas gru mo mtshan nyid ces pa ni rkang lag gi bya ba ste/ ji skad du/ bud med pus mo bkug pa ni// pho yi gru mor bzhag par bya// bus mo gru mor bshad pa ste// de dag dga’ ba’i rim par ‘dod// ces te/ rnam par lta bar bya ba’o// (Sukusuma, D 107b.4-5; P 129b.1-2). 239 kāmaśāstric work,145 or an indication that the text Vaidyapāda cites here is actually not a kāmaśāstric text, but rather a Buddhist one. I have been unable to determine at this point which is more likely to be the case. However, even in the sections of the Dvitīyakrama that appear to show some influence from the kāmaśāstric tradition, borrowings from there are directed toward the soteriological aims of yogic practice. Several verses after the one describing the series of postures, cited above, another verse from the Dvitīyakrama—still in the section that Vaidyapāda identifies as instructions on “physically cultivating passion”—mentions a number of locations on his partner’s body that the yogin should stimulate: While sucking and making the sound ṣīt,146 He plays with her breasts, the tips of her fingers, Her throat, lower lip, cheeks, and earlobes, Her eyes, the crown of her head, and her secret place— Kissing these with his mouth. |110|147 Vaidyapāda explains that the stimulation described in this verse is performed in order to “invoke the places that are the sources of bodhicitta,”148 thus connecting acts that produce erotic pleasure directly with the stimulation of parts of the subtle body related to the soteriological function of sexual yoga. Indeed the lines from this verse (b-d) listing the locations on the body that are to be stimulated are mostly parallel with a verse cited in the only surviving manuscript of Kalyāṇavarman’s later (late 9th century?)149 Catuṣpīṭhapañjikā, where these lines appear combined together with verses 171-174 from the Dvitīyakrama and with some other verses that are not derived from the Dvitīyakrama, but are clearly associated with the practice of sexual yoga.150 Verses 171-174 from the Dvitīyakrama deal with the practice of nyāsa, in which the practitioner visualizes the placement of sixteen syllables on specific places in the body during the yoga of the indestructible bindu. The fact that the citation in the Catuṣpīṭhapañjikā combines the verses from these two separate sections of the Dvitīyakrama, the latter of which is clearly related to perfection stage yogic practices, provides further corroboration of Vaidyapāda’s assertion that these physical locations mentioned in verse 110 of the Dvitīyakrama, in addition to their function as erogenous zones the stimulation of which incites pleasure in the yogin’s partner, are indeed also connected to yogic practices involving the manipulation of elements of the subtle body. 145 It seems that a reference to a sequence of blisses in a non-Buddhist kāmaśāstric work would at the very least be unusual (Harunaga Isaacson and Mattia Salvini, personal communications), but as the passage contains but a vague reference to “stages of bliss” (dga’ ba’i rim pa), without any further specifications that would make it clear that this is a reference to the typical progression of blisses from the Buddhist tantric tradition, I think it is difficult to completely rule it out the possibility that this work that Vaidyapāda refers to is a kāmaśāstric one. 146 zid sgra. A sound used in Indian literature to indicate sexual arousal and pleasure. 147 gzhib cing zid sgra brjod nas kyang// nu ma lag rtse mgrin pa dang// ma mchu ‘gram pa rna ba’i rtsa// mig dang spyi bo gsang bar yang// kha yis ‘o byas rtse bar bya// |110| (Dvitīyakrama, verse 110). 148 de nas byang chub kyi sems ‘byung ba’i gnas rnams bskul ba’i phyir (P + ro) / nu ma lag rtse zhes pa la sogs pa ‘o// (Sukusuma, D 108a.3; P 130a.1-2). 149 Szántó (2012a, 15) suggests a late 9th century date for Kalyāṇavarman’s work. The manuscript itself dates to 1012 CE. 150 The passage from the Catuṣpīṭhapañjikā, which combines verses from these two different sections of the Dvitīyakrama along with some other verses that are not from the Dvitīyakrama, is cited in that commentary as coming from the “Aṣṭāṣṭaka,” (the verses are preceded with the statement “uktañ ca aṣṭāṣṭake”) which may perhaps be the title of a text, though this is not certain. Thanks to Péter Szántó for sharing his diplomatic transcript of these verses with me and for pointing out the parallels with the Dvitīyakrama vv. 171-74. The lines that are parallel with verse 110 (which I was able to notice only because Péter kindly shared the Sanskrit passage from the Catuṣpīṭhapañjikā with me) read: pīnastane karāgre ca grīvāyāṃ adhare tathā | gaṇḍākṣikarṇṇamūle ca mūrdhni sarvāṅgam eva ca | (I have edited the text slightly following Harunaga Isaacson’s suggestions, for which I am grateful.) 240 In fact, the verses from the Dvitīyakrama that present the list of sixteen locations on the body and the installation there of syllables—the sixteen vowels of the Sanskrit alphabet—are yet another place where we can see evidence of a relationship between Buddhajñānapāda’s writings and kāmaśāstra, and where recourse to kāmaśāstric texts can help us to more clearly understand some rather opaque statements in Buddhajñanapāda’s and Vaidyapāda’s writings. The Dvitīyakrama describes the sixteen syllables located at the sixteen points in the body as “the sixteen forms of bodhicitta in one’s own interior.”151 After listing each of the syllables and their locations in the body, the Dvitīyakrama states, somewhat cryptically, “These are completed at the time of the sixteenth.”152 Vaidyapāda explains: “Also, one should know that this is with regard to the stages of the first day [of the month] and so forth. They are completed at the time of the sixteenth, means that at the time when the outer moon comes to fullness, these are also perfected. One must understand that this is then reversed. Regarding being perfected at the time of the sixteenth, the sixteen places that are stirred up through practice also become “the sixteen.” These then [become] the bindu and this becomes like the moon, which produces the blissful consciousness. The previous light rays hook, means that they hook the sixteen syllables and draw them into the bindu. By slightly holding one’s mind, like the first wisdom, there for a moment, what happens? [The text then says] Meditate with determination/ On the great [maṇḍala]-cakra of deities together with its support.”153 This is very similar to some comments that Vaidyapāda made earlier in the Sukusuma: “Moreover, through practicing, by means of the agitation of the locations, the sixteen syllables appear, and these, then, become the sun and moon. Having transformed into a bindu like that, they go to the tip of the vajra. This itself, in a form which blazes with thousands of light rays, is meditated upon by the yogin in accordance with the ritual that will come below. When this happens, the suchness that has been spoken of will be realized, [and that is the] purpose [of this practice.]”154 155 151 bdag nyid nang// byang chub sems gzugs bcu drug po// | (Dvitīyakrama, verses 170c-d). 152 bcu drug dus su rdzogs ‘gyur ba// |174| (Dvitīyakrama, verse 174d). 153 de yang tshes gcig la sogs pa’i rim par shes par bya’o// bcu drug dus su rdzogs gyur pa zhes pa ni phyi’i zla ba rdzogs par’i dus su de yang rdzogs pa’o// de ne bzlog ste shes par bya’o// bcu drug dus su rdzogs par ‘gyur ba ni sgrub (sgrub] P, bsgrub D) pas dkrugs pa bcu drug/ de (de] D, ste P) yang bcu drug par ‘gyur/ de yang thig le/ de yang zla ba lta bur song nas/ dga’i ba’i shes pa ‘byung ba’o// gong gi ‘od kyis rnam pa bkug ste zhes pa ni yi ge bcu drug po rnam par bkug nas thig le’i nang du bcug la der rang gi sems dang po’i ye shes ltar bag zhad bzung bas cir ‘gyur zhe na/ lha’i ‘khor lo che/ rten dang bcas pa mos pas bsgom/ (Sukusuma, 115a.6-115b.1; P 138b.6- 139a.2). 154de yang bsgrub pas gnas rnams dkrugs pa las yi ge rnams bcu drug par gyur/ de yang nyi zlar gyur/ de lta bu’i thig ler gyur nas rdo rje rtse mor ‘gro ba ste/ de nyid ‘od zer stong du ‘bar ba’i gzugs su rnal ‘byor pa rnams kyis ‘og nas ‘byung ba’i cho gas bsgoms nas/ ji skad du gsung pa’i de kho na nyid rtogs par ‘gyur pa’i phyir ro/ (Sukusuma, D 88a.4-5; P 105b.6-8). 155 Tāranātha, who reports having received initiation into and teachings on the Jñānapāda lineage from his master Buddhaguptanātha, in his much later instruction manual on the perfection stage rituals of the Jñānapāda School, gives an ever-so-slightly more clear presentation of this practice that does not seem substantially different from what is already here in Buddhajñānapāda’s text, with Vaidyapāda’s clarifications. He writes, “For the second part, the light from the bindu illuminates the jñānasattva, and from that light radiates forth and illuminates the interior of the foundational body. Like holding up a lamp in darkness, one sees clearly the sixteen bindus, which are the white substance.... [He lists here the syllables at all of the locations on the body, exactly as they are described in the Dvitīyakrama]...All of these are white and radiate white light. Think of them as being of the nature of bliss. The light from the heart center, either in stages or all at once, as one prefers, dissolves those syllables into the indestructible bindu at the heart center, and [it] then blazes with light and causes a strong increase in the essence of bliss. Contemplate thus.” gnyis pa ni thig le’i ‘od kyis ye shes sems dpa’i sku gang / de las ‘od ‘phros gzhi lus kyi nang gsal zhing gang bar byas/ mun khung du sgron me bteg pa ltar dkar cha thig le bcu drug po rnams gsal bar mthong ba ni/ ...... thams cad kyang kha dog dkar po ‘od zer dkar po ‘phro ba/ bde ba’i rang bzhin can du bsam/ snying ga’i ‘od kyi yi ge de rnams rim pas sam cig car gang mos kyis snying ga’i mi shigs pa’i thig ler bstims pas/ ‘od zer ‘bar zhing/ bde ba’i ngo bo lhag par rgyas par bsam mo// (Dpal grol ba’i thig la’i khrid yig, 243-244). 241 Vaidyapāda’s comments about the phase of the moon in relation to these syllables in different parts of the body can, I believe, be understood more clearly with reference to the kāmaśāstric doctrine of candrakalā in which Kāmadeva was understood to dwell in different parts of the body at different points in the moon’s phases.156 As described in Kokkoka’s Ratirahasya and in Padmaśrī’s Nāgarasarvasva, this involves Kāma moving gradually through the left side of the body in the moon’s waxing phase, pervading the entire body for two days during the moon’s fullness, and traveling down the right side of the body during the waning phase.157 A man is meant to stimulate these specific locations on his lover’s body at particular days in the lunar calendar in order to please her, and the texts even prescribe the visualization of the vowels of the Sanskrit alphabet (i.e. precisely the syllables listed in the Dvitīyakrama), along with a candrabindu, at these various places on the body, on the appropriate dates.158 While the doctrine of candrakalā described in these works is several centuries later than the Dvitīyakrama—as noted above the Ratirahasya is likely not earlier than the 10th century, and perhaps as late as the 13th, and the Nāgarasarvasva dates to the 12th century—and pertains to the genre of erotics rather than tantric practice, as I noted above, Ali has shown clearly that kāmaśāstra authors from this period, including Padmaśrī (who perhaps not incidentally was a Buddhist), were drawing on tantric Buddhist ideas in their writings.159 A similar practice is described also in *Surūpa’s Kāmaśāstra, which may be earlier than the Ratirahasya and the Nāgarasarvasva, but as I noted above, that work is difficult to date.160 In any case, as we saw earlier, Buddhajñānapāda and Vaidyapāda both appear to show familiarity with kāmaśāstra. While the specific association with Kāmadeva is unlikely to be relevant here in the Dvitīyakrama—the syllables are specified in the Dvitīyakrama as being present within the yogin’s own body, and the practice of candrakalā in a kāmaśāstric context seems always to pertain specifically to a woman’s body161–the idea of syllables or bindus in the practitioner’s body becoming fully “perfected” at the time of the full moon (and perhaps otherwise individually “perfected” on the waxing or waning days of the moon) may be related to a more widely shared conception of specific areas of the body being associated with the progression of the lunar calendar. Such a conception does indeed appear to be a more broadly Indic idea, as it is also found in Indian medical traditions, where both the life force and the pulse are also said to travel through the body on specific days of the lunar calendar, and are likewise associated with the Sanskrit vowels located at sixteen different places on the body, as described in the Dvitīyakrama.162 Moreover, the passage on the syllables at the sixteen places found in Kalyāṇavarman’s Catuṣpīṭhapañjikā that is parallel with Dvitīyakrama verses 171-174 makes reference to the waxing and waning phases of the moon, and its context is clearly one of sexual yogic practice. While this work is also slightly later than the Dvitīyakrama, it further confirms the connection between the syllables and locations described in the Dvitīyakrama and the lunar phases within a sexual (and in this case also a yogic) context. Thus, in several passages in Buddhajñānapāda’s writings we see further evidence of the relationship between tantric Buddhist writings and kāmaśāstra that Ali’s work has already begun to explore. As tantric Buddhist traditions developed and refined the practices of sexual yogas that came to characterize and even define later tantric Buddhism in India, it is only natural that they 156 See Ali 2011, 47. 157 ibid. 158 ibid., 47-48 159 Ali 2011, esp. pp. 53-54 160 See Vogel 1965, 24. 161 See Desmond 2011, 26. 162 Somānanda Dharmanātha, personal communication. 242 should have drawn on the rich Indian tradition of erotology as a support for such practice. It increasingly appears to be the case, however, that tantric Buddhist traditions themselves may have contributed to the development of kāmaśāstra. Further study of the relationship between the these two genres, the authors who wrote both types of texts, and their communities will certainly contribute to our understanding of Buddhist tantra and kāmaśāstra. The Yoga of Utkrānti, the Yogic Ejection of Consciousness One further practice that is connected in the Dvitīyakrama with the perfection stage is utkrānti (‘pho ba), the yogic ejection of consciousness at the time of death.163 The Dvitīyakrama’s presentation of utkrānti is one of the earliest, if not the earliest, found in a Buddhist work, and appears to have served as an inspiration for the instructions on utkrānti given in the Catuṣpīṭha-tantra.164 In his research on the Catuṣpīṭha-tantra, Péter Szántó has suggested that the technique of utkrānti is likely not of Buddhist origin, and he provides a number of references to instances of the practice, or its parallels, in a range of non-Buddhist texts, both Brahmanical and Śaiva.165 Whatever tradition Buddhajñānapāda may have been drawing on in incorporating these instructions into his work, it is worth noting that in the Dvitīyakrama the practice of the ejection of consciousness, in addition to being referred to as utkrānti, is also called the practice of svādhiṣṭhāna (“self-consecration”), a term that is used in a number of tantric works to refer to what seem to be several different types of practices; to my knowledge, the use of the term svādhiṣṭhāna to refer to the practice of utkrānti is unique to the Dvitīyakrama.166 Utkrānti is connected in the Dvitīyakrama with the perfection stage inasmuch as its practice appears to be contingent on a practitioner’s already having received tantric initiation, and having thereby “obtained suchness” from the guru. Indeed, the Dvitīyakrama prescribes utkrānti specifically for a practitioner who has already “obtained suchness” from his guru and “realized the secret and supreme secret,” but who has been unable to train in, and therefore to complete, the other practices taught in the Dvitīyakrama during his lifetime.167 Vaidyapāda explains, “Having in this way taught the stages [of practice] for attaining nirvāṇa in 163 This practice is described in the Dvitīyakrama, verses 326-353. 164 Szántó 2012a, 455-56. Szántó contends that the Catuṣpīṭha, which he asserts clearly draws on the Dvitīyakrama, is likely the earliest Buddhist scriptural source for the practice of utkrānti. See Szántó 2012a, 455-68 for a translation and brief analysis of the passage on utkrānti from the Catuṣpītha, including its parallels with the Dvitīyakrama. 165 ibid., 456-57. 166 The term svādhiṣṭhāna is used in several works of the Guhyasamāja system, as well as in the corpi of later tantras. For example, it is mentioned in the Samājottara (verse 77), in reference to what appears to be a practice within the context of the generation stage, and is also used within the literature of the Ārya School to describe the third of the five stages of that tradition’s perfection stage practices, called the svādhiṣṭhānakrama, and also termed the practice of the illusory samādhi (māyopama-samādhi), or of the illusory body (māyādeha) (see Wedemeyer 2007, 68 and Tomabechi 2006, 79-81). The Hevajra-tantra uses the term in what has been interpreted by commentators as just a reference to utpannakrama practice more generally (see Isaacson and Sferra 2014, 267 n 74). None of these usages of the term relates to utkrānti. However, the practice of svādhiṣṭhāna according to the Ārya School is the method by which the yogin produces the body or form of an awakened buddha (ibid.), and indeed, as we shall see below, the utkrānti instructions given here in the Dvitīyakrama seem to serve precisely this same function of generating a saṃbhogakāya form, which is done here by means of first bringing the mind into the dharmakāya at the time of ejecting the consciousness in the moment of death. Once the saṃbhogakāya form is achieved, the Dvitīyakrama contends, one will naturally take birth in the next life in a nirmāṇakāya form (see Dvitīyakrama verses 351-353). 167 I discuss more about the prerequisites for utkrānti practice and its function within the structure of the tantric path in Buddhajñānapāda’s practice tradition below. 243 this life [i.e. the three bindu yogas of the perfection stage that were explained in the earlier section of the text], now he teaches the stages [of practice] for attaining nirvāṇa in the intermediate state with the verse beginning Now....”168 The Dvitīyakrama begins: Now for the stage of svādhiṣṭhāna, This will be explained To a few yogins Who are fortunate due to their actions. |326| Someone who has pleased the guru And received the vase [initiation] and the others Together with the samayas and vows given by him And has thus obtained the suchness169 |327| That is found through the guru’s words, And has realized the secret and supreme secret [But] is not able to genuinely train by means of the activities In the way explained [above]— |328| He should train in this stage Of suchness, just as it is. At some time in the future One will see the signs of death. |329| When the time of death has arrived And one is not completely overcome by illness Engage in the yoga of utkrānti. |330|170 The yoga of utkrānti is thus, like the yogas of the perfection stage described earlier in the Dvitīyakrama, identified here as a practice of “suchness, just as it is,” which also explains why it is necessary for the practitioner to have first “obtained suchness” from the guru before being able to take up the practice of utkrānti. After introducing the yoga in the verses above, the Dvitīyakrama goes on to describe what happens when consciousness leaves the body at the time of death. The text mentions the various apertures of the body through which consciousness (referred to consistently in the section of the text as “wisdom,” (ye shes, jñāna)) might depart at the time of death, and the different realms into which an individual will be reborn if the consciousness departs from those various apertures.171 As the Dvitīyakrama explains, the egress of consciousness from any of the apertures mentioned will result in rebirth in one among the realms within cyclic existence. The yoga of utkrānti is thus intended to prevent this, and instead to bring the consciousness directly into union with suchness at the moment of the body’s physical death, resulting in attainment of the three kāyas of perfect awakening. The yoga itself begins with the practitioner’s blocking the nine apertures through which he does not wish his consciousness to depart with syllables—the anus and urethra are blocked 168 de ltar mthong ba’i chos la mya ngan las ‘da’ ba’i rim pa bstan nas/ da ni bar ma dor mya ngan las ‘da’ ba’i rim pa gsungs pa/ da ni zhes pa la sogs pa’o (Sukusuma, D 129b.6-7; P 156a.7-8). 169 Vaidyapāda specifies that this refers to having received the instructions on suchness together with the sādhana for accomplishing suchness via the seven yogas (Sukusuma, D 130a.3; P 156b.3-4). 170 da ni rang nyid byin brlabs170 pa’i// rim pa ‘di ni rnal ‘byor pa// bya bas bskal pa ‘ga’ zhig la// yang dag tu ni bshad par bya// |326| gang zhig bla ma mnyes byas nas// des gnang dam sdom170 bcas ba ru// bum pa la sogs rab thob ste// bla ma’i zhal las rnyed170 pa yi// |327| de bzhin nyid ni rab thob cing//gsang dang mchog tu gsang rigs170 pas// ji skad bshad pa’i bya ba yis// yang dag bsgom ni mi nus pas// |328| ji bzhin pa yi de kho na// rim pa ‘di [14b] yis bsgom par bya// gang zhig dus ni phyi zhig la// ‘chi ba’i mtshan ma bdag gis mthong// |329| ‘chi bar gyur pa’i dus byung na// nad kyis yang dag ma rnyogs par// ‘pho ba’i sbyor ba yang dag bya// |330| (Dvitīyakrama, verses 236-30). 171 Dvitīyakrama, verses 331-336. 244 with suṃ and kṣuṃ, respectively, and the crown, forehead, eyes, ears, nose, mouth, and navel are all blocked with hūṃ.172 Then the aggregates, elements, and sense sources are to be meditated upon “as they are explained in the Yoga tantras”—which Vaidyapāda clarifies to mean that they should be contemplated being the buddhas, the buddha consorts, and the bodhisattvas of the maṇḍala, respectively.173 The yogin then generates himself in the form of the deity, following the usual generation stage procedure, and then visualizes a five-colored nine-pronged vajra above the crown of his head. He should then imagine his consciousness in the form of a smaller vajra with a wisdom bindu at its center, yellow in color, the size of five chickpeas. He should imagine that all phenomena disappear, focusing only on himself in the form of the deity, and then imagine that he himself dissolves until there is only mind. This mind then shoots up like an arrow and enters into the vajra visualized above his head, and dissolves into the bindu, which has the nature of the tathāgatas and the goddesses. The practitioner is to hold his mind there until it becomes dissipated. When that happens it emerges from the top of the nine-pronged vajra onto a moon disc on top of a lotus, where it transforms into the body of Vajrasattva,174 who is unclothed but ornamented, possesses the major and minor marks of an awakened being. Vaidyapāda adds that Vajrasattva is to be visualized embracing his consort.175 Then, presumably from the point of their union, he is to visualize the emanation and absorption of a great maṇḍala-cakra “arisen from the blessings of nondual union.” The yogin is instructed to engage in this meditation repeatedly, for as long as he is able, by means of which his mind “enters into the [dharma]dhātu” and he “realizes that which is luminous and perfectly joyful, like the sky,” which Vaidyapāda explains is the dharmakāya. Then, the yogin accomplishes “the form of a five-year-old child,” which Vaidyapāda explains as the sambhogakāya, and “realizes unparalleled perfect bliss.” Finally, when the practitioner moves on to the next rebirth he will “genuinely realize the nirmāṇakāya.” Vaidyapāda’s commentary on this section of the text explains the verses that describe the practitioner’s attainment of the three kāyas also in terms of their relationship to the ordinary death process for a non-practitioner. He states that that what, for the practitioner, is the attainment of the dharmakāya is referred to as the “death state” (shi ba’i srid pa) by “proponents of karma who do not know the nature of mind.”176 Thus, at the time when ordinary beings experience the moment of death the practitioner of utkrānti realizes dharmakāya; at the time when an ordinary being would be in the intermediate state (for seven days, etc.) the practitioner actualizes the saṃbhogakāya; and at the time when an ordinary being would be reborn into another body the practitioner accomplishes the nirmāṇakaya.177 In this way the ordinary death process is transformed by this yoga into a process for attaining the final accomplishment of perfect awakening. The Dvitīyakrama praises the yoga of utkrānti as a practice by means of which even 172 My description of the practice here is a summary of Dvitīyakrama verses 337-353. 173 Sukusuma, D 131a. 5-6; P 158a.2-4. This equation of the buddhas with the aggregates, the buddha consorts with the elements, and the bodhisattvas with the sense sources is made in Chapter 17, vv. 51-52 of the Guhyasamājatantra. 174 Generally in Buddhajñānapāda’s and Vaidyapāda’s writings the term “Vajrasattva” is an epithet used to describe the causal or progenitor deity in generation stage practice, not to refer to any specific deity. In that role as the causal or progenitor deity, “Vajrasattva” may also function as a sort of representative of primordial awakening, an ādibuddha of sorts. Here in the context of the practice of utkrānti neither the Dvitīyakrama or Vaidyapāda’s commentary specifies what precisely is meant by Vajrasattva, but given that this practice seems to be about merging the yogin’s mind with suchness itself at the moment of his passing, and that Vajrasattva is described as naked, it seems likely that he is meant precisely as an embodied representation of primordial awakening. 175 Sukusuma, D 132a.5; P 159a.5. 176 sems kyi rang bzhin ma shes pa’i las su smra ba (Sukusuma, D 132b.2; P 159a.1). 177 Sukusuma, D 132b.1-4; P 159a.1-5. 245 someone who has committed the gravest of sins is able to attain accomplishment, and through which the accomplishment of the three kāyas is certain. Yet, the text explains, even if one does not accomplish the three kāyas (immediately?) by means of its practice, at the very least he will “become the leader of the vidyādharas and gradually transform into the mahāmudrā.” Thus even if his accomplishment of awakening through arising in the form of the deity (we must again remember that the term mahāmudrā in the 8th and 9th centuries refers to the practitioner taking on the form of the deity) is not immediate, it seems, the final result of awakening is certain to eventually transpire for a practitioner of the yoga of utkrānti. As I noted above, the Dvitīyakrama appears to be among the earliest Buddhist texts to present and advocate for this unique method of utkrānti, specified in this work as a means for bringing about the attainment of the three kāyas of perfect awakening during the death process. The function of this particular yoga within Buddhajñānapāda’s system of tantric practice seems to be as a sort of failsafe for a yogin who has received initiation and “obtained suchness,” and thus has all of the prerequisites for taking up the practice of the perfection stage, but has not been able to fully or perfectly do so during his lifetime. However, beyond having received suchness during initiation, precisely what training is needed in order to be able to successfully practice utkrānti at the time of death is not made entirely clear in Buddhajñānapāda’s or Vaidyapāda’s writings. As we saw in the passage cited from the Dvitīyakrama above, Buddhajñānapāda appears simply to suggest that utkrānti may be practiced by someone who received initiation and “obtained suchness” from the guru, “and has realized the secret and the supreme secret” but was unable to train (fully?) in the practices set forth in the Dvitīyakrama. Vaidyapāda explains that this refers to a disciple who has received suchness from the guru by means of the seven yogas, but who has been unable to genuinely train in it, meaning that he has begun with the generation stage, but been unable to train in accordance with both stages.178 Vaidyapāda’s subsequent explanation gives many options for the type and frequency of practice that a yogin who wishes to perform utkrānti may have engaged in before undertaking this final practice, but the very fact that he includes such a list indicates that he understood some type of training in suchness by means of the generation and perfection stages as a necessary prerequisite for performing the yoga of utkrānti at the moment of death.179 In this sense utkrānti indeed seems to function in Buddhajñanapāda’s system as a “second chance” at awakening during the death process if a practitioner has not managed to awaken during his lifetime by means of the practice of the perfection stage, but it is an option that is only available to a practitioner who has already made a connection with the second stage through “receiving suchness” from his guru during tantric initiation. Perfecting Awakening: Some Conclusions As we have seen, Buddhajñānapāda’s writings provide a window into the perfection stage and its practices as they were understood during the early period of their development. For Buddhajñānapāda the perfection stage was equated or identified with suchness itself, and was first recognized through the guidance of the guru during the sexual yogas undertaken in the context of tantric initiation. It was associated with the progression of the three blisses that a practitioner experienced during the sexual yogic practices of Buddhajñānapāda’s tradition. The three bindu yogas, performed with the support of a tantric consort, served as the procedures through which the yogin was to cultivate the perfection stage, and the yoga of utkrānti served as 178 Sukusuma, D 130a.3-4; P 156b.4-6. 179 Sukusuma, D 130a.4-7; P 156b.6-157a.1. 246 a sort of fail-safe—as long as he had already “obtained suchness” from his guru—if the practitioner was unable to gain accomplishment by means of these other practices during his lifetime. The systems of yoga that were emerging in this time, including the practice of several among the branches of what came to be an important tradition of the six-branch yoga, as well as the typology of blisses in sexual yogic practice are shown in the early stages of their development in Buddhajñānapāda’s writings. The centrality in Buddhajñānapāda’s perfection stage system of sexual yogic practices performed with a consort appears to have entailed his drawing upon, but also likely contributing to, the rich Indian tradition of kāmaśāstra. However, even the sexual practices that seem to have been drawn from kāmaśāstric sources were always tailored to the specific soteriological purposes of the systems of the yogic manipulation of winds and energies in the subtle body that were developing in his time. All of these diverse factors, visible in the Dvitīyakrama’s and Muktitilaka’s instructions on the practices of the perfection stage, point to a vibrant tradition in which an increasing variety of human experiences— including the intensity of both sex and death—were being drawn into the yogin’s repertoire as techniques for bringing about a direct experience of suchness, and thus put into practice in the service of the soteriological aims of tantric Buddhist practice. 247 Chapter Seven Revealing Reality: Tantric Initiation in Buddhajñanapāda’s and Vaidyapāda’s Writings In order to accomplish great awakening you must experience great bliss with the girl who liberates and gives joy. Nothing else can bring about buddhahood—this girl is the genuine supreme. Thus, throughout endless saṃsāra you must never separate from her. -Mañjuśrī’s instructions to Buddhajñānapāda, Dvitīyakrama One of the main characteristics that distinguishes the practice of Buddhist tantra from non-tantric paths is that a practitioner is required to receive initiation (abhiṣeka) prior to undertaking tantric practice. Unlike in Śaiva traditions, where initiation (dīkṣā) was understood to be liberative in and of itself, the Buddhist understanding of karma meant that in Buddhist tantra the ritual of initiation could not be held to function as soteriologically efficacious in that same way.1 Buddhist tantric initiation is nonetheless understood as an essential prerequisite that prepares the initiate by giving him the necessary permission and blessing to take up tantric practice. But it is only through personally engaging in those practices, for which he has become qualified through the initiatory rites, that a practitioner is able to attain liberation. In the late eighth century, precisely at the time when Buddhajñānapāda was active, the system of tantric Buddhist initiation was developing from an earlier five-fold series of initiations that characterized the Yoga tantras, to the addition of a set of higher initiations that prepared the practitioner for the newly emerged second stage, the perfection stage, of tantric practice. To be permitted to take up the practices of this second stage—the yogic manipulations of the winds and energies of the subtle body through sexual yogas—the initiate had to receive these higher tantric initiations, which were themselves sexual in nature. With the advent of the higher initiations, the first five initiations from the Yoga tantra system, which were associated with the five wisdoms, were collapsed into a single “first” initiation, the kalaśābhiṣeka (the “vase initiation”),2 to which a second, third, and eventually a fourth initiation were added. 3 These “higher” initiations first appear in the literature of the Guhyasamāja-tantra tradition, and Buddhajñānapāda’s writings serve as an important early description of the rituals for more than one of the higher initiatory sequences. By Buddhajñānapāda’s time the series of tantric initiations numbered three: the first, the kalaśābhiṣeka (the “vase initiation”), which, as just noted, combined the earlier five initiations from the Yoga tantras; the second, the guhyābhiṣeka (the “secret initiation”), and the third, the prajñājñānābhiṣeka (the “wisdom-through-insight initiation”4). The Dvitīyakrama is important for containing early descriptions of the rituals for both the second and the third tantric initiations. 1 Isaacson 2010b, 263. 2 The kalaśābhiṣeka seems to be generally understood to additionally include the ācāryābhiṣeka, the so-called “master initiation,” through which an initiate received permission to act as a tantric ritual officiant. I discuss this point in brief below. 3 See Isaacson 2010b and Sakurai 1996 (the latter is in Japanese) on the historical development of tantric Buddhist initiation. 4 The name of this initiation is a bit difficult to translate. My rendering here follows the presentation of the term in Vaidyapāda’s Yogasapta, which explains the name of the third initiation as follows: “Because one realizes the not [yet] realized wisdom (jñāna)/ Through the actions of the “insight” (prajñā; i.e. the consort) it is called [the prajñājñāna] initiation.” Shes rab las kyis (kyis] P, kyi D) ye shes te// ma rtogs rtogs phyir dbang yin no// (Yogasapta, D 70b.6; P 84a.7) The term prajñā (“insight”) is employed in this verse in the sense of its commonlyused tantric meaning, to refer to the female consort. 248 In terms of scriptural sources for the higher initiations, there appear to be multiple opinions within even traditional commentarial sources about whether the guhyābhiṣeka is already indicated in the root Guhyasamāja-tantra. While Candrakīrti in his 10th-century commentary on the Guhyasamāja, the Pradīpoddyotana, does claim that Chapter Eight of the tantra sets forth the ritual for the guhyābhiṣeka, later commentators on the tantra, like Ratnākaraśānti in the 11th century, show no indication that they read Chapter Eight as referencing the guhyābhiṣeka at all.5 Indeed, both Vaidyapāda, in the 9th century, and Ratnākaraśānti in the 11th, make statements that suggest they do not find reference to the guhyābhiṣeka in the Guhyasamāja root tantra, but only in the Samājottara, which we know to have first circulated separately—and after the root tantra—before being added to the Guhyasamāja-tantra as its eighteenth chapter. In his commentary to the passage on the guhyābhiṣeka from the Samājottara (vv. 114-117) Vaidyapāda writes: “Regarding the ritual for the kalaśābhiṣeka it is explained in the root tantra. As for the explanation of the ritual for bestowing the guhyābhiṣeka, [it is found in the verse in the Samājottara] starting with The wide-eyed one…6 Ratnākaraśānti, commenting on the same passage from the Samājottara writes, “From among these [initiations], the bestowal of the kalaśābhiṣeka is set forth quite clearly in the sixteenth chapter [of the root tantra] itself. The second [initiation] is explained [in the verse from the Samājottara] beginning with The wideeyed one...”7 My own reading of Chapter Eight of the root tantra tends to follow Vaidyapāda’s and Ratnākaraśānti’s, in not finding the guhyābhiṣeka there. The guhyābhiṣeka is, however, found clearly in the Samājottara. Likewise, in terms of scriptural sources, it is only in the Samājottara that we find the prajñājñānābhiṣeka.8 As I have already mentioned, and will present in more detail in the next chapter, while Buddhajñānapāda knew the full seventeen chapters of the Guhyasamāja root tantra, his writings appear to precede the circulation of the Samājottara. This makes the Dvitīyakrama’s description of the ritual for both the second and the third initiations the earliest with which I am familiar.9 5 See Pradīpoddyotana, pp.112-19. and Kusumāñjali, D 271a-276b. Ratnākaraśānti not only makes no mention of the guhyābhiṣeka anywhere in his comments on Chapter Eight of the Guhyasamāja-tantra, but he characterizes the chapter as a whole as describing an extensive ritual for worship (pūjā) (de la mchod pa’i cho ga rgyas pa yang dam tshig yin la de ni le’u ‘di’i don yin no// (Kusumañjali, D 271a.7)), and glosses the single mention in the chapter of the “guru” as referring to the “form of one’s own cakravartin,” (bla ma zhes bya ba ni rang gi ‘khor los sgyur ba’i gzugs so// ibid., D 276b.1), a term that he has earlier in the chapter equated with the form of the practitioner’s yidam deity (bdag po’i phyag rgya zhes bya ba ni dkyil ‘khor gyi bdag po ste rang gi ‘khor los sgyur ba’i phyag rgya chen po’o// ibid., D 276a.1). Modern scholars have similarly made both observations on the presence or absence of the guhyābhiṣeka in Chapter Eight of the root tantra. Isaacson (2010b 264n13) notes, with reference to Chapter Eight of the Guhyasamāja-tantra, that the guhyābhiṣeka may have been originally performed as a separate ritual sequence, suggesting that he may read the guhyābhiṣeka there, while J. Dalton (2004, 16n41) suggests that Chapter Eight describes only a pūjā, rather than an initiatory rite. Dalton goes into further detail on his arguments that Chapter Eight does not include the guhyābhiṣeka, and that Candrakīrti’s reading of the chapter “runs against the grain of the tantra” in his forthcoming study of tantric ritual manuals at Dunhuang (Jacob Dalton, personal communication). 6 bum pa’i dbang bskur ba’i cho ga rtsa ba’i rgyud du bshad nas/ gsang ba’i dbang bskur ba’i cho ga gsungs pa/ mig yangs zhes pa la sogs pa’o// (Samyagvidyākaraṇa, D 192a.2-3) 7 de la bum pa’i dbang bskur ba ni le’u bcu drug pa nyid du yongs su gsal bar mdzad pa yin no// gnyis pa ni mig yangs zhes bya ba la sogs pa gsungs pa la… (Kusumāñjali, D 106b.3). It is worth noting, though, that the term “the wide-eyed one” (viśālākṣīṃ) used in verse 114 of the Samājottara may be alluding to the use of the same term in Chapter Eight, verse 26 of the root Guhyasamāja-tantra, where the term is employed to refer to the consort in the ritual described there. 8 Isaacson 2010b, 264. I am not aware of any traditional commentators, or modern scholars, who suggest that the prajñājñānābhiṣeka is found in the root tantra. 9 There is some evidence suggesting the presence of the guhyābhiṣeka in manuscripts from Dunhuang (see ITJ579, ITJ754, and especially PT321) which reflect a period of ritual development similar to that of Buddhajñānapāda’s time. However, none of the manuscripts contains an especially clear reference to this initiation (Jacob Dalton, 249 In this chapter I will examine the topic of tantric initiation as it occurs in Buddhajñānapāda’s writings, which provide us with a helpful snapshot, as it were, of the state of development of tantric Buddhist initiation in the late 8th and early 9th centuries when he was writing. The Dvitīyakrama in particular appears to have been very influential as the source of a set of popular liturgical verses for the ritual of the third initiation (and the accompanying vidyāvrata10) that were incorporated into at least fourteen (!) later scriptural and authored works.11 I will first give an overview of the initiatory sequences found in the Dvitīyakrama, focusing on several features of this section of the text, including the question of the ritual context of the initiatory sequence in Buddhajñānapāda’s tradition, particularly in terms of the relationship of the lower and higher initiations, respectively, to the two stages of tantric practice, as well as the function of the higher initiations. I will then take up the question of the pointing out of suchness in the context of initiation, which I discussed briefly already in Chapter Three. I examine this topic here as it is found in Buddhajñānapāda’s works, as well as in Vaidyapāda’s short composition on tantric initiation, the Yogasapta. I consider both the role of the seven yogas in this process, as well as the use of the term “the fourth,” to refer to this pointing out of suchness and the relationship of this procedure to what eventually became known as a fourth initiation. As we shall see, Buddhajñānapāda’s writings, as well as those of his immediate disciples, capture a moment in the development of tantric Buddhist initiation where the higher initiations had recently been added on to the set of earlier initiations from the Yoga tantras, and served to provide the disciple with the crucial direct experience of suchness that provided the basis for its cultivation and full realization on the path of the perfection stage.12 Tantric Initiation in Buddhajñānapāda’s Writings: Initiatory Ritual in the Dvitīyakrama Initiatory Ritual Sequence: Lower and Higher Tantric Initiations, Together or Apart? Despite the fact that tantric initiations in Buddhajñānapāda’s system numbered three, the section of the Dvitīyakrama that sets forth initiatory rituals describes only the rituals for the two higher initiations: the second, the guhyābhiṣeka, and the third, the prajñājñānābhiṣeka. The Dvitīyakrama’s presentation of these rituals appears, in terms of the text’s narrative structure, following a general presentation of the nature of suchness and a description of the qualified guru, personal communication). Likewise, it is possible that there may be some reference to the guhyābhiṣeka in Chapter Ten of the Guhyagarbha-tantra, but although later commentators are clear about its presence there, the reference in the tantra itself is not clear (J. Dalton 2004, 22 and 22n59). 10 I address the topic of the vidyāvrata in relation to the third initiation below, with reference to a paper by Christian Wedemeyer delivered at a conference on Buddhist tantra at UC Berkeley in 2014. An updated version of that paper is in the process of publication (Wedemeyer forthcoming). 11 I discuss these verses and the later sources that incorporate them below. 12 This chapter would certainly be much improved if I were able to read Munenobu Sakurai’s 1996 publication, Indo Mikkyō Girei Kenkyū: Kōki indo Mikkyō no Kanchōshidai (A Study on the Ritual of Indian Esoteric Buddhism: Initiation Procedures in Late Indian Esoteric Buddhism), in Japanese, as well as his 2007 article, also in Japanese, on the seven yogas in the Jñānapāda School. The table of contents of the 1996 book, which is given there in English translation, indicates that Sakurai has done a significant amount of research on the development of the ritual of tantric Buddhist initiation, and quite a lot of it with specific reference to the Jñānapāda School. Unfortunately, I do not read Japanese and therefore could not take Sakurai’s research into consideration in writing this chapter. I have, however, benefitted from the Sanskrit edition of Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi included in his 1996 publication, which also contains a Tibetan edition of the section of the Dvitīyakrama on initiation. Unfortunately, I had already completed my own edition of the whole Dvitīyakrama before I became aware of Sakurai’s partial edition in this book, and I was therefore unable to take his edition of this section of the text into consideration in my own. 250 disciple, and consort—all three of whom are necessary in order for a disciple to come to a direct encounter with suchness. The initiatory sequences described thus appear to set forth the process by which the pointing out of suchness takes place, and then the text later goes on to describe the superiority of the view of suchness gained through initiation, and then the yogas by means of which the disciple should train in the cultivation of the suchness that was first encountered during the initiatory sequence. In fact, neither of the two initiations whose rituals are found in the Dvitīyakrama are referred to by their names (or even as initiations at all) in the text itself, but the rituals described are clearly those for the guhya- and the prajñājñānābhiṣeka, as known from later works, and Vaidyapāda explicitly identifies them as such in his commentary. The fact that only the rituals for the second and third initiations are described in the Dvitīyakrama suggests a specific relationship between these two initiations and the second stage of tantric practice that is the Dvitīyakrama’s primary topic. However, it also raises the question of the ritual structure for the initiatory sequence in the early Jñānapāda tradition: since the first initiation is not explicitly mentioned at this point in the Dvitīyakrama, should we understand that it was to have been conveyed in a separate ritual context? In later traditions, it appears that there may have been options for bestowing the ritual for the kalaśābhiṣeka either separately from or together with the higher initiations,13 and this may have been the case in the early Jñānapāda tradition, as well. Buddhajñānapāda mentions the first initiation, the kalaśābhiṣeka, in a verse in the Dvitīyakrama that seems to reference tantric initiations as a set, and the receiving of suchness in that context,14 which suggests that he did understand the group of tantric initiations from the kalaśābhiṣeka up to the higher initiations during which suchness was “received” to constitute a unified set (as was definitely the understanding of the later tradition),15 but in the passage of the Dvitīyakrama that actually sets forth initiatory rituals for the second and third initiations, the kalaśābhiṣeka receives no direct mention. Vaidyapāda, however, adds a rather extensive explanation of the kalaśābhiṣeka in his commentary on that passage. And yet several features of Vaidyapāda’s comments suggest that he may still have had a separate ritual context in mind for the bestowal of the kalaśābhiṣeka. The point in the Dvitīyakrama where this issue arises is in regard to the first verse describing the ritual for the guhyābhiṣeka. It follows the text’s description of the tantric consort (which I discussed in Chapter Six), and the claim that it is by means of practicing with such a female partner that a yogin is able to attain accomplishment. The Dvitīyakrama then begins the description of the ritual for the guhyābhiṣeka stating, Additionally, together with the ordinary, Perform the gaṇapūjā Then, having searched for a girl [who fits the description] that has been taught, 13 See, for example, Vāgīśvarakīrti’s 11th-century Saṃkṣiptābhiṣekavidhi, where it is mentioned that the culminating rituals for initiation, the anujñā and so forth, may optionally be bestowed following the five initiations that constitute the vidyābhiṣeka or after the fourth initiation. amī cānujñādayaś catvāraḥ paṃcasekānantaraṃ caturthasekānantaraṃ va dīyante (ed. Sakurai 1996, 416). Presumably the former option would be for a disciple who was only to receive the first set of inititations, and not (or only later?) to receive the higher inititations. I am grateful to Harunaga Isaacson for drawing my attention to this passage. 14 “Someone who has pleased the guru/ And received the vase [initiation] and the others/ Together with the samayas and vows given by him/ And thus obtains the suchness/ |327| That is found through the guru’s words/…” gang zhig bla ma mnyes byas nas// des gnang dam sdom bcas ba ru// bum pa la sogs rab thob ste// bla ma’i zhal las rnyed pa yi// |327| de bzhin nyid ni rab thob cing// (Dvitīyakrama, verses 327-328a). 15 There is also a set of three verses at the conclusion of the Dvitīyakrama, verses 395-97, that refer to the three initiations in sequence. The reference to the kalaśābhiṣeka there, though, is somewhat veiled. I discuss these verses below. See also note 28. 251 She must be offered to the guru. |83|16 Vaidyapāda explains that the referent of “the ordinary” here is, in fact, the kalaśābhiṣeka.17 According to Vaidyapāda the ritual for the guhyābhiṣeka begins only in the second line of the verse, with the instructions to engage in the gaṇapūjā. He works the presentation of the kalaśābhiṣeka into his commentary on this verse from the Dvitīyakrama as follows: first Vaidyapāda explains “the ordinary” (thun mong), in the first line, to refer to “that which is attained through [the vows of a] bodhisattva, and so forth, and through the ordinary vows, that is to say the vidyābhiṣeka.”18 The term vidyābhiṣeka refers to the set of five (or six) initiations known from the earlier Yoga tantra tradition.19 Vaidyapāda goes on to interpret the next part of the first line of this verse, “together with that” ([de] dang bcas pa) to mean “the extraordinary,” and explains that this refers to “that which is obtained by means of the extraordinary vows—the irreversible ācārya initiation,”20 a tantric initiation authorizing the practitioner to serve as a tantric ritual officiant. He then gives a rather extensive explanation of the ritual for purifying the land in preparation of making an initiatory maṇḍala (sa dag par bya ba’i cho ga), followed by a rather detailed description of the rituals of the vidyābhiṣeka, and the ācāryābhiṣeka, which rites he takes together to constitute the kalaśābhiṣeka.21 In short, then, Vaidyapāda’s commentary on the first line of this verse from the Dvitīyakrama includes a presentation of the kalaśābhiṣeka with respect to a verse that does not, on the surface of things, appear to refer to that ritual. It is, of course, possible that the term “the ordinary” would have been an obvious reference to the kalaśābhiṣeka, though I am not familiar with the use of the term precisely in that way. Vaidyapāda in his Yogasapta makes a reference to the kalaśābhiṣeka as being “given in order to render [oneself] an appropriate recipient/ Of the ordinary and other vows,/”22 so perhaps in this period the term “the ordinary” (*sādhāraṇa?, *sāmānya?) was indeed commonly used with reference to the kalaśābhiṣeka. 23 Yet even if it were the case that the term “the ordinary” was a 16 de yang thun mong dang bcas pa// tshogs kyi mchod pa yang byas te// gang gsungs bu mo btsal nas kyang// bla ma la ni dbul bar bya// |83| (Dvitīyakrama, verse 83). 17 Another possible reading is that “together with the ordinary” in the first line of the verse ought to be read as meaning “together with the ordinary [offerings/worship].” This reading would make sense when one takes the second line of the verse, in which the noun mchod pa, “offerings/worship (pūjā),” is supplied, into account. This is clearly not, however, how Vaidyapāda reads the line, and in the Yogasapta Vaidyapāda also makes a reference to the kalaśābhiṣeka as being “given in order to render [oneself] an appropriate recipient/ Of the ordinary and other vows/. de yi thun mong gzhan pa yi// sdom pa snod du rung phyir sbyin// (Yogasapta, D 69b; P 83b). 18 thun mong zhes pa ni byang chub sems dpa’ la sogs pa dang thun mong du gyur pa’i sdom pas thob pa ste/ rig (rig] D, rigs P) pa’i dbang ngo// (Sukusuma, D 102b.6; P 123b.5-6). 19 Abhayākaragupta uses this term to encapsulate six consecrations, beginning with the garland consecration, which determines the buddha family to which the disciple is connected, along with the five consecrations from water to name that are correlated with the five families (Mori, n.d, 100). 20 de dang bcas pa ni thun mong ma yin pa’i sdom pas thob pa ste/ rdor je slob dpon phyir mi ldog pa’i dbang ngo// (Sukusuma, D 102b.6; P 123b.6-7), 21 At the conclusion of the descriptions of the rituals for the vidyābhiṣeka and the ācāryābhiṣeka, Vaidyapāda writes that he has just given a brief description of the kalaśābhiṣeka, and that more detailed ritual procedures should be obtained from elsewhere. This comment suggests that he takes the vidyābhiṣeka and the ācāryābhiṣeka (which he refers to sometimes as the “irreversible ācāryābhiṣeka” (phyir mi ldog pa’i slob dpon gyi dbang) and other times just as the “irreversible abhiṣeka” (phyir mi ldog pa’i dbang)) together as the kalaśābhiṣeka (Sukusuma, D 104b.5; P 126a.3). The rituals for these initiations are described in Sukusuma D 102b.6-104b.5; P 123b.6-126a.4. The consideration of all of the initiations up to, and also including, the ācāryābhiṣeka as the kalaśābhiṣeka also appears in the Vajrāvalī and the Kriyāsammucaya (Sanderson 1994, 90). 22 De dag dang ldan pa’i slob ma la gsang dbang bskur bar gsungs pa/ (Sukusuma, D 104b.6; P 126a.4). 23 There appears to be at least some evidence for a use of sādhāraṇa in a somewhat similar way in a later work, Kṛṣṇa/Kāṇha’s Yogaratnamālā on the Hevajra-tantra (...kriyātantrādisādhāraṇāvaivartikābhiṣekalābhamātreṇa…; see Snellgrove 1959, 108), and sāmānya is attested for vows in the 11th -century Vajrāvalī (Mori 2009, Vol. 2, 429). 252 commonly used and obvious reference to the kalaśābhiṣeka, the allusion to the first initiation in this verse of the Dvitīyakrama is at best a cursory one. Moreover, Vaidyapāda’s presentation of the kalaśābhiṣeka at this point in his Sukusuma also seems to indicate that he may have had a separate ritual context in mind for that initiation. He ends the description of the ritual for the kalaśābhiṣeka with the series of rites, such as the vyākaraṇa, anujñā, and āśvāsa, that normally conclude an initiatory sequence, and he mentions these same rites again at the conclusion of his presentation of the third initiation later in the Dvitīyakrama, suggesting that he may understand these as two separate ritual sequences.24 He also begins his explanation of the guhyābhiṣeka, by noting that it is to be given to “a disciple who [already] has those [earlier initiations].”25 Vaidyapāda even states in the Sukusuma that his presentation of the kalaṣābhiśeka there has been made “quickly, just in order to uphold/remember it (skyus kyis ‘dzin tsam du smros),” with the advice that the details for the ritual should be found elsewhere.26 These comments suggest that despite the fact that he has included a presentation of the kalaśābhiṣeka in his commentary on the Dvitīyakrama, Vaidyapāda may not—or at least not necessarily—have intended this first initiation as part of the same ritual sequence of the guhya- and prajñājñānābhiṣekas, which are clearly detailed in the Dvitīyakrama and appear intended to be bestowed as a pair. Given that the Dvitīyakrama is concerned primarily with the perfection stage, its lack of an explicit (or at least a detailed) mention of the kalaśābhiṣeka in conjunction with the ritual sequence for the guhya- and prajñājñānābhiṣekas, is suggestive of several things. First, as I have showed here, it may indicate that in the early Jñānapāda tradition the kalaśābhiṣeka was (or at least optionally could be) bestowed in a separate ritual sequence from the higher initiations; that is, the Dvitīyakrama may assume a student who has already received the kalaśābhiṣeka separately—we may here recall Vaidyapāda’s statement that the guhyābhiṣeka is to be given to “a disciple who [already] has those [earlier initiations]”27—and is thus now ready to receive the guhya- and prajñājñānābhiṣekas in order to begin the practice of the second stage of tantric practice. But also, as I have noted above, the fact that the Dvitīyakrama only outlines the rituals for the guhya- and prajñājñānābhiṣekas clearly indicates a special relationship between the higher tantric initiations and the second stage of tantric practice, which is the Dvitīyakrama’s main focus. This special relationship, as we will see below, entails the pointing out of suchness Thanks to Harunaga Isaacson both for the suggestions of possible Sanskrit terms for thun mong and for the references to similar usages of the term in the Yogaratnamālā and the Vajrāvalī. 24 Sukusuma D 104b.6; P 126a.4 and Sukusuma D 110a.5. In the later tradition, at least, it seems that these concluding rituals were to be given after the kalaśābhiṣeka only in the instance where the ritual sequence ended with that initiation; if the initiatory sequence were to combine both the lower and higher initiations, these concluding rituals would be given only at the very end. See note 13. 25 de dag dang ldan pa’i slob ma la gsang dbang bskur bar gsungs pa/ (Sukusuma, D 104b.6; P 126a.4). Vaidyapāda’s commentary on Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi identifies the candidate for the guhyābhiṣeka as “the very one who received the vidyābhiṣeka and the irreversible [ācārya] abhiṣekam,” (rig pa’i dbang dang phyir mi ldog pa’i dbang thob pa de nyid) which might be taken to suggest a single ritual sequence, but in that same work he also mentions that the maṇḍala for the guhyābhiṣeka could be “either the very same maṇḍala as previously [employed] or another one,” (sngar gyi dkyil ‘khor ‘di nyid dam gzhan yang rung) suggesting perhaps the possibility of both the option of a single ritual sequence for all of the initiations or a separate sequence for the higher initiations (Guhyasamājamaṇḍalopāyikā-ṭikā, D 110a.5). A study of Dīpaṃkarabhadra’s ritual manual and Vaidyapāda’s commentary on it would be a tremendous contribution to our understanding of initiation in the early Jñānapāda School. The doctoral dissertation of Daisy Cheung of Hamburg University on Dīpaṃkarabhadra’s work is thus very much awaited. 26 De rnams ni bum pa’i dbang bskur ba’i cho ga bsdus pa ste/ rgyas par ni gzhan du shes par bya ste/ ‘dir ni dkyus kyis ‘dzin tsam du smos so// (Sukusuma, D104b.5; P 125a.3). 27 de dag dang ldan pa’i slob ma la gsang dbang bskur bar gsungs pa/ (Sukusuma, D 104b.6; P 126a.4). 253 in the context of these higher initiations, such that the practitioner is able to experience it directly, and thus knows what to train in during the subsequent cultivation of the perfection stage through the practice of the bindu yogas. With the development of the higher consecrations that were associated with the second stage of tantric practice, it seems that the kalaśābhiṣeka, as a “lower” set of initiations thus began to be associated with the first stage, the generation stage, of tantric practice. Indeed, the association of the five initiations of the Yoga tantra tradition with the five wisdoms, and the similar association of the five stages of the process of deity generation according to the pañcākarābhisaṃbodhi with the five wisdoms makes such a connection quite logical. While neither Buddhajñānapāda’s nor Vaidyapāda’s works mention the kalaśābhiṣeka as a ritual that gives the practitioner permission to train in the generation stage, specifically, there is one verse in the Dvitīyakrama that appears to associate the kalaśābhiṣeka with the first stage of tantric practice.28 The later tradition did come to make the association of the kalaśābhiṣeka with the generation stage more explicit, and to specify that the function of this first initiation was to give permission for the practice of the generation stage.29 28 This is verse 396, which is part of the final dedication at the end of the work. This verse is the first of a series of three verses focusing on each of the three initiations. The association of verses 396 and 397 with the guhyā- and prajñājñānābhiṣekas, respectively, while not mentioned in Vaidyapāda’s commentary, is very obvious; even the language of these verses closely mirrors the language used to describe those two initiations in the earlier section of the Dvitīyakrama. The first line of verse 395 clearly references the ācāryābhiṣeka, and Vaidyapāda explains (perhaps precisely because it is less obvious than the association of the subsequent two verses with the subsequent two initiations!) that final line of this verse is a reference to the kalaśābhiṣeka. The penultimate line mentions the first stage of tantric practice, but the verse seems to make reference to having already realized the first stage of practice, rather than to being initiated in such a way that one receives permission to practice it. The verse reads: “When one has been cleansed and sprinkled and made pure, and thus become a great ācārya/ Who holds all of the tantras, and brings others to connect with all tantras,/ And having perfectly realized the first stage and purified all stains,/ May the yogin become a suitable vessel for illusory wisdom!/” |395| blugs dang gtor dang dag par byas pas slob dpon cher ‘gyur te// thams cad kun kyi rgyud ‘dzin gzhan rnams rgyud kun la sbyor ba’i// dang po’i rim pa rab rtogs dri ma rnams ni dag byas te// ye shes sgyu ma’i snod du rung bar rnal ‘byor de ‘gyur shog// |395| (Dvitīyakrama, verse 395). Vaidyapāda explains that becoming a “suitable vessel” for wisdom is a reference to the kalaśābhiṣeka—clearly referencing the fact that a kalaśa, a vase, is a type of vessel (Sukusuma, D 138b.4). In his Yogasapta Vaidyapāda bases his description of the etymology of the kalaśābhiṣeka on precisely this verse from the Dvitīyakrama. He writes: “Because [it makes one into] a suitable vessel it is [called] the “vase.”/ Because one is sprinkled, consecrated, and purified it is [called] “initiation.”/ |13| snod du rung phyir bum pa ste// blugs gtor dag par byed pas dbang// |13| (Yogasapta, D 70a; P 83b). 29 See, for example, Lakṣmī’s Svādhiṣṭhānakrama: phyir mi ldog pa’i dgang bskur ba yang bskyed pa’i rim pa’i dbang bskur ba yin te/ (cited in Tomabechi 2006, 147n190). The later Tibetan tradition likewise makes the association of the kalaśābhiṣeka with the generation stage and the role of the kalaśābhiṣeka as a ritual giving permission to practice that stage even more explicit, and further specifies the different aspects of perfection stage practice that are associated with the higher initiations. An 18th-century Tibetan liturgy that includes a description of the function of each of the four initiations states, with respect to the kalaśābhiṣeka: “Having in this way received the kalaśābhiṣeka the stains of the body are purified, one is empowered to train in the path of the generation stage, the wisdom that perceives everything just as it appears as the divine maṇḍala is actualized, and one gains the fortune to attain the resultant [state] of the nirmāṇakāya.” (de ltar bum pa’i dbang bskur bas lus kuyi dri ma dag/ lam bskyed rim bsgom pa la dbang/ ji ltar snang bat hams cad lha’i ‘khor lor ‘char ba’i ye shes mngon du byas/ ‘bras bu sprul pa’i sku thob pa’i skal pa can du byas pa’o// (Rdo rje theg pa sngags kyi gso sbyong bdud rtsi’i rol mtsho zhes bya ba bzhugs so, 69a.3-6). The same manual also associates the second, third, and fourth initiations with different aspects of the practices of the perfection stage but this is not the place to get into those distinctions. (A brief aside on the remarkable continuity of tradition: this same 18th-century Tibetan liturgy includes—not as a citation, but simply as part of the liturgy to be chanted as part of this extensive group sādhana practice—a section on the fourteen root downfalls of the tantric samayas that is also found in Vaidyapāda’s 9th-century Samyagvidyākara, his commentary on the Samājottara. I do not know the source of the passage—I imagine it is likely a scriptural one— but nonetheless I do find the continuity of tradition across the centuries worth remarking on!) 254 However, as we saw already in Chapter Four, in Buddhajñānapāda’s tradition the generation stage and the perfection stage seem not to have been completely separated out from one another in a practical sense. Indeed, Buddhajñānapāda’s Caturaṅga/Samantabhadrasādhana, generally known as the generation stage sādhana in his Guhyasamāja practice system, includes several practices that appear to be connected in some way to the perfection stage. Thus, although the passage from the Dvitīyakrama dealing with tantric initiation and Vaidyapāda’s commentary on it both suggest the likelihood that in their tradition the kalaśābhiṣeka was, or at the very least could be, given in a separate ritual context from the two higher initiations, the fact that the Caturaṅga/Samantabhadra-sādhana includes practices that seem likely to have been permitted only for a practitioner who had also obtained the second and third initiations raises some questions. That is, if the kalaśābhiṣeka was already, at the turn of the 9th century, associated specifically with permission to practice only the generation stage, and if it were bestowed in a separate ritual from the higher initiations, simply on the basis of the Guhyasamājarelated practice texts that survive in Buddhajñānapāda’s own oeuvre, it is not clear what ritual manual would have been suitable for practice by a disciple who had received only the kalaśābhiṣeka and not the higher initiations. Given that there seems to be some evidence pointing in both directions, it may not be possible to resolve the question of the ritual context of the higher initiations in the early Jñānapāda School with certainty.30 The Dvitīyakrama’s Initiatory Sequence: The Guhyābhiṣeka The Dvitīyakrama’s presentation of the ritual for the guhyābhiṣeka is quite brief, and yet it is clear and direct both with respect to the ritual procedure, as well as to the function of this initiation. Later manuals for bestowing tantric initiation like Abhayākaragupta’s Vajrāvalī provide more detail about the ritual procedures for the guhyābhiṣeka. But that manual, for example, outlines the details of the ritual as it was practiced at the turn of the 12th century, several centuries after its practice seems to have emerged. The Dvitīyakrama, dating to the turn of the 9th century, is, as I noted above, the earliest work of which I am aware to clearly set forth procedures of the ritual for bestowing the guhyābhiṣeka,31 and thus provides us with an important early picture of this initiation. The ritual for the guhyābhiṣeka, along with some comments about its function and purpose, is presented in the Dvitīyakrama in just three verses: Additionally, together with the ordinary, Perform the gaṇapūjā 30 There is also some evidence from the later Jñānapāda tradition that the higher initiations may have been given in a separate ritual context. In Kṣitigarbha’s Daśatattva, only these two initiations are described, and included as one among the ten fundamentals that are the principal topic of this text. The absence of a presentation of the kalaśābhiṣeka in this work may also indicate that it was bestowed in a separate ritual context. And, in fact, just like Vaidyapāda’s commentary on the Dvitīyakrama, the Daśatattva also specifies the candidate for the two higher initiations as someone who has already received the vidyābhiṣeka and the ācāryābhiṣeka. The text itself gives the following explanation for its presentation of only these two initiations (in Klein-Schwind’s translation): “Now the two initiations will be explained in detail. [The dual form] “two initiations” refers to the secret and the wisdom [- gnosis initiation], which are explained because they are the foremost. It is precisely (eva) these two that are explained in the form of fundamentals (tattvarūpeṇa), since the other [initiations] are subordinated to them/function as their limbs (etadaṅgabhūtatvāt) in that they are exoteric in as much as they purify the body. The mantrin who has received the [five] vidyā-initiations and the initiation to become an officiant duly worships an excellent teacher…” (Schwind 2012, 282). 31 As I noted above this is the case if one considers, like Vaidyapāda and Ratnākaraśānti seem to, that the ritual procedure described in Chapter Eight of the Guhyasamāja-tantra is not the guhyābhiṣeka, and if I am correct in my assessment that Buddhajñānapāda’s works precede the Samājottara (See Chapter Eight). 255 Then, having searched for a girl [who fits the description] that has been taught, She must be offered to the guru. |83| Then, when the guru is pleased, He engages in union with her Due to which the sugatas melt and become the sixteenth part; This is dropped in the mouth of the disciple, |84| And having descended, it enters the lotus at his heart. Through this the field is purified And the twelve [experiences]—[perceiving] all phenomena as illusions, and so forth— Are realized in actuality. |85|32 We have already examined the first line of this verse, which Vaidyapāda explains as a reference to the kalaśābhiṣeka. The ritual for the guhyābhiṣeka thus appears to begin with the performance of a gaṇapūjā.33 As the Dvitīyakrama straightforwardly explains, in this initiation, the guru engages in sexual union with the consort, and the resultant bindu of bodhicitta, described in the text as the “sixteenth part,”34 is dropped into the mouth of the disciple, and descends to his heart.35 The text follows this direct presentation of the ritual procedure with some comments on the function of the guhyābhiṣeka—it is through the ritual of consuming this bindu of bodhicitta received from the guru, which is identified as the embodiment of the sugatas, that the disciple’s “field is purified.” Vaidyapāda explains that “the field” here means the disciple’s own aggregates.36 As we saw earlier, in the Guhyasamāja-tantra the five aggregates are identified 32de yang thun mong dang bcas pa// tshogs kyi mchod pa yang byas te// gang gsungs bu mo btsal nas kyang// bla ma la ni dbul bar bya// |83| de nas bla ma de mnyes nas// de dang snyoms par zhugs pa yis// bde gshegs zhu gyur bcu drug char// gyur pa slob ma’i kha ru ltung// |84| ltung bas snying gi padma zhugs// de yis zhing ni dag byas te// chos kun sgyu ma la sogs pa’i// bcu gnyis don du rtogs par ‘gyur// |85| (Dvitīyakrama, verses 83-85). 33 In his commentary on the Samājottara and his commentary on Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, Vaidyapāda likewise states that a gaṇapūjā is to be performed at the outset of the guhyābhiṣeka. In that commentary Vaidyapāda describes the ritual as follows: “Having offered the wide-eyed one together with the gaṇapūjā to the guru, he bestows the initiation, as [will be] described, upon the disciple” (de yang ji skad du gsungs pa’i mig yangs tshogs kyi mchod pa dang bcas te bla ma la phul nas des ji skad du gsung pa’i dbang bskur ba slob ma la sbyin par bya’o//) (Samyagvidyākara, D 103a.3). The ‘wide-eyed one’ is a term used to refer to the consort in Chapter Eight of the root Guhyasamāja-tantra, as well. That chapter is interpreted by some commentators as describing initiatory practices. See also Vaidyapāda’s Guhyasamājamaṇḍalopāyikā-ṭīkā D 110a.6. 34 The term “the sixteenth part” bcu drug cha and bcu drug phyed cha is used at several points in the Dvitīyakrama (see also verse 1 and verse 122) to indicate the bindu of bodhicitta in the context of initiation and sexual yogic practices. I believe bcu drug cha may be a translation of ṣoḍaśakalā, understood in each of these contexts as “the sixteenth part,” referencing the sixteenth phase of the moon in the lunar month, the day when the fullness of the fifteenth phase is perfectly complete, and is here in all three verses used to indicate the bindu/bodhicitta drop. Alternatively, the term could be ṣoḍaśakala, referring more generally to the moon itself as “that which has sixteen parts.” In either case the association with the drop of bodhicitta remains relevant, as the moon in general is frequently used as a metaphor for bodhicitta. See note 9 in my translation of the Dvitīyakrama for further discussion of this point. 35 Vaidyapāda clarifies this statement: “Having become fully impassioned, [the guru] enters into union [with her]. The sugatas, who have been invoked by the seed [syllable], enter into the mouth and one should think that having melted as the moon they become the sixteenth part, that is, the essence of the bindu.” rjes su chags pa’i mtha’ la thug pas snyoms par zhugs te/ sa bon gyis bskul ba’i bde bar gshegs pa rnams zhal du zhugs te zla bar zhu bar (zhu bar] D, P om.) gyur nas bcu drug cha zhes te thig le’i ngo bor gyur bar bsam mo// (Sukusuma, D 105a.5-6; P 126b.4-5). He further explains the visualization for this procedure: “That [drop] itself is dropped into the disciples’s mouth means that from the bindu comes a syllable and from that arises a maṇḍala and that [maṇḍala] itself, which has melted due to the heat of great passion, is [then] given into the mouth of the disciple.” de nyid slob ma’i kha ru ltung/ zhes te thig le las yi ge/ de las dkyil ‘khor de nyid ‘dod chags chen po’i mes bzhus pa slob ma’i khar sbyin pa’o// (Sukusuma, D105a.6; P 126b.5-6). 36 de yis zhing dag byas te/ zhes te zhing ni de’i phung po la sogs pa’o// (Sukusuma, D 105a.7; P 126b.7-8). 256 with the five buddhas, a doctrine that is ritually re-enacted in Buddhajñānapāda’s generation stage sādhana. Here in the guhyābhiṣeka the drop of bodhicitta that the disciple receives from his guru is identified with the buddhas, and its ritual consumption thus serves as a method for purifying the disciple’s aggregates. The effect of this purification, the Dvitīyakrama continues, is that the disciple realizes, in actuality, the illusory nature of phenomena.37 This presentation in the Dvitīyakrama of the function of the second initiation thus indicates that the ritual serves to both purify the disciple as well as to bring him or her to an experience that approaches the direct experience of suchness (the actual direct experience of which, as we will see, will take place only in the third initiation). In Mahāyāna scriptures the post-meditative experience of a bodhisattva is often said to entail a perspective in which he perceives everything as illusory or dream-like, while the meditative equipoise of suchness itself—the direct experience of emptiness—is a state is described as space-like.38 Here, the Dvitīyakrama’s description of the function of the second initiation parallels that of the bodhisattva’s post-meditative experience; the initiation brings the disciple into a state that is similar to a direct experience of suchness, or at least closer to that experience than is his ordinary perception of reality. Indeed, Vaidyapāda’s comments on this passage link the experience of the disciple directly to that of the bodhisattvas. He writes that having received this initiation and having his “field” thereby purified, the disciple “attains equal fortune to bodhisattvas such as Maitreya, and will thus travel from buddhafield to buddhafield. Thus, having entered this path he will swiftly attain accomplishment.”39 Vaidyapāda’s comments here appear to be echoing verse 396 in the Dvitīyakrama, which is a summary of the guhyābhiṣeka in the form of a dedicatory verse.40 Certainly the transgressive nature of the rite for the guhyābhiṣeka—involving both sex, which would in and of itself have been transgressive, along with the ritual consumption of impure substances—was part of what was meant to bring the disciple into a state that would transcend his ordinary perceptions of reality. Indeed, the use of transgressive and antinomian acts as a method for cultivating states of nonduality or nonconceptuality is an important feature of Buddhist tantric practice.41 However, the way that the guhyābhiṣeka is described in the Dvitīyakrama suggests that the experience that this ritual evoked was only a step in the direction of an experience of genuine reality, not yet the full experience. This second initiation, however, is then followed by the third, the prajñājñānābhiṣeka, which, according to the Dvitīyakrama, brings the disciple even further, to a direct experience of suchness or dharmakāya itself. The Dvitīyakrama’s Initiatory Sequence: The Prajñājñānābhiṣeka 37 Vaidyapāda clarifies that the twelve experiences mentioned in the root text are the twelve examples that show phenomena to be illusory: an illusion, a mirage, an echo, a spinning firebrand, a delusion, a dream, a city of gandarvas, a bubble on water, a flash of lightening, an emanation, a rainbow, and a cloud (Sukusuma, D 105b.1-2). 38 See, for example Subhūtighoṣa’s Sarvayānālokakaravaibhāṣya: rnam par mi rtog pa’i ye shes kyis nam mkha’ dkyil ltar rtogs la, rjes thob kyis sgyu ma ltar rtogs te/ (Sarvayānālokakaravaibhāṣya, D 313a.3) and Jñānavajra’s Laṅkāvatāra-vṛtti: byang chub sems dpa’ rjes thob kyis gnas skabs su ‘khor gsum sgyu ma lta bur rtogs pas’I shes rab kyis… (Laṅkāvatāra-vṛtti, D 32b.2-3). 39 de byams pa la sogs pa’i byang chub sems dpa’ rnams dang skal pa mnyam pas na sangs rgyas kyi zhing nas sang rgyas kyi zhing du ‘gro ste/ de’i lam la ‘jug pa mnyur bar thob par ‘gyur ro// (Sukusuma, D 105b.2-3; P 127a.2-3). 40 That verse reads: “Through respectfully [serving at] the feet of a compassionate guru/ And by means of that which has the rabbit-holder’s form, may one’s mindstream be perfectly ripened/ So that the field is purified, and one perfectly realizes the reality of phenomena to be illusory and the like:/ In this way may all beings, like Maitreya and others, arrive [in that state].” |396| snying rje ldan pa’i bla ma’i zhabs la gus par rab ldan pas// ri bong ‘dzin pa’i gzugs kyis rang rgyud rab tu smin byas te// zhing dag byas pas chos kun sgyu sogs don du rab rtogs nas// byams pa la sogs bzhin du sems can kun gyis ‘gro bar shog// |396| (Dvitīyakrama, verse 396). 41 See Wedemeyer 2012 (especially chapters 5 and 6) on the tantric Buddhist use of transgression as a method to evoke states of nonduality and nonconceptuality. 257 The Dvitīyakrama’s presentation of the ritual for the prajñājñānābhiṣeka is much more extensive than that of the guhyābhiṣeka, comprising thirty-nine verses compared to the guhyābhiṣeka’s three. While the ritual procedure for the third initiation as it is described in the Dvitīyakrama is also more complex than that of the guhyābhiṣeka, thus necessitating a longer presentation, this attention to the third initiation is also indicative of the crucial importance of the prajñājñānābhiṣeka, in particular, in Buddhajñānapāda’s practice system. The section of the Dvitīyakrama that sets out this ritual is also noteworthy in that it contains the earliest instance of which I am aware of a set of quite popular liturgical verses for the bestowal of the prajñājñānābhiṣeka as well as the vidyāvrata, the so-called “consort observance.” The liturgical passage includes verses spoken by the guru as he gives the female consort to the male disciple with the command never to separate from her for the duration of saṃsāra, along with a set of verses recited by the initiate partners to one another, as a sort of dialogue. Several lines from a later passage describing the culmination of the third initiation likewise appear to have been quite popular. Parts of the section of the Dvitīyakrama on the prajñājñānābhiṣeka were thus incorporated into quite a number of later tantric works, both scriptural and authored,42 including the Samājottara,43 the Caṇḍamahāroṣaṇa-tantra, Vaidyapāda’s Guhyasamājamaṇḍalopāyikāṭīkā, Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, Nāgabodhi’s Maṇimālā, Advayavajra’s Samkṣiptābhiṣekaprakriyā, Kṛṣṇācārya’s Śrīguhyasamājamaṇḍalopāyikā, Prajñāgupta’s Abhiṣekaratnāloka, Prajñāśrī’s Abhiṣekavidhi, Vagīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, Kṣitigarbha’s Daśatattvasaṃgraha, Ratnākaraśānti’s Ratnāvalī, Abhayākaragupta’s Vajrāvalī, and Kuladatta’s Kriyāsamgraha.44 42 It is of course possible that Buddhajñānapāda incorporated these verses from an earlier source, but the Dvitīyakrama is the earliest source I am aware of to include these verses. 43 See Chapter Eight for my arguments that the Samājottara circulated only after Buddhajñānapāda’s writings. 44 Some of the works listed here contain only the liturgy for the guru, others contain only the liturgy for the initiate couple, and some contain both (see details below). Given the obvious popularity of these verses from the Dvitīyakrama, I imagine that there are other works that include them, as well. The verses were studied by Christian Wedemeyer in a paper on the vidyāvrata presented at a conference at UC Berkeley in 2014. I have Wedemeyer’s work to thank for pointing out that these verses are found in Vaidyapāda’s Maṇḍalavidhiṭīkā, Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, Advayavajra’s Samkṣiptābhiṣekaprakriyī, Kṛṣṇācārya’s Guhyasamājamaṇḍalopāyikā, Kuladatta’s Kriyāsamgraha, Prajñāgupta’s Abhiṣekaratnāloka, and Prajñāśrī’s Abhiṣekavidhi; I have since discovered parallel passages in the other sources listed above. Wedemeyer was not aware at the time of that presentation of the Dvitīyakrama as the source of these verses (all of the works from which he cited these verses in his 2014 paper are later than the Dvitīyakrama). In our communications since that time, I have shared this fact with him and his forthcoming publication on the vidyāvrata has now been updated to include this information. For parallels of the liturgical verses for the guru’s bestowal of the consort on the disciple see, for verses 87b-88d of the Dvitīyakrama: Vajrāvalī (Mori 2009 Vol. 2, 444); for verses 87b-89d of the Dvitīyakrama: Daśatattva V.14 (Klein- Schwind 2012, 209); for verses 87b-88c of the Dvitīyakrama, Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi (verses 365c-366b) and Vāgiśvarakīrti’s Saṃkṣiptābhiṣekavidhi (see Sakurai, 218); and for verse 89 of the Dvitīyakrama: Samājottara 125c-126d. For parallels with the liturgical verses involving the initiate couple’s calland- response see, for verses 91a-95d from the Dvitīyakrama, Caṇḍamahāroṣaṇa-tantra III.26-27; Vaidyapāda’s Guhyasamājamaṇḍalopāyikāṭīkā, which gives the first part of these verses in a very garbled Sanskrit transliteration, and the second part in Tibetan translation (!) (211a.4-5); Ratnākaraśanti’s Piṇḍīkṛtasādhanopāyikāvṛtti-ratnāvalīnāma (91b.6-7); Nāgabodhi’s Pañcakramaṭīkā-maṇimālā-nāma (130b.5-7); Advayavajra’s Samkṣiptābhiṣekaprakriyā (131b.5-7); Kṛṣṇācārya’s Śrīguhyasamājamaṇḍalopāyikā (258a.4-6); Prajñāgupta’s Abhiṣekaratnāloka (299a.7-b.2); and Prajñāśrī’s Abhiṣekavidhi (48b.4-5). For parallels with verses 91a-94d of the Dvitīyakrama see Kṣitigarbha’s Daśatattvasaṃgraha (V.17-20) (Klein-Schwind 2012, 210). Please see also the notes to verses 87-95 of my translation of the Dvitīyakrama in this dissertation for more details on the parallels, including comments on the variations found within these various parallel passages. I address the works that have parallels with the lines from the verses on the culmination of the third initiation below. 258 The Dvitīyakrama begins its description of the initiatory sequence for the prajñājñānābhiṣeka by clearly stating the purpose of this third initiation.45 The text reads: And then, in order to bring about the realization Of the self-arisen dharmakāya, great joy That is equal to space, called the *adhideva, The girl is given to him [i.e. the disciple]. |86|46 If the second initiation was described in the Dvitīyakrama as bringing about an experience of phenomena as illusory that parallels descriptions in Mahāyāna texts of the bodhisattva’s postmeditative experience, the third initiation’s function according to the Dvitīyakrama parallels the traditional Mahāyāna descriptions of the bodhisattva’s experience in meditative equipoise—a direct experience of emptiness, or suchness, that is, precisely as it is described in the Dvitīyakrama, “equal to space” or sky-like.47 As I discussed in Chapter Three, in this verse Buddhajñānapāda uses a term, the *adhideva, that describes the result of awakening in a uniquely tantric way, and which he links here with the dharmakāya, the fundamental and formless “body” of a buddha’s awakening. It is in order to bring about this realization—that is to say a direct, unmitigated experience of suchness, the result of awakening itself—that the consort is given to the initiate for the ritual of the prajñājñānābhiṣeka. In his commentary, Vaidyapāda makes the link between the purification achieved through the guhyābhiṣeka and the direct experience of dharmakāya by means of the third initiation explicit, suggesting that the purification of the disciple’s aggregates by means of the earlier ritual serves as a preliminary foundation for the direct realization of dharmakāya.48 The ritual for the third initiation as it is described in the Dvitīyakrama begins with the guru’s giving over of the female consort to the male disciple with the instruction that in order to attain awakening he must experience bliss arising from practicing the maṇḍala-cakra ritual together with her. The guru specifically tells the initiate that “nothing else can bring about 45 In this section I summarize the third initiation ritual as described in verses 86-125 of the Dvitīyakrama, and draw attention to the structure of the ritual and some of its key features. My notes to the translation of this section of the Dvitīyakrama, however, include quite a bit more information about this section of the text than I have included here; I would direct the interested reader to the translation of and notes to verses 86-125 of the Dvitīyakrama. 46 de nas de la rang ‘byung gi// chos sku rab dga’ mkha’ mnyam pa// lhag pa’i lha zhes bya ba ni// rtogs bya’i ched du bu mo byin// |86| 47 This metaphor is repeated during the dialogue between the yogic partners during the prajñājñānābhiṣeka. The yogin says to his consort: “Sweet-faced one, come play with me/ And [we] will have an experience that is like the sky!” zhal bzang khyod ni nga dang lhan cig tu// rnam par rtsen pas mkha’ ‘dra myong bar bya// (Dvitīyakrama, verse 100ab). 48 He writes, “In order that the disciple, whose field has been purified, now comes to realize the dharmakāya, the prajñājñānābhiṣeka is presented.” da ni zhing dag par byas pa’i slob ma la chos kyi sku rtogs par bya ba’i phir shes rab ye shes kyi dbang bskur ba gsungs pa/ (Sukusuma, D 105b.3; P 127a.3). It is worth noting that the position that the direct experience of the dharmakāya itself occcurs during the third initiation is explicitly and strongly rejected in the writings of some later tantric authors, like Rāmapāla in his Sekanirdeśapañjikā. This position is there refuted as duḥseka, a wrong or unwholesome abhiṣeka. As Isaacson and Sferra summarize the problem, from Rāmapāla’s perspective, of holding this position, “the error of the duḥseka proponents is not just that they claim (or at least imply) the ultimate reality of non-dual mind, but that they claim that reality is directly experienced in sexual union. Such a position, amounting to the reification of sexual bliss as an absolute, is not only philosophically inadequate (in a manner parallel to the Yogācāra position) but is highly dangerous, entailing as it does that sexual intercourse is all that is necessary to experience (at least temporarily) true Awakening, and is also dangerously similar to the views of the Śaiva Kaula tradition, at least as understood by Maitreyanātha and Rāmapāla” (Isaacson and Sferra 2014, 104). While I would argue that these errors do not follow (!) from Buddhajñānapāda’s claim that the dharmakāya is experienced directly during the third initiation (though I will resist my impulse to compose a defense of his position in regard to these specific objections here), it is certainly possible that his writings were among those that Rāmapāla considered incorrect, and even dangerous, in this regard. 259 buddhahood,” and that therefore he is not to separate from his partner until the end of saṃsāra. Vaidyapāda’s commentary on the Dvitīyakrama specifies that the consort is given to the disciple’s right hand, and later liturgical sequences for this ritual mention this detail explicitly. Several scholars have remarked on the similarity between this practice and the traditional Indian marriage rite, in which the bride and grooms’ hands are joined in this same manner.49 When this ritual element is considered together with the guru’s command that the student should not separate from his partner “throughout endless saṃsāra,” the rite is indeed rather evocative of a marriage ceremony. Following the consort’s being handed over to her partner in this way, the Dvitīyakrama sets forth a series of verses in which the initiate couple address each other directly in a rather playful dialogue. Here, the female consort, addressing the male initiate with a term of endearment (vatsa, bu), questions her partner about his willingness to engage in the ritual consumption of several samaya substances—substances that are ordinarily considered quite impure—and to kiss her bhaga (vagina). The yogin, addressing his partner as “goddess” and acknowledging that she is deserving of respect, is to reply—joyfully and with laughter—that he is absolutely ready to do all of these things without a second thought. The consort then directly shows him her “lotus” and recites some verses of praise to the lotus itself, lauding it as the abode of self-arisen great bliss, the place of veneration of all the buddhas, and that which brings about the accomplishment of one’s aims. Following this spirited exchange between the initiate couple are three short sequences that Vaidyapāda identifies as the processes by which the couple mentally, verbally, and finally physically “cultivates passion” ahead of the sexual union that constitutes the main part of the third initiation. The mental cultivation of passion involves a visualization of circulating the deities of the body maṇḍala between both partners. The body maṇḍala deities are stimulated by light rays emerging from the seed syllable at the practitioner’s heart, and thus impassioned so that they emerge from the yogin’s vajra into the consort’s lotus. This causes the maṇḍala deities within her body to likewise become impassioned and to emerge from her mouth and enter into the yogin’s mouth. Repeating this visualization of the circulation between the two partners is said to bring about intense mental passion in the maṇḍala deities, and Vaidyapāda adds that this creates mental passion in both practitioners, as well. The second stage of “impassioning” is a verbal one, in which the initiate couple speak passionate words to each other, calling on each other to “come play” and to “sport” together, in order to bring about a sky-like experience and to realize “that which is not realized by other [means].” In this way, through speaking “illusory 49 Klein-Schwind (2012, 291n 1062) cites Isaacson on this point, (noting that pāṇigrahana, “taking by the hand” is a common term for marriage); on this point see also Wedemeyer (unpublished 2014 and forthcoming), who likewise remarks on the pāṇigrahana element, and Onians (2003, 176), who refers to this ritual of the vidyāvrata as a sort of “sacred marriage.” In his 2014 paper at UC Berkeley on the vidyāvrata, Wedemeyer examined the relationship between the vidyāvrata and the prajñājñānābhiṣeka and concluded that the literature generally shows one of two ritual paradigms: one in which the vidyāvrata, the main element of which is precisely this handing over of the consort to the disciple and the guru’s command for the disciple to remain with her, is conducted separate from the initiations, and another in which it is joined with, or even identical to, the prajñājñānābhiṣeka. The Dvitīyakrama appears to adhere to the latter paradigm in that a ritual for the bestowal of a vidyāvrata is not separated out from the prajñājñānābhiṣeka, and Vaidyapāda’s commentary on the Dvitīyakrama likewise makes no reference to the bestowal of the vidyāvrata. However, in his commentary on the Samājottara, which does make mention of bestowing the vidyāvrata, Vaidyapāda identifies the ritual described in verses 124-126 of the Samājottara as the “vidyāvrata initiation” (rig pa’i brtul zhugs kyi dbang) (Samyagvidyākara, D 192a.6). Interestingly, a full four pādas of those verses from the Samājottara are parallel with verse 89 from the Dvitīyakrama, which as we have seen Vaidyapāda does not, in his commentary on the Dvitīyakrama, refer to as the vidyāvrata. The passage on the ritual for bestowing the vidyāvrata in Vaidyapāda’s commentary includes several liturgical verses, ten pādas of which are taken directly from the vidyāvrata section of the Samājottara (125c-127d), but without attribution (Guhyasamājamaṇḍalopāyikā-ṭīkā, D 211b.6-212a.1; P 540a.4-7). 260 words of desire” the couple heighten their passion for one another. The passionate expressions that they are to utter are, however, always clearly soteriologically oriented; the ultimately liberative aim of the practice is never far from mention. The final means for arousing passion in one another is physical. This short section of the Dvitīyakrama reads much like a kāmaśāstric manual describing different postures of embrace (though all of this is still preliminary to the sexual union that constitutes the main part of the third initiation), and sexual acts, primarily those through which the female partner is stimulated, though the text specifies that performing these acts on his partner should arouse passion in the yogin, as well. Then the consort is to again show her “lotus” to her partner, this time telling him that, “The king of natural great bliss Abides in this lotus |114| Because it is realized by means of the channels and winds You should search for the cakra.”50 The yogin is then instructed to use his fingers, in accordance with the oral instructions from his guru, to search for and ascertain several of the subtle channels in his partner’s body; the left, right, and central channels are all mentioned specifically in the text. After this final preliminary step, the partners are to embrace in sexual union. The process of the yogic partners’ union is described in this section of the Dvitīyakrama in terms of its correspondence with the same four branches (caturaṅga) that are often used, including in the practice of Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, to describe the stages of generation stage practice. Here the branches of sevā, upasādhana, sādhana, and mahāsādhana are linked to the stages of sexual union: the vajra touching the lotus is sevā, the vajra entering the lotus is upasādhana, moving repeatedly is sādhana, and the final moment when the bindu of bodhicitta remains for a moment at the tip of the vajra, and is then emitted51 is mahāsādhana. The final stage of mahāsādhana is described in the Dvitīyakrama in a way that clearly references the yogic manipulations of the winds and energies of the subtle body; indeed, Buddhajñānapāda’s is the eariest work with which I am familiar to set forth such techniques, which became an essential part of the perfection stage practices in the later Yoginī tantras. This passage in the Dvitīyakrama describes the yogin’s moving the winds in order to bring about a blazing of wisdom fire that melts the elements and causes the dripping of the bindu of bodhicitta, which is then “offered by unifying the winds.” While it appears not yet to be referred to as such—neither the Dvitīyakrama nor Vaidyapāda’s commentary uses the term—this practice very much resembles what came to be called the yoga of caṇḍālī (gtum mo) in the Yoginī tantras, and later commentators do identify it as such.52 The final verses from the section of the Dvitīyakrama on the prajñājñānābhiṣeka explain that the union of the two partners brings about bliss. This is the first reference in the Dvitīyakrama to the three blisses in Buddhajñānapāda’s system, which were discussed in some detail in the previous chapter and are later in the Dvitīyakrama linked with the three bindu yogas. Here we find the line, already discussed in Chapter Six, describing the culminating experience of the prajñājñānābhiṣeka, in which 50 rang byung bde chen rgyal po ni// padma ‘di la50 rab tu gnas// |114| rtsa dang rlung gis rtogs ‘gyur bas// khyod kyis rtsa yi ‘khor lo tshol// (Dvitīyakrama, 114c-115b). 51 The text seems to associate mahāsādhana both with the holding of the bindu at the tip of the vajra, as well as with its emission, or perhaps just with the precise moment in which the bindu begins to be emitted. See note 54 below, note 261 in my translation of the Dvitīyakrama, and the discussion in Chapter Six with regard to the question of emission in relation to perfection stage practice in Buddhajñānapāda’s system. 52 Tsongkhapa identifies this passage as an instance of caṇḍālī practice in his commentary on the Five Stages of the Ārya School’s Guhyasamāja practice system (Kilty 2013, 324). 261 “between the cessation of bliss, and bliss, an absence is seen and should be stabilized.”53 This absence, which is observed when “the mind,” that is, the drop of bodhicitta is “observed within the jewel,” that is, within the tip of the vajra (the head of the penis), is identified both as wisdom and as the perfection stage.54 The experience is also described with a verse that has been incorporated into the Dvitīyakrama from the Sarvabuddhasamāyoga-tantra stating that “neither passion, dispassion, nor something in between is perceived.” That wisdom should continue to be experienced, the Dvitīyakrama advises, for increasingly long periods of time, presumably as the yogin continues to engage in perfection stage practice following the initiation.55 Finally, as the conclusion of the ritual, the text states that the practitioner should take up and drink the “liquid nectar that abides in the lotus,” that is the bodhicitta that has been emitted there during the act of sexual union. Vaidyapāda’s commentary indicates that it is at this point that the concluding initiatory rituals of the vyākaraṇa, aśvāsa, and anujñā are to be performed, along with the taking of oaths and the performance of a pacifying homa.56 These final verses from the section of the Dvitīyakrama on the third initiation describe the moment identified in an earlier verse on the prajñājñānābhiṣeka as the function or purpose of this initiation—the direct recognition of the dharmakāya, suchness itself, which, as we have just seen, takes place through the process of union with a consort under the guru’s guidance. It thus appears that the elements of the ritual that precede this crucial moment—the guru’s handing over of the consort to the disciple in a marriage-like rite; the lively dialogue between the initiate couple regarding the practice of transgressive samayas; the practices of arousing mutual passion mentally, verbally, and finally physically; and the male partner’s ascertainment of the channels within his partner’s body—serve as preliminaries or supports that allow for this experience to take place. A later verse from the Dvitīyakrama states clearly that sexual union is one of the natural human experiences in which suchness can be glimpsed: The dharmakāya, perfect bliss equal to the sky, Is experienced for just an instant At death, when fainting, falling asleep, When yawning, and during intercourse. Therefore, by training in this, embodied beings purify their minds. |355|57 Vaidyapāda explains that because these experiences are so short, without the instructions of a guru one is unable to recognize the great bliss that arises in these moments. However, he continues, when one does receive instructions from a guru, and thus comes to recognize and 53 gang gang yang dag dga’ byed chags bral dga’ gnyis bar du ‘ben nyid mthong byas brtan par gyis// (Dvitīyakrama, verse124b). 54 As noted above, an earlier verse from this section of the Dvitīyakrama appears to identify both the momentary arrest of the bindu at the tip of the vajra and the moment of its emission (or perhaps just the precise moment in which emission begins?) as mahāsādhana, the culminating moment of the sexual yogic practice in the context of the prajñajñānābhiṣeka. See note 51 above and note 261 in my translation of the Dvitīyakrama. Later discussions of the prajñājñānābhiṣeka would make minute distinctions with respect to the exact location of the bindu in relation to the innate bliss glimpsed at this moment (see, for example, the Sekanirdeśapañjikā (Isaacson and Sferra 2014, 104-5)). As we saw in Chapter Six, the bliss of cessation, which certainly is also to be experienced in the prajñājñānābhiṣeka, is clearly linked at several places in Buddhajñānapāda’s writings with emission. 55 These final two verses on the prajñājñānābhiṣeka (Dvitīyakrama verses 124-125) are also found in at least one later source, Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, and several lines from these verses are cited in a number of later sources including the Caturmudrānvaya, the Abhiṣekanirukti, the Ratnāvalī, the Abhayapaddhati, the Āmnāyamañjarī, the Sekanirdeśapañjikā, the Kriyāsaṅgrahapañjikā, the Yamāritantramaṇḍalopāyikā. For further details of these correspondence see notes 263, 268 and 273 of my translation of the Dvitīyakrama. 56 Sukusuma, D 110a. 57 chos sku rab dga’ mkha mnyam pa// shi dang brgyal dang gnyid log dang// glal dang ‘khrig dus skad cig tsam// myong bar ‘gyur bas rab bsgoms na// lus can rnams ni yid ni sbyong// |355| (Dvitīyakrama, verse 355). 262 habituate oneself to the suchness that is thereby experienced, then the mind becomes purified, leading to realization.58 The ritual for the third initiation, then, makes use of the bliss of sexual union, and the natural experience of suchness that is said to occur in that moment. With the guru’s guidance this can be ascertained by the student, such that he experiences the wisdom of the nonconceptual state of suchness directly. The transgressive elements—sexual and otherwise—of the ritual sequence of the higher initiations seem intended to create a context within which the initiate couple are already brought outside of their ordinary mental frameworks, and thus perhaps made even more ready to experience and ascertain the moment of suchness that occurs during their sexual union. The aspects of the ritual directed towards the cultivation of passion—in particular the kāmaśāstra-like passage of the Dvitīyakrama on the “physical cultivation of passion,” detailing sexual postures and acts to increase the couple’s sexual arousal—are clearly intended to intensify the passion experienced in union, presumably as a means to heighten the sexual bliss and therefore sharpen the wisdom that is ascertained during sexual union. However, as I discussed in the previous chapter, there are also indications that some of the acts described in this section of the text are connected with the stimulation of specific points in the body associated with the channels and bindus of the subtle body and therefore connected to the yogic manipulations of the winds and bindus in the context of the initiation ritual, as well as in post-initiatory perfection stage practice.59 In any case, in the context of the ritual for the prajñājñānābhiṣeka, all of the elements of the ritual appear to be focused upon supporting the disciples’ coming to a direct realization of suchness in reliance upon the conditions of both sexual union and the guru’s instructions. The Dvitīyakrama itself does not elaborate further on the ritual procedures of the higher initiations beyond what I have summarized here. However, as we have seen briefly already in Chapter Three, in the context of examining the structure of the tantric path according to his writings, Buddhajñānapāda’s work does have more to say about the process of the disciple’s coming to a direct experience of suchness in the initiatory context by means of the guru’s oral instructions. Let us take a closer look at this issue now. Initiation in The Early Jñānapāda Tradition: Obtaining Suchness Through the Third Initiation, the Issue of “The Fourth,” and the Seven Yogas Obtaining Suchness As we already saw in Chapter Three, Buddhajñānapāda’s writings repeatedly mention the process of a “transference” of suchness, which is “obtained” or “received” by the disciple from the guru. Several passages indicate that this takes place in the context of initiation, with the third initiation specifically suggested by a verse from the Muktitilaka which states, with respect to the “inner yoga” of suchness that: It is only said to come from elsewhere Though [in fact] it is realized by self-aware bliss, 58 Sukusuma, 132b.6-7; P 159b.8-160a.1. 59 The practices for cultivating passion are mentioned in the Dvitīyakrama only in the context of the third initiation, and not in the context of post-initiatory sexual yogas. It is therefore unclear whether they are meant to be practiced also in that context, as well. The first verse of the section on physically cultivating passion, however, may suggest that they are meant also to be practiced in a post-initiatory context. That verse reads, “Then, with great passion/ Engage in physical practice with her;/ Practicing this play in an isolated place/ You should examine the joys./ |105| de nas rab tu chags ldan pas// de dang lhan cig lus kyi ni// spyod pas dben pa’i gnas su spyad// rol pas dga’ ba brtag par bya// |105| (Dvitīyakrama, verse 105). 263 It is thus explained as “bestowing initiation.”60 Again, it is the “self-aware bliss” mentioned in this verse that specifically suggests the context of the prajñājñānābhiṣeka as the initiation in which suchness is directly realized. And, as we just saw above, the Dvitīyakrama’s presentation of the prajñājñānābhiṣeka describes the third initiation as having precisely the function of bringing the disciple to a direct realization of dharmakāya, or suchness itself. Thus there are multiple indications in Buddhajñānapāda’s work that the third initiation provides the context for the disciple’s “obtaining” of suchness. Certainly, as we just saw in the description of the prajñājñānābhiṣeka, this is related to the blissful experience of sexual union with a partner during that initiation, and the fact that, according to the Dvitīyakrama, the dharmakāya is anyway naturally experienced briefly during sexual union, even outside the context of initiation. But, as Vaidyapāda’s comments cited above indicate, without the instructions of a guru one is unable to recognize as one’s innate nature the great bliss that arises in that moment, and thus the guru’s instructions are crucial to the recognition of suchness that takes place during initiation. Indeed, as I discussed in Chapter Three, Buddhajñanapāda’s writings mention, on several occasions, the fact that suchness is “received” specifically through, or by means of “the guru’s words.” As we saw in the account of Śākyamuni Buddha’s awakening from the Muktitilaka that parallels the process of tantric initiation, Buddhajñānapāda does not even mention the sexual aspects of initiation—though, we may remember, Vaidyapāda does bring in Śākyamuni’s being “blessed by the sahaja ācārya,” that is, uniting with a tantric consort, and Buddhajñānapāda’s account very likely also implies Śākyamuni’s sexual union with a consort. But his account of the awakening narrative directly states only the essential aspect of Śākyamuni’s being shown the suchness of “nondual profundity and luminosity” by his gurus, all of the sugatas, and thus highlights the guru’s role in indicating suchness as the most crucial element in the process of a disciple’s being brought face-to-face with that reality. Buddhajñānapāda’s writings, however, do not state precisely how the communication of suchness in the context of tantric initiation takes place, or the details of its content, and this is certainly a purposeful omission. The transference of suchness is specifically identified as being conferred by means of the oral instructions of a master who holds the lineage,61 and such pithy instructions, passed on “from ear to ear” within the context of a direct and personal relationship between guru and disciple, are not something generally committed to writing. In fact, we see many references in tantric texts to crucial information on tantric practices being conveyed by the guru’s oral instructions, making it quite clear that the textual record alone is an incomplete source for a study of these traditions, and 60 ming tsam gyis ni logs (logs] P, log D) ‘byung yang// rang rig bde bas rtogs byas na// dbyang bskur shes ni bshad pa yin// (Muktitilaka, D 50b.2; P 60b.4) 61 On the importance of the master who conveys suchness holding the lineage see the Muktitilaka: “By constantly revering a lineage guru/ Who knows this reality/ Occasionally, like the [appearance of] the uḍumbara flower,/ Those with merit will know it in an instant.” ‘di yi don shes brgyud pa yi// bla ma dam pa rtag bkur bas// u duṃ bar (u duṃ bar] P, u dumbār D) ltar brgya lam na// bsod nams can gyis skad cig shes// (Muktitilaka, D 48b.6; P57a.7-8); the Muktitilaka-vyākhyāna: the student should please “a sublime guru who knows nondual reality and possesses the great pith instructions of the lineage that has been passed from ear to ear” gnyis su med pa’i don shes shing rna ba nas rna bar brgyud pa’i man ngag chen po dang ldan pa’i bla ma dam pa (Muktitilaka-vyākhyāna, 50a.5-6; P 336a.2-3); the Dvitīyakrama: “He [i.e. the disciple] should please a guru who is genuine and venerable,/ Who possesses the lineage of supreme oral instructions” de yis rje btsun yang dag pa//mchog gi gdams ngag rgyud la ldan// (Dvitīyakrama, verse 45a-b), and “Through relying upon a genuine lineage teacher,/ And one’s own previously gathered accumulation of merit—/ One will come to realize this [reality].” |389| brgyud pa’i bla ma yang dag rab bsten dang// rang gi bsod nams tshogs ni sngon bskyed pas// rtogs par rab tu ‘gyur ba ma gtogs par// |389| (Dvitīyakrama, verse 389). 264 certainly for their practice; the latter, it seems, is precisely the reason for the omission of such instructions from the textual record, or at least one important reason behind it.62 However, despite the fact that the instructions on this crucial point are passed on orally, in this case the textual record does still have more to tell us about this process of the communication of suchness. A passage from the Dvitīyakrama mentions both the receiving of suchness in an initiatory context and the receiving of suchness “through the guru’s words:” Someone who has pleased the guru And received the vase [initiation] and the others Together with the samayas and vows given by him, And thus obtains the suchness |327| That is found through the guru’s words...63 As I noted already in Chapter Three—but I will recapitulate the key points here in order to frame the ensuing discussion—in this passage Buddhajñanapāda singles out “obtain[ing] suchness that is found through the guru’s words” as something particularly important that takes place in the initiatory context. In his comments on this passage, Vaidyapāda makes the distinction between the receiving of initiation and the obtaining of suchness even more strongly, adding the phrase “and then...” between his comments on initiation and vows and his comments on obtaining suchness. With regard to the latter, Vaidyapāda writes, “And then, And thus obtains the suchness/ That is spoken by the guru,64 means that the suchness of the seven yogas, together with the method for accomplishing that, is received.”65 While the verse cited above from the Dvitīyakrama does not explicitly ritually separate out the pointing out of suchness to a disciple from the earlier initiatory sequence, the indication of suchness to a disciple after, and thus ritually separately from, the bestowal of the third initiation is mentioned explicitly in Buddhajñānapāda’s direct disciple Dīpaṃkarabhadra’s Guhyasamājamaṇḍalavidhi, which says: “having bestowed the guhya and prajñā [initiations], suchness should be fully pointed out.”66 Vaidyapāda’s Yogasapta, a short treatise on tantric initiation, gives even further details about this pointing out of suchness after the third initiation. The Seven Yogas and “The Fourth” 62 The fact that in this case the instructions being conveyed have to do with suchness, something which is frequently described in Buddhist texts as ineffable, also suggests that in addition to their being kept out of the textual record in order to keep them secret so that they cannot be practiced without having been received directly from a guru (and even to prevent any attempt to do so), it may not even be considered possible to convey these particular instructions outside of the context of a personal interaction with a guru. Vaidyapāda’s comments on the topic of “the fourth,” which I address below, indeed suggest something along these lines. 63 gang zhig bla ma mnyes byas nas// des gnang dam sdom bcas ba ru// bum pa la sogs rab thob ste// bla ma’i zhal las rnyed pa yi// |327| de bzhin nyid ni rab thob cing// (Dvitīyakrama, verse 327a-328a). 64 Bum pa la sogs pa’i dbang gong du gsungs pa ltar rab tu thob par byas te/ de nas/ bla ma’i zhal nas gsungs pa yi// de bzhin nyid ni rab thob cing// zhes pa ni sbyor ba bdun gyis de kho na nyid sgrub pa’i thabs dang bcas pa rnyed pa… (Sukusuma, D 130a.3; P 156b.4-5). Vaidyapāda’s commentary preserves a slight variant on one line from the Dvitīyakrama. The pādas as found in the Sukusuma read: bla ma’i zhal nas gsungs pa yi// de bzhin nyid ni rab thob cing// (Sukusuma, D 130a. 3; P 156b.4) as opposed to bla ma’i zhal nas rnyed pa yi// de bzhin nyid ni rab thob cing// in the Dvitīyakrama. However, in his comments on these two lines Vaidyapāda mentions the word “received” (rnyed pa) which is absent in the verse as translated in his commentary but present in the verse as translated in the Dvitīyakrama, so I suspect the variant arose in the context of translating the Sukusuma into Tibetan, rather than in the citation of the verse in Vaidyapāda’s commentary itself. 65 de nas/ bla ma’i zhal nas gsungs pa yi/ de bzhin nyid ni rab thob cing/ zhes pa ni sbyor ba bdun gyi (gyi] P, gyis D) de kho na nyid bsgrub pa’i thabs dang bcas pa rnyed pa [/] (Sukusuma, D 130a. 3; P 156b.4). 66 The full verse reads, maṇḍalaṃ devatātattvamācāryaparikarma ca | saṃkathya guhyaprajñābhyāṃ siktvā tattvaṃ samuddiśet | (Guhyasamājamaṇḍalavidhi, v. 367.) 265 As we just saw, Vaidyapāda explains that the “suchness that is found through the guru’s words” mentioned in the Dvitīyakrama refers specifically to the suchness of the seven yogas and the method for their accomplishment. Vaidyapāda’s comment, like Dīpaṃkarabhadra’s statement about suchness being fully pointed out after the bestowal of the second and third initiations, appears to refer to a set of oral instructions that are given to the disciple following, and with reference to the experience encountered within, the third initiation. Vaidyapāda’s Yogasapta, which incorporates a number of verses directly from the Dvitīyakrama, describes these instructions in much more detail. In that text, Vaidyapāda identifies the seven yogas—perfect example-less bliss (dpe med bde rdzogs), nonduality (gnyis su med pa), great bliss (bde ba chen po), lacking nature (rang bzhin med pa), unfolding compassion (thugs rjes rgyas pa), unbroken continuity (rgyun mi chad pa), and non-cessation (‘gog pa med pa)—both with the state of perfect awakening itself, and with what he refers to simply as “the fourth.” “The fourth” (caturtha) is a term found in the Samājottara (which text, we will recall, Vaidyapāda certainly knew—he composed a commentary on it—although it seems that Buddhajñānapāda did not) in a rather perplexing passage on tantric initiation where it is explicitly stated that initiation is threefold, but after listing the three initiations the work then makes a rather cryptic reference to “the fourth.” This passage, despite the opacity of its final two pādas, is the much-cited locus classicus of the three (or four!) initiations, and the source, it seems, of a significant amount of debate on the topic of what might constitute “the fourth.” As I noted already in Chapter Three debates on the topic of a “fourth initiation,” which took place over a number of centuries, seem to have centered on—and indeed probably sprung from—the meaning of precisely this reference to “the fourth” (caturtha) in the Samājottara.67 Isaacson and Sakurai have argued, partly on the basis of the absence of a fourth initiation in early Jñānapāda School and early Ārya School works, that a separate fourth initiation was not likely intended in the passage on initiation in Samājottara, but arose later out of the debate on the what was indeed meant by “the fourth” in that passage.68 Vaidyapāda’s Yogasapta, which is one of the earliest works to address the topic of “the fourth”—indeed, to my knowledge it is the earliest extant work that does so—appears to confirm Isaacson’s and Sakurai’s suspicions with respect to the way that “the fourth” was understood in the earliest period of the circulation of the Samājottara. That is, in this work Vaidyapāda certainly holds “the fourth” as something separate from the three initiations, and indeed something absolutely essential—he equates it there with the suchness of the seven yogas that is bestowed upon the disciple by means of an oral instruction after the prajñājñānābhiṣeka—but he does not appear in the Yogasapta to hold it to be a separate initiation.69 The passage from the Samājottara that served as the source of so much debate on the topic of “the fourth” reads: 67 Isaacson 2010b, 268-271. 68 ibid., 269. 69As I mentioned already in Chapter Three, Vaidyapāda’s Yogasapta-nāma-caturabhiṣekaprakaraṇa, The Seven Yogas: An Explanation of the Four Initiations (Sbyor ba bdun pa zhes bya ba dbang bzhi’i rab tu byed pa), mentions the “four initiations” in the title, but throughout the work itself the term “fourth initiation” is never used; the first three initiations are clearly called initiation but “the fourth” is only ever referred to as simply “the fourth” (bzhi pa). Given this fact, along with the unreliability of the Sanskrit titles in the Tibetan canon, some of which (like Dvikrama for the *Dvitīyakrama!) appear to be incorrect Sanskrit reconstructions made by the redactors of the Tibetan canon, we may be inclined to raise doubts about the “fourth initiation,” mentioned in the title of the Yogasapta. However, the Tibetan translation of the title likewise makes reference to “four initiations” (dbang bzhi). Moreover, in his Guhyasamājamāṇḍalopāyikā-ṭīkā Vaidyapāda does indeed refer to a “precious fourth initiation” (dbang bskur ba rin po che bzhi pa (pa] P, D om.) that consists of the guru’s oral instructions with respect to union (Guhyasamājamaṇḍalopāyikā-ṭīkā, D 211b.3-4; P 539b.6-7). The issue of whether Vaidyapāda considered “the fourth” an initiation or not therefore appears to be a slippery one. I address this point briefly below. 266 Initiation in this tantra Is understood as three-fold: The kalaśābhiṣeka is the first, The second is the guhyābhiṣeka The prajñājñānā is the third And the fourth is again like that. |113|70 Vaidyapāda’s Yogasapta, provides a short synopsis of the different tantric initiations in which he incorporates this passage from the Samājottara, but adds a crucial final line explaining that “the fourth” is to be known from the words of the guru.71 He later elaborates on the meaning of “the fourth,” which he identifies both with perfect awakening and with the seven yogas, which are the main topic of his work. In his discussion of “the fourth” in the Yogasapta, not only does Vaidyapāda not ever refer to “the fourth” as an initiation (he refers to it only as “the fourth”), he even specifically states that in the context of the fourth “there is no initiation.” However, in another one of his works, the Guhyasamājamaṇḍalopāyikā-ṭīkā, Vaidyapāda does refer to the guru’s oral instructions on union as a “precious fourth initiation,”72 and even in the Yogasapta he discusses “the fourth” very much in relation to his discussion of the three initiations, suggesting that he views it in similar terms. It is difficult to understand Vaidyapāda’s presentation of “the fourth” in the Yogasapta without reference to the manner in which he presents the first three initiations, since, as we shall see below, he describes the fourth in a way that is essentially an inversion of the presentation of the previous three initiations. Thus, in order to provide the crucial context for this important early passage on “the fourth,” and also because the earlier part of the text gives an excellent overview of the first three initiations and their functions as Vaidyapāda understands them, I will give a somewhat lengthy citation consisting of the first few pages of Vaidyapāda’s Yogasapta: I pay homage to the omniscient one! To he who is incomparable Nondual, great bliss, natureless, With vast compassion, free from obstruction, and unceasing— I pay homage to the buddha.73|1| To that which pervades all that moves and does not, And remains in the center of the lotus at the heart, The indestructible bindu, abiding at the center— I pay homage to the dharma that cannot be overcome. |2| To the one who teaches this, Whose mind is placed in suchness, The great saṅgha who liberates [beings] from the treacherous path— I bow at the feet of my sublime guru. |3| The preliminary stages of the path Are the maṇḍala, samayas, initiation, and the rest. I will explain [here] a little bit about the different initiations 70 abhiṣekaṃ tridhā bhedam asmi tantre prakalpitam/ kalaśābhiṣekaṃ prathamaṃ dvitīyaṃ guhyābhiṣekataḥ/ prajñājñānaṃ tṛtīyaṃ tu caturtham tat punas tathā/ (Samājottara, verse 113; Matsunaga 1978, 121). 71 See the passage, which I cite at length, below. 72 de nas des rgyas btab nas dbang bskur ba rin po che bzhi pa (pa] P, D om.) bla ma’i man ngag gis mkha’ gnyis kyi sbyor ba’i bshad pa sbyin par bya ste/ (Guhyasamājamaṇḍalopāyikā-ṭīkā, D 211b.3-4; P 539b.6-7). 73 This first verse references each of the seven yogas—perfect example-less bliss (dpe med bde rdzogs), nonduality (gnyis su med pa), great bliss (bde ba chen po), lacking nature (rang bzhin med pa), unfolding compassion (thugs rjes rgyas pa), unbroken continuity (rgyun mi chad pa), and non-cessation (‘gog pa med pa)—identifying all seven as characteristics of the buddha. 267 [Conveyed] by means of the vase (kalaśa) and the rest. |4| [To say that] by means of the maṇḍala ritual and so forth Great liberation [is accomplished] is deceptive. I will correctly explain the reality of the fourth,74 Since without suchness [great liberation] is not accomplished. |5| The omniscient one Taught the Mahāyoga tantras And explained in this tantra That initiation is three-fold. |6| The kalaśābhiṣeka is the first; The second is the guhyābhiṣeka; The third is the prajñājñānābhiṣeka; And in that way, likewise, the fourth, as well, Should be known from the words of the guru.75 |7| With respect to the initiations I will first explain a little bit about their distinctions. [The Kalaśābhiṣeka] It is asserted that the [first initiation] is given By the great causal ācārya76 |8| One’s lands, palaces, home, cattle, and so forth Are suitable [offerings] and should be offered to him. This is given in order to render [oneself] an appropriate recipient For the ordinary and other vows. |9| That [initiation] is obtained within the Drawn[-maṇḍala], and the body-maṇḍala The rituals for the vidyā [initiations] and their cause,77 the ācārya initiation, Are explained in other78 [texts]. |10| Its essence is The nature of vajra body.79 The [negative actions of body] like killing and the rest, past and present, 74 bzhi pa’i don 75 This is a direct parallel of the Samājottara’s verse on initiation, but here Vaidyapāda crucially adds an extra line explaining that “the fourth” is known from the words of the guru. 76 The “causal ācārya” is described in both Vaidyapāda’s Sukusuma and Muktitilaka-vyākhyāna by means of a citation that I have been unable to trace, but which Vaidyapāda identifies as a text called The Precious Garland (rin chen phreng ba (phreng ba] P, phrod pa D). That citation explains that the causal ācārya is the guru who gives vows and commitments and who purifies the disciple’s mind by means of the stages of initiation, beginning with the water initiation (Sukusuma, D 88a; Muktitilaka-vyākhyāna, D 47b). See Chapter Three, note 149 for more details on the three ācāryas. 77 This statement is perplexing, but the text as it reads now does seem to say this. There are a number of textual problems in the Yogasapta. A more comprehensive study of the work would surely improve some of the readings here. 78 gzhan du] P, bzhin du D. This and other instances of variation in the Tibetan translations sometimes make it appear that scribes for translators were sometimes taking dictation, rather than copying from a written source. 79 de yi rang gi ngo bo ni// sku yi rdo rje’i rang bzhin no// I am tempted to emend these, and the parallel lines below, to read de yis rang gi ngo bo ni// sku yi rdo rje’i rang bzhin no// and to read them as “By means of this [initiation] one’s own essence// [Is recognized] as having the nature of vajra body.” This is certainly the way that these first three initiations are understood to function in the later Tibetan tradition. However, none of our sources reads yis in this or any of the parallel lines below (for vajra speech and mind) so I have, for now, held back from emending the text in this way, despite the fact that the translation of the Yogasapta is generally somewhat problematic and does require emendation in a number of other instances. 268 Are purified and their negativities will not transpire.80 |11| Because of superior methods This is explained as [placing the initiate at the level of?] the eighth bhūmi.81 The [first?] three yogas and their branches82 Are to be explained here. |12| One comes to realize that appearance and existence Are free from objects, like form and the rest. Because [it makes one into] a suitable vessel it is [called] the “vase.”83 Because one is cleansed, sprinkled, and purified it is [called] “initiation.”84 |13| This is explained definitively In the scriptures that have been passed down through the lineage.85 Having explained the distinctive [features] of the first one I will explain the second [initiation], as well. |14| [The Guhyābhiṣeka] The initiation is given by the causal and conditional ācāryas86 80 I am unsure of the meaning of this line, and the parallel lines with regard to the guhya- and prajñājñānābhiṣekas below. Does it mean that the negative results of past actions will not come about or that the negativity associated with such acts will not transpire because the acts themselves have been purified? 81 The corresponding verse with respect to the second initiation says, “One abides on the ninth and tenth bhūmis.” (dgu dang bcu la yang dag gnas) (Yogasapta, verse 17d), and for the third that “One abides in the vajra-like [samādhi],” (rdo rje lta bur yang dag gnas//) Yogasapta, verse 25b), referring to the final meditation at the end of the tenth bodhisattva bhūmi prior to attaining awaking in the Mahāyāna. For “the fourth” the text says, “[Since] it is the state of a vajra-holder, there is no bhūmi” (rdo rje ‘dzin gnas sa med de//) (Yogasapta, verse 30d), but the attainment is nonetheless associated with the state of a vajra-holder. Since in the Buddhist tradition initiation is not taught to have a directly liberative function, these lines may suggest that by means of the practices that these specific initiations permit an initiate to practice, he or she can swiftly progress to the levels that are mentioned in Yogasapta with respect to each of these initiations. The Dvitīyakrama includes several references to a practitioner arriving at different bodhisattva bhūmis by means of the practices outlined in that text. A practitioner of the third bindu yoga, also known as vajrajapa, is said, through that practice, to “share the fortune of the lords of the tenth bhūmi,” (sa bcu pa’i// dbang phyug rnams dang skal mnyam ste// (Dvitīyakrama, verse 233, c-d)), and, as we saw in Chapter Three, another passage from Dvitīyakrama associates the progressive stages of sexual union in a yogic context with each of the ten bhūmis, and identifies the result of tantric practice with the thirteenth bhūmi (Dvitīyakrama, verses 298-313). Tomabechi (2006, 146-47; 147n190) notes that a statement in the Pañcakrama to the effect that “through the practice of a beginner one attains the eighth bhūmi and the one who has a vision of the three lights is established on the tenth bhūmi” (adhikarmikayogena cāṣṭamīṃ bhūmim āpnuyāt/ ālokatrayadarśī ca daśabhūmyāṃ pratiṣṭhitaḥ//) may also be understood in terms of the initiatory sequence, in the sense that the “practice of a beginner” may refer to the generation stage practices, which one is permitted to practice by means of the kalaṣābhiṣeka, which itself culminates with the ācāryābhiṣeka, also known as the “irreversible abhiṣeka (avaivartikābhiṣeka). (As we saw above, Vaidyapāda even uses the two terms together: “irreversible ācāryābhiṣeka” (phyir mi ldog pa’i slob dpon kyi dbang)). Tomabechi points out that in traditional Mahāyāna theory according to the Daśabhūmikasūtra, the eighth bodhisattva bhūmi is the stage at which the bodhisattva reaches a stage on the path that is, precisely, “irreversible.” Thus, the connection between the “irreversible ācāryābhiṣeka” and the eighth bodhisattva bhūmi may be made in relation to this particular feature of the eighth bhūmi. Note that the passage in the Pañcakrama with respect to which Tomabechi makes these observations is, however, stating that it is through the “beginner’s” practices—i.e. those associated with the kalaśābhiṣeka—that the yogin attains the eighth bhūmi, not by means of the initiation itself. As I have suggested, it may be best to read the references to the various bhūmis in Vaidyapāda’s Yogasapta in this way, as well. 82 The grammar of this line is unclear. It is difficult to make sense of function of the particle ru at the end of the line. 83 ~C.f. Dvitīyakrama, verse 395d. 84 ~C.f. Dvitīyakrama, verse 395a. 85 rgyud] sugg. em. based on parallel verses below, rgyu D P. 86 On the causal ācārya see note 76. The conditional ācārya is identified in the same source mentioned in note 76, The Precious Garland, as the “great goddess” with whom one engages in play and who purifies the field of one’s mind by means of the “sixteenth part” (Sukusuma, D 88a; Muktitilaka-vyākhyāna, D 47b). In this case it seems that 269 And is thus genuinely obtained. To him87 the appropriate [offerings] [70b] Of gold, silver, and the rest are to be offered. |15| It is given in order to make one a suitable vessel For the common vows regarding [not] killing and the rest.88 This [initiation] is obtained In the maṇḍala of superior bodhicitta. |16| The particulars of what to imagine During the ritual for conferring this [initiation] are explained elsewhere.89 Its essence is The nature of vajra speech.90 |17| The [negative deeds of speech], lying and so forth, past and present, Are purified91 and their negativities92 will not arise. Because its methods are powerful93 One abides on the ninth and tenth [bhūmis].94 |18| By means of the essence of vajrajāpa The māyopama samādhi is to be taught. One realizes appearance and existence to be free of The eyes and the rest of the inner phenomena. |19| Since it is not to be [widely] proclaimed it is [called] “secret” Because one is cleansed, sprinkled, and purified, it is [called] “initiation.”95 This, as well, is explained in the Scriptures that have been passed down through the lineage. |20| [The Prajñājñānābhiṣeka] Having taught about the second one I will now explain about the third. It is obtained by means of The causal, conditional, and sahaja ācāryas—all three.96 |21| the term refers to the consort in her role as the guru’s partner in the guhyābhiṣeka. See Chapter Three, note 149 for more details on the three ācāryas. 87 Although the text mentions both ācāryas as the means by which the initiation is given, the offering (gurudakṣina) is made presumably just to the causal ācārya, the guru who is bestowing the initiation. However, in Kṣitigarbha’s Daśatattva, the passage on the guhyābhiṣeka notes that after offering the consort to the guru for the purposes of this initiation, the disciple “worships the guru and the consort in manifold ways of worship” (nānāpūjayā saprajñaṃ gurum pūjayitvā…) (Schwind 2012, 283; 208), so it is possible that some kind of offering is made to the “conditional ācārya,” as well, in this context. 88 This seems to be a repetition of the verses for the kalaśābhiṣeka above. I imagine that this is an error in the text, and that something else is intended with regard to the guhyābhiṣeka. 89 gzhan du] P, bzhin du D. 90 See note 79. 91 dag] sugg. em. ngag D P 92 nges pa] P, nyen pa D 93 rab tu phye ba, *prabhāvita? (see Negi). 94 See note 81. 95 ~C.f. Dvitīyakrama, verse 395a. 96 See notes 76 and 87 on the causal and conditional ācāryas. The sahaja ācārya is identified in the same source mentioned in notes 76 and 86, The Precious Garland, as the one from whom one receives that (bindu?) and by means of whom and through whose blessing one realizes innate joy (Sukusuma, D 88a; Muktitilaka-vyākhyāna, D 47b). In this case it seems that this term refers to the consort in her role as the disciple’s partner in the prajñājñānābhiṣeka. In Buddhajñānapāda’s system it seems that the conditional ācārya and the sahaja ācārya are, in fact, the same individual—that is the same consort is the guru’s partner in the guhyābhiṣeka and the disciple’s 270 Rubies, amber97 and so forth, as appropriate [for this initiation] Are offered to him.98 Drawing the maṇḍala and so forth are stopped And the supreme vow is given. |22| This [initiation] is obtained Within the maṇḍala of the “completely pure bhaga.” The stages of the ritual for the prajñā initiation and the jñāna initiation Are explained elsewhere.99 |23| Its essence is The nature of vajra mind.100 The [mental negative acts] of covetousness and the rest, past and present, Are purified and their negativity will not arise. |24| Because the methods are unique One abides in the vajra-like [samādhi].101 In order to accomplish the supreme siddhi The vrata102 ritual is taught here. |25| One realizes that all phenomena are free Of the essence of subject and object. Because one realizes the not [yet] realized jñāna Through103 the actions of the prajñā it is [called the prajñājñāna] initiation. |26| This, as well, is explained in the scriptures104 That have been passed down in the lineage. |27| [The Fourth] Now, I will present that which is to be correctly explained, The fourth, perfect awakening, That which is obtained from the mouth of the guru, The suchness of the seven yogas. |28| Since it abides naturally there is no ācārya. Since it is priceless there is no offering. Because it is beyond binding, there is no vow.105 Since it is [bestowed within] the maṇḍala of suchness, there is no maṇḍala. |29| [71a] Since it is obtained by power, there is no initiation. partner in the prajñājñānābhiṣeka—so this is perhaps why it is stated that the third initiation is received on the basis of all three ācāryas. In other later initiation manuals it is mentioned that the disciple’s partner for the third initiaton can be the same or a different consort as the one offered to the guru in the context of the guhyābhiṣeka (Isaacson 2012b, 262). See Chapter Three, note 149 for more details on the three ācāryas. 97 pursha. This term occurs in Buddhajñānapāda’s and Vaidyapāda’s writings as an offering for the third initiation, but it is unclear what it refers to. Khenpchen Chodrak Tenphel suggested amber (personal communication, Febuary 2016). 98 Although the text mentions all three ācāryas as the means by which the initiation is given, the offering is made presumably just to the causal ācārya, the guru who bestows the initiation. See also note 87. 99 gzhan du] P, bzhin tu D 100 See note 79. 101 This is the final meditative state that a practitioner enters after the tenth bodhisattva bhūmi and before the final moment of awakening according to the Mahāyāna system. See also note 81. 102 Presumably this refers to the vidyāvrata, which often accompanies the third initiation, as, for example, according to Vaidyapāda’s Guhasamājamaṇḍalopāyikā-ṭīkā. 103 kyis] P, kyi D 104 lung] D, lus P 105 sdom pa bas ni sdom pa med// 271 Since it is universal, it has no essence of its own. [Since] it is the antidote to all things, there is no antidote. [Since] it is the state of a vajra-holder, there is no bhūmi.106 |30| Since there is no nature, there is no ritual, And no view, and no etymology— It is correctly explained by the guru to be like this. The wisdom (jñāna) that arises from the prajñā Should be understood as the fourth. |31|107 The first part of this passage from the Yogasapta provides a clear and helpful overview of the first three initiations, and this passage (along with the entirety of this text), certainly deserves 106 See note 81. 107 thams cad mkhyen pa la phyag tshal lo// gang zhig dpe med tshul ldan pa// gnyis med bde chen rang bzhin med// thugs rje rgyas shing bar chad med// ‘gog med sangs rgyas de la ‘dud// |1| rgyu dang mi rgyu kun khyab pa// gang gi snying gi pad dbus su// mi shigs [D 70a] thig le dbus gnas pa// gzhom med chos la phyag ‘tshal lo// |2| de ‘dra rab tu ston phyed pa// sems ni de nyid la bzhag pa// g.yang sa’i lam sgrol dge ‘dun che// bla ma dam pa’i zhabs la ‘dud// |3| lam gi rim pa sngon ‘gro ba// dkyil ‘khor dam tshig dbang sogs te// bum pa la sogs rim pa yis// dbang gi dbye ba cung zad brjod// |4| dkyil ‘khor la sogs cho ga (cho ga] P, rim pa D) yis// thar chen yang dag slu ba ste// de nyid med pas mi ‘grub pas// bzhi pa’i don ni yang [P83b] dag bshad// |5| thams cad mkhyen pas ji skad du// rnal ‘byor chen po’i rgyud bstan la// dbang ni rnam pa gsum dag tu// rgyud ‘di las ni rab tub shad// |6| bum pa’i dbang ni dang po ste// gnyis pa la ni gsang ba’i dbang// gsum pa shes rab ye shes dbang// de ltar de bzhin bzhi pa yang// bla ma’i bka’ las shes par bya// | 7| dbang ni dang po dbye pa yang// cung zad kyang ni bzhad bya ste// rgyu yi slob dpon chen po yis// de ni yang dag bskur bar ‘dod// |8| yul mkhar khyim dang glang po sogs// rjes mthun de la dbul bar bya// de yi thun mong gzhan pa yi// sdom pa snod du rung phyir sbyin// |9| ri mo dang ni lus kyi ni// dkyil ‘khor du ni de nyid thob// rig dang ‘di rgyu slob dpon dbang// cho ga’i rim pa gzhan (gzhan] P, bzhin D) du bshad// |10| de yi rang gi ngo bo ni// sku yi rdo rje’i rang bzhin no// srog gcod la sogs snga phyi yi// dag dang nyes par mi ‘gyur ro// |11| thabs kyis khyad par ‘phags pa’i phyir// sa brgyad pa ru yang dag bshad// sbyor gsum yan lag bcas pa ru// de la yang dag bstan par bya// |12| gzugs la sogs pa’i yul rnams kyi// snang srid dben par rtogs pa’o// snod du rung phyir bum pa ste// blugs gtor dag par byed pas dbang// |13| ‘di ni yang dag rgyud (rgyud] sugg. em. based on parallel verses below; rgyu D P) rim pa’i// /lung gis nges par bshad pa’o// dang po’i dbye ba bstan nas ni// gnyis pa yang ni bshad bya ste// |14| rgyu dang rkyen gyi slob dpon gyis// dbang bskur de ni yang dag thob// gser dang dngul la sogs pa [D 70b] yi// rjes mthun de la dbul bar bya// |15| srog gcod la sogs thun mong pa’i// sdom pa snod du rung phyir sbyin// lhag pa byang sems zhes bya ba’i// dkyil ‘khor du ni de nyid thob// |16| blo yi bye brag gis bskur ba’i// cho ga’i rim pa gzhan (gzhan] P, bzhin D) du bshad// de yi [P 84a] rang gi ngo bo ni// gsung gi rdo rje’i rang bzhin no// |17| brdzun (brdzun] D, rdzun P) la sogs pa snga phyi yi// dag (sugg. em., ngag D P) dang nyes (nyes] P, nyan D) par mi ‘gyur ro// thabs kyis rab tu phye ba’i phyir// dgu dang bcu la yang dag gnas// |18| rdo rje bzlas pa’i ngo bo yis// sgyu ‘dra’i ting ‘dzin bstan par bya// mig sogs nang gi chos rnams kyi// snang srid dben par rtogs pa’o// |19| bsgrags min phyir na gsang ba ste// blugs gtor dag pas byed pas dbang// ‘di yang brgyud pa’i rim pa yi// lung gis yang dag bshad pa’o// |20| gnyis pa’i dbye ba bstan nas su// gsum pa yang ni bstan par bya// rgyu dang rkyen dang lhan cig byed// slob dpon gsum gyis thob pa’o// |21| padma rā ga pur sha sogs// rjes mthun de la dbul bar bya// dkyil ‘khor bri ba la sogs ‘gog/ sdom pa mchog ni sbyin pa’o// |22| bha ga rnam dag ces (ces] D, zhes P) bya yi// dkyil ‘khor du ni de nyid thob// shes rab dbang dang ye shes dbang// cho ga’i rim pa gzhan du bshad// |23| de yi rang gin go bo ni// thugs kyi rdo rje’i rang bzhin no// brnab sems la sogs nga phyi yi// dag (dag] D, ngag P) dang nyes (nyes] sugg. em., nye D P) bar mi ‘gyur ro// |24| thabs kyi khyad par phye ba’i phyir// rdo rje lta bur yang dag gnas// mchog gi dngos grub sgrub (sgrub] D, grub P) pa’i phyir// brtul zhugs cho ga de la bstan// |25| gzung dang ‘dzin pa’i ngo bo yis// chos kun dben (dben] D, dbyen P) par rtogs pa’o// shes rab las kyis [kyis] P, kyi D) ye shes te// ma rtogs rtogs phyir dbang yin no// |26| ‘di yang brgyud pa’i rim pa yi// lung (lung] D, lus P) gis yang dag bshad pa’o// |27| da ni yang dag bshad bya ba// bzhi pa mngon par byang chub pa// bla ma’i zhal nas nges thob pa// sbyor ba bdun gyi de nyid bshad// |28| rang bzhin gnas phyir slob dpon med// rin thang med phyir yon med de// [P 84b] sdom pa bas ni sdom pa med// de nyid dkyil [D71a] ‘khor dkyil ‘khor med// |29| stobs kyis thob phyir dbang med de// spyi pas rang gi ngo bo med// chos kun gnyen po gnyen po med// rdo rje ‘dzin gnas sa med de// |30| rang bzhin med phyir cho ga dang// lta ba nges tshig med ba ru// bla ma yis ni yang dag bshad// shes rab las skyes ye shes ni// bzhi pa yin par shes par bya// |31| (Yogasapta, D 69a.7-71a.2 ; P 83a.5-84b.3). 272 further consideration, but now is not the occasion to get into those details.108 I would like to focus here on Vaidyapāda’s presentation of “the fourth,” which, as is quite clear from simply reading the text itself, is made very much in contradistinction to his presentation of the first three initiations. Using the sort of apophatic language preferred in many Buddhist traditions—like the literature of the Prajñapāramitā, Madhyāmaka, and also the Great Perfection and Mahāmudrā traditions—as a way of describing suchness, awakening, or ultimate reality, Vaidyapāda sets “the fourth” apart from the three initiations, stating that not a single one of the features that characterize the three initiations pertains to “the fourth,” which he identifies as “perfect awakening” itself. We can see in his presentation precisely why, for him, “the fourth” is not, and indeed in some sense cannot be, “an initiation” (though we should remember that Vaidyapāda does call it the “precious fourth initiation” in his Guhyasamājamaṇḍalopāyikā-ṭīkā); it is identical with the result of awakening, and therefore goes beyond the boundaries of such ordinary characterizations. However, each of the negations that Vaidyapāda applies here is also complemented by positive language describing “the fourth” as more exalted than the earlier initiations—it “abides naturally,” is “universal,” “priceless,” “the state of a vajra-holder.” Vaidyapāda’s use of language that is parallel to, but contrasts with, the language that he uses to describe the three initiations indicates that, while he is not—in the Yogasapta, at least— willing to call it an “initiation,” Vaidyapāda does regard “the fourth” as something that is more or less parallel with initiation. Nonetheless, partly for the doctrinal reasons described above, and perhaps partly because he did not have a scriptural basis to do so—precisely because, it seems, the Yogasapta captures the fourth initiation in the process of emerging—Vaidyapāda says that “there is no initiation” when it comes to “the fourth.” In stating this in the way that he does, though, within the framework of its (non)parallels with the three initaitions, Vaidyapāda is essentially acknowledging “the fourth” as something that is on par with initiation, and, as we saw above, he also uses language that shows “the fourth” to be superior, even, to the three initiations. Regarding his lack of a scriptural source for the ritual that he describes in this work—and, for that matter, for referring to “the fourth” as an initiation—Vaidyapāda appears to be aware that he is recording in the Yogasapta instructions that are not part of a previous scriptural tradition, but were previously passed down only in an oral lineage. He explicitly states, with respect to each of the first three initiations, that their rituals have been “explained in the scriptures that have been passed down through the lineage.”109 But with regard to “the fourth,” he instead writes—at a place in the text that is precisely parallel to his statements about the scriptural souces for the rituals of the three initiations—simply that the fourth “is correctly explained by the guru to be like this.”110 And, indeed, there were scriptural sources for the rituals of the first three initiations at the time that Vaidyapāda was writing: Chatper Sixteen of the Guhyasamāja-tantra describes the kalaśābhiṣeka and the Samājottara—the circulation of which, I will argue in the next chapter, follows Buddhajñānapāda’s writings but precedes Vaidyapāda’s—describes the rituals for the guhya- and prajñājñānābhiṣekas. But despite the fact that the Samājottara mentions “the fourth,” there is no explanation in that tantra of what “the fourth” might mean (or even be!), nor any indication of what its ritual might entail. Vaidyapāda’s presentation of “the fourth,” in the Yogasapta, however, culminates with a very clear and specific statement about what “the fourth” is: “The wisdom (jñāna) that arises 108 I tried to draw brief attention to just a few points of interest in the notes to the translation here. I have nearly completed a draft translation and edition of the complete Yogasapta, which I hope to publish, along with a short study of the text, once it is completed. 109 See Yogasapta verses 14 (with respect to the kalaśābhiṣeka), 20 (with respect to the guhyābhiṣeka), and 27 (with respect to the prajñājñānābhiṣeka). 110 Yogasapta, verse 31. 273 from the prajñā should be understood as the fourth.” This final statement employs a sort of word play, in which the term prajñā can be—and I would suggest that it was indeed meant to be— taken in two ways. The term prajñā generally has the meaning of insight, and the statement taken in this sense thus identifies the fourth with wisdom—the wisdom, we will remember, that is identical with awakening itself—that arises based on insight. Presumably this wisdom, the direct knowing of suchness itself, is brought about through the essential oral instructions of the guru that are given in this initiatory context, immediately following the third initiation. However, in a tantric context the term prajñā is very commonly used to refer to the tantric consort herself. Thus, just like in the earlier statement from the Yogasapta about the prajñājñānābhiṣeka, where it was stated that Because one realizes the not [yet] realized jñāna Through111 the actions of the prajñā it is [called the prajñājñāna] initiation |26|112 here in describing “the fourth” as “the wisdom (jñāna) that arises from the prajñā” Vaidyapāda is directly linking “the fourth” with the prajñājñānābhiṣeka. Taken in this second sense, this description of the fourth as “the wisdom (jñāna) that arises from the prajñā” identifies it as wisdom that arises in reliance on uniting with the tantric consort (the prajñā), which takes place during the third initiation. “The fourth,” then, is not only intimately connected with the prajñājñānābhiṣeka, it is in some sense the very outcome of that initiation—suchness, or awakening itself, which is realized in reliance upon union with the consort. But, in this same passage Vaidyapāda also identifies the fourth as “that which is obtained from the mouth of the guru” and as the seven yogas. So, at the same time that “the fourth” is the outcome of the prajñājñānābhiṣeka, a wisdom that arises in reliance upon the tantric consort, it is also something obtained from the guru’s words, and something that is, or consists of, the seven yogas. The next verses of the Yogasapta (continuing on in the text from the end of verse 31, where we left off above) elaborate on the relationship between the fourth, the prajñājñānābhiṣeka, and the guru’s oral instructions, incorporating a number of verses from the section of the Dvitīyakrama on the third initiation (I italicize the verses incorporated from the Dvitīyakrama, just to make it easier to see the correspondences) : Having joined the two spaces113 Suchness is examined [through] the oral instructions. When realized through effort In reliance upon a well-practiced mudrā114 |32| Desire, freedom from desire, and something in between— 115 The characteristics of the seven yogas—[are realized] |33| 111 kyis] P, kyi D 112 shes rab las kyis [kyis] P, kyi D) ye shes te// ma rtogs rtogs phyir dbang yin no// |26| (Yogasapta, verse 26cd). 113 mkha’ gnyis. Presumably this refers to the union of the two sexual organs. The more commonly used term is, precisely, the union of the “two organs” (dbang po gnyis). Vaidyapāda also uses the term mkha’ gnyis in this same way in his description of the “precious fourth initiation” in the Guhyasamājamaṇḍalopāyikā-ṭīkā, (D 211b.3-4; P 539b.6-7). 114 Mudrā is understood here as (yet another!) term referring to the consort. The term legs par goms pa (svabhyasta), which is occurs also in verse 37 of the Yogasapta, is also used in Dīpaṃkarabhadra’s maṇḍalavidhi to describe the yogiṇī in the second initiation, who is there consecrated as a mudrā (Guhyasamājamaṇḍalavidhi, verse 362). These statements suggest that the success of the initiation is understood to be in some measure reliant on the consort’s having practice experience. 115 As I have noted earlier, the phrase “desire, freedom from desire, and something in between” is excerpted from a line of the Sarvabuddhasamāyoga-tantra that Buddhajñānapādada incorporates into the section on the third initiation in the Dvitīyakrama (see also Dvitīyakrama, 125b and my notes on the translation at that point). 274 116From uniting the realm of space and117 the vajra Great bliss that has genuine vision arises Which brings about genuine bliss. Between bliss and the bliss of cessation an absence is seen.118 |34| In the lotus maṇḍala the jewel And the heart of the lotus join And in vajra posture The mind is observed within the jewel, |35| And that itself is wisdom. This explained by the guru As the perfection stage; This is known as the fourth,119 The stage of120 the seven yogas. |36|121 Here, in incorporating verses from the Dvitīyakrama on the prajñājñānābhiṣeka, Vaidyapāda directly links the wisdom discerned during the third initiation, described in the Dvitīyakrama as “the perfection stage,” with “the fourth,” and with the seven yogas. The subsequent section of the Yogasapta, which makes up the majority of the text, addresses each one of these seven yogas in some detail, presenting each yoga both in terms of the way it is experienced seen during the prajñājñānābhiṣeka, and then in terms of its being fully actualized during post-initiatory practice. Buddhajñānapāda himself makes reference to the seven yogas twice in the Muktitilaka, once equating them with the “perfection stage of the perfection stage,”122 and once in reference to their being realized instantaneously during the yogin’s post-initiatory practice.123 Although Buddhajñānapāda’s writings do not mention the seven yogas in the initiatory context, the 116 The next eleven lines are incorporated from the Dvitīyakrama, verse 124a-125a. However, either the text transmission is corrupted here in the Yogasapta, as several words or short phrases have been omitted, or Vaidyapāda has rephrased the Dvitīyakrama. I surmise that he may have rephrased the passage, because there is only one instance where the Yogasapta’s version has been rendered unintelligible. Parallel verses are extant in Sanskrit in Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, which allow us to understand the sometimes opaque Tibetan of the Dvitīyakrama’s verses, and thus also those of the Yogasapta, more clearly, though there are places where the Saṃkṣiptābhiṣekavidhi does clearly differ from the Dvitīyakrama. See also notes 263-274 in my translation of the Dvitīyakrama. 117 Cf. Dvitīyakrama, verse 124, which has dang where the Yogasapta reads ni. 118 gang gang yang dag dga’ byed chags// dga’ gnyis rab tu dben nyid mthong//. When compared with the parallel line from the Dvitīyakrama (gang gang yang dag dga’ byed chags bral dga’ gnyis bar du dben nyid mthong byas brtan par gyis//, Dvitīyakrama 124b), the first of these two lines in the Yogasapta omits the word bral, thus significantly changing the meaning of the line. Given that Vaidyapāda clearly upholds Buddhajñānapāda’s system of three blisses in his writings, I presume this omission is an error and though I have not re-edited the Yogasapta to add in this syllable, since it would render the verse unmetrical, I have translated as if the line read chags bral, like it does in the Dvitīyakrama. This line also survives in Sanskrit in Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, which confirms the reading of chags bral (here = viramānanda). See note 263 in my translation of the Dvitīyakrama for the full citation from Vāgīśvarakīrti’s text. 119 bzhi pa] P, bzhin pa D 120 gyi] P, gyis D 121 mkha’ gnyis rab tu sbyor ba yis// man ngag tu ni de nyid brtag/ legs par goms pa’i phyag rgya las// ‘bad pa’i bya pas rtogs pa na// |32| ‘dod chags chags bral bar ma ste// sbyor ba bdun gyi mtshan nyid do// |33| rnam mkha’ khams ni rdo rje sbyor// yang dag spyan can bde chen ‘byung// gang gang yang dag dga’ byed chags// dga’ gnyis rab tu dben nyid mthong// |34| padma’i dkyil du rdo rje nor// pad snying la ni ‘byor ba ste// rdo rje’i skyil mo krung nang sems// nor bu bar du (bar du] P, rab tu D) mthong byas pas// |35| gang de nyid ni ye shes te// ‘di nyid rdzogs pa’i rim par ru// bla ma’i zhal lnga nas kyis bshad// de la sbyor ba bdun gyi (gyi] P, gyis D) ni// rim pa bzhi (bzhi] P, bzhin D) pa shes par bya// |36| (Yogasapta, D 71a.2-4; P 84b.3-6). 122 Muktitilaka, D 52a.2. 123 Muktitilaka, D 51b. 275 “perfection stage of the perfection stage,” with which he does explicitly equate them in the Muktitilaka, is itself identified with suchness, which is precisely what the initiate is meant to realize directly as part of the third initiation and/or by means of the oral instructions of the guru which are given as part of or subsequent to that. As such, although Vaidyapāda’s presentation of each of the seven yogas and their relationship to initiation—specifically their being experienced in the third initiation—as well as their perfection in post initiatory practice, is much more elaborated than what we find in Buddhajñānapāda’s own writings, Vaidyapāda’s presentation of the seven yogas in the Yogasapta and in his Sukusuma and Muktitilaka-vyākhyāna does correspond precisely within the way in which we find the seven yogas referenced, though more briefly, in Buddhajñānapāda’s own writings. A discussion of each of the seven yogas according to the Yogasapta is beyond the scope of the present chapter, but it will not take us too far afield, I think, and will provide a better sense of what the seven yogas are in this system, to include here the passage on just the first yoga from the Yogasapta. This passage continues directly from where we left off above, at the end of verse 36, where Vaidyapāda identified the wisdom experienced through the prajñājñānābhiṣeka with “the fourth” and with the seven yogas. (Again, we find here several lines incorporated from the Dvitīyakrama’s section on the prajñajñānābhiṣeka, which I have again italicized to make them easy to identify.) Vaiyapāda continues, So, what is that like? Together with a well-trained [consort]124 Following the instructions that have been given One should ascertain the cakra from the interior And come to know it a bit with one’s fingers.125 |37| 126The heart quivers127 and attentiveness wanes The hair on the crown falls loose and garments are cast off Beads of sweat cover the whole body Losing steadiness, |38| Even all of one’s [body] hairs stand on end,128 Looking with bloodshot eyes The wisdom of bliss genuinely arises And one becomes adorned with the complexion of Vairocana—129 |39| That is perfect example-less bliss. The space-like kāya Adorned with the major and minor marks Is luminous like a rainbow; |40| When one takes up this great body [i.e. the sambhogakāya], Which is luminous and yet unoriginated, The dharmakāya and nirmaṇakāya, as well, Come about due to [its] strength. |41| This excellent body of the great vajra holder, 124 See also verse 30 of the Yogasapta which mentions a well-practiced mudrā (i.e. consort) (legs par goms pa’i phyag rgya). This term legs par goms (svabhyasta) is also used in Dīpaṃkarabhadra’s maṇḍalavidhi to describe the yogiṇī in the second initiation, who is there consecrated as a mudrā (Maṇḍalavidhi, verse 362). 125 See Dvikrama, verse 118. 126 C.f. Dvitīyakrama, verse 120b-121a. I have again italicized the lines incorporate directly from the Dvitīyakrama simply to make them easier to identify here. 127 ‘dar] P, ‘dir D. Dvikrama verse 120b supports ‘dar 128 g.yo. Literally, they “move.” 129 The intent of this line remains unclear to me. 276 Which one never tires of seeing, Appears with three faces and six arms, Embraced by the vidyā, who is endowed with all supreme [qualities]. |42| At the time of the prajñājñānābhiṣeka One abides with those features. [Later] the [yogin] who has brought wisdom under control, Abiding in that state during the practice of the vratas and the rest, |43| [Fully] accomplishes that reality. At that time one [fully] obtains this [perfect example-less bliss] |44| This was the first chapter on the attainment of perfect example-less bliss.130 The initial part of the description of the first among the seven yogas, the yoga of perfect example-less bliss, closely follows the description of the union of the yogic partners during the prajñājñānābhiṣeka in the Dvitīyakrama, again incorporating verses directly from that text. Vaidyapāda then identifies the arising of the wisdom of “bliss” (ānanda), the first of the three blisses according to Buddhajñānapāda’s system, in the context of the prajñājñānābhiṣeka as the first yoga of “perfect example-less bliss.” As he goes on to explain, this corresponds with the yogin’s remaining self-visualized in the form of the deity during the sexual yoga of the third initiation. Later, a yogin “who has brought wisdom under control”—this is a reference to a yogin who has achieve the third among three (or four) stages of progress in practice131—actualizes this 130 ci lta zhes na/ legs par bsgoms dang lhan gcig tu// ji skad bshad pa’i cho ga yis// nang nas ‘khor lo gsal bya ste// sor mo cung zad go byas te// |37| snying ‘dar (‘dar] P ‘dir D) dran pa nyams pa dang// spyi bor skra grol gos kyang ‘dor// rngul chu thigs pas lus kun khyab// brtan pa’i ngang tshul shor nas ni// rang gi ba sbu kun kyang g.yo// mig dmar phra bas rab bltas pa’i// dga’ ba’i ye shes yang dag ‘byung// rtag pa’i mdog gis yang dag brgyan// de ni dpe [P 85a] med bde (bde] P, bda D) rdzogs so// nam mkha’ lta bu’i sku la ni// mtshan dang dpe byad kyis brgyan pa// ‘ja’ tshon lta bur rab gsal ba// gsal ba nyid ni (ni] D, na P) ma skyes pa’i// sku chen dang du blangs gyur na// chos kyi sku dang sprul sku yang// stobs kyi ‘byung bar ‘gyur pa’o// rdo rje ‘dzin pa chen po yi// sku che blta bas mi ngoms pa’i// zhal gsum phyag drug ltar snang ba// mchog kun ldan pas rig mas ‘kyud// [D 71b] // shes rab ye shes dbang dus su// mtshan nyid de yi tshul du gnas// ye shes dbang du gyur pa yis// brtul zhugs la sogs rim pa yis// gnas nas de yi don sgrub tshe// de y idus su de ‘thob pa’o// dpe med bde rdzogs kyi le’u ste dang po’o// (Yogasapta, D 71a.4-71b.1; P 84b.6-85a.4). 131 This refers to the third of three (or four) stages of a yogin’s progress that are referenced in Buddhajñanapāda’s and Vaidyapāda’s writings. In his Muktitilaka-vyākhyāna Vaidyapāda writes of “the instant blazing of nondual wisdom due to the observance of post-initiatory practices (vratacaryā) at the time when one [has reached the state of being a] third [-level] yogin.” rnal ‘byor pa gsum par ‘gyur pa’i tshe/ brtul zhugs kyi spyod pas skad cig tsam la gnyis su med pa’i ye shes ‘bar bas... (Muktitilaka-vyākhyāna, D 51b.1; P 337b.2-3). Such a “yogin of the third level” is explained later in the Muktitilaka-vyākyāna, commenting on a verse in the Muktitilaka that likewise indicates “the occasion of being a third[-level] yogin,” as the time when one should engage in the various postinitiatory practices described in the text. Vaidyapāda there explains the “third[-level] yogin” as someone who has not only gone beyond being a beginner (the first level), but also having gone beyond the intermediate level of a yogin who has “control over limited wisdom,” (the second level) to the level of a yogin who has “control over the wisdom that brings oneself and others to behold the illusory maṇḍala” (Sukusuma, 57b.2-3). The fact that here in the Yogasapta the yogin who has “control over wisdom” refers to the third, rather than the second, of these levels is indicated by the fact that in the context of several among the seven yogas the “third-level yogin” is explicitly mentioned, along with the fact that all of the references in the Yogasapta that mention a practitioner at this level of progress are made in the context of the practice of the vratas, and, as we have just seen, both Buddhajñānapāda’s and Vaidyapāda’s writings repetatedly indicate that it is specifically a third-level yogin who is to engage in these kinds of practices. This three-level schema of the development of a yogin’s meditative progress, mentioned at several places in Buddhajñānapāda’s and Vaidyapāda’s works, seems to either have been later expanded into four levels of progress in later works of the Jñānapāda School or (perhaps more likely?) simply have been mentioned only in an an abbreviated form in these earlier works. (In any case I have seen no reference to a fourth level in any of Buddhajñānapāda’s or Vaidyapāda’s writings). Sabine Klein Schwind writes of the distinctive practice instructions included in Kṣitigarbha’s Daśatattvasaṃgraha that are connected with generation stage pratice for the yogins on each of four levels, and also references the four-fold schema also in Dīpaṃkarabhadra’s 277 state of the sambhogakāya (which, as this passage explains, naturally brings about the actualization of the dharmakāya and nirmāṇakāya, as well) by means of the procedures of the post-initiatory vratas, at which point he fully obtains this aspect of the result of awakening.132 While the individual characteristics of each of the seven yogas differ, their presentation in terms of an experience in the prajñājñānābhiṣeka which is then perfected or fully actualized at the time of post-initiatory practice is consistent throughout. For example, with regard to the second yoga, non-duality, Vaidyapāda writes: For the illusory male and female, At the time of the prajñājñānābhiṣeka |48| When the wisdom of bliss [is experienced] A unique self-arisen nonduality comes about— That is explained as nonduality. |49| [Later] when the yogin of the third [level] Is practicing the vratacāryā He likewise [experiences] nonduality, And [at that time] this [experience] is said to be fully resolved.133 In this way, Vaidyapāda consistently sets forth each of the seven yogas in reference to their being experienced at the time of the prajñajñānābhiṣeka and fully obtained, resolved, or fully manifested by means of the practice of the post-initiatory conduct of a yogin of the third level. This moment of the full manifestation or full realization of the seven yogas by means of postinitiatory practices is mentioned in Buddhajñānpāda’s Muktitilaka, as well: The one who engages in these [types of] practices Realizes the seven yogas in a single instant. And for as long as existence persists He will have the eight characteristics of the taste.134 This passage seems to describe the moment of awakening itself, since the “eight characteristics of the taste,” are listed and identified in the Dvitīyakrama as eight characteristics of the awakened state; these eight are further elaborated in several of Vaidyapāda’s works, including the Yogasapta, as well.135 Based on the way they are presented in both Buddhajñānapāda’s and Vaidyapāda’s writings, taken together, it seems that in the early Jñānapāda tradition the seven yogas are experiences that take place during the prajñājñānābhiṣeka, which are pointed out orally by the guru’s instructions immediately following that initiation, and finally realized all at once in the moment of awakening, which takes place during post-initiatory practice. They are different aspects of the state of suchness, or awakening itself, and as such are glimpsed directly during Guhyasamājamaṇḍalavidhi, and in Ratnātaraśānti’s works (Klein-Schwind 2012, 87-92). According to Schwind’s work there is some degree of variation in the terminology used to refer to yogins on the varying levels of progress. 132 While a more thorough investigation into the works addressing the seven yogas would further clarify this issue, it appears from Vaidyapāda’s work that this first yoga is connected both with the experience of a moment bliss and the practitioner’s self-visualized form as the deity in the prajñājñānābhiṣeka, and, at the time of awakening, with his experience of (presumably lasting) bliss along with manifesting in the form of the sambhogakāya. 133 sgyu ma’i skye bu bud med las// shes rab ye shes dbang dus su// dga’ ba’i ye shes dus su ni// khyad par gnyis med rang ‘byung ba// [P 85b] gnyis su med bar de la bshad// rnal ‘byor gsum pas brtul zhugs kyi// spyod pa’i tshe ni de bzhin du// gnyis med ‘gyur bar ‘gyur ba ru// nges pa nyid du bshad pa yin// (Yogasapta, D 71b.5-6; P 85a.8-85b.1). 134 de ‘dra’i spyod pa la gnas pa// skad cig gis ni sbyor ba bdun// rtogs nas ji srid bar du ni// ro myang mtshan nyid brgyad ldan par// (Muktitilaka, D 51b. 5-6; P 62a.5-6). 135 The eight are listed in the Dvitīyakrama as: permanent, free from torment, cool, singular, blissful, stainless, joyful, and mentally joyful (Dvitīyakrama, verses 292-3). Vaidyapāda elaborates them in the Sukusuma (D 127a.4-7) and gives a similar presentation in the Yogasapta (Yogasapta, D 74a.1-3; P 88a.2-5). 278 initiation, but not fully perfected until the completion of post-initiatory practices. However, it is only in Vaidyapāda’s works that the relationship of the seven yogas and tantric initiation is made explicit, and where they are also referred to as “the fourth;” Buddhajñānapāda’s surviving writings do neither of these. Even within Vaidyapāda’s works, it is only in the Yogasapta that the dual relationship of each of the seven yogas to both initiation and the final moment of awakening is made clear. While Buddhajñānapāda’s own writings do not make direct reference to the seven yogas in an initiatory context, as noted above, he does identify them in the Muktitilaka with “the perfection stage of the perfection stage,” suchness, which is precisely what is pointed out by the guru and realized by the student in the context of the third initiation. Buddhajñānapāda also clearly references suchness being pointed out by an oral instruction, most likely one following the third initiation. This suggests that Vaidyapāda’s presentation of the seven yogas in the Yogasapta most likely represents a system that was already in practice in Buddhajñānapāda’s lifetime. Why, then does the Dvitīyakrama not make direct reference to these seven (or to “the fourth”) in its rather detailed instructions on the higher initiations? Perhaps this is because, as a set of essential oral instructions given in the initiatory context, Buddhajñānapāda did not see fit to describe, or even reference, them in a written text. Why Vaidyapāda chose to write about these ritual details is not clear—and as I noted above he seems to be conscious of the fact that he was putting into writing something that had not been previously recorded—but his work certainly gives us a much fuller picture of initiation in the early Jñānapāda School, and indeed a much clearer picture of the issue of “the fourth” in this early period, than we would otherwise have. “The Fourth” versus “The Fourth Initiation:” Points of Continuity and Divergence As for the relationship of this early tradition of the seven yogas and “the fourth” to later tantric initiatory traditions, which did come to assert a “fourth initiation,” it appears that there is quite a bit of continuity. Regarding what eventually came to be the standard idea of what constituted the “fourth initiation,” in both the systems of the Guhyasamāja-tantra and the later Yoginī tantras, Isaacson has written that this position136 holds that the Fourth empowerment is one which is bestowed verbally, i.e. by the initiating guru giving a verbal instruction to the initiand. Now some texts seem to indeed refer to this consecration as only verbal—Vāgīśvarakīrti in his Saṃkṣiptābhiṣekavidhi, calls it the vacanamātrābhiṣeka (cf. Sakurai 1996, 419, l. 11 and l. 13), and in this text at least does not indicate that anything more is involved than this speech by the guru. But it is also clear that in fact usually, if not always, the Fourth empowerment was seen as having, theoretically at least, another component as well. Indeed, had it been otherwise, that which as the final one one expects to be the culminating or crowning empowerment or consecration could well seem an anti-climax. This no doubt was as clear to these tantric authors as it is to us. The way that this added element is sometimes expressed is that the prajñājñānābhiṣeka and the Fourth empowerment are said to be related to each other as mark/characteristic and that which is marked, or ultimate goal. In theory, the verbal instruction received from the guru is supposed to cause the bliss experienced, for an instant, without sensation of duality in the prajñājñānābhiṣeka to become strong or steady. 136 Note that Isaacson here explains that what he is here expressing as a single position is indeed two slightly different positions, and he refers the reader to Sakurai’s 1996 book (in Japanese) on a tentative distinction between the two (Isaacson 2012b, 270). 279 Treating the Fourth empowerment as representing or being, in some way, the ultimate goal, means that explanations of its nature may vary according to just how that goal is envisaged. In the—as far as I can tell at present very influential—works of Vāgīśvarakīrti, the transcendental or goal-aspect of the Fourth empowerment is explained as being seven-fold or having seven aspects, the so-called seven aṅgas of mahāmudrā.137 Based on this description of the way that the fourth initiation was presented in works of the later tradition—Isaacson is referencing works from around the eleventh century—we can see that quite a number of its features are already found in the system of “the fourth” as described in Vaidyapāda’s writings, and which, as I have argued above, I believe we can take to be reflective of Buddhajñānapāda’s own tradition. This includes “the fourth” being a verbal instruction; something that follows, as the fourth initiation does, the prajñājñānābhiṣeka; and its “representing, or being in some way, the ultimate goal.” Given all of these similarities, we may not be surprised to discover that the seven aṅgas of mahāmudrā that Isaacson mentions are set forth in Vāgīśvarakīrti’s writings as the “transcendental or goal-aspect” of the fourth initiation are none other than the same seven aspects described in Buddhajñānapāda’s and Vaidyapāda’s writings as the “seven yogas.”138 The only aspect of the fourth initiation that Isaacson highlights in his article that is not explicitly found in Buddhajñānapāda’s and Vaidyapāda’s writings (though without a closer reading of the entire Yogasapta I am hesitant to completely rule out its being there) is the relationship between the prajñājñānābhiṣeka as the mark or characteristic and the fourth initiation (or “the fourth”) as what is marked, the goal of awakening itself. Indeed, as I read Buddhajñānapāda’s works, he states rather clearly that what is experienced during the prajñajñānābhiṣeka is precisely suchness itself—or “the perfection stage” itself, as Buddhajñānapāda refers to it in the Dvitiyakrama’s final verse on the prajñājñānābhiṣeka—and not just an example of it. And yet, this very suchness is also said in his writings to be received from the words of the guru. In other words, there does not appear, in Buddhajñānapāda’s or Vaidyapāda’s writings, to be a clear distinction between what is experienced in the prajñājñānābhiṣeka and what is pointed out after it (during “the fourth”), apart from the simple fact that without its being pointed out by the guru’s words, the disciple would be unable to recognize the experience in the third initiation as suchness.139 In the Yogasapta, while Vaidyapāda describes “the fourth” as something “known from the guru’s words,” he also identifies it with the seven yogas, and the individual descriptions of the yogas, as we have seen, consistently specify both a particular experience within the context of the prajñajñānābhiṣeka and the full realization of that aspect of awakening in the moment of obtaining the final result of perfect awakening. While the experience in the prajñājñānābhiṣeka is certainly not the full experience of awakening itself, it does not seem to be distinguished qualitatively from the aspect of the result, nor, as is more directly relevant to the discussion at hand, does it appear to be 137 Isaacson 2012b, 270-71. 138 A citation from Vāgīśvarakīrti’s Saptaṅga listing the seven aṅgas is found in Rāmapāla’s Sekanirdeśapañjikā. The seven aṅgas are listed in that citation from Vāgīśvarakīrti’s work as sambhoga, sampuṭa, mahāsukha, niḥsvabhāva, kāruṇyanirbhara, nirantara, aniroda (Isaacson and Sferra 2014, 271). This list is almost identical with Vaidyapāda’s list: dpe med bde rdzogs, gnyis su med pa, bde ba chen po, rang bzhin med pa, thugs rjes rgyas pa, rgyun mi chad pa and ‘gog pa med pa. Only the second member of the list of seven bears a different name—but sampuṭa seems to be understood in this context to refer to union, and it is not such a stretch to understand union and nonduality (gnyis su med pa) as referring to a similar idea. In any case a fuller study of Vāgīśvarakīrti’s Saptaṅga alongside Vaidyapāda’s Yogasapta would certainly be illuminating. 139 We may also recall here verse 355 in the Dvitīyakrama, cited above, where it is mentioned that the dharmakāya is experienced briefly when falling asleep, during sexual union, and at other moments, but Vaidyapāda specifies that these moments are so brief that without a guru’s instructions a disciple is unable to recognize them as such (for Vaidyapāda’s comments on this point see Sukusuma, D 132b.6-7; P 159b.8-160a.1). 280 distinguished from the “suchness” that is transferred in the verbal communication from guru to disciple that constitutes “the fourth.” There is, however, in the Dvitīyakrama, a passage that mentions the yogin’s remaining within the three blisses “in the manner of the mark (mtshan nyid)” (or the “characteristic” to use the other term that Isaacson mentioned above) during what does appear to be a reference to the prajñājñānābhiṣeka. But this “remining in the manner of the mark” in the three blisses during the prajñājñānābhiṣeka is contrasted there, not with an experience of “the goal” (or “that which is marked”) during “the fourth,” but rather with the full attainment of the three blisses—which Vaidyapāda equates with the three kāyas—at the time of the yogin’s complete awakening by means of post-initiatory practice.140 In his comments on this passage, Vaidyapāda links the three blisses not only to the three kāyas, but also to the seven yogas, and notes that the yogin remains in these “in the manner of the mark” during the prajñājñānābhiṣeka, and that they will “in their own time” be fully attained as explained in the tantras.141 Even though the terminology that is used here—that of remaining in the three blisses “in the manner of the mark” during the prajñājñānābhiṣeka—could possibly be interpreted to mean that what is experienced in the third initiation is not identical with result of awakening itself—even if experienced only momentarily—I think the preponderance of the evidence in Buddhajñānapāda’s writings suggests that we should not read the passage that way. He has clearly stated that the purpose of the third initiation is to bring the disciple to a direct experience of the dharmakāya 140 That passage reads, “Just as the yogin of the higher stage/ |289| Having put forth tireless effort/ Remained there [in the consort’s lotus] for a moment / In the manner of the example / In bliss, middling bliss, / |290| And the bliss of cessation, / [Likewise,] in time, he will attain, just as has been taught, / The three blisses just as they are. / 291| Then until saṃsāra’s end/ He will remain, free from torment, / Cool, singular, / Blissful, stainless, / |292| Joyful, and mentally joyful—/ These are the eight [signs] of having tasted great bliss./” gong ma’i rim pa rnal ‘byor pas// |289| ‘bad pa che thang rab byas pas// de ltar de ru mtshan nyid kyi// tshul du cung zad rab gnas pa’i// dga’ dang dga’ ba bar ma dang// |290| dga’ dang bral ba’i dus su yang// ji bzhin gnas pa’i dga’ ba gsum// ji skad gsungs pa thob par ‘gyur// |291| de nas mtha’ med ‘khor ba’i bar// rtag pa dang ni mi gdung dang// bsil ba dang ni gcig pa dang// bde ba dang ni dri med dang// |292| dga’ ba dang ni yid dga’ ba// ‘di ni bde chen ro myang brgyad// (Dvitīyakrama, verses 289d-293b). The context of this passage in the Dvitīyakrama, just after a description of post-initiatiory practices, makes clear that this is the context in which the full realization described in the passage will take place. See also Sukusuma, D 128a.1-4; P 153a.1-6. Vaidyapāda’s Sukusuma mentions both the “actual” and the “example wisdom” in commenting on an earlier passage from the section of the Dvitīyakrama on the third initiation, which has been incorporated into many later liturgies as part of the vidyāvrata ritual. The lines from the Dvitīyakrama state: “Thus, throughout endless saṃsāra/ You must never separate from her,” de bas mtha’ med ‘khor ba’i bar// khyod kyis ‘di dang bral mi bya// |89| (Dvitīyakrama, verse 89cd). Vaidyapāda comments: You must never separate from her means that since she is the seal of the perfection of wisdom you must examine the actual and the example wisdom together with her in order that the continuity of wisdom is not severed.” khyod kyis ‘di dang ‘bral mi bya/ zhes pa ni shes rab kyi pha rol tu phyin pa’i phyag rgya bas na de dang lhan cig tu mngon sum dang dpe’i ye shes brtag par bya ba ste/ ye shes rgyun mi ‘chad pa’i phyir ro// (Sukusuma, D 106a. 3; P 107b.4-5). While this passage could be understood to refer to the “example wisdom” of the third initiation and the “actual wisdom” of the fourth, as would be common in the later tradition, given the passage in the Dvitīyakrama just cited that contrasts the remaining in the three blisses “in the manner of an example” in the third intitaiton with their full attainment in post initiatiory practice—rather than with their direct experience in “the fourth”—I would suggest that it makes more sense to read Vaidyapāda’s comments about the “example wisdom” and “actual wisdom” in a similar way. This is especially the case since I have thus far not seen any indication of such a relationship between the third initiation and “the fourth” evidenced elsewhere in Buddhajñānapāda’s or Vaidyapāda’s writings. The question, though, should remain an open one, as I have not read all of Vaidyapāda’s oeuvre. 141 Sukusuma, D 128a.1-4; P 153a.1-6. Vaidyapāda here links bliss with two of the seven yogas, middling bliss with two, and the cessation of bliss with one. He does not specify which yogas are linked to the different blisses. He makes this same link in his Muktitilaka-vyākhyāna (D 58a.1), and there also notes that the remaining two of the seven yogas pertain to all three blisses, but in that statement he also does not identify which of the yogas pertain to which of the blisses. I imagine that this point is clarified in the Yogasapta, but did not yet have the opportunity to check this. 281 (Dvitīyakrama, verse 86); that what is experienced in the third initiation is “bliss,” “wisdom,” (Dvitīyakrama, verse 124), and the “perfection stage” itself (Dvitīyakrama, verse 125), which as we have seen in Chapter Six, Buddhajñānapāda repeatedly equates with suchness itself; and even that the dharmakāya is experienced briefly during the non-initiatory contexts of falling asleep, fainting, yawning, and during intercourse (Dvitīyakrama, verse 355), though Vaidyapāda is careful to specify that it is not recognized in those moments due to lack of oral instructions from a guru. Thus, I still contend that for Buddhajñānapāda what is experienced in the third initiation with the guidance of the guru’s oral instructions (and therefore as part of, or by means of “the fourth,”) is nothing other than a direct glimpse of the result of awakening itself, but an experience that is brief, and therefore must be cultivated and stabilized through training in the practices of the perfection stage, in order that it is fully realized during post-initiatory practice. It is also worth emphasizing, I think, that despite the fact that I believe “the fourth”—in the sense of the bestowal of an oral instruction on the suchness of the seven yogas by the guru to the disciple in the context of the third initiation—to have been a feature of Buddhajñānapāda’s own system, it is clear that Buddhajñānapāda, at least, did not understand this to be a separate initation. As I have shown above, he does appear to separate out this verbal communication of suchness as an important aspect within the initiatory context, but even though it seems likely that this instruction was referred to as “the fourth” in his lifetime (though, again, this term does not appear in his extant writings), such an oral instructions was, for him, still intimately connected with, and not fully separated out from, the third initiation. In the concluding section of the Dvitīyakrama, Buddhajñānapāda includes precisely three verses that summarize the three initiations; there is no fourth verse representing a fourth initiation, and the term “great oral instruction” (man ngag chen po) is included in the verse on the prajñājñānābhiṣeka. These three verses, which make reference to the three initiations, the kalaśābhiṣeka (including the ācāryābhiṣeka), the guhyābhiṣeka, and the prajñājñānābhiṣeka, respectively, in the form of a dedication read: When he has been cleansed and sprinkled and made pure, and thus become a great ācārya Who holds all of the tantras, and brings others to connect with all tantras,142 And having perfectly realized the first stage and purified all stains, May the yogin become a suitable vessel for illusory wisdom! 143 |395| Through respectfully [serving at] the feet of a compassionate guru And by means of that which has the rabbit-holder’s form,144 may one’s mindstream be perfectly ripened So that the field is purified,145 and one perfectly realizes the reality of phenomena to be illusory and the like: 142 Vaidyapāda here specifies that all the tantras refers to “the Kriyā, Cārya, Yoga and Yoganiruttara mantras and tantras” (bya ba dang / spyod pa dang/ rnal ‘byor dang/ rnal ‘byor bla na med pa’i sngags dang rgyud) (Sukusuma, D 138b.3; P 167a.2-3) 143 Here Vaidyapāda associates the vessel mentioned here with the vase (kalaśa) initiation, and identifies this passage as referring to the receiving of the kalaśābhiṣeka (Sukusuma, D 138b.4). The next two verses refer to the guhya- and prajñājñānābhiṣekas, respectively. 144 This is a reference to the moon, and therefore a metaphor for the bindu of bodhicitta. 145 I believe this is yet another instance in which Buddhajñānapāda takes a Mahāyāna concept—here the concept of zhing sbyang ba, the “cultivation/purification of the [buddha]field,” and reenvisions it according to a tantric paradigm. Vaidyapāda explains the zhing here as referring to the aggregates of the yogin himself (Sukusuma, D 138b.6-7). Thus the field that is purified here is indeed the body of the yogin himself. This is an internalization of the concept of the purification of the field, directing it towards the locus of the yogin’s body—the macrocosm having become microcosm. This supports Jacob Dalton’s (2004) analysis of the interiorization of ritual during precisely this period. 282 In this way may all beings, like Maitreya and others, arrive [in that state]!146 |396| Through the blessings of the sahaja [ācārya and] the great compassionate revered master,147 [One] encounters bliss, through which one [realizes] the undeciving truth, just as it is, The supreme, great pure essence of all things, the drop which is the sixteenth part, Achieved through resting,148 the great instruction —may you come to encounter this! |397|149 Initiation in the Early Jñānapāda School: Concluding Reflections While further study of Buddhajñānapāda’s and Vaidyapāda’s writings (as well as those of other direct disciples of Buddhajñānapāda, like Dīpaṃkarabhadra) will certainly reveal more about initiation in the early Jñānapāda tradition, this brief and preliminary study of initiation in their writings indicates that in Buddhajñānapāda’s system of tantric initiation, the details of which are made clearer in Vaidyapāda’s writings, we find not only an early presentation of the guhyābhiṣeka and the prajñājñānābhiṣekas, but also an early instance of the ritual that eventually came to be known as the “fourth initiation,” but at this time seems to have been referred to more commonly simply as “the fourth,” and to have still been very closely connected with the prajñājñānābhiṣeka. Vaidyapāda’s list in the Sukusuma of Buddhajñānapāda’s compositions, in fact references a work of Buddhajñānapāda’s called The Method for Engaging in the Fourth (bzhi pa la ‘jug pa’i thabs). Most unfortunately this work seems not to be extant in either its original Sanskrit nor in Tibetan translation (as we saw in Chapter One, many of the works in this list are extant), but the possibility that Buddhajñānapāda may have composed a work on “the fourth” remains a very interesting one. Despite the fact that this work is not extant, Vaidyapāda does seem to be quite a reliable witness of Buddhajñānapāda’s system, and I see little reason to doubt him with respect to his list of Buddhajñānapāda’s compositions. Yet, the fact remains that we have no reference at all in any of Buddhajñānapāda’s surviving writings to the use of that term. And despite the fact that in his Yogasapta Vaidyapāda is careful not to describe “the fourth” as an initiation, there are indications that already by Vaidyapāda’s time the term “the fourth initiation,” rather than just “the fourth,” had already begun to be used. As I noted above, in his Guhyasamājamaṇḍalopāyikā-ṭīkā, a commentary on Dīpaṃkarabhadra’s Maṇḍalavidhi, Vaidyapāda makes a brief reference to the bestowal of a “precious fourth initiation” (dbang bskur ba rin po che bzhi pa) consisting of the guru’s oral instructions with respect to the union of the third initiation.150 If we take seriously Vaidyapāda’s claim that Buddhajñānapāda himself 146 This line is a bit gramatically unclear. Khenchen Chodrak Tenphel explains that the basic sense is the aspiration for all beings to follow in Maitreya’s footsteps (Khenchen Chodrak Tenphel, personal communication, March 2016). 147 Vaidyapāda explains that this refers to the causal guru, which in this system is the guru from whom one receives initiation and instruction (Sukusuma, D 139a.1-2). 148 Vaidyapāda clarifies that it is “attained through resting” since it is encountered through the winds resting in the central channel (Sukusuma, D 139a.3-4). 149 blugs dang gtor dang dag par byas pas slob dpon cher ‘gyur te// thams cad kun kyi rgyud ‘dzin gzhan rnams rgyud kun la sbyor ba’i// dang po’i rim pa rab rtogs dri ma rnams ni dag byas te// ye shes sgyu ma’i snod du rung bar rnal ‘byor de ‘gyur shog// |395| snying rje ldan pa’i bla ma’i zhabs la gus par rab ldan pas// ri bong ‘dzin pa’i gzugs kyis rang rgyud rab tu smin byas te// zhing dag byas pas chos kun sgyu sogs don du rab rtogs nas// byams pa la sogs bzhin du sems can kun gyis ‘gro bar shog// |396| lhan cig byed pas byin brlabs rje btsun thugs rje chen po yis// dga’ ba brnyed pas chos kun dag pa’i ngo bo chen po mchog// ji bzhin gnas pa’i don la mi slu bcu drug thig le cha// ngal gso las thob man ngag chen po rab tu rnyed par shog// |397| (Dvitīyakrama, verses 395-97). 150 de nas des rgyas btab nas dbang bskur ba rin po che bzhi pa (pa] P, D om.) bla ma’i man ngag gis mkha’ gnyis kyi sbyor ba’i bshad pa sbyin par bya ste/ (Guhyasamājamaṇḍalopāyikā-ṭīkā, D 211b.3-4; P 539b.6-7). 283 wrote a text on “the fourth,” then we can say that there seem to be, in the early 9th century, a circle of texts around Buddhajñānapāda, Vaidyapāda, and the Samājottara in which the term “the fourth” was common parlance; in the communities using these texts we presumably also find the practice of the bestowal of this essential oral instruction closely connected with the prajñājñānābhiṣeka. The observations on the topic of “the fourth” and its status in the early Jñānapāda tradition that I have set forth here are, however, far from conclusive, and a more extensive examination of Buddhajñānapāda’s and Vaidyapāda’s writings on this topic will surely reward further study. Buddhajñānapāda was writing in a vibrant time for tantric Buddhist traditions, when the higher initiations had been newly added to the set of Yoga tantra initiations in order to prepare practitioners for the second stage of tantric practice. As we have seen, his writings provide us an early window into the initiatory sequences for the second and third initiations, and even a glimse of what came to be known as “the fourth,” though there we need to turn to Vaidyapāda’s works for more clarification, and the topic still deserves further study. What these writings do suggest, however, is that already by Buddhajñānapāda’s time—the late 8th and early 9th centuries—the basic features of the initiatory sequence that would characterize tantric Buddhism all the way through its late period in India were already in place, even if certain aspects had yet to be fully fleshed out, as it were, into their more mature forms. As I have noted, the material that I have presented in this chapter constitutes just a preliminary inquiry into this topic. A more thorough study of initiation in Buddhajñānapāda’s writings, and perhaps even more so a careful study Vaidyapāda’s and Dīpaṃkarabhadra’s works on initiation, will certainly shed more light on initiation in the early Jñānpāda School, and indeed on the development of tantric Buddhist initiatory practices on the whole. 284 Conclusion 285 Chapter Eight Buddhajñānapāda and Beyond: Buddhajñānapāda’s Thought Moving into the Later Tradition The dharma taught by the buddhas Abides authentically in two stages: The generation stage And the perfection stage. -Buddhajñānapāda, Muktitilaka The dharma taught by the vajra-possessors Is taught authentically in two stages: The generation stage And likewise the perfection stage. -Samājottara As we have seen in the earlier chapters, the 8th and 9th centuries marked a period of creativity and development in which new methods and techniques found their way into tantric Buddhism, and the practitioners and exegetes of the time incorporated these into the already rich tapestry of rituals and doctrines pertaining to that tradition. Buddhajñānapāda appears to have been at the forefront of many of these developments, as he set forth an integrative system of tantric theory and practice inspired not only by his many human teachers, but also by a vision of Mañjuśrī himself, whose instructions, delivered during their visionary encounter in a forest near Vajrāsana, constitute the core of Buddhajñānapāda’s practice tradition. The ideas and practices found in Buddhajñānapāda’s writings would go on to influence and inspire the tantric Buddhist tradition as it continued to develop in India and spread to other regions of Asia. In this final chapter I would like to look briefly to some of the pathways through which his thought spread and influenced later tantric traditions. A comprehensive study of that influence would certainly require more than a short chapter; here I will just focus on the relationship of Buddhajñānapāda’s writings with what I believe to be one of the earliest vehicles by means of which his thought made its way into later tantric traditions: the Samājottara. In earlier chapters, I have already referenced a number of instances in which it appears that Buddhajñānapāda’s writings have influenced the Samājottara rather than the other way around, and here I will provide some of the evidence behind those claims. I have not been able to complete a full study of the Samājottara, which I imagine would likely turn up further evidence of this relationship (and could, of course, turn up evidence contrary to what I present here!).1 Nonetheless, I will examine in this chapter several passages that I believe suggest that Buddhajñānapāda’s works likely preceded, and indeed influenced, the Samājottara. The Samājottara: The “Eighteenth Chapter” of the Guhyasamāja-tantra The Samājottara is widely known as the eighteenth chapter of the Guhyasamāja-tantra, though it appears to have first circulated separately from the seventeen chapters of the Guhyasamāja root tantra, and was only later appended as its final chapter.2 To wit, the Samājottara itself is comprised of a series of questions posed by a group of bodhisattvas to all the tathāgatas that address a number of topics presented in the root Guhyasamāja-tantra. In their 1 I am currently preparing a translation of the Samājottara for the 84000, Translating the Words of the Buddha project, and hope to complete that translation and study of the Samājottara in the coming months. 2 The first to have made this observation seems to have been Matsunaga (1977b, 116). 286 answers to these questions, the tathāgatas reference the first seventeen chapters of the root tantra by their chapter numbers,3 and the Samājottara even appears to refer to itself as an uttaratantra, a “supplementary tantra,” to the Guhyasamāja tantra.4 The text also very much appears to have been understood by Indian commentators as a separate work from the root Guhyasamāja-tantra; among the sixteen extant Indic5 commentaries on the Guhyasamāja-tantra of which I am aware, eight address only the root tantra (chapters 1-17), while four are dedicated exclusively to the Samājottara, and four address all eighteen chapters in a single commentary.6 However, all of the commentaries that address both the root tantra and the Samājottara together make a clear break between their comments on the two, and use the terms “root tantra” (rtsa ba’i rgyud) and “supplementary tantra” (rgyud phyi ma) to distinguish the two sections. Two of the commentaries that address the full eighteen chapters even begin their comments on the Samājottara with a separate homage. What is more, the Samājottara is still preserved as an independent text in the Derge edition of the Tibetan Kangyur (where it is entitled the Rgyud phyi ma (Tōh. 443)), and the translation of the Guhyasamāja-tantra preserved at Dunhuang lacks the Samājottara.7 Thus it seems clear both that the Samājottara was composed later than the 3 Samājottara, vv. 25-28. Technically speaking these verses only reference Chapters 2-17 by chapter number, but my point—that the text references the full root tantra—remains the same. 4 “That which is spoken in accordance with Bodhivajra/ Is called the supplementary tantra.” sahoktir bodhivajrasya sottaraṃ tantram iṣyate/ (Samājottara, 36cd). Vaidyapāda explains: “Because of being said to be similar to the words that were spoken by Bodhicittavajra, who is the teacher in the context of the root tantra, [this] is called the supplementary tantra” rtsa ba’i rgyud la ston pa byang chub kyi sems rdo rje yis gsungs pa’i tshig dang ‘dra bar gsungs pas de rgyud phyi mar ‘dod ces te/ (Saṃyagvyākaraṇa, D 184b.6-7). 5 The commentary by Viśvamitra (Tōh. 1844) is included in the Tibetan canon as a translation from the Sanskrit, but several of its features have led me to conclude that it is mostly likely a Tibetan composition. In addition to lacking both a Sanskrit title at the beginning and a translator’s colophon at the end (which would not in and of itself preclude its being an Indic text), the commentary, which deals only with the Samājottara and not with the root tantra, is nearly twice the length of most Indic commentaries on the tantra and shows a number of linguistic features that I believe could only have arisen in an indigenous Tibetan composition commenting on a Tibetan translation of the Samājottara, rather than on the Sanskrit text. 6 The sixteen commentaries are: Nāgārjuna’s Śrīguhyasamājatantrasya-tantraṭīkā and his Aṣṭādaśapaṭalavistaravyākhyā (the two are preserved together as Tōh. 1784, though they present themselves as two separate texts with separate titles); Candrakīrti’s Pradīpoddyotana-nāma-ṭīkā (Tōh. 1785), which also survives in its original Sanskrit, Yaśobhadra’s Sarvaguhyapradīpaṭīkā (Tōh. 1787) (this commentary is sometimes attributed to Nāropa, but the colophon says Yaśobhadra); *Praśāntajñāna’s Upadeśaniścaya-nāma-śrīguhyasamājavṛtti (Tōh. 1843); Viśvamitra’s *Śrīguhyasamājatantropadeśasāgarabindu (Tōh. 1844); Thagana’s Śrīguhyasamājatantravivaraṇa (Tōh. 1845); Cilupa’s Ratnavṛkṣa-nāma-rahasyasamājavṛtti (Tōh. 1846); Jayadatta Śrīguhyasamājatantrapañjikā (Tōh. 1847); Vimalagupta’s Śrīguhyasamājālaṃkāra (Tōh. 1848); Vimalagupta’s Aṣṭādaśapaṭalavyākhyāna (Tōh. 1849); Vaidyapāda’s Samyagvidyākāra-nāma-uttaratantravyākhyāna (Tōh. 1850); Ratnākaraśānti’s Kusumāñjaliguhyasamājanibandha (Tōh. 1851); *Pramuditākaravarman’s Śrīguhyasamājatantrarājaṭika-candraprabhā (Tōh. 1852); Vajrahāsa’s Tantrarājaśrīguhyasamājaṭīkā (Tōh. 1909); Vilāsavajra’s *Guhyasamājatantranidānagurūpadeśabhāṣya (Tōh. 1910); Smṛtijñānakīrti’s Śrīguhyasamājatantrarājavrtti (Tōh. 1914); and Ānandagarbha’s Śrīguhyasamājapañjikā (Tōh. 1917). Of these sixteen the commentaries by Nāgārjuna (he wrote two; one on the root tantra and one just on the Samājottara), Candrakīrti, Praśāntajñāna, Cilupa, Vimalagupta (he also wrote two separate commentaries, one on the root tantra and another just on the Samājottara), Pramuditākaravarman, Vajrahasa and Ānandagarbha address only the root tantra (chapters 1-17); the commentaries by Nāgārjuna, Viśvamitra, Vimalagupta, Vaidyapāda are dedicated exclusively to the Samājottara, and those by Thagana, Jayadatta, Ratnākaraśānti, and Smṛtijñānakīrti address all eighteen chapters in a single commentary. 7 Regarding the Samājottara’s translation into Tibetan, while the Guhyasamāja-tantra was translated during the early translation period, it appears that the Samājottara was not translated until the later translation period. The Rnying ma rgyud ‘bum translation of the Samājottara states, confusingly, that it was translated by the translators Buddhaguhya, who lived in the 8th century, and Drogmi Palgyi Yeshe (‘Brog mi dpal gyi ye shes) who lived in the 11th—while the translation of the Samājottara preserved in the Derge Kangyur states that it was translated by Śraddhākaravarman and Rinchen Zangpo (Rin chen bzang po) (958-1055), of the later translation period. 287 Guhyasamāja-tantra and with reference to it, and that even after the Samājottara was appended to the root tantra it was considered a separate but connected work by Indian commentators.8 What is unclear, however, is how long of a period passed between the circulation of the complete root tantra and the circulation of the Samājottara. The evidence that I will present below suggests that this may not have been a very long period. The root Guhyasamāja-tantra is known to have reached its final form in the last half of the 8th century, and I believe the Samājottara likely began to circulate in the early to mid 9th century, perhaps some time between 830-850. Relating Buddhajñānapāda’s Works and the Samājottara Though Matsunaga has argued convincingly that even the root Guhyasamāja-tantra appears to have compositional layers,9 Buddhajñānapāda seems to have known the full Guhyasamāja-tantra, as he incorporates several verses from the seventeenth and final chapter of the tantra into the Caturaṅga/Samantabhadra-sādhana.10 Importantly, these verses from Chapter Seventeen of the Guhyasamāja-tantra are adopted quite faithfully—that is, without any alteration—into Buddhajñānapāda’s sādhana.11 However, while it thus seems that he knew the full Guhyasamāja root tantra, there is no clear evidence in Buddhajñānapāda’s writings that he knew the Samājottara. In fact, a careful comparison of several passages in his writings with parallel passages in the Samājottara seems to show that Buddhajñānapāda’s writings likely preceded that tantra. The first piece of evidence for this relationship is a passage in the Muktitilaka that is parallel with the important verse from the Samājottara distinguishing the two stages of tantric practice.12 Harunaga Isaacson has identified the verse in the Samājottara as the scriptural locus classicus of the two stages—which, even if I am correct that the Muktitilaka’s verse is earlier, remains the case, as the Muktitilaka is, of course, an authored work rather than a scriptural one.13 Isaacson has further remarked on the undoubtedly intentional parallels between the Samājottara verse and Nāgārjuna’s well-known verse from the Mūlamadhyāmakakārikās that sets forth the distinction between the relative and ultimate truths.14 If we accept Isaacson’s suggestion that the author(s) of the verse on the two stages were consciously evoking Nāgārjuna’s statement about the two truths, as I believe we should, a comparison with the verse from the Mūlamadhyāmakakārikās while keeping this point in mind can provide a small clue that suggests 8 Perhaps the more interesting question is not whether it was originally or continued to be considered a separate work, both of which seem clearly to be the case, but how the Samājottara gained the distinction of being appended to the root tantra as a final chapter. There are quite a number of supplementary or explanatory tantras of the Guhyasamāja-tantra, but the Samājottara is the only one that came to be considered, in some way, a part of the tantra itself. 9 Matsunaga 1980. I discussed this point also in Chapter One. 10 Matsunaga 1980, xxv. 11 See Guhyasamāja-tantra, 17.72-75 and Caturaṅga/Samantabhadra-sādhana, 46-49. The Sanskrit for these verses in the Caturaṅga/Samantabhadra-sādhana fortunately survives and is edited in Kano 2014, allowing for a careful comparison of the verses in question. 12 The argument about these verses that I present here is based on a paper that I gave some years ago at a conference at UC Santa Barbara (C. Dalton 2014). 13 Isaacson 2002a, 468-9. 14 ibid. Building on Isaacson’s work, Christian Wedemeyer (2007, 40-41) has interpreted the parallelism between Nāgārjuna’s verse and the Samājottara’s as an attempt to link the Guhyasamāja-tantra with the Madhyamaka tradition and its commentators, a central project of the Ārya School of Guhyasamāja exegesis, which is generally recognized by scholars to be later than the Jñānapāda School. Jacob Dalton (2004, 24) also suggests that the Samajottara may have been a later composition associated with the Ārya School and its legitimizing project. However, given the presence of this verse in Buddhajñānapāda’s Muktitilaka, and the close association of the Samājottara with Buddhajñānapāda’s writings, this seems unlikely to be the case. 288 a philological basis for establishing Buddhajñānapāda’s verse about the two stages as closer to Nāgarjuna’s text, and thus potentially earlier than the Samājottara’s parallel verse. Let us take a look at all three verses. The Sanskrit is extant for both Nāgārjuna’s and the Samājottara’s verses, but with the verse from the Muktitilaka we have to make do with only the Tibetan translation. I provide the Tibetan translations of all three verses here, though, as an additional tool in our comparative endeavor. Here is Nāgārjuna’s verse on the two truths: dve satye samupāśritya buddhānāṃ dharmadeśanā/ lokasaṃvṛtisatyaṃ ca satyaṃ ca paramārthataḥ//15 sangs rgyas rnams kyis chos bstan pa// bden pa gnyis la yang dag brten// ‘jig rten kun rdzob bden pa dang// dam pa’i don gyi bden pa’o//16 The buddhas taught the dharma in reliance upon the two truths The relative truth of the world and the ultimate truth. Here is the verse from the Muktitilaka on the two stages: sangs rgyas rnams kyis chos bstan pa// rim pa gnyis la yang dag gnas// bskyed pa yi ni rim nyid dang// rdzogs pa’i rim pa kho na’o//17 The buddhas taught the dharma [As] abiding in two stages: The generation stage And the perfection stage. And here is the Samājottara’s verse on the two stages: kramadvayam upāśritya vajriṇāṃ dharmadeśanā/ kramam autpattikaṃcaiva kramam autpannakaṃ tathā//18 rdo rje can gyis chos bstan pa// rim pa gnyis la yang dag brten// bskyed pa yi ni rim nyid dang// de bzhin rdzogs pa’i rim pa’o//19 The vajra-holders taught the dharma in reliance upon the two stages The generation stage and the perfection stage. 15 I cite Nāgārjuna’s verse as given in Isaacson 2002a, 469. 16 Mūlamadhyamakakārikās, D 14b.7-15a.1. 17 Muktitilaka, D 52.1-2; P 61b.1-2. 18 I here cite the verse as provided by Isaacson (2002a), which he has emended slightly from Matsunaga’s edition. 19 Rgyud phyi ma, D 152.2 289 An examination of these three verses shows that there are essentially two places where they differ significantly from one another—apart, of course, from the fact that Nāgārjuna’s verse is about the two truths while the later verses concern the two stages of tantric practice. One is in the verb used in the first pāda of the Sanskrit verse (which is rendered, in all three cases, as the second pāda in the Tibetan translation). In Nāgārjuna’s verse the verb is saṃupāśritya, and in the Samājottara’s verses, upāśritya. These verbs are synonymous—both mean “rely,” and are even translated identically into Tibetan as yang dag brtan; the variation in the Sanskrit appears to be based just on metrical considerations. The Muktitilaka verse varies slightly here, using the verb yang dag gnas, “abide,” rather than yang dag brten, “rely.” We unfortunately do not have the original Sanskrit of this verse, but I suspect that yang dag gnas may actually be a translation of samupāśritya (the longer verb would be necessary here for metrical purposes), given that this verb can also mean “to abide in.”20 Even if that is not the case, and if the Muktitilaka did use a different verb here, this minor difference does not significantly alter the meaning of the passage.21 The second difference that we notice in the verses pertains to who it is that taught the dharma in terms of the two truths or the two stages. Buddhajñānapāda’s verse follows Nāgārjuna’s verse in its use of the subject “the buddhas” (buddhānāṃ, sangs rgyas rnams) rather than the “vajra-possessors” (vajriṇam, rdo rje can) mentioned in the Samājottara verse. This difference is more significant and, supposing that one of the verses on the two stages is directly derived from the other,22 it would suggest that the Samājottara verse derives from Buddhajñānapāda’s, and not the other way around. Assuming that the original intent of the author(s) of this verse was to evoke Nāgārjuna’s words, leaving “the buddhas” as the agent of the dharma teaching, is significant. Moreover, making the change of the agent to the “vajrapossessors” can be seen as an increased “tantrification” of the verse. If one of the two verses on the two stages of tantric practice derives from the other, it is unlikely that the derivative verse would change the agent of the proclamation about the two stages from “the vajra-possessors” to “the buddhas;” the other way around is much more likely. Additionally, the Samājottara is a tantra, meaning that it is a scriptural source, considered within the tradition to be buddhavacana, the word of the buddhas, and therefore not something that a tantric exegete would be likely to willfully alter. Buddhajñānapāda does not hesitate to incorporate lines and verses from tantras directly into his writings without attribution. We see instances of the incorporation of lines from the Sarvabuddhasamāyoga-tantra23 and the 20 While yang dag brten is certainly a more expected translation of (sam)upāśritya, the Yogācārabhūmi attests to yang dag par gnas pa as a translation of samāśraya, which derives from the same verbal root, √śri, as samupāśritya. Thanks to Ryan Damron for pointing out this attestation to me. 21 The fact that this difference is so minor is more difficult to show in the English translation, because the use of the verb gnas has forced me to change the structure of the English in a way that makes this verse appear less parallel than it actually is. That is partly why I provided the Tibetan translation—even someone who does not read Tibetan can nonetheless see the parallels between the verses more clearly there than in my English translations. If indeed, as I suspect, yang dag gnas is translating samupāśritya, the English translation of the first two lines of the Muktitilaka would be precisely parallel with those lines from the Mūlamadhyāmikakārikās; they would then read: “The buddhas taught the dharma/ In reliance upon the two stages.” 22 It is, of course, possible that both the Muktitilaka and the Samājottara are relying upon an earlier source—either written or oral—that was itself based on Nāgārjuna’s verse. While this is not impossible, it seems preferable to try to work out the relationship between the materials at hand without positing a third, no-longer-extant source. And, indeed, I think we can presume it to be likely that one of these two verses on the two stages is based upon the other, and that whichever was the earlier of the two—which I take to be Buddhajñānapāda’s—was composed based on Nāgārjuna’s verse. 23 See Dvitīyakrama verses 50, 125, and 313 and my notes on these verses for the incorporation of lines from the Sarvabuddhasamāyoga-tantra. 290 Guhyasamāja-tantra24 directly into the Dvitīyakrama and the Caturaṅga/Samantabhadrasādhana, respectively. What Buddhajñānapāda does not do, however, is change those lines or verses that he incorporates—he may add to their meaning by preceding or following them with a line of his own that provides a different context than the one from the original tantra, but he does not alter the lines that he incorporates directly from these tantras.25 This fact, combined with the fact that there is a significant variant in the verses on the two stages in the Muktitilaka and the Samājottara, suggests that it is unlikely that the verse on the two stages in the Muktitilaka represents an instance where Buddhajñānapāda incorporated a verse from a tantra he already knew into his writings. Rather, it seems much more likely that the Samājottara’s verse came about on the basis of Buddhajñānapāda’s. Moreover, there is another passage in his oeuvre suggesting that Buddhajñānapāda may have enjoyed composing verses that echoed important verses from earlier non-tantric authored works, but with some modifications that made the verses more tantric. In his Ātmasādhanāvatāra, Buddhajñānapāda writes, with respect to the practice of deity yoga: tasmān nirastasaṃkalpaṃ samantaspharaṇatviṣam/ Samantabhadram ātmānaṃ bhāvayann eva bodhibhāk//26 Therefore, he who meditates upon himself as Samantabhadra, shining with full radiance, Having abandoned concepts, he alone partakes in awakening. This seems to be an intentional echo of the first verse of Dharmakīrti’s Pramāṇavārttika: vidhūtakalpanājālagambhīrodāramūrtaye/ namaḥ samantabhadrāya samantasphuraṇatviṣe//27 Homage to Samantabhadra who shines with full radiance And whose form, vast and deep, has cast off the net of thought. While the parallels here are not quite as striking as they are with Nāgārjuna’s verse above, it is nonetheless clear that they are intentional.28 Again, this suggests that Buddhajñānapāda may have had a proclivity for echoing earlier authored verses, with modification. Combined with the other evidence given above, this therefore makes it even more likely that Buddhajñānapāda himself composed the verse on the two stages of tantric practice on the basis of Nāgārjuna’s verse. 24 As I noted above, verses 46-49 of the Caturaṅga/Samantabhadra-sādhana, faithfully incorporates verses 72-75 from Chapter Seventeen of the Guhyasamāja-tantra. The Sanskrit for these verses in the Caturaṅga/Samantabhadra-sādhana fortunately survives and is edited in Kano 2014, allowing for a careful comparison of the verses in question. 25 See also, for example, Vaidyapāda’s incorporation of the verse on the three initiations (and the fourth!) from the Samājottara into his Yogasapta. Vaidyapāda incorporates the Samājottara’s verse wholesale, but then adds a final pāda which clarifies how he understands the final pāda of the Samājottara’s verse. Again, he does not alter the verse from the tantra itself. dbang ni rnam pa gsum dag tu// rgyud ‘di las ni rab tub shad// |6| bum pa’i dbang ni dang po ste// gnyis pa la ni gsang ba’i dbang// gsum pa shes rab ye shes dbang// de ltar de bzhin bzhi pa yang// bla ma’i bka’ las shes par bya// | 7| (Yogasapta, D 70a.3-4; P 103b.1-2) 26 Szántó unpublished, 147. I am grateful to Péter Szántó for sharing his draft Sanskrit edition of the Ātmasādhanāvatāra. 27 Pramāṇavārttika, 1.1. 28 I am grateful to Harunaga Isaacson for drawing my attention to the echoes of the Pramāṇavārttika in this verse. 291 Another verse from Buddhajñānapāda’s writings, this time from the Dvitīyakrama, has a parallel in the Samājottara, and yet again a comparison of the two suggests that Buddhajñānapāda’s is likely earlier. This verse occurs in the section of the Dvitīyakrama on the third initiation, and in the section of the Samājottara on the ritual for the vidyāvrata.29 Again we have only the Tibetan translation of the Dvitīyakrama verse to compare with the Sanskrit from the Samājottara. The verse from the Dvitīyakrama reads: gzhan kyis sangs rgyas mi nus pa// bu mo ‘di ni yang dag mchog// de bas mtha’ med ‘khor ba’i bar// khyod kyis ‘di dang bral mi bya// |89| Nothing else can bring about buddhahood This girl is genuinely supreme Thus, throughout endless saṃsāra You must never separate from her.” |89| The parallel passage from the Samājottara has an intervening two pādas between the first and second half of the Dvitīyakrama’s version of the verse: Nānyopāyena buddhatvaṃ tasmād vidyām imāṃ varāṃ// |125| Advayāḥ sarvadharmās tu dvayabhāvena lakṣitāḥ/ Tasmād viyogaḥ saṃsāre na kāryo bhavatā sadā// |126| Nothing else can bring about buddhahood, therefore this consort (vidyā) is the most excellent. All phenomena are nondual but they are marked by duality Therefore you must never separate from her [throughout] saṃsāra. As we can see here, apart from the addition of the two intervening pādas on nonduality, the main difference between the Dvitīyakrama’s verse and the one in the Samājottara, is, like in the passage we examined above on the two stages of tantric practice, simply the use of a different noun. (The other, quite minor, differences are, I believe, easily attributable simply to the translation of the verse from Sanskrit into Tibetan.) The Dvitīyakrama mentions a “girl” (bu mo,

• kanyā30) who is genuinely supreme, while the Samājottara uses the term “consort” (vidyā).31

While the latter is no doubt related to the fact that these verses in the Samājottara are part of the tathāgatas’ answer to a question about the vidyāvrata, the “consort observance,” posed by the bodhisattvas in the earlier part of the text,32 just like in the case of the use of “vajra-possessors” rather than “buddhas” in the verse we examined above, the use of the term “consort” (vidyā) rather than “girl” (*kanyā) amounts to a sort of “tantrification” of the verse. As such, it is more likely that the “more tantric” version of the verse is the later of the two. Also, as we saw above, 29 As we will recall from Chapter Seven, the Dvitīyakrama adheres to the paradigm suggested by Wedemeyer (in his 2014 paper at Berkeley, an updated version of which is forthcoming for publication in 2019) in which the vidyāvrata is not separated out from, but occurs as an integral part of, the third initiation. The Samājottara follows the paradigm where the vidyāvrata constitutes a separate ritual. 30 Bu mo is attested as a translation of kanyā. See Negi. 31 Vidyā literally means knowledge, but in tantric Buddhist texts it is frequently used as a term for the female consort. See Wedemeyer (forthcoming) for a helpful discussion of the development of the use of this term, especially with reference to the vidyāvrata. 32 “And likewise, what of the vidyāvrata?” kathaṃ vidyāvrataṃ tathā/ (Samājottara, 19b). 292 if Buddhajñānapāda were incorporating a verse from a tantra that he knew, like in the case of the lines and verses he incorporates from the Sarvabuddhasamāyoga-tantra and the Guhyasamājatantra, he would be likely to do so faithfully, without making a change to the text. In this case, moreover, the presence of an intervening two pādas about the nondual nature of reality in the Samājottara also suggests that if one of these passages is based upon the other33 the Samājottara’s passage is probably later than Buddhajñānapāda’s, as it is unlikely that Buddhajñānapāda would incorporate only part of a scriptural passage, leaving out the middle two pādas.34 A third piece of evidence that I believe suggests that Buddhajñānapāda was writing prior to the circulation of the Samājottara comes not in the form of a parallel passage, per se, but rather in the inclusion in Buddhajñanapāda’s writings of only part of a set of practices that are listed in full in the Samājottara: the practices of the six-branch yoga (ṣaḍaṅgayoga). I have already discussed this point in brief in Chapter Six, but I will review those arguments here. The Samājottara (vv. 141-154) appears to be the Buddhist locus classicus for the six-branch yoga, though versions on this system are found in many non-Buddhist traditions, as well.35 Francesco Sferra has shown the importance of the Samājottara’s presentation of the six-branch yoga in later Buddhist tantric literature by demonstrating that those works frequently cite the Samājottara and invariably draw the list of six yogas from that source, even when they describe the actual practice techniques for the yogas differently than their presentation in the Samājottara.36 Buddhajñānapāda’s presentation in the Dvitīyakrama of the second of the three bindu yogas of the perfection stage makes reference to just three—the third, fourth, and fifth—among these six yogas that are listed in the Samājottara, and refers to one of them by a different name than the standard term for that yoga used in the Samājottara. Given the popularity of the Samājottara as a source for these yogas that Sferra’s work has demonstrated, it seems unlikely that Buddhajñānapāda would have taught only three of its six branches, which usually come as a full set, had he known Samājottara. The three are, however, referred to in the Dvitīyakrama as “branches,”—they are even once mentioned as a set, “the three branches”—which may suggest that Buddhajñānapāda was familiar with a larger set of practices to which they pertained. But the fact that he refers to the first of the three branches that are mentioned in his system (which corresponds with the third among the six branches according to the Samājottara) with the unusual term, the “branch of emptying” (gtong pa’i yan lag),37 rather than the its more common name used in the Samājottara and later tantric literature, “breath control,” (prāṇayāma),38 33 That is, if they are not both drawing from some separate earlier source. Again, as I noted before, I believe it is better to try to understand the relationship of the sources that are available to us rather than to posit the existence of a theoretical third source. 34 These verses are included in a number of later sources—they are part of the section from the Dvitīyakrama many verses from which, as I documented in Chapter Seven, have been incorporated into many tantric liturgies for the third initiation and/or the vidyāvrata. Interestingly, most of the sources that cite this verse appear to follow the Dvitiyakrama, rather than the Samājottara, in the sense that they lack the two pādas on nonduality. 35 Sferra 1990, 15. 36 ibid. 37 As I noted in Chapter Six, there appears to be some confusion with regard to the name of this branch in Buddhajñānapāda’s system. While all of the available recensions of the Dvitīyakrama itself identify the practice as “the branch of emptying” (stong pa’i yan lag), Vaidyapāda Sukusuma reads “the branch of casting out” (gtong pa’i yan lag) (as does Vaidyapāda’s Samantabhadrī-ṭīkā, which also mentions the practice). Tāranātha’s later commentary on the perfection stage practices of Buddhajñānapāda’s system follows the Dvitīyakrama in reading stong pa’i yan lag for this practice—Tāranātha calls it the “branch of emptying which stops the breath” (dbugs dgag stong pa’i yan lag) (Dpal grol ba’i thig le’i khrid yig, 247). 38 I have already given in Chapter Six the reasons behind my identification of Buddhajñānapāda’s “branch of emptying (gtong pa’i yan lag) with the yoga of prāṇayāma, so I will not repeat them here. 293 suggests that he was unfamiliar with the Samājottara as a scriptural source. Had he known it, it seems quite likely that he would have chosen to hew to its terminology. Possible Evidence to the Contrary As far as I have been able to determine, there is but a single piece of evidence in his writings that might suggest that Buddhajñānapāda did know the Samājottara. That is a statement in the Muktitilaka made in reference to the arising of the signs that come about during the practice of the perfection stage yoga of the secret bindu. Buddhajñanapāda writes: Then, from the illuminated bindu, Following the previous procedure emanate, absorb, and [then] hold. When you reach this [realization] The five [signs] described in the tantra[s?] will occur.39 It is the final statement here, that mentions “the five signs described in the tantra[s?]” that might be interpreted as a reference to the Samājottara. These five signs that appear to the yogin as the elements dissolve into one another as a result of his perfection stage practices are not listed in the Guhyasamāja-tantra, but they are found in the Samājottara, and I am unfamiliar with their being set forth in a Buddhist tantra prior to the Samājottara. However, the five signs are also outlined in the Dvitīyakrama, where they form part of Mañjuśrī’s instructions to Buddhajñānapāda on the practice of the perfection stage. The question, then, comes down to what Buddhajñānapāda meant here in the Muktitilaka by “the tantra[s?].” Vaidyapāda seems to take this to refer to the Samājottara, or at least he explains it by citing the passage in the Samājottara that sets forth these five signs.40 This is not surprising; we know that Vaidyapāda knew the Samājottara—he composed an entire commentary on it. We also know that Vaidyapāda seems to have a preference for citing the Samājottara over Buddhajñānapāda’s works in places where they have parallel content, and where we know that Vaidyapāda knew both texts, since he wrote commentaries on both. For example, Vaidyapāda cites the passage from the Samājottara on the two stages in his Sukusuma (a commentary on Buddhajñanapāda’s own Dvitīyakrama!), even though we know that Vaidyapāda knew the Muktitilaka passage on the two stages, since he also wrote a commentary on the Muktitilaka.41 However, it makes sense that Vaidyapāda would give preference to a scriptural citation over a non-scriptural one when he had a choice between two passages with the same content. Therefore, even Vaidyapāda’s citation of the Samājottara in reference to Buddhajñānapāda’s statement about the “five signs described in the tantra[s?]” does not necessarily constitute evidence that the Samājottara was indeed Buddhajñānapāda’s referent here. Were this passage from the Muktitilaka the single indication of a relationship between Buddhajñānapāda’s writings and the Samājottara, I would indeed probably take the passage as a reference to the five signs as set forth in the Samājottara, specifically, and understand that text to be the referent of “the tantra[s?]” here. However, given the other evidence I have cited above suggesting that the Samājottara circulated only after the period when Buddhajñānapāda was writing—evidence that I myself, at least, find convincing—I believe it is unlikely that this 39 de la snang ba’i thigs pa las// gong gi rim pas spro bsdu bas// bzung ste de la reg ‘gyur bas// rgyud gsung lnga po yang dag ‘byung// (Muktitilaka, D 49a.6; P 69a.4) 40 Muktitilaka-vyākhyāna, D 53b.2-3. 41 Vaidyapāda’s Muktitilaka-vyākhyāna, though, makes no mention of the Samājottara’s parallel passage on the two stages in commenting on that passage in the Muktitilaka. He simply introduces it by saying, “Now, in order to teach the distinction between the two stages he writes…The buddhas taught the dharma…” da ni rim pa gnyis kyis dbye ba bstan par byas pa’i phyir sangs rgyas rnams kyis zhes pa la sogs pa’o// (Muktitilaka-vyākhyāna, D 58b.2- 3). 294 reference to “the tantra[s?]” in the Muktitilaka is, in fact, a reference to the Samājottara. I am at a loss, however, for providing a satisfying answer regarding the work to which Buddhajñānapāda may have been referring here.42 Some Initial Conclusions Despite this one perplexing passage from the Muktitilaka, I still believe the preponderance of the evidence suggests clearly that Buddhajñānapāda did not know the Samājottara and that his writings preceded, and indeed influenced, the Samājottara. If true, this would give a rather narrow window for the circulation of that tantra, given that Vaidyapāda, who did know the Samājottara, was likely a direct disciple of Buddhajñānapāda’s. As I discussed in Chapter One, I believe that Vaidyapāda likely became Buddhajñānapāda’s disciple when the latter was already well on in years and Vaidyapāda was still young, since it seems that Vaidyapāda was also a disciple of some others among Buddjñānapāda’s direct disciples, including Dīpaṃkarabhadra and perhaps also Praśāntamitra. Nonetheless, we are speaking of just a matter of decades in the early-to-mid part of the 9th century as a window for the initial circulation of the Samājottara, perhaps some time between 830 and 850. If my contention that Buddhajñānapāda did not know the Samājottara is true, this would mean that many elements already found in his writings made their way into the later Buddhist tantric tradition through their articulation as buddhavacana in that tantra. This includes the two stages of tantric practice, the rituals for the guhyābhiṣeka and the prajñājñānābhiṣeka (and for the vidyāvrata, for that ritual was separated out from the prajñājñānābhiṣeka in the Samājottara), the practice of “the fourth,” and perhaps even the inclusion of the five signs in connection with perfection stage practice. I certainly do not mean to argue that all of these are Buddhajñānapāda’s own innovations, simply that they are elements that seem both not to be found in the extant tantric literature prior to his writings, and to have first entered Buddhist scripture in the Samājottara. As I noted at the beginning of this chapter, I have not yet completed a study of the Samājottara. A full study of that work in comparison with Buddhajñānapāda’s writings will certainly turn up more evidence—one way or the other—on the relationship between the two. Nonetheless, I feel that the passages I have analyzed here provide enough evidence to essay these initial conclusions on the direction of influence, and to suggest that it was through the Samājottara that many important elements found in Buddhajñānapāda’s thought and ritual systems began to make their way onwards and outwards into the world of later Buddhist tantric systems. 42 Could it be that he is referring to Mañjuśrī’s speech as recorded in the Dvitīyakrama itself as a “tantra”? This seems implausible, but perhaps not impossible? 295 Concluding Reflections Through the blessings of the sahaja ācārya and the great compassionate revered master, one encounters bliss, through which undeceiving truth is realized, just as it is: the supreme, great pure essence of all things, the drop which is the sixteenth part, achieved through resting, the great instruction —may you come to encounter this! -Buddhajñānapāda, Dvitīyakrama Over the course of these chapters we have met and gotten to know—at least a little bit— the yogin and tantric exegete Buddhajñānapāda, whose extraordinary autobiographical record in combination with his other surviving writings have enabled us to behold a much more intimate portrait of an 8th-9th century Indian figure than is ordinarily available to us, and to open a window into a period of tantric Buddhist history that has otherwise remained obscure. As we have seen, his life was both ordinary and extraordinary. Buddhajñānapāda’s travels throughout the Indian subcontinent meeting and receiving instruction from different teachers and putting their teachings into practice hint at the perhaps ordinary life of a mendicant tantric Buddhist yogin of the time. However, his visionary encounter with his most exalted guru, who was none other than Mañjuśrī, and Buddhajñānapāda’s recording and passing on in the Dvitīyakrama of the instructions received from that master—within an autobiographical frame narrative, no less— indicate his life as something beyond the ordinary. Making use of the unique genre of a mukhāgama and recording Mañjuśrī’s instructions directly as they were delivered, in Mañjuśrī’s own words, allowed Buddhajñānapāda to set forth a system of doctrine and practice that essentially had scriptural legitimacy. Mañjuśrī’s command to Buddhajñānapāda to compose further related works gave nearly that same seal of authenticity to the remainder of Buddhajñānapāda’s oeuvre. The contents of his writings, as we have seen, include many of the major doctrinal and ritual developments of the early period of what is often referred to as “late,” or “mature,” Indian Buddhist tantra, thus shining light on the still obscure period in which these traditions began to emerge. Aspects of the vocabulary, ideas, and ritual practices in his works show that Buddhajñānapāda was writing in an eclectic milieu in which he had contact with and knowledge of both non-Buddhist philosophical and religious systems, like the Sāṃkhya and Śaiva traditions, as well as secular traditions like that of kāmaśāstra. Buddhajñānapāda’s writings are among the early tantric works to place a strong emphasis on nondual wisdom, which came to hold a major doctrinal role in later tantric traditions. He gave the term a tantric resonance, describing it as the “wisdom of the nonduality of the profound and the luminous,” the empty aspect of the mind integrated with its expression as the illusory form of the deity. This nondual wisdom was, for Buddhajñānapāda, the very identity of the mind and of all phenomena, and even served as the source of the phenomenal world. Nondual wisdom, identified with the awakened state, suchness itself, was something that could be—and indeed had to be—"transferred” by a guru to his disciple during the higher tantric initiations, so that a disciple could experience it directly and cultivate it by means of yogic practice, in order to come to the final state of perfect awakening. Because of the unique methods that allowed the disciple to directly “receive,” during initiation, and train in suchness by means of the sexual yogas of the perfection stage, this second stage of Buddhist tantra was, for Buddhajñānapāda, a superior path that led to a result higher than that 296 attainable through the practice of the exoteric Mahāyāna, or even through the practice of the lower tantras. In advocating non-action, while simultaneously composing elaborate ritual liturgies for the lower stage of tantric practice, Buddhajñānapāda appears to have understood the path of non-action to pertain specifically to the perfection stage, in which the complex outer rituals of the earlier stage of tantric practice were set aside. However, he may also have been referencing the conceptual non-action of the “perfection stage of the perfection stage,” identified by Vaidyapāda with the “great perfection,” nonconceptual wisdom itself, free from any sort of activity or elaboration. Though the use of the term “the great perfection” (rdzogs pa chen po) in the Tibetan translations of his writings most likely does not represent Buddhajñānapāda’s use of a semantically equivalent Sanskrit term, many of the doctrinal claims in his works are indeed parallel with the doctrines evident in early works of the Great Perfection. His writings therefore represent an Indic system that accords in many ways with the early tradition of the Great Perfection that was emerging around the same time in Tibet. As an early author to make the division of tantric practice into two stages, and perhaps the earliest to have composed still-extant manuals of practice for both, Buddhajñānapāda’s writings provide us an invaluable window into the early ritual articulations of both the generation and perfection stages. His generation stage sādhanas, the most important of which, the Caturaṅga/Samantabhadra-sādhana, also includes practices connected to the perfection stage, indicate that for Buddhajñānapāda the generation stage served both the function of achieving desired worldly aims, as well as the soteriological function of acting as a framework and support for the liberative practices of the perfection stage. A distinctive feature of the Caturaṅga/Samantabhadra-sādhana, in which the constituent aspects of world and beings are purified through consecration in the maṇḍala and re-emerge in the form of maṇḍala deities indicates that Buddhajñānapāda not only employed generation stage sādhana in its usual function as a method of reconstructing personal identity in the purified form of the deity, but also as a ritual re-enactment of doctrines set forth in the Guhyasamāja-tantra that viewed the entire phenomenal world as divine. For Buddhajñānapāda, though, it was only through the yogas of the second stage, or the perfection stage—terms that he used synonymously with suchness itself—that a yogin could reach the final fruition of perfect awakening. The three bindu yogas of the perfection stage, trained in with a consort, were the means by which a practitioner of Buddhajñānapāda’s system of the perfection stage came to repeatedly experience the three blisses in sexual yogic practice, and therefore to repeatedly cultivate the nonconceptual experience of suchness itself, leading to final awakening. Several passages from Buddhajñānapāda’s writings on these perfection stage sexual yogic practices suggest that he was familiar with kāmaśāstric traditions, and the Dvitīyakrama, in its discussion of the types of tantric consort, contains what appears to be the earliest instance in an extant Indian text of the classical four-fold typology of women so important in later kāmaśāstra. For a yogin who was not able to awaken in this lifetime by means of the perfection stage bindu yogas, Buddhajñānapāda’s system provides the “failsafe” option of the practice of utkrānti, the yogic ejection of consciousness at the moment of death. Engaging in the liberative practices of the perfection stage, including the yoga of utkrānti, was only possible for a practitioner who had received the higher tantric initiations through which he was first introduced to the “suchness of the second stage” that was cultivated by means of perfection stage practice. These initiatory rituals themselves involved sexual practice, and therefore necessitated the participation of a consort. Buddhajñānapāda’s writings include early descriptions of the rituals for several of the higher initiations, including probably the earliest extant description of the prajñājñānābhiṣeka, in which a yogic couple were introduced to the suchness of the perfection stage by means of sexual yogic practice conducted 297 under the tutelage of a guru who held the lineage of these special liberative oral instructions. Buddhajñānapāda’s writings hint at the practice of bestowing a verbal instruction on suchness immediately after the prajñājñānābhiṣeka, and Vaidyapāda’s Yogasapta gives much more detail on this bestowal of what seems, at the time, to have been primarily referred to just as “the fourth,” an instruction on the so-called seven yogas that constitute seven different aspects of the state of awakening. The experiences of these seven aspects during the prajñājñānābhiṣeka were intensified or stabilized through the sexual yogic practices of the perfection stage until, for a yogin who had progressed to the “third level” of post initiatory practice, they were fully realized in the moment of perfect awakening. In including the guhyābhiṣeka, the prajñājñānābhiṣeka, and quite possibly also “the fourth,” Buddhajñānapāda’s initiatory system already contained all of the main elements of the higher initiations that came to characterize the initiatory sequence of the “mature” form of late Indian tantric Buddhism. This study of his writings and thought shows that Buddhajñānapāda was, I believe, an even more influential figure than has been heretofore acknowledged. In addition to the fact that his works serve as source texts for the Jñānapāda School of Guhyasamāja doctrine and practice, the early exemplars of many of the ritual practices and ideas found in his work seem to have formed the basis for their later iterations in quite a number of later systems of tantric practice. We see verses from his writings incorporated into no less than fourteen later tantric works, both authored works and scriptures, and in particular the ritual structures that appear in early forms in his writings are found throughout later tantric Buddhist literature. The Samājottara, it seems, may have been an early vehicle through which many of these ritual and practical frameworks made their way into the later tantric tradition. Buddhajñānapāda’s writings are therefore important both for showing us a picture of this unique individual, as well as giving us a window into his world—the world of Indian tantric Buddhism at the close of the 8th century and the opening of the 9th. Yet what I have been able to show here, both of the individual and of his world, is just a preliminary glimpse at both. As I have emphasized repeatedly throughout the dissertation, essentially every topic addressed here deserves further inquiry and analysis. As the first book-length study of Buddhajñānapāda’s life, writings, and thought, what I have written here cannot be more than a first look into these matters, and it is my hope that this dissertation will serve as the basis for further study of Buddhajñānapāda and his oeuvre. There are a several specific points brought up in my study, however, that I would like to mention here as particularly worthy, or in need, of further research. I have only been able to barely touch on Buddhajñānapāda’s philosophical perspectives, and these—especially when taking into consideration his non-tantric writings, which I have not considered at all here—certainly warrant further study. There are also surely other aspects of his doctrinal positions that I did not think to mention, as I only selected for discussion a few of the points that stood out to me, personally, as particularly interesting and worthy of comment. An inquiry into Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana in relation to other early generation stage sādhanas, and indeed a study of the early development of sādhana as a genre would give scholars a much-enhanced perspective on tantric ritual. But it is particularly in terms of his perfection stage system and the related higher tantric initiations where I believe further study of Buddhajñānapāda’s writings is needed. In trying to understand these practice systems, which—like all of the material studied in this dissertation—I wanted to present in the context in which they were understood in Buddhajānapāda’s own time, I strove not to be unduly influenced by the presentations of these systems according to the later tradition, in which they had certainly undergone some development and change. However, it has become increasingly clear to me that without reference to the later tradition we simply do not have enough materials to fully understand and appreciate these earlier traditions. Further study of these materials with more 298 reference to the later developments in perfection stage and initiatory practices will certainly shed further light on the early iterations of these practices and ritual systems. In particular, with reference to the early understanding of what constituted “the fourth,” an in-depth study of Vaidyapāda’s Yogasapta will certainly be illuminating. Buddhajñānapāda’s world, that of tantric Buddhist north India at the turn of the 9th century, was one in which we find the intricacies of Yogācāra and Madhyamaka philosophy right alongside visionary encounters with Mañjuśrī and the practice of sexual yogas. Buddhajñānapāda’s writings resist our tendencies to categorize, including philosophical analysis right in the middle of practical ritual sequences, and even blurring the boundary between scripture and authored treatise. His works reveal to us a world where reality itself could be received from the words of a guru, with the assistance of the bliss that arose in union with a tantric consort, and where that very suchness itself could be cultivated and actualized, bringing about a uniquely tantric experience of Buddhist awakening. The yogic practices that were being newly developed in his time—cutting edge techniques involving the manipulation of internal winds and energies—were viewed as important, crucial even, to the attainment of that final awakening. Buddhajñānapāda draws us into this world in a direct and personal way through his use of autobiography, allowing us unusual access into a very human dimension of a world that might otherwise seem very remote, opening a door for us and making—for a moment at least— his world our own. 299 Part II: Tibetan Edition and English Translation of Buddhajñānapāda’s Dvitīyakramatattvabhāvanā-mukhāgama 300 The *Dvitīyakramatattvabhāvanā-mukhāgama of Buddhajñānapāda: A Critical Edition of the Tibetan Translation by Kamalaguhya and Lha Yeshe Gyaltsen This Tibetan critical edition is based on all available recensions of the Tibetan translation of the

• Dvitīyakramatattvabhāvanā-mukhāgama, from the Cone, Derge, Narthang, Peking, and

Sertrima Tengyurs. Significant variants are noted in the footnotes, whereas insignificant variants have been relegated to the endnotes. Vaidyapāda’s Sukusuma (again available only in Tibetan translation) sometimes, but not always, provides direct citations of the Dvitīyakrama. I have recorded variants in direct citations of passages in the Dvitīyakrama that appear in Vaidyapāda’s Sukusuma (from both the Derge and Peking editions of the Sukusuma) only when 1. there is a significant variant in at least one recension of the Dvitīyakrama itself, and Vaidyapāda’s commentary includes a direct citation of the passage (i.e. dzā] D C V(D), dza P N S, ‘dza’ V(P)); or 2. there are no variants in the recensions of the Dvitīyakrama, but I suggest emending the text based on its direct citation in Vaidyapāda’s commentary (i.e. bar du] sugg. em. based on V (D and P), rab tu D C P N S). There are other cases where Vaidyapāda does not directly cite a particular passage, but I suggest an emendation based on the way he comments on, or glosses, that passage. In those cases I provide the passage from the commentary upon which I based my suggestion emendation (i.e. rgyu] sugg. em., rgyur D C P N S. Vaidyapāda’s commentary seems to support this reading: rgyu ba dang mi rgyu ba ni snod dang bcud do//). There are a number of lines and verses from the Dvitīyakrama that are incorporated from earlier works and/or incorporated into later works, both scriptural and commentarial. I have referenced the parallels of which I am aware (certainly there are more, of which I am not yet aware) in the notes to my translation, rather than the notes to this edition. Abbreviations: D Derge Tengyur C Cone Tengyur P Peking Tengyur N Narthang Tengyur S Sertrima Tengyur V Vaidyapāda’s Sukusuma commentary on the Dvitīyakrama V (D) Vaidyapāda’s Sukusuma, Derge Tengyur V (P) Vaidyapāda’s Sukusuma, Peking Tengyur 301 /rgya gar skad du/ dvia kra ma ta ttva bhāb wac nad nā ma mu khāe gaf mag/ bod skad du/ rim pa gnyis pa’ih de kho na nyid sgomi pa zhes bya ba’i zhal gyi lung/ thams cad mkhyen pa la phyag ‘tshal lo// dpal ldan bde rdzogs zab gsal gnyis med ‘od zer ldan// de yi rang bzhin zhi ba bcu drug phyed cha ‘bar// don dam snying po de ston bla ma gsum po rab// de la bya byed gsum mnyam pa yis rtag tu ‘dud// |1| ‘jig rten gsum gyi sgron ma kun gyis rab bsngags pa// chos rnams kun gyi snying po dngos po’i de bzhin nyid// srid pa’i dug chu zlog byed gsum khong waṃj steng ‘bar// ‘jam dbyangs bla ma’i lung gis rtogs phyir bshad par bya// |2| dbu kyi yul chen kha pirk grong khyer rdo ‘jog tu// bzang po seng ge zhes byar grags pa rab thob pa// bla ma de ni mnyes byasl lung thob gzhung mang thos// de la rnam dpyad rig ‘byung shrīm na lāndarn// |3| yon tano bshes gnyen zhes bya rigs can de yi ngor// blun blosp rab tu byed pa phyogs ‘ga’ rtsom byed pa’i// brod pas der gnas rnams la gzhung des phan gdags par// bsams nas der gnas rtsom dang ston sogs rab tu byasq// |4| de nas yon tan kun ‘byung u rgyanr yul du bgrod// a dvi] P N S, dva D C. I believe that this is likely a mistaken Tibetan back-translation of the Sanskrit title of the work, which should have here read dvitīya, rather than dvi. See Note 3 in my Dvitīyakrama translation, and Chapter Two, where I discuss this point in more detail. b bhā] D C, bha P N S c wa] D C, ba P N S d na] D C, sa P N S e khā] D C S, khī P N f ga] D C, saddhya P N S g ma] D C, nya P N S h pa’i] P N S, ba’i D C i sgom] P N S, bsgom D C j waṃ] D C, dam P N S k pir] D C S V (P) , bir P N V (D) l byas] D C P N, S om. m shrī] P N S, shī len D C; The reading from S, P, and N is unmetrical, but more correct Sanskrit. The readings from D and C are metrical, but otherwise unsatisfactory. n lāndar] P N S, lendrar D C o Here and in a number of places S, as is common with manuscript versions of texts, uses abbreviations. Since these abbreviations are obviously indicating the same reading as the other recensions, I will not continue to note them. Moreover, there is not anyway, to my knowledge, a convention for transcribing such abbreviations using roman letters—the Wylie system does not really work for doing so, and even Tibetan typing programs will not allow one to stack a ya upon a ta capped by a naro and followed by na—all as a single syllable (yton?)—which is the abbreviation used here for yon tan. p blos] D C, bro P N S q byas] D C, bya P N S r S abbreviates to oyan. 302 ‘jo sgeg rdo rje zhes byar mtshan gsol rab tu grags// de la mang thos rnam par dpyad cing de nyid du// gu ne rua grags bla ma de yang mnyes byas te// |5| lung thob gnas de’i byang phyogs chab sgo bdag// dzāb thig dzāc lād zhes bya’i bu mo bcu drug pa// lakṣmī chen mo mnyes byas zla ba brgyad kyi bar// de yi lung ni rab thob mnone nasf grub pa thob// |6| de nas dzā lendhag rar grong khyer ko no dzer// phyin nas bā li pā dah zhes byar grags pa rab thob gang// mnyes byas gzhung thos lung ni mang du thos gyur nas// lho phyogs nam mkha’ shing ldan kong kai na ru bgrod// |7| grub pa’i dbang phyug bā li pā tarj rab grags pa// rdzu ‘phrul ldan pa’i slob ma’i tshogs dang rab tu bcas// de kun yo byad gos zas nor rgyun gyis sbyor ba// bla ma dam pa de drung lo dgu rab tu btud// |8| ‘dus pa’i rgyud chen ‘grel bcas bco brgyad bar duk mnyan// bdag gis ma rtogs bla ma chen pos de1 bzhin gsungs// ‘dil ma rtogs par gzhan ni don med par bsams nas// glegs bam mgul btags byang phyogs yul du rab bgrod de// |9| rdo rje gdan rgyab ku ba rtsam zhes bya ba’i tshal// stag dang dred sogs mang zhing shin tu ‘jigs pa’i sar// zla ba drug gnas pa yisn chos rnams de nyid rtogs// sprul pa’i dge slong bla ma gnyis dang bcas dang phrad// |10| mgo dang lag gnyis yol dang tsheso brgyad rgyal la bab// ston zla ra ba’i tho rangs skya rengs shar dus su// ‘jam dpal2 dbyangs kyip dkyil ‘khor ‘khor loa sprul ba la// a ru] S P V(D and P) , nu D C N. b dzā] D C V (D), dza P N S, ‘dza’ V (P) c dzā] sugg. em. based on V (D); dza D C P N S, dzva V (P) d lā] sugg em; la D C P N S e mnos] D C, gnon P N S f nas] D C, gnaP N S S g lendha] D C, lāndha P N S h bā li pā da] D C, ‘ba’ mo pa ta P N S. Vaidyapāda’s commentary has the name translated as byis pa chung ba’i zhabs which supports the reading from D and C. i ka] D P N S, kaṃ C j bā li pā dar] D C, ba li pa tar P N S. We now know from Sanskrit sources that this guru’s name is Pālitapāda, but to include his full name would be problematic for the meter so I have left it here. k bar du] sugg. em. based on V (D and P), rab tu D C P N S. l ‘di] sugg. em. based on V (D and P), ‘dir D C P N S. m rtsa] D C, tswa P N S n yis] D C, yi P N S o tshes] C P N S, ches D p kyi] P N S, kyis D C 303 bdag gis don ‘di blang phyir gsol ba rab tu btabb// |11| khyod ni sems can kun gyi yab ste yum yang yin// khyod kyis bdag la sogs te ‘jigs pa che las skyobs// ‘gro3 ba’i dbang po rje btsun khyod kyis sdug bsngal sel// khams gsum stong byed chec ba che khyod skye bo skyongs// |12| thog ma med khyod nga ro thigs pa’i steng cha med// je btsun yigd me bdud rtsi rab dga’ bde stonge byed// ‘gro la phan phyir bde baf sangs rgyas rnams kun la// rab dgyes bde bag mgon chen khyod kyis gtongh bar mdzod// |13| skyon gyes ma gos bla na med pa’i byang chub lam// sna tshogs sdug bsngal zhi mdzad skom pa ngoms mdzad cing// ‘khor ba dba’i klong thar mdzad bde la ‘god mdzad paj// dngos kun rgyal bas mi dpogs lam ‘di bstan par mdzod// |14| zhi ba’i zhi ba rdul dang rab bral ‘jungsk pa spangs// bsgom pa las ‘das mkha’ ‘dra dri mas ma gos pa// sgyu ma kun ‘das ‘dod pa kun la rgyab kyis phyogs// de ‘dra’i rje btsun khyod la bdag ni skyabs su ‘chi// |15| dregs med ‘jigs med byang chub chen po thob kyi bar// ‘khor bar ‘tsho bar ‘jig rten dmyal rtsub ma ltungl bar// khyod gsung dam tshig sdom pa thun mong ma yin rnams// bdag la sogs pas rtag tu rab tu bkyangs bar bgyi// |16| gar chen rnam par rol pas gar mdzad cing// sna tshogs phyag ni gcum dang bsgrimsn pa yis// padma’i lo ma ‘jam pa brgyad phye nas// gnyis med bde rgyu rdo rje gzhag par mdzad// |17| gsang ba’i de nyid mi nyams gsal bar ‘gyur// rdo rje ‘dab skyes zla ba’i yang dag ‘du// thabs shes las byung chos kun de nyid mchog// rje btsun [3a] bdag la phan phyir sbas pa shod// |18| a lo] P N S, lor D C b btab] D C, gdab P N S c che] N V, chi D C, cha S P d yig] P N S, yid D C e stong] sugg. em based on V (D and P), gtong P N S, btang D C f ba] sugg. em based on V (D and P), ba’i P N S D C g ba] sugg em based on V (D and P), ba’i P N S D C h gtong] P N S V (D and P), btang D C i dba’] D C, rba P N S j pa] D C S P, cing N k ‘jungs] P N S V (P), ‘jum D, ‘dzum C; ‘jums, V (D) l ltung] P N S V lhung D C m gcu] V, bcu D C P N S n bsgrims] P N S V, bskyings D C 304 de nas ‘jam dbyangs byang chub sems dpa’ chen po yis// bdag la ‘dzum pa’i bzhin bltas legs zhes lan gsum gsungs// rol pa’i gar dang chos kun de bzhin nyid// sgra brnyan lta bur rdo rje glu yis bdag la bstan// |19| a bi yaṃ raṃ baṃ laṃ hūṃa a la la lab ho// rje btsun thugs rje che rnams kyis// ‘di rtogs pa yis bde gshegs kyi// phun sum tshogs pa ‘thob ‘gyur pa’i// |20| ‘das dang da ltar ma byon pa’i// rdor rje ‘dzin pas snod ‘ga’ la// gsungs shing gsung dang gsung ‘gyur ba// yang dag don rab rab rtogs phyir// |21| nga yis khyod la bstan par bya// yid gcig bsdus la mnyan par gyis// |22| chos rnams gszugs la sogs pa rnams// kun mkhyen bar gyi rang bzhinc ni// nam mkha’ dkyil ltar rnam dag pa’i// zab gsal gnyis med ye shes te// |23| de ni dngos med bsgom dngos med// dngos po thams cad dang bral ba’i khams dang skye mched kyis ma zin// rang bzhin gyis ni ‘od gsal ba// |24| ‘dod nas dag pa nam mkha’ bzhin// med nas chos rnams mtshan nyid spang// chos dang chos nyid yang med de// dngos po med pas mkha’ dang mtshungs// |25| tshig dang yi ge kun las grol// de ni dus dang phyogs rnams dang// chos kun gyi ni ngo bor song// lus ma yin te ngag dang ni// |26| yid kyang ma yin ‘dod khams dang// gzugs dangd gzugs med khams ma yin// ‘byung chen bzhi yang ma yin te// de bas de ni gang na yang// |27| a +// S P D b a la la la] P N S V(D and P), a la la D C c bzhin] P N S C, gzhan D d S om. 305 mi gnas pas na mnyam pa nyid// de ‘dra rdo rje ‘chang chen po// chos rnams kun gyi chos nyid mchog// thabs kyis bsgrub par bya ba’i dngos// |28| brtag pa kun las rab dben pa// phun sum tshogs rgyu dpag dka’ ba// phyag rgya chen por rab snang ba’i zer gyis rang gzhana smin byed pab// |29| de gnyis med pa’i rang bzhin mchog// rdo rje ‘dzin pa chen pos kyang // brjod du med pa’i sku chen po// de nyid rgyal sras zhing bcas dang// |30| srid gsum rgyu dang mi rgyuc bcas// thams cad kun gyi bdag nyid de// dngos kun gtso bo [3b] rang sems kyid// ngo bo nyid de de rtogs na// |31| sangs rgyas byang chub de nyid do// ‘jig rten gsum yang de nyid do// ‘byung chen rnams kyang de nyid do// ci yi phyir na chos thams cad// |32| sems la gnas te de nyid kyang// nam mkha’ la gnas nam mkha’ ni// gang du min gnas ‘od gsal ba// gzod nas rnam dang stong pa’o// |33| sangs rgyas mya ngan ‘das pa’i khams// skye med rdo rje mngon byang chub// bder gshegs kun kyi snying po mchog// gnyis med rtog bral don chen te// rim pa gnyis par rab tu bshad// |34| de ‘dra’i don des dngos po kun// rnam par khyab stee rnam gnas kyang// thog med dus nas de las kyang rnam rtog chen po tshul byung ste// |35| a gzhan] D C V(D), nyid S P V(P). Vaidyapāda’s comments on the verse (rang phul du byung zhing gzhan rgyud dang par byed pa’o//) also make it clear that he was reading rang gzhan, despite the fact that the actual citation of the verse in the Peking edition of his commentary reads rang bzhin (Sukusuma, 95b.7). b pa] P N S, pa’i D C. Vaidyapāda’s commentary (Sukusuma, 95b.7) also seems to support P, N, and S. c rgyu] sugg. em., rgyur D C P N S. Vaidyapāda’s commentary seems to support this reading: rgyu ba dang mi rgyu ba ni snod dang bcud do// d kyi] D C, kyis S P e ste] D C, te P N S 306 de lasa yang ni ‘byung ba che// rlung gi dkyil ‘khor nyid byung ste// de las kyang ni me yi khams// chen po byung nasb khyab mdzad de// |36| de las chu khams chen po yang// byung ste khyab mdzad de las kyang// sa byung bhzi bsdus ngo bo las// ri sogs sems can thams cad kyang// |37| sna tshogs phra ba sbom po dang// skyes pa bud med ma ning dang// gzhon nu dang ni rgan po dang// lha dang klu dang gnod sbyin dang // |38| gdon dang skar ma gshin rje dang// chu bdag rgya byin dmyal ba dang// yi dags dud ‘gro dang de kun// spang dang shes tsam rab brten pa’i ‘gro ba kun du rgyas par gnas// |39| de bas de yi gong ma yi// gnyis med rtog bral rab bsgribsc te// ma rtogs pas na ‘gro ba kun// ‘khor bar rab tu ‘khor bar ‘gyur// |40| de ni ‘khor ba’i sdug bsngal che// dug chu lta burd ‘khrul pa yis// rnam bsgrub thag pa’i sbrul lta bur// snang zhing med pa nyid du ‘dod// |41| de bas gnyis med don chen po// rang rig dkyil ‘khor chen po mchog// rgyun dang rnam grangs kyis gnas pas// de yi ngo bo rtogs par bya// |42| gang zhig de rtogse rab ‘dun pa// sems kyi smon lam yang dag can// sngon du bsod rams mthu bskyed dang// skyes bu yang dag gzhif la brten// |43| sbro dang gusa dang bsnyen bkur dang// a las] D C, la P N S b nas] D C P S, gnas N c bsgribs] P N S V(P and D), bsgrims D C d bur] D C, bu P N S e rtogs] P N S C, rtags D f gzhi] sugg. em. based on V (D and P), bzhi D C P N S 307 ‘dun bar ldan zhing sgom pa dang// brtan po som nyi rnam spangs pa4// rjes mthun gtong phod des shes ‘gyur// |44| [4a] de yis rje btsun yang dag pa// mchog gi gdamsb ngag rgyud la ldan// theg chen lam gyi spyod pa dang// bsgom pa la ni lhur len dang// |45| de nyid bcu la sogs pa yi// gsang dang rab gsang shin tu gsang// shes shing de yi don gab par// ston byed bla ma mnyes par bya// |46| yul dang mkhar dang khyim dang rta// glang po che sogs mal stanc dang// chung ma sdug dang yid ‘ong bu// bu mo sring mo tsha mo sogs // |47| gzhan yang gser dang dngul gyi rdzas// zangs lcags sogs las byas pa yid// mdzes pa’i rdzas dang mu tig gi// phreng dang padma rā ga dang// |48| pu shele indra nī la dang// mar gadf g.yu sogs mchod pa’i tshogs// zhing ‘di mkhas pas rab bkang ste// de ‘dra’i rje btsun dbang po mchod// |49| gsal shing rab dga’ mkha’ mnyam pa// gzhan du5 rigg par mi ‘gyur bas// bud med sgyu ma’i phyag rgya ni// sgyu ma kun las khyad par ‘phags// |50| sgyu ma de yang ‘jig rten ‘dir// spyan la sogs par rnam dag pash// rigs ni rnam pa bzhir ‘gyur te// ming dang mtshan nyid bshad par bya// |51| a gus] C P N S, kus D b gdams] D C, gdam P N S c stan] D C P S, bstan N d yi] P N S, yis D C e shel] sugg. em., shar D C P N S. This suggested emendation is based on the oral commentary of Khenchen Chodrak Tenphel who suggests that the text should read pu shel, “amber,” rather than pu shar, which does not yield any sense (Khenchen Chodrak Tenphel, personal communication, February 2016). Vaidyapāda’s commentary does not mention the term at all. f mar gad] P N S, ma rgad D C g rig] P S V(P), rigs D C N V(D) h pas] P N S, par D C 308 ka ma līa b dang shangkhīc nīd// tsi trie nīf dang ha stīg nīh// dang bo klu mo’ii rigs yin te// gnyis pa stag dang seng ge’ij rigs// |52| gsum pa ri dwagsk kṛṣṇal tsogsm// bzhi pa glang po che rigs so// de la ka ma līn yi ni// dbyibs dang mtshan nyid bstan par bya// |53| bu mo padma’i dri bro zhing// ngo zlum sna rtse til ‘dra ste// seno mo dmar zhing rgyab sgur dang// rkang mthil kun gyis sa la reg |54| pa spu ‘khyil ldan serp mo can// dkar ‘chang til gyi ‘bras bu ‘dra// gsus pa’iq gnyer ma gsum ldan zhing // brang mdzas glang chen ltar ‘gros dang// |55| rākta’ir ro ni skyurs ba’o// sha mdog dmar te lha mo ‘di// mā ma kī yis rnam par dag// shangkhīt nīu yi mtshan nyid dang// |56| a lī] D C, li P N S b All versions of the text read simply kamalī, which is an unusual and unexpected form. Kamalinī would be the expected feminine form that would correspond with the other names given in the verse. Is is possible that that Tibetan translators may have simply shortened the form for metrical reasons. Vaidyapāda’s text also reads kamalī, but this may again simply be because the translators of the commentary were referencing the Tibetan translation of the root text. I have not taken the liberty of changing the text in my edition, however, as this would render unmentrical all verses in which it occurs. c shangkhī] D C, shang ki P N S d nī] sugg. em., ni D C P N S e tri] D C, tra P N S f nī] sugg. em., ni D C P N S g stī] D C, sti P N S h nī] sugg em., ni D C P N S i mo’i] D C, mo P N S j seng ge’i] D C S N, sengge’i P k dwags] sugg em., dags D C P S, dag N l kṛṣṇa] D C, kriṣṇa P N S m sogs] D C V(P and D), tshogP N S S n lī] D C, li P N S o sen] D C, se P N S p ser] sugg. em. based on V(D and P), sen D C, se P N S q pa’i] P N S V(D and P), pa D C r rākta’i] P N S, rakta’i D C s skyur] C P N S V(D and P), skar D t shangkhī] D C, shang ki P N S u nī] sugg. em., ni D C P N S 309 dbyibs ni khyod la bstan par bya// bu mo ba tsha’ia dri bro zhing// skra ni ring zhing sna yang ring// dkar ‘chang nā ga rab ‘dra ba// |57| ‘o ma zho la rab tu sred// rākta’ic ro ni mngar ba’o// mdog ni dkar ser gnas pa yi// gos dkar mo yis rnam par dag// |58| de bzhin tsi tri nīd yie ni// dbyibs dang mtshan nyid bshad par bya// bu mo sha rlon dri bro [4b] zhing// lus chung brla ni shin tu mdzes// |59| dkar ‘chang shing thog pa la ‘dra// ngo tsha med cing khro ‘dod pa// rtag tu thab mo dag la dga’// rje ngar bya rog rkangf ‘dra zhing// |60| mchug ‘phyang gan rkyal du nyal ba// phugh ron skad ‘dra sha mdog ni sngo bsangs rākta’ii ro tsha ba’o// de ni lha mo sgrol mas dag// |61| hastī nīj yi mtshan nyid dang// dbyibs ni nga yis bstan bar bya// bu mo chang gi dri bro zhing// byin sbom sna ni cung zad yok// |62| dri nga zhing ni lus sbom dang// spyod pa rtsing zhing mdog rdul skya// dkar ‘chang g.yas g.yon du dgye ba// sha mdog smug cing rākta’il ro// |63| bsngal zhing de ‘dra’i bu mo ni// sangs rgyas spyan mas dag ba’o// a tsha’i] D C V(D and P), tshī P S, tshwi(?) N b nā ga ra] sugg. em., na ra ga D C P N V(P), ma ra ga S, na ga ra V(D) c rākta’i] P N S, rakta’i D C d tsi tri nī] sugg em., tsi tri ni D C, tsi tra ni P N S e yi] P N S, ya D C f rkang] D C, dang P N S g mchu] D C P S, chu N h phug] D C P S, phu N i rākta’i] P N S, rakta’i D C j hastī nī] sugg. em. hastī ni D C, hasti ni P N S k yo] D P N S, po C l rākta’i] P N S, rakta’i D C 310 da ni de dag thams cad kyi// rung dang mi rung spyod pa bstan// |64| rol tshea rgyab kyis phyogs pa dang// khong du dga’ zhing tshig mang dang// bsgo ba bzhin du bka’ yang gcog// ‘o byas kha ni ‘phyi byed de// |65| bla ma’i yon tan shes ‘gyur yang// gzhan du de la cung zad bsnyadb// bla ma ma mthong ltar byed cing// ‘ong kyang de6 la phyag mi ‘tshal// |66| rje btsun de dang mi mthun pa’i// mi gzhan dang ni rtse dga’ byed// de sogs rab tu mi rung ste// mkhas pas rab tu spang bar bya// |67| rung ba ‘di ni mkhas pas btsal// mig gi rtsa mdangs mig rtsa rgod// slar zhing skad cig bsam pa dang// skad cig rgod pa’i ‘phro bzhagc nas// |68| bsam zhing yang na smra‘am ‘dzumd// rje btsun bla ma mthong gyur na// dga’ zhing mdzes lta ‘dzum7 par byed// de yi bka’ stsal nyan zhing mos// |69| steg pa’i tshul8 gyis zhe sar ldan// gzhan du de la snyan par brjod// de mthong ‘grogse la ‘khyud cing bsnyen// phag tu ‘khri zhing sor tshigsf nyedg// |70| skra ni bkrol nas slar ‘ching byed9// gos ‘chos shing ni bar bar ‘dzum10// bla ma ma mthong ltar byed cing// gos ‘chos pa dang lus nyed dang// |71| rkang pas sa la ‘dudh cing ‘drii// bus pa’i kha la ‘o byed cing// a tshe C P N S, che D b bsnyad] D P N S, bsnyed C c bzhag] D C P S, gzhag N d ‘dzum] D C P S, mdzum N e ‘grogs] D C P S, ‘grog N f tshigs] D C, tshig P N S g nyed] D P N S, nyid C h ‘dud] P N S, ‘drud D C i ‘dri] D C, ‘bri P N S 311 nu ma ston cing ske rags ‘grol// rje brtsun de yisa ma mthong na// |72| glu dbyangs len cing mthong bar byed// de ni gang lab gnas pa der// don med bsnyad kyis ‘gro ba dang// drag tu gad mos rab stegc sted// |73| lu zhing glal ba bltar byed cing// [5a] mi mo ‘di yang stene par byed// |74| daf ni gsang gnas dbyibs kyig ni// mtshan nyid rung dang mi rung ba// nga yis yang dag bstan par bya// yid gcig bsdus la mnyan par gyis// |75| gyo mo lta bu kha sbub dang// khung ring shin tu mi sdug dang // rtag tu skam por gnas pa dang// rtag tu zil pa ‘dzag pa dang// |76| sbal pa’i rgyab ltar rtsub pa yi// gnas ni spang bya mi rung ngo// rus sbal rgyab ltar ‘phang mtho zhing// stengh mnyam shin tu ‘jam po dang// |77| ba glangi gi ni ltag pa ltar// steng mnyam thog gi dbyibs kyang ring// padma’i snying po bzhin du zlum// cung zhig ‘dzag min cung zad ‘byung// |78| spu nyung sgyu ma’i phyag rgya ni// mkhas pas thabs kyis shin tu btsal// |79| mtshan nyid de dang mi ldan yang// rigs dang gzugs dang lang tshor ldan// rgyan mdzas yid ‘ong bu mo ni// longs spyod gsum ldan gzung bar bya// |80| chos zab la mos sems dang ni// a yis] D C P S, yi N b la] D C, na P N S c bstegs] D C, bsteg P N S d ste] C , sta D, te P N S e sten] D C, bstan P N S f da] D, de C g kyi] D C, kyiP N S S h steng] D P N S, stong C i glang D C, lang P N S 312 las ‘brel gyis ‘khyud ma byas dang// sngon gsungs grogs kyis shes bslus pa’i// mi mo yang ni mkhas pas bsgrubs// |81| de ltar de sogs bud med kyi// sgyu ma’i phyag rgya chen pos ni// ‘jig rten gsum du rnyed dka’ ba’i// lhag pa’i lhaa shes bya ba bsgrubb// |82| de yang thun mong dang bcas pa// tshogs kyi mchod pa yang byas te// gang gsungs bu mo btsal nas kyang// bla ma la ni dbul bar bya// |83|c de nas bla ma de mnyes nas// de dang snyoms par zhugs pa yis// bde gshegs zhu gyur bcu drug char// gyur pa slob ma’i kha ru ltung// |84| ltung bas snying gi padma zhugs// de yis zhing ni dag byas te// chos kun sgyu ma la sogs pa’i// bcu gnyis don du rtogs par ‘gyur// |85| de nasd de la rang ‘byung gi// chos sku rab dga’ mkha’ nmyam pa// lhag pa’i lha zhes bya ba ni// rtogse bya’i ched du bu mo byin// |86| lha mo ‘di ni khyod dang mthun// sems chenf khyod kyis ‘dod pa gyis// yid ‘ong bu mo ‘di nyid ni// sangs rgyas kun gyis spyod du gnangg// |87| dkyil ‘khor ‘khor lo’ih cho ga yis// bu mo sgrol byed dga’ byin ma// a lha] C P N S V(P and D), lnga D b bsgrub] P N S, sgrub D C c Verses 83-125 of the Dvitīyakrama are edited in Sakurai 1996, 531-35. I only became aware of Sakurai’s edition of these verses after completing my own (his book is in Japanese, which I do not read, though of course the Sanskrit and Tibetan editions that it contains remain accessible to me), and have not taken his edition into consideration here. d nas] D C, laP N S S e rtogs] D C, togP N S S f chen] P N S V(P), can D C V(D); Vaidyapāda’s commentary also suggests that sems chen is the better reading: sems chen (chen] P, can D) zhes pa ni sems can bsgral ba’i sems gang la yod pa’o// g gnang] D C V(D and P), snang P N S; The Peking edition of Vaidyapādas commentary cites the line from the verse with snang, but then in the explanation of the verse uses the correct spelling, gnang. h lo’i] D C, lo P N S 313 byang chub chen po bsgrub pa’i phyir// khyod [5b] kyis bde chen myong bar gyis// |88| gzhan kyis sangs rgyas mi nus ba// bu mo ‘di ni yang dag mchog// de bas mtha’ med ‘khor ba’i bar// khyod kyis ‘di dang bral mi bya// |89| de nas sems11 chena de yis ni// yid ‘ong bu mo blangs byas te// de yis dam tshig sdom pa’i mchog// tshig ‘dis rab tu smra bar bya// |90| smros shig bu khyod rdo rje can// sha khrag khu ba la sogs pa// dri chen dri chu sogs za’am// bha ga ‘o byed mi rtog gam// |91| gzungs ma nga la khyod brod dam// de yis bzhadb gad dang ldan par// kye12 lha mo ci phyir bdag mi brod// dri chen dri chu sogs za’o// |92| lha mo khyod la bkur sti dgos// bha ga ‘o13 byed nga mi rtog// de nas bu mos ga shac phud// padma gsal bstan tshig ‘dis bstod// |93| e ma ho bdag gi padma ‘di// bde ba tham cad dang ldan pa// cho ga’i rim pa gang gis spyad// de yi mdun na rtag tu gnas// |94| don byed padma dam pa ‘di// sangs rgyas kun gyis bkurd b’ai gnas// rang ‘byung bde ba chen po ni// ‘di na rtag par bzhugs pa yin// |95| bhajae mokṣa ho// de nas rang gi sa bon las// a chen] P N S, can D C b bzhad] D C P S , gzhad N c sha] D C, zha P N S d bkur] D C P S, bskur N e bhaja] sugg. em., bhanydza D C, bhaṃdza P N S. I suggest this emendation based on the Vajrāvalī which reads bhaja here. This also makes more sense, understood as the second person imperative of bhaj: “Grant liberation, hoḥ!” Vaidyapāda makes it clear that this mantra is spoken by the disciple to the consort (Sukusuma, 106b.6). 314 ‘od zer stug pos nang gsal nas// lha yi ‘khor lo chags byas nas// rdo rje nang ne phyir byung ste// |96| padma’i nang du zhugs byas nas// rig ma’i lha yi ‘khor lo rnams// chags pa chen po bskyed nas kyang// zhal nas zhal zhugs ‘khor lo rnams// |97| chags chen bskyed nas rdo rjer phyung// sngon bzhin zhugs sogs rab byas pas// yang nas yang du lha’ia dkyil ‘khor// yid kyis shin tu chags bar bya// |98| de nas ‘dod pa’i sgyu ma’i tshig// brjod pas bu mo chags byar bya// gnyis su med pa bde ba chen po mchog// lha mo khyod ni sgyu ma’i phyag rgya ste// |99| zhal bzang khyod ni nga dang lhan cig tu// rnam par rtsenb pas mkha’ ‘dra myongc bar bya// des kyang sgyu ma lta bu yi// ‘dod pa’i tshig gis gsol btab nas// |100| bdag nyid chags pa chen po yang// bskyed nas rtsa yi ‘khor lo btsal// |101| bde mchog rgya chen bdag la dgongs su14 gsol// bsam mi khyab pa’i bde chen ngagd dang ldan// mi g.yo [9a] tshig dang bral ba’i dbyangs snyan pas// bdag la khyod kyis rol cig bzhenge su gsol// |102| ngo tshar che ba’i rol pas snang mdzad cing// byang chub snying por ‘gro ba’i15 lam mchog ‘di// rgyud rnam yang dag gang la mi rtenf pas bde chen yid ‘ong nga yis grol bar gyis// |103| ho ho ho// a lha’i] DC, lha P N S b rtsen] P N S V(P), brtson] D C V(D). Vaidyapāda’s commentary, which glosses the term as rnam par rol pa also supports the reading of rtsen (Sukusuma, 107a.2) c myong] P N S, myang D C d ngag] sugg. em. based on V(D and C); dag D C P N S. Not only do both the Derge and Peking editions of Vaidyapāda’s commentary here read ngag, his comments also make it clear that this is his reading: bsam mi khyab pa’i bde chen te// zhes te thams cad la khyab bdag tu gnas pa’i bde ba chen po ni mi slu bas mtson par byed pas ngag dang ldan zhes bya ‘o// (Sukusuma, 107a.3-4). e bzhengs] D C P S, bzhe N f rten] D C, brten P N S 315 rol cig rol cig bdag la rol// ‘dod pa gyis shig dga’ gsola ba// khyod kyis gzhan du mi rtogs pa// rtogs pas the tsom ma byed cig// |104| a la la lab ho// de nas rab tu chags ldan pas// de dang lhan cig lus kyi ni// spyod pas dben pa’i gnas su spyad// rol pas dga’ ba brtag par bya// |105| dang por ‘dus pac byas nas su// de nas gru mo mtshan nyid dang// yang ni brkyangd par bya ba dang// de bzhin yar bteg mtshan nyid dang// shin tu brkyange pas lnga ru ‘gyur// |106| dpung pa rab tu brkyangf byas nas// gru mo mdud pas mgul nas ‘khyud// bcingg bas dam ‘khyud g.yas g.yon skra// mgo ni mi g.yo zung nas blta// |107| de nas brla gnyis barh bltas te// bung ba lta bu zid sgra’i glu// blangs nas mchu yis rtse bar bya// |108| ‘gying bag stabs kyis ‘khyud nas ni// g.yon gyisi spyi bo’ij skra bzung ste// g.yas kyis lkog ma brten nas kyang// ma mchu’i sbrang rtsi gzhibk par bya// |109| gzhibl cing zidm sgra brjod nas kyang// nu ma lag rtsen mgrin pa dang// ma mchu ‘gram pa rna ba’i rtsa// a gsol] D C, gsal P N S b la] P N S, om. D C c pa] D C, ma P N S d brkyang] D C, bskyang P N S e brkyang] D C, bskyang P N S f brkyang] D C, bskyang P N S g bcing] D C P S, cing N h bar] D C S, par P N; Although Vaidyapāda’s commentary also reads par (in both D and P), his comment suggests bar: brla gnyis par bltas te zhes pa ni ‘og gi padma la bltas na// (Sukusuma, 108a.1-2). i gyis] D C P S, gyi N j bo’i] D C, bo P N S k gzhib] D C, bzhib P N S l gzhib] D C, bzhib P N S m zid] D C, sring S, srid P N n rtse] P N S V (D and P), rtsa D C 316 mig dang spyi bo gsang bar yang// kha yis ‘o byasa rtseb bar byac// |110| rna ba gnyis dang mchan khung dang// mgrin pa gnyis dang sum mdo rnams// sen mo ri mos gdabd par bya// |111| nu ma gnyis dange mchanf khungg gnyish// gtso gnyis dang ni ‘gram pa gnyis// lag mthil gnyis dang rkang mthil gnyis// mnye pas mdza ba chen po ‘grub// |112| g.yon gyis padma’i dkyil ‘khor ni// mnye zhing lce yis skyod par bya// steng ‘og tu ‘angi blta byas nas// sems kyis de la chags par bya// |113| de nas dga’ ba’i bu mo des// padma bstan nas ‘di skad smra// rang byung bde chen rgyal po ni// padma ‘di laj rab tu gnas// |114| rtsa dang rlung [6b] gis rtogs ‘gyur bas// khyod kyisk rtsa yi ‘khor lo16 tsholl// de nas de yism sorn mo yis// nang17 gnaso rtsa yi ‘khor lo che// |115| padma la gnas snying po lngas// brgyanp byas ze’uq ‘bru ge sar dang// ‘dab ma brgyad pa gsal byas nas// ā li kā li mantra dang// |116| a byas] D C, bya P N S b rtse] D C, brtse P N S c bya] D C, yang P N S d gdab] D C N, gdag S, gtab P e dang] D C P N, S om. f mchan] D C N, mtshan P, S om. g khung] C P N, khud D, S om. h gnyis] D C P N, S om. i tu ‘ang] P N V(D and P), tu‘ang S, tu yang D C j la] D C V(D and P), na P N S k kyis] P S, kyi D C l tshol] P N S, rdzol D, tshal C m yis] sugg. em. based on V (D and P), yi D C P N S n sor] D P N S, yir C o gnas] sugg. em based on V (D and P), nas D C P N S. Vaidyapāda reads de nas nang na gnas pa’i rtsa’i ‘khor lo chen po btsal bar bya’o// (Sukusuma, 108b.1). p brgyan] P N S V (D and P), rgyan D C q ze’u] D C, ze S N, bre P 317 kura ma ka dang shab shāngc ka// nā ḍid gsum po btsal bar bya// rdo rje dyings kyi dbang phyug rtsa// gzung ‘dzin bral ba bha ga’i dbus// |117| bla ma’i man ngag stobs kyis ni// sor mos go bar bya dgos so// de nas sae bcu dang ldan pasf// lha mo de la ‘khyud bar bya// |118| rdo rje padmar reg pa ni// bsnyen pa’i de nyid yin par bshad// rdo rje padmar zhugs pa ni// nye bar sgrub pa’i de nyid do// |119| de nas bsgul zhing bskyod tsam gyis// snying ni ‘dar zhing dran pa nyams// spyi bo’i skra grolg gos kyang ‘dorh// rnguli gyis lus khyab mdog dmar te// |120| mig dmar phra bas bdag la blta// yang du bsgulj bas sgrubk pa’o// de bas sems khral med pa ru// sdom brtson gzhu dbyibs gyo ba yis// |121| sum mdo ye shes mel sbar nas// khams bzhusm nas ni bcu drug char// gyur ba me tog kundan ‘dra// rlung gi sbyor bas phul bar bya// |122| rang bzhin gyis ni rab zhi ba// chos kun zhi ba de kho na// a kur] D C, kun P N S V (D and P) b sha] D C, shang P N S V (D and P) c shāng] sugg. em based on V (D and P) which read shang (I suggest adding the long ā), sha D C P N S. d ḍi] sugg. em., li D C P N S e sa] D C, P N S om. f pas] sugg. em., pa’i D C, pa yi P N S. The reading from P, N, and S gives the correct grammar, but is metrically incorrect, so I have suggested emending to pas. g grol] D C, ‘grol P N S h ‘dor] D C, ‘dor bar byed] P N S i rngul] sugg. em., rdul D C P N S V (D and P). This emendation is based on the line from the parallel verse in Vaidyapāda’s Yogasapta which reads rngul chu thigs pas lus kun khyab//// (Yogasapta, D 71a.5; P 84b.7) j bsgul] sugg. em. based on V, bskul D C P N S. Buddhajñānapāda’s text here reads bskul, but given the fact that earlier the text read bsgul ba, as well as the fact that this is glossed in Vaidyapāda’s commentary as yang dang yang du bskyod pa suggests that it is bsgul that is is meant. k sgrub] P N S, bsgrub D C, l me] D C S, ma P m bzhus] D C P S, gzhus N n kunda] D C, kun da S N, ku da P 318 bde ba de nyid nor bur ‘dug// skad cig dran med g.yo bar byed// sgrub pa chen po de nyid do// |123| nam mkha’i khams dang rdo rje sbyor ba las ni yang dag spyan can bde chen ‘byung byar ‘gyur// gang gang yang dag dga’ byed chags bral dga’ gnyis bar du ‘bena nyidb mthong byas brtan par gyis// padma’ic mkha’ la rdo rje nor bu pad snying gnyis la ‘byord dang rdo rje skyil krung sems// nor bu’ie bar du mthong byas gang de bde ba ‘byung ba nges par de nyid ye shes te// |124| ‘di ni rdzogs pa’i rim pa yin par bla ma mchog rnams kun gyis yang dag bshad// ‘dod chags chags bral bar ma mif dmigs ye she lha ni skad cig de ru gsalg// chu tshod brgyad dam nyin gcig dang nih zla ba gcig tu a’m// lo cig bskal paa’m bskal pa stong du ye shes de myong bya// padma la gnas bdud rtsi khu ba blo gros can gyis kha yis blangs nas btung bya’o// |125| de ‘dra’i dngos pos [7a] thams cad kyi// mtha’ yi de nyid zab gsal ba// thog med dus nas so so yi// skye bo nga dang ngar ‘dzin pas// ma brtags par ni bdag tu bzung// |126| dus las byung dang ‘dod rgyal dang// gnam gyis bskos dang gyi na dang// ‘dzin pa pos ni sprul pa dang// khams kun las ni nges ‘byung dang// |127| dbang phyug dang ni ‘byini pa’o// dus dang gtso bo byed pa dang// byed pa po ni ma yin dang// rnal ‘byor dang ni tshad ma dang// |128| dag pa dang ni ma dag pa// nang gnasj brjod du med pa’i bdag// skyes bu dang ni khyab bdag dang// a ‘ben] sugg. em. based on the parallel Sanskrit verse from Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, dben V (D and P), bden D C P N S. The Yogasapta parallel verse cited in the next note also supports reading dben. b nyid] sugg. em., gnyis D C P N S. This emmendation is likewise based on the parallel verses (precisely these same verse from the Dvitīyakrama) in Vaidyapāda’s Yogasapta where they read nyid in both P and D: dga’ gnyis rab tu dben nyid mthong (Yogasapta, D 71a. 3; P84b.4). In his comments on the Dvitīyakrama verse, Vaidyapāda’s Sukusuma simply reads dga’ ba gnyis gyis dben pa, which also suggests he was reading dben nyid rather than dben gnyis (Sukusuma, D 109b.5). c padma’i] D C, padma P N S d ‘byor] D C P S, sbyong N e bu’i] D C, bu P N S f mi] D V (D and P), ma C P N S g gsal] P N S V (D and P), bsam D C, h dang ni] D C, P N S om. i byin] D C, byon P N S j gnas] P N S, nas D C 319 srog dang gang zag rnam shes dang// |129| kun gzhi dang ni shes pa poa// mthong ba po dang gzung ‘dzin dang// shes pa dang ni shes bya dang// shed las skyes dang shad bu dang// |130| srog dang gso ba la sogs pa// mu stegs rnams kyis rnam par rtogs// |131| nam mkha’ dang ni ‘gog pa gnyis// shin tu ‘dus ma byas pa ste// shin tu brtan byas ‘dus byas kun// skag cig ma ste bdag po med// |132| phra rab rdul las grub pa ni// blo yi rnam pa med pa ru// kha che bye brag sma bas rtogs// |133| rang gi rig pa skyed byed pa// blta bya sypod yul rnam par ni// ‘dus ma byas pa gsum po dagb// mo gsham bu bzhin med pa dang// |134| ‘du byed thams cad bems sdigsc dang// dus gsum rgyu ba ma yin dang// dman mind thogs med gzugs su ni// mdo sde pa yis rab tu rtogs// |135| yan lag can yang don dam min// rdul phran dag kyang de bzhin no// so sor snang ba mi dmigs pa// rmi lam lta bu mi snang ste// |136| gzung ‘dzin spangs pa’i ye shes te// don dam shel ltar dag pa ru// rnal ‘byor spyod pas rab tu rtogs// |137| so sor snang ba’i gzhung thams cad// don dam min te gcig pa dang// du ma’i rang bzhin bral ba’i phyir// rnam mkha’i chu skye bzhin du ni// |138| a pa po] D C, pa’o P S b dag] P N S, dang D C c bsdigP N S S, rig] D C. I am unsure of what this word should be, but all of the texts, including Vaidyapāda, have some form of bsdigs, sdigs, or rdeg, so it seems that the root text in the Derge may have been corrected to read rig. The meaning of the passage remains unclear. d min] D C V (D and P) , med P N S 320 gnyis med gnyis su med min zhi// shin tu dri med nam mkha’ ltar// blo ldan dbu ma pa yis rtogs// |139| de la sogsa te mtha’ yas pa// don de kho na la gnas nas// tha dad so sor rtogb pa’o// de bas de dag thams cad kyang// |140| [7b] yang dag min pas bla dang bcas// ‘og ma’i ‘og ma’i rnal ‘byor blo// gong ma gong ma’ic khyad par ‘phag// ‘og ma’i blo ni gong ma yi// |141| shes rab kyis ni sun ‘byin no// de bas gong ma’i rim pa yis// lhan cig pa yi slob dpon gyis// byin gyis brlabs pa yang dag bya// |142| gsal zhing rab dga’ nam mkha ‘dra// rang byung18 lhag pa’i lha chen po// lhan cig skyes pa’i ye shes kyis// bla ma’i kha las rtogs par bya// |143| de ltar dam tshig sdom ldan pa’i// snod du rtse gcig chu ‘dra las// ye shes gzugs brnyan lta bu ru// rnal ‘byor pa yis rab tu brtag// |144| de thob phyin chad rnal ‘byor pas// khor ba nyid na gnas pa’i tshe// de yi nyes pas mi gos so// ji ltar snags dang sman ldan pas// |145| sbrul dag gsod par byed pa bzhin// de ‘dra’i rnal ‘byor dbang phyug che// ye shes sman gyis rgyas btab pas// nyon mongs gyis ni ci byar yod// |146| gang zhig lag na gdugs thogs la// de la char pas ci byar yod// de bzhin gnyis med ye shes kyi// gdugs thogs19 la ni rtogd pa yi// |147| a sogs] P N S V (D and P), rtogs D P b rtog] P N S, rtogs D C, brtags V (D and P) c ma’i] D C V (D and P), ma P N S d rtog] D C P S, rtogs N 321 char pa rab tu ‘bab ‘gyura yang// de la de yis ji ltar gnod// de ‘dra’ib rab mchog ye shes ni// so so skye bos ga la shes// |148| nyan thos rnams kyis mi shes so// rang sangs rgyas kyis kyang mi shes// rnal ‘byor spyod dang dbu ma pa// byang chub sems dpas mi shes so// |149| bla bcas sangs rgyas kun gyis kyang// ‘di ni cung zad mi shes so// ‘di yi don shes ma ‘ongs pa’i// rdo rje ‘dzin pa mnyes byas nas// |150| rang gi bsod nams chen stobsc kyis// yi ge med par rnam par ‘pho// de la dkyil ‘khor sbyin sreg dang// gtor ma bzlas pa bgrang phreng dang// |151| skyil mo krung20 dang stang stabs sogsd// spros bral rnam bar slue ba ste// bya ba ma yin dgag pa min// lhag pa’i lha yis sprul phyir ro// |152| bya ba rnams la rnal ‘byor pa// lam chen dag tu yongs ‘dzin pa// ri dwags21 smig rgyu snyeg pa ltar// rtag tu snang yang ma zin no// |153| bya ba’i nad chen gyis zin la// ye shes g.yo med sman chen gyis// gso byed skyes bu dam pa’o// |154| de bas lus ngag yid gsum gyi// [8a] sdom pa mchog la gnas byas nas// nga’o snyam pa’i gzi bskyed la// rim pa gnyis pa de nyid bsgrub// |155| dgon pa’m yang na grong gi mthar// gang rung cif rung de lta bur// a ‘gyur] P S, gyur D C b ‘dra’i] D C, ‘dra P S V (D and P) c stobs] D C P S V (D and P), bstobs N d sogs]D P N S V (D and P), so C e slu] D C, bslu P N S f ci] D C N, cing P S 322 bya bar ‘os pa’i las kun yang// byas nas bde ba’i stan la ‘dug// |156| de nas sems can thams cad kun// tshangs spyod chen po bzhis dmigs22 te// de la sogs pa rang rgyud kyi// las kyi sgrib pa dag byas nas// |157| sems tsam23 la ni blta bas te// phi rol rang bzhin stong24 bar bya// sems tsam de yang stonga byas nas// rang rig tsam25 du gnas par bya// |158| rang rig de yang zla ba sogs// gdan du brtags nas yi ge yib// smyu gus bsnun byas mtshan mar byur// de las rang nyid lhar bskyed la// |159| lha yi nga rgyal dang ldan par// phyag rgya bzhis ni rgyas btab ste// dkyil ‘khor ‘khor lo sprul bar bya// de mnyes de la goms bslabs pas// |160| phyi rol lus ni log26 byed de// sangs rgyas rnam kyic spyod yul mchog// mtha’ yi de nyid bsgom par bya// rang snying dam tshig phyag rgya yid// |161| thugs kare mtshan ma la gnas pa// de bzhin gshegs kun ye shes mchog// rim pa dang po pa rnams la// yi ge gzugs su rab snang ba// |162| gzhom du med pa rang lha yi// sa bon ‘od zer lnga ‘bar baf// de yi kha dog rnam lnga yig// ‘od zer steng27 gi sgo yi ni// |163| g.yas nas rnam par spro bar bya// de yi rtse lash dkyil ‘khor gyi// ‘khor lo ldan pa’i bde gshegs rnams// a yang stong] D C S N, yang stong yang stong P b ge yi] sugg. em based on V (D and P) which read yi ge’i, ge’o D C P N S c kyi] D C V (D and P), kyiP N S S d yi] sugg. em. following V (D and P), yis D C P N S e kar] D C P S V (D and P), dkar N f ba] D C V (D and P), baP N S S g yi] sugg. em. following V (D and P), yis, D C P N S h rtse las] sugg. em following V (D and P), rtsa la D C P N S 323 phyogs bcu’i ‘jig rten bkang byas nas// |164| dbugs kyi kun rtog las byung ba’i// sems can thams cad sangs rgyas skura// bskyed nas zla bar zhu byas ste28// thim nas rnam par dag byas nas// |165| de rnams ye shes ngo bo ru// g.yon pa’i ha sar zhugs par bya// de yi mtshan ma dbus gnas pa// sa bon la ni gzhug byas par// |166| de nyid sangs rgyas thams cad kyi// yon tan kun bskyed rin po che// dngos po kun la khyab pa’i bdag// mi shigs pa yi thig le che// |167| ‘od zer lnga dang rab ldan pa’ib// tsa na ka yi ‘bru tsam29 du // rang gi sems su chos thams cad// bsdus nas rnam par bsam [8b] par bya// |168| de yi ‘od zer rim pa yis// rang gi gnas ni khyab par bya// de las phyir byung dam tshig rgya// nang gsal de yis phyi yi lus// |169| gsal byas dkyil ‘khor gsal byas nas// de yi gnas kyang gsal bar byac// de las phyir byung bdag nyid nang// byang chub sems gzugs bcu drug po// |170| mthe ba’i rtsa bar yi ge a// rje ngar gnyis la de bzhin ā// brla gnyis la ni i yi gzugs// gsang ba la ni ī yi gzugs// |171| lte ba’i rtsa ru u gzugs gnasd// de bzhin gsus par ū yi gzugs// nu ma gnyis su ṛe gzugs gnas// de bzhin lag par ṝ f gzugs te// |172| a skur] sugg. em. following V (D and P), sku D C P N S b pa’i] P N S V (D and P), pa D C c bya] D C S N, ba P d gnas] D C P S, pas N e ṛ] D C P, ri S N f ṛ] C N S, rā D, rī P 324 mgrin pa la ni ḷ yi gzugs// de bzhin ma mchur yi ge ḹ// mgram pa gnyis su e gzugs so// de yi mig gnyis yang ni ai// |173| rna ba’i rtsa bar o yi gzugs// sphyi bor au gzugs yang dag gnas// aṃ aḥ yia gzugsb tshigs kun la// bcu drug dus su rdzogs ‘gyur ba// |174| gong ma’i ‘od kyis rnam bkug ste// thig le ru ni gzhugc bar bya// de ru sems ni cung zad tsam// gzung bas lha yi ‘khor lo che// |175| rten dang bcas pa mos pas brgom// de dbus bdag po thugs ka rud// gong ma’i rim pas mi shig pa’i// thig le chen po bsgom byas nase// |176| yungs kar tsad tsam de las spro// rang gi gnas dang phyag rgya yi// nang gsal kdyil ‘khor gsal byas nas// de yi gnas dang bdag po yi// |177| nang gsal phyi lus gsal byas nas// dkyil ‘khor rten bcas gsal bar bya// de nas bcu drug gnas pa yi// ‘od zer gzhag pa phyung byas te// |178| phyi yi lus ni gsal bar bya// de yis dkyil ‘khor ‘khor lo dang// rten dang bcas pa gsal byas nas// phyogs bcu‘i ‘jig rten bzhugs pa yi// |179| bde gshegs kun gyi spyan sngar gshegs// bdud30 rtsi zhu nas zhal zhugs te// thig ler bdud rtsi blangs byas nas// rdo rje’if lam du byung nas su// |180| ‘o ma’i rgyun ltar ‘ongg byas nas// a aḥ yi] P N S, aḥ’i D C b gzugs] P N S, gzugs su D C c gzhug] P N S, bzhug D C d ru] D C P S , rus N e nas] D C S N, P om. f rje’i] D C, rje P N S g ‘ong] D C, sngon P N S; ‘ongs V (D and P) 325 thabs kyi ye shes g.yas zhugs te// de bzhin gzhan ni g.yon du zhugs// de nas ye shes thig le la// snang ba’i [9a] thigs pas thim par ‘gyur// |181| de ni ‘dzag pa’i ngang tshul can// ‘od zer lnga ‘bar dkar ‘tsher ba// cung zad dmar ba’i nang du sems// mi g.yo zhing ni gzung bar bya// |182| de lasa gong ma’i rim pa ltar// spro pa dang ni bsdus pa dang// gzung ba yang dag byas gyur pasb// ‘gog pa’i rdo rje sems nyid du// |183| nam zhig de la reg gyur pasc// sangs rgyas kun gyid mchog ‘dzin pa// yid bzhine dpal dang ‘dra ba yi// rtags rnams yang dag skye bar ‘gyur// |184| phung po khams dang skye mched kyi// lha rnams ye shes me yis su// thog mar snying gar sdud byed pasf// cho ga ‘di ni dang por bshad// |185| de la bsgrub pas byin brlabs pas// sna yi rtse mor ‘gro bas na// mi shigsg thig le’i rjes thogs la// gsang ba’i thig le bsgom par bshad// |186| gong ma’i ye shes thig le las// ‘od zer lcags kyu ‘dra ba rnams// phyog bcur spros pas bde gshegs rnams// dkyil ‘khor ‘khor lo dang bcas kun// |187| bkug nas ye shes ngo bo ru// rang gi snying gar rab tu btsud// de rnams zhu nas thig le la// zhugs pas rdul dang mun pa dang// |188| snying stobs tshul du thig le yang// a las] sugg. em. following V (D and P), la D C P N S b gyur pas] D C, ‘gyur ba P N S c gyur pas] D C, ‘gyur ba P N S d gyi] P N S, gyis D C e bzhin] D C (and Muktitilaka, parallel verse), byin P N S, sbyin V (D and C) f pas] sugg. em. following V (D and P) pa’i D C, pa P N S g shigs] D C P S, shig N 326 snying gi padma lasa babs nas// rang gi rdo rje nor bu yi// dkyil du rnam par gnas pa las// |189| ‘byung ba lnga yi gzugs can mchog// ye shes lnga yi ngo bo nyid// ‘od zer lnga ‘bar rang lha yi// mtshan ma phra mo rnamb bsams nas// |190| de yi bum parc rten gyi tshad// yungs kar tsam dbus ‘khor lo che// thig led bcas pa mos par bsgom// gal te de la rang gi sems// |191| bying bar song ngam skyo ba na// rdo rje las ni phyir byung ste// sna rtse rnam par gnas byas nas// dga’ bral dga’ bas brtag par bya// |192| ‘di yi thig le rnal ‘byor las// gong gi rim pas ‘od zer lnga// steng gi sgo las byunge ba dang slar yang bdud rtsi ‘gugsf pa dang// |193| de bzhin gzhi la gzhug pa dang// dbugs dgag stong pa’i yan lag go// de nas gsang ba’i thig le che// rab tu phra ba gong ma bzhin// |194| rang sems dal bu dal bus gzung // dngo po med pa’i gnas su [9b] ‘gyur// |195| de yang gzung31 ba goms pa yis// phying log sa nih rnam par log32// smig rgyu lta bu snang ba ni// rtags ni dang por shes par bya// |196| de bzhin du ni chu log pas// du ba lta bur ‘byung ba nii// a las] sugg. em., la D C P N S. The reading las seems to have been transmitted in Tsongkhapa’s commentary on the five stages (see Kilty 2013: 174-75). b rnam] D C P, rnams S N c par ] D C, pa P N S d le] sugg. em. following V (D and P), ler D C S P N e byung] D C, ‘byung P N S f ‘gugs] D C P S , ‘gug N g phyin] P N S, phyin ci D C h sa ni] P N S, ni D C i ni] D C P S, naṃ N 327 gnyis pa yin par shes par bya// me ni rnam par log gyur pas// |197| mkha’ snang ‘dra ba gsum pa’o// rlung ni de bzhin log pas naa// mar me lta bub rab snang basc// rtags ni bzhi par shes bar bya// |198| de bzhin phyin log rnam par shes// log pas zab gsal gnyis med kyi// don ‘drad sprin med nam mkha ltar// gsal ba rtags ni lgna pa’o// |199| mtshan ma lnga po de yis ni// mi gnas mya ngan ‘das thob pas// rnal ‘byor pas ni ‘di la ‘bad// gzung ba yi ni yan lag go// |200| de ltar de la mtshan ma lnga// mthong byas don dese khyab bya’i phyir// rdo rje yi ni lam nas su// nam mkha’i khams su spro bar bya// |201| spros pa de las sngonf mthong ba’i// so sor snang ba bskyed par bya// sangs rgyas rjes su dran pa dang// chos ni yang dag rjes dran dang// |202| rdo rje rjes su dran pa dang// rigs ni yang dag rjes dran dang// khro bo rjes su dran pa ni// snang mdzad dang ni tshe dpag med// |203| mi bskyod rin chen ‘byung ldan sogs// gshin rje gshed sogs khro bo yig// dkyil ‘khor sku sogs ji rnyed po// spro bdsu las ni gsal bar bya// |204| de bzhin dam tshig rjes dran ni// rjes su chags dang cho ga bzhi// dang po spro bsdu las gsal bya// a na] D C, ni P N S b bu] D C P S, bur N c bas] D C, ba P N S d ‘dra] sugg. em. based on V (D and P), dran D C P N S e des] D C P S V (D and P), de N f sngon] sugg. em. based on V (D and P), mngon D C P N S g bo yi] D C, bo’o P N S 328 dkyil ‘khor rjes su dran pa ni// |205| cho ga gnyis pa la sogs pa’i// dkyil ‘khor spro bsdu lasa gsal bya// sku ni rjes su dran pa dang// gsung dang thugs ni rjes dran dang// |206| sems can rjes su dran pa ni// sku dang gsung thugs bsgoms pa yi// rdo rje gsum ni spro bsdu las// rnam par smonb cing dran pa’o// |207| de bzhin sngags ni thams cad kyi// spyi gzugs rjes su dran pa ni// ye shes lnga yis rdo rjec sems// bskyed pa dang ni rang gi lha// |208| longs spyod bzhi ldan spro bsdu las// rjes su gsal bad rnyad pa’o// dam tshig rjes su dran pa ni// bdud rtsi myang ba la sogs pa// |209| spro bsdu las ni gsal [10a] byed pa’o// shes rab pha rol phyin pa dang// ma skyes rjes su dran pa ni// spros ba’i dkyil ‘khor de dag kyang// |210| ci yang yod pa ma yin te// de ni gnyis med ngo bo ru// spro bsdu las ni dran pa’o// zhe sdang rigs la sogs pa yi// |211| mchod pa rjes dran sbyor ba ni// erang gi rigs kyi bu mo mchog// bar ma’i longs spyod bzhi dag pa// rjes su chags pas mnyes bya ba// |212| spro bsdu las ni gsal byed pa’o// rjes su dran pa’i yan lag go// de ltar yan lag gsum gyis ni// gsang ba’i thig le bsgom byas nas// |213| nor bua las ni byung nas kyang// a las] D C, la P N S b smon] P N V (D and P), smin D C S c rdo rje] D C, ye sheP N S S d ba] D C, dang P N S e + rang gi rigs la sogs pa yi// mchod pa rjes dran sbyor ba ni// P N S 329 khams gsum khyab par byed pas ni// sprul ba’i thig le bsgom parb bshad// rang gi spyi bo’ic dbus su ni// |214| sku yi rdo rje’id thugs ka ru// mtshan ma la gnas rlung dkyil ‘khor// dud ka ‘dra bar zla ba la// eoṃ dkar ‘ong ba’i mtshan nyid bsam// |215| de bzhin mgrin pa’i dbus su yang// gsung gi rdo rje thugs ka yi// mtshan ma’if dbus su chu dkyil ‘khor// dkar po dbus su zla ba la// |216| āḥg dmar gnas pa’i rang bzhin bsgom// thugs kar thugs kyi rdo rje yi// thugs kar mtshan ma la gnas pa’i// me33 yi dkyil ‘khor dmar po la// |217| zla gnas dbus su hūṃ nag po// ‘gro ba’i rang bzhin bsams nas kyang// nu ma gnyis kyi bar du yang// dam tshig phyag rgya’i thugs ka ru// |218| mtshan ma la gnas dbang chen ni// dkyil ‘khor ser po zla ba la// rang gi lha34 yi sa bon che// ‘byung dang ‘jug dang gnas pa la// rnam par grol ba’i sa bon bsamh// |219| de ltar phyag rgya bzhi chen dang// bdag nyid ‘brel bar byas nas su// rim pa ‘di yis mi shigs pa’i rdo rje bzlas pa’i ngo bo bya// |220| g.yas pa nas ni ‘byung ba’i khams// rlung gi dkyil ‘khor chen po ste// kha dog dud ka rgyu ba las// ‘byung ba hūṃ gi ngo bo yis// |221| snang srid chos kun khyab byas nas// a bu] P N S V(D and P), du D C b par] D C, pa P N S c bo’i] C, po’i D, bo P N S d rje’i] D C, rje P N S e + sa N f ma’i] P S, pa’i D C g āḥ] D C, ā P S h bsam] D C, bsams P S 330 chos kun dag byas bdud rtsira ‘gyurb// oṃ gyi ngo bos ‘ongc byas nas// rang gid rgyud kyi bag chags ni// |222| bkrus nas dag byas de la zhugs// āḥe yi ngo [10b] bos gnasf byas nas// de la lha yi de bzhin nyid// gzung ba dal bus dal busg byas// |223| dngos po med pa’i gnas su ‘gyur// g.yon pa nas ni ‘byung ba’i khams// chu yi dkyil ‘khor chen po ste// kha dog dkar po rab gnas pa// |224| de las byung ba’i hūṃ dang ni// oṃ gyi rang bzhin ‘dus ba dang// āḥh yi rang bzhin gnas pa dang// de nyid gzung bai snga ma bzhin// |225| gnyi ga las ni drag tu ‘byung// me yi dkyil ‘khor nyid yin te// kha dog dmar po’ij ngo bo las// huṃ gis spros dang oṃ gyis bsdus// |226| āḥk yis gnas pa de nyid gzung// snga ma bzhin du shes par bya// de bzhin gnyis las dal bu ‘byung// dbang chen gyi ni dkyil ‘khor rgyu// |227| kha dog gser ltar rab snang ba// de la huṃ gis ‘byung ba dang// oṃ gyis sdudl dang āḥm yis gnas// de la de nyid rab bzung nas// |228| ‘jug dang gnas dang ldang ba las// grol ba’i ye shes rnyed ‘gyur ba// a rtsir] D C , rtsi P N S b ‘gyur] P N S, gyur D C c ‘ong] D C, ‘ongP N S S V (P), ‘od V (D) d S om. e āḥ] D C V (D), a P N S, aḥ V (P) f gnas] P N S V (D and P), dag D C g dal bus] P N S, dbus lus D C h āḥ] D C, āh P N S i gzung ba] P N S, bzung bas D C j po’i] D C, po P N S k āḥ] D C, āh P N S l sdud] D C P sdu S, bsdud N m āḥ] D C P N, aḥ S 331 som nyi yid gnyis ma byed cig// |229| phyag rgya bzhi ‘brel bzlas pa ni// nyis brgya nyi shu rtsa nga ste// de ni dgu brgya rnam pa bsgres// zhag gcig tu ni nyi shu bzhis// |230| khri phrag gnyis dang gcig stong dang// brgya phrag drug tu dus thams cad// rnal ‘byor dbang phyug chen po yis// nyin mtshan bzlas brjod grangs byed do// |231| ‘di yis rnam pa’ia chos thams cad// sgyu ma dang ni smig rgyu dang// brag ca mgal me’i ‘kho lo dang// ‘khrul pa dri za’i grong khyer dang// |232| chu bur dang ni mig ‘phrul dang// gzugs brnyan chu zla la sogs par// thams cad shes shing sa bcu pa’i// dbang phyug rnams dang skalb mnyam ste// |233| rang bzhin bzlas pa ‘di la ni// rnal ‘byor bas ni rab tu ‘bad// ‘di ni thogs ma med dus nas// rang bzhin yang dag rtag zlo yang// |234| bla ma yang dag ma brtenc bar// don ‘di rtogs par mi ‘gyur ro// de ni yang dag shes nas ni// bsgom pa’i bar chad phyi rol gyi// bzlas pa yongs su spang bar bya// |235| bzlas pa’i ngo bo chen po mchog// rdo rje ‘chang ba brjod med sku// bsam brjod yul las rab ‘das bas// ngag gis [11a] ji ltar bzlas pa byed// |236| de bas glang chen rnyed gyur pas// rjes kyi de la ci zhig bya// |237| don de la gnas rnal ‘byor che// phyi rol yang ni rten mi ‘gal// glang po che yi stobs kyis ni// bud shing ‘byung ba lta bu’o// |238| a yis rnam pas] P N S, yi rnam pas D C b skal] D C V (D and P), bskal P N S c brten] P N S, bsten D C 332 de bas rdo rje bzlas pa mchog// ‘bad pas bde gshegs thams cad kyia// byin brlabs gnyidb ni ‘ong bas na// phyag rgya chen po ‘grub pa yi// mtshan ma yang dang yang ‘byung ‘gyur// |239| de ‘byungc cho ga la gnas pas// skad cig gis ni kun mkhyen che// rgyud lad rab tu ‘pho bar ni// the tshom med par nges35 pas na// cho ga ‘di la rab tu ‘bad// |240| de ltar dga’ gsum bye brag gi// thig le rnam gsum bsgom pa yi// cho ga yang dag bstan pa’o// |241| da ni gnyis med ye shes nyid// kun rdzob gzugs la brten nas kyang// mi g.yo la sogs g.yo dang bcas// tshangs sogs lha dang lha min sogs// |242| thams cad rab tu mi snang yang// thig le ‘gog par mi ‘gyur te// rgyu dang mi rgyue bcas pa kun// slar yang de las ‘byung bar byed// |243| de las skye bo rnams kyis ni// yang dag ma rtogs pa yi phyir// tshangs pa’i sgo nga las skyes par// ‘khrul pas chos rnams kyis mi bsgul// |244| gang gis kyang ni mi shigs pa// ji srid ‘di ni lus gnas pas// las dang las min rab byed pasf// brtan po ‘jigs pa rab spangs pa’i // mi shigs thig le bsgoms par bshad// |245| tshangs pa’i spyod pa bzhi po dang// phyi rol stong byasg la sogs pa// rang gi lha ru skyed nas kyang// a kyi] D C, kyis P S b gnyid] sugg. em. based on V (D and P), nyid D C P N S c ‘byung] D C, byung P S d la] sugg. em. based on V (D and P), las D C P N S e rgyu] sugg. em. based on V (D and P), rgyur D C P N S f pas] sugg. em. based on V (D and P), pa D C P N S g byas] P N S, bcas D C, 333 phyag rgya bzhis ni rgyas btab pa// |246| snga ma bzhin du shes byas nas// rim pa ‘di yisa thig le la// rang gi sems ni yang dag bzungb// rang gi ye shes sems dpa’ yi// |247| mtshan ma’i dbus gnas sa bon las// ‘od zer lnga ‘phros de yi rtse// dkyil ‘khor ‘khor lo dpag med sprul// de yis khams kun khyab byas nas// |248| dbugs kyi rta zhon kun rtog rnams// rnam par bsal te de dag kyang// dkyil ‘khor ‘kho lor rab byas te// gnyis pac dkyil ‘khor bsdu [11b] byas nas// |249| ye shes ngo bo sna bug g.yon// zhugs byas mtshan ma’i dbus gnas bya// sa bon la ni thim pas kyang// de ni phyogs dus gnas pa yi// |250| sangs rgyas yon tan ci gsungs pa// yang dag bskyedd byed mi shig pa’i// thig le ‘od zer lnga ‘bar ba// rnam par gsal byas de las kyang// |251| ‘od zer ‘byung bas rang gnas kyang// ye shes sems dpa’i nang gsal te// de yi phyi36 dange dkyil ‘khor dang// gnasf ni gsal bya de la byung// |252| mthe bo’i rtsa dang rje ngar dang// brla gnyis dang ni gsang ba’i gnas// lte ba dang ni gsus pa dang// nu ma gnyis dang lag rtse dang// |253| mgrin pa dang ni mchu gnyis dang// ‘gram pa gnyis dang mig gnyis dang// rna ba’i rgsa dang spyi bo dang// tshigs kun thig le āḥg sogs kyi// |254| a yis] P N S V (D and P), yi] D C b bzung] D C P N V (P), gzung S V (D) c pa] sugg. em. based on V (D and P), pa’i D C P N S d bskyed] P N S, skyed D C edang] P N S V (D and P), yi] D C f gnas] P N S V (D and P), nang D C g āḥ] D C, a P N S 334 gzugs su gnas pa bkuga byas nas// mi shigs thig ler thim par bya// der ni rang gi sems cung zad// gzungb bas rten dang rtenc can gyi// |255| ‘khor lor mos37 byas de dbus su// mi shigs thig le ye shes la// snang ba’i ngo bo bsgom par bya// de yi ‘od gyis rang gnas sogsd// |256| gsal byas ye shes thig le yi// nang gsal gzhage pa’i ‘od zer gyis// phyogs bcu ‘jig rten khams bzhugs pa’i// spyang ngar bdud rtsi zhu byas nas// |257| zhal du zhugs nas snying ga yi// mi shigs thig ler rab song ste// kṣuṃf gzugs bdud rtsig rab blangs nas// rdo rje’ih lam nas phyung nas su// |258| ‘o ma’i rgyun ltar phyogs bcu nas// ‘ongs pa yis ni sems can dang// sangs rgyas rnams dang mi rgyur bcas// yang dag bsdus nas gur sogs zhugs// |259| de bzhin chos dbyings phyag rgya dang// rten gyi ‘khor lor zhugs byas nas// de dag yang ni rab bsdus te// dkyil ‘khor du ni rab bsdus nas// |260| de yang bdag la bsdus byas te// bdag nyid kyang ni mi snang bar// ye shes sems dpa’i gnas la dmigs// de yang ‘khor lo ‘khor lo yang// |261| ye shes sems dpar zhugsi par bya// ye shes sems dpa’ mi snang nas// mtshanj ma ‘ba’ zhig la [12a] dmigs bya// a bkug] sugg em. based on V (D and P), bsgrub D C, sgrub P N S b bzung] P N S V (D and P), gzung D C c rten] C V (D), rtan D, brten P N S V (P) d sogs] D P N S, so C e bzhag] P N S V (P), gzhag D C V (D) f kṣuṃ] D C P N, kṣu S g rtsi] sugg. em. based on V (D and P), rtsir D C P N S h rje’i] D C, rje P N S i zhugs] D C V (D and P), gzhug P N S j + pa N 335 mtshan ma de yang bsdu byas nas// |262| mi shig thig ler sems bzung ste// de las yang ni nang gi tshogs// kun bsdus rang snang thig le che// ‘ba’ zhig la sems yang dag gzunga// |263| ci ltar nus bzhin bsgom pas der// sems ni yang dang yang du zhugs// yang ni de las phyir byung ste// de yi gnas dang rang gi lus// |264| dmigs nas dkyil ‘khor ‘khor lo dang// rten dang khams gsum snang bar bya// de nas yang ni gong ma ltar// rim parb bcug nas thim byas la// |265| sems ni thig ler gzhagc par bya// rang gi dbang po der bzung nas// sa yi dkyil ‘khor chu la zhugs// chu de me la de bzhin zhugs// |266| me yang rlung la rab tu zhugs// rlung yang sems la zhugs par gyur// sems ni gnyis med ye shes su// cung zad zhugs pas rtagsd gnas pa// |267| smig rgyu ‘dra dang du ba dang// mar me mkha’ snang rab ‘dra ba// sprin med name mkha’ ltar snang ba’i// mtshan ma lnga po ‘byung bar ‘gyur// |268| rdo rje sems dpa’ zhugs phyir ro// thig le la ni sems bzung ste// glal zhing dgodf dang ‘dar la sogs// rnal ‘byor de la gang tshe ‘byung// |269| thig le yang dag spro byas nas// gong ma’i rim pas kun khyab bya// ‘di ni yang dag goms byas pas// mi gnas mya ngan ‘das pa che// |270| a gzung P N S V (D and P), bzung] D C b par] D C V (D and P), paP N S S c gzhag] D C V (D), bzhag P N S V (P) d rtags] D C, brtagP N S S e C om. f dgod] D C V (D) , rgod P N S V (P) 336 thabs kun gyia ni bsgrub bya mchog// rdo rje ‘chang ba chen po thob// ‘di ni lhan cig skyes pa yi// ye shes ‘ba’ zhig dbang byas nas// rim pa gnyis pa bsgom pa’i thabs// mi shigs thig le38 bsgom pa’o// |271| de ltar chos kun de bzhin nyid// dpag med dam pa’i phun sum tshogs// de dmigs yang dag skyes ‘gyur ba// bsgom pa’i cho gab rnam bshad nas// |272| rnam grangs dag kyang yongs su bstan// de bzhin nyid dang yang dag mtha’// bsam gyis mi khyabs pa yi dbyings// chos nyid dang ni chos skyon medc// |273| stong pa nyid dang mtshan ma med// smon pa med pa nyid dang yang// nyon mongs khur chen ‘bord byed pa// skye ba med dang ‘od gsal [12b] ba// |274| mngon par byang ni chub pa dang// gzhan gyi sems shes byed pa dang// lha yi rna ba ster ba dang// de bzhin lha yi mig ster39 dang// |275| rdzu ‘phrul dpag med sprul chen dang// dngos po mthar ni phyin pa dang// don dam pa yi bden pa dang// rdzogs pa yi ni rim pa dang// |276| yongs su dag pa’i sku gtsang dang// kun gyise bsten bya nyid dag dang// mkha’40 ltar rnam par dag pa dang// glo bur dri mas mi gos dang// |277| gdod nas ‘od gsal ba nyid dang// gang gis kyang ni mi shigs dang// dngos po med pa nyid dang ni// rgyu sogs bcu gnyis ‘byung byed dang// |278| dpal chen ye shes yongs dag dang// a gyi] P N S, gyis D C b cho ga] D C P S, mchog N c + ngag N d ‘bor P N S V (D), ‘dor D C, por V (P) e gyis] D C V (D and P), gyi P N S 337 thig le chen po yongs dag dang// sangs rgyas kun gyi gsang chen dang// nam mkha’ nam mkha’i spyod yul dang// |279| bsgom pa med pa nyid dang ni// rje41 btsun man ngag chen po dang// rnaa nas rna bar ‘pho byed dang// nyan thos kyis ni shes min dang// |280| rang sangs rgyas sogs42 mi shes dang// yi ge med pa nyid dang ni// tshig dang bral dang brjod med sogs// de la mdo dang rgyudb rnams las// |281| de ‘dra rnam pa mtha’ yas pa// gsungs shing yang dag gsung ‘gyur ba// der ni de bzhin nyid ‘di las// gzhan ni ci yang ma gsungs so// |282| de bas dad pa sngon ‘gro ba’i// sems kyisc chos kun de bzhin nyid// zab gsald gnyis med don chen po// rim pa gnyis pa’ie de kho na// |283| bla ma’i gsung ni yang dag gzung// de gzungf gong ma’i rim pa yis// rtag tu gomsg byed skyes bu gang// de la brtenh pa’i rtags skyes nas// |284| sa nas sar ni ‘phar ba ltar// bloi yi rnam pa gong du ‘phel// rig ‘dzin la sogs sprul pa yis// gzhan ni don de la sbyor byed// |285| de yis brtanj ‘gyur rnal ‘byor pa// sgrub pa’i bdag nyid cank du ‘gyur// bdag nyid can des bsgrub pa’i mchog// ji skad gsungs pa brtsam par bya// |286| a rna] P N S V (D and P), sna D C b rgyud] D C P S, brgyud N c kyis] D C P S V (D and P), kyi N d zab gsal] D C, gsal zab P N S V (D and P) e pa’i] P N S V (D and P), pa D C, f gzung] P N S V (D and P), bzung D C g goms] D C S N, gom P h brten] D C, brtan P N S i blo] P N S V (D and P), de D C j brtan] D C, bstan P N S k can] D C P S, tsam N 338 lha mo klu mo gnod sbyin mo// mi mo mi ‘am ci mo’am// mkha’ ‘gro ma la sogs pa rnams// rang gi nus pas brtsama par bya// |287| smyon pa’i brtul zhugs la sogs pas// zla drug sogs par rab tu ‘bad// de yis rje brtsun [13a] yigb med pa// dpal ldan sangs rgyas kun ngo bo// |288| rdo rje ‘dzin pa thams cad gnas// chos kun de bzhin nyid zab pa// sangs rgyas rnams kyi bsgrub bya mchog// gong ma’i rim pa rnal ‘byor pasc// |289| ‘bad pa che thang rab byas pas// jid ltar de ru mtshan nyid kyi// tshul du cung zad rab gnas pa’i// dga’ dang dga’ ba bar ma dang// |290| dga’ dang bral ba’i dus su yang// ji bzhin gnas pa’i dga’ ba gsum// ji skad gsungs pa thob par ‘gyur// |291| de nas mtha’ med ‘khor ba’i bar// rtag pa dang ni mi gdung dang// bsil ba dang ni gcig pa dang// bde ba dang ni dri med dang// |292| dga’ ba dang ni yid dga’ ba// ‘di ni bde chen ro myang brgyad// bdag po rdo rje ‘change ba mchog// rdzogs pa’i rim pa’i rnal ‘byor paf// |293| bya ba byas pa byed pa byas// dbang phyug chen po khur chen bor// tsher ma dkrugs pa thams cad mkhyen// skyes bu dpa’ bo cangg shes pa// |294| a brtsam] D C P N, btsam S b yig] P N S, yid D C. This is also strongly supported by Vaidyapāda’s commentary, which provides a gloss of yi ge med pa (Sukusuma, 126b.7). c pas] D C, pa P N S d ji] sugg. em. based on V (D and P), de D C P N S e ‘chang] D C P S, chang N f pa] P N S, pas D C g cang] D C P N, kyang S 339 glang po chen po dul ba’oa// ‘khor ba’i mtha’ yib pha rol son// don dam dang ni kun rdzob kyi// bden pa gnyis med rnal ‘byor che// |295| tha bac spangs pa’am lam kun rdzogs// yon tan ‘byung gnas kun du bzang// thams cad ‘dus pa’i de kho na// sku sogs khams kun ‘gengs pa’o// |296| gong na med pa yang dag gnas// de ltar de sogs rab grags pa// ming ni rnam pa mtha’ yas pa// gnyis med ye shes mtson pa ste// mdo dang rgyud rnams thams cad du// blo ldan rnams kyis rtogs par bya// |297| sa bcu dang ni rab ldan pasd// gong43 ma’i lha mo rdo rje ‘dzin// dge ba bcu gsum la gnas pa// skad cig gis ni yang dag tu// sgrib bral de yis rtogs par byed// |298| phreng pa ha ra nu pure sogs// mdog dang ku tsa la sogs pa// padma rtags pa’i dga’ shes pa// dang por rab tu shes par bya// |299| ṣa dzdza ṛī ṣa ni ṣāf sogs// glu byangs bstod dang sidg sgra yi// snyan pa’i dbyangs kyis dga’ ‘gyur bas// gnyis pa ru ni shes par bya// |300| tsandanh la sogs sna tshogs dris// lus byugs lhan cig [13b] rtsen byed tshe// dga’ ba yang dag thob byed pa// gsum pa ru ni shes par bya// |301| a pa’o] P N S V (D and P), pa ‘o C, pa po D b yi P N S, ni] D C. This is supported by Vaidyapāda who glosses the phrase ‘khor ba’i mtha’i pha rol son (Sukusuma, D 127b.3) c tha ba] P N S V (D and P), thab D C d pas] sugg. em., pa’i D C P N S. My emendation here is based on the same phrase used earlier in the text in verse 118. e pur] D C, phur P N S, f ṣa dzdza ṛī ṣa ni ṣā] sugg. em. following V (D), ṣa dzdze rī ni ṣā na D, ṣa dzdza rī ni ṣā na C, sha rdzas gri tra gri na P N S), sha rdza gri ta ghri na V (P) g sid] P V (D), sing D C S N V (P). This same term is used in verse 110 of the text for a kind of buzzing sound that is made, presumably with the mouth against the partner’s body, to produce pleasure. h tsandan] D C, tsan dan P S, tsadna na N 340 ma mchu sbrang rtsi gzhiba byas pas// mgor gnas byang chub sems ‘ju bas// ro ‘thung dga’ bas bdag mnyes pas// bzhi par rab tu shes par bya// |302| lus la byug cing sna tshogs kyi// spyod pas rtsen tshe reg bya yis// yang dagb dga’ bar byed pas na// lnga pa ru ni shes par bya// |303| de yis ye shes rnam pa gsum// rig par byed cing rang gi yid// yang dag dga’ bar rab byed pa// drug par shes bya rnal ‘byor pas// |304| de yi lus ni bdag gi ni// rten du gnas pa sra ba yis// yang dag dga’ bar byed pas na// bdun pa ru ni shes par bya// |305| de yi padma’i zil sogs dang// byang sems rlan gyis rang gi sems// rab tu dga’ bar byed pas na// brgyad par rab tu shes par bya// |306| gsang gnas drod sogs tsha ba yis// bdag gi yid ni yang dag par// dga’ byar byed pa’i mchog yin pas// dgu pa ru ni shes par bya// |307| de nas bskyod pas ye shes mes// phung po khams sogs sreg byed pas// yid ni yang dag dga’ ‘gyur pas// bcu parc rab tu shes par bya// |308| bcu po de yis dang po dang// phyis kyi ‘bras bu mchog ‘gyur ba// gongd du gsungs pa rab thob ste// de bas de ‘dra’i don chen la// |309| yang dag ‘jug par mi nus pa’i// gdul bya rnams la bde gshegs kyis// rab tu dga’ sogs mtshan nyid du// a gzhibs] P N, bzhibs S, gzhib D C b C om. c par] D C P S, pa N d gong] P N S, gang D C 341 bstan nas de yi don zhugs pas// |310| de yis rtogs kyang bla dang bcas// de nyid dang po’ia rnal ‘byor la// rten dang rtenb can dkyil ‘khor gyi// ‘khor lo rab tu bstanc byas nas// |311| de zhugs de la brtan byas pas// rtogs kyang don ‘di ma shes na// yang dag sangs rgyas ma yin no// |312| ‘did ni rang byunge bcom ldan ‘das// gcig pu rab tu che ba’i lha// lhag pa’i lha zhes bya ba ni// bcu gsum saf zhes bya bar bshad// |313| de ltar rim pa gnyis pa yi// de bzhin nyid ni thabs bcas pa’i// bdud rtsi mchog ‘di rnal ‘byor gang// ‘thung bar byed pa sangs rgyas kyi// |314| de yi sras su nges pa ste// byang chub sems dpa’ rnam kyi [14a] grogs// rig pa ‘dzin pa’i dpon po ste// mkha’ ‘gro ma yi khyo ru ‘gyur// |315| nyan thos rang ‘dren rnams kyi ni// ‘dren par byed pa’i gtso bo ste// sems can phal pa’i rje brtsun no// de la phyogs bcu khams bzhugs pa’i// |316| sangs rgyas byang chub sems dpa’ dang// rig pa’i lha dang khro bo sogs// me tog la sogs mchod par bcas// zhag gcig skyag rengsh dang po dang// |317| phyed ni dros pa’i dus nyid dang// chal chil mtshams su nam mkha’ las// mchod pa byas nas sngags brjod cing// rang gi zhing du ‘gro bar ‘gyur// |318| a po’i] D C, po P N S b rten] D C V (D and P), brten P N S c bstan] P N S V (P), brtan D C V (D) d ‘di] D C S N, de P e byung] D C, ‘byung P N S f sa] P N S, pa D C g skya] D P N S, skye C h rengs] D C, reng P N S 342 de bas dag pa’i lha rnams kyang// de bzhin du ni mchod byed na// ma dag pa yi lha rnams ni// de la ci yi phyir mi mchod// |319| sems can phal pa gzhan gyis ni// me tog snyim pa bkang nas ni// mgrin paa dma’ bas zhabs ‘dud cing// rtag tu bkurb bar shin tu ‘o// |320| de ni skyes bu rkang gnis gtso// kun mkhyen rnams kyis bzhag pa ste// phyin log khur chen bor byed pas// ma ‘ongs rdo rje ‘chang ba ste// |321| de la brnyas byed sems can gang// sgrub pa po de glang dang ‘dra// de yis nga la brnyas byed pas// ngas ni de kun dus kun spong// |322| cig shos lus la nga gnas phyir// mchod cing brjod pas lus rnams kyi// grib pa de yis dag byed do// |323| rim pa gnyis ‘dir sgrub pa po// rtse gcig pa yis rnam par gzung// de gzung tshe ‘di nyid la yang// brtul zhugs cho ga nyid kyis ni// |324| lha mos rkyen ni rab byas te// phyag rgya chen po rang rgyud la// ‘pho bar the tshom mi bya’o// |325| da ni rang nyid byin brlabsc pa’i rim pa ‘di ni rnal ‘byor pa// bya bas bskal pa ‘ga’ zhig la// yang dag tu ni bshad par bya// |326| gang zhig bla ma mnyes byas nas// des gnang dam sdom44 bcas ba ru// bum pa la sogs rab thob ste// bla ma’i zhal las rnyedd pa yi// |327| de bzhin nyid ni rab thob cing// a pa] D C P S, par N b bkur] D C, bskur P S c brlabs] D C V (D), brlab P N S V (P) d rnyed] D C P N, rnyes S 343 gsang dang mchog tu gsang rigsa pas// ji skad bshad pa’i bya ba yis// yang dag bsgom ni mi nus pas// |328| ji bzhin pa yi de kho na// rim pa ‘di [14b] yis bsgom par bya// gang zhig dus ni phyi zhig la// ‘chi ba’i mtshan ma bdag gis mthong// |329| ‘chi bar gyur pa’i dus byung na// nad kyis yang dag ma rnyogs par// ‘pho ba’i sbyor ba yang dag bya// |330| dpral ba dang ni ste ba dang// spyi gtsug dang ni mig dang ni// rna ba dang ni sna dag dang// chu yi gnas dang chu min gnas// |331| kha yi gnas dang ye shes kyi// ‘gro ‘ong gib ni rtags shes bya// dbral ba gzugs kyi khams kyi ni// rtags skyes ba ru shes par bya// |332| lte bar ‘dod khams lha rnams kyi// rtags byung de ru skye bar nges// spyi gtsug gzugs med khams rnams kyi// rtags byung de ru shye bar ‘gyur// |333| snac gnyis ye shes ‘pho ba na// gnod sbyin gnas su skye bar ‘gyur// rnad gnyis rig pa ‘dzin pa yi// gnas su nges par ‘gro ba’o// |334| mig gnyis mi rnams rgyal por ‘gyur// rtags ni yang dag skye bar ‘gyur kha ru ye shes ‘pho ba na// yi dwagse rtags su shes par bya// |335| chu gnas dud ‘gro rnams kyi nif// rtags su yang dag rab shes bya// chu min bu gur ye shes ‘gro// dmyal ba’ia rtags su shes par bya// |336| a rigs] D C, rig P N S b gi] P N S, gis D C c sna] P N S V (D and P), rna, D C d rna] P N S V (D and P), sna D C e dwags] sugg. em. , dags D C P N S V (D and P) f ni] D C, rtagP N S S 344 de ltar ye shes ‘pho ba yi// rnamb ba’i cha ni shes byas nas// sgrogs pa lnga yi ye ge yis// steng gi sgo bdun dgag par bya// |337| chu yi sgo ru suṃc gisd dgag// chu min kṣuṃ gis dgag bar bya// de ltar sgo dgu bkag nas kyang// rim pa ‘di yis rang sems kyie// |338| gnas ni rab tu btsal byas pas// dbyings nyid du ni nges par ‘gro// phung po skye mched khams rnams su// rnal ‘byor rgyud kyis dpyad pa ltar// |339| rnam par bsgom byas ye shes gyisf// gong na med pa rtogs par ‘gyur// bdag nyid ji skad gsungs pa bzhin// lha yi lus su gyur byas lag// |340| chos kyi dbyings dang rnam shes dang// bdag nyid sangs rgyas skur byas pa’i// gnas ni rin chen sna tshogs pa’i// kha dog sna lngas rab brgyan pa’i// |341| rdo rje rtse mo dgu pa mchog// bdag gi spyi bo’i steng du ni// bsams nas de ru rang gi sems// rdo rje dkar po bcu tshal tsam// |342| rtse lnga bsams [15a] nas brtag byar bya// yar gyi rtse lnga thabs lnga ste// de bzhin mar bltas shes rab lnga// dbus kyi bum barh ri bong can// |343| bsams nas de ru ye shes che// ser po byang chub sems ‘dra ba// ‘dzag pa’i ngangi tshul dang ‘dra ba// a ba’i] D C, ba P N S b rnam] D C P N V (D and P), rna S c suṃ] D C V (D), sūṃ P N S, sum V (P) d gis] D C S, gi P N e kyi] D C V (D and P) , kyiP N S S f gyis] sugg. em. based on V (D and P), so D C P N S g la] P N S, pa D C. Vaidyapāda also supports this reading: lha’i lus su gyur par byas la// (Sukusuma, 131a.7). h bar] sugg. em. based on Vaidyapāda’s commentary, ba D C P N S. Vaidyapāda reads dbus kyi bum pa ni de’i dbus kyi bum par ro// (Sukusuma, 131b.4). i ngang] D C S, dang P N 345 thig le tsa na ka yi ‘bru// |344| lnga ‘byor tshad du bsgom par bya// de nas chos kun thams cad kun// med par bsams nas bdag dmigs te// bdag kyang bdag la thim byas na// |345| sems tsam de yang mda’ bzhin du// spyi bor rdo rje ‘og cha yi// sbua gur zhugs nas thig lerb thim// de ni de bzhin gshegs pa rnams// |346| lha mo rnams kyi rang bzhin gyi// gnas su shes par rab tu bya// de ni gdod nas grub pa’i gzugs// de ru shes pa rab zhugs pa// |347| sems de yang dang yang du gzung// de las sems de yar ba ltar// rdo rje rtse dgu steng cha yi// sbuc gur byung nas sna rtsogs kyi// |348| padmar zla ba gnas pa la// rdo rje sems45 dpa’i skur gyur te// mtshan dang dpe byad kun brgyan pa// rgyan kun brgyan pa’i na bza’ dang// |349| bral nas shin tu gsal dmigs te// de gnyis med pa’i sbyor ba yis// byin rlabsd las byung ‘khor lo che// rnam par ‘phro zhing ‘du bar bsam// |350| de ltar rnam par bsgom pa mchog// yang dang yang du ci nus bsgom// bsam pa de yi rkyen gyis ni// bdag semse dbyings su zhugs pa’i tshe// |351| gsal zhing rab dga’f mkha’ ‘dra rtogs// de nas rdzu ‘phrul shugs ldan pag// lo lnga lon pa’i byis pa’i gzugs// ‘grub tshe dpe med bde rdzogs rtogs// |352| a sbu] D C V (D and P), bu P N S b ler ] D C V (D and P), le P N S c sbu] D C, bu P N S d rlabs] D C, brlabP N S S e + can N f dga’] P N S V (D and P), dag D C g pa] D C V (D and P), pa’i P N S 346 de las skye gnas gzhan du ni// ‘phen para byed tshe sprul pa’i gzugs// yang dag tu ni rtogs par ‘gyur// |353| de lta bas na sems can gyi// bsam pa gang dang gang dag gis// mi rnams yid ni yang dag sbyong// de dang de ni mdo46 bzhin du// sna tshogs nor bu lta bur ‘gyur// |354| chos sku rab dga’ mkha mnyam pa// shi dang brgyal dang gnyid log dang// glal dang ‘khrig dus skad cig tsam// myong bar ‘gyur bas rab bsgoms na// lus can rnams ni yid ni sbyong// |355| cho ga ‘di yis sems can gang// [15b] mtshams med byed cing rmongs pa yi// skye bo bram ze gsod pa yang// de yis mi ‘grub cung zad med// |356| de bas bka’ lung rab thob nas// dam sdom rnam par bsrung byas te// sku gsum ‘grub par the tshom med// gal te sku gsum ma grub na// |357| rig pa ‘dzin pa’i gtsor ‘gyur te// rim gyis phyag rgya chen por ‘gyur// de ltar rnam pa gsum gyis ni// dngos grub de ni bsgom pa bshad// |358| gang zhig ‘di ni ma rtogs par// gsang ba smra bar byed pa ni// de nib nga dang de bzhin gshegs// nam zhig de la sbyar byas pas// byin gyis rlob par mi byed do// |359| ‘di ni don ldan ‘ga’ zhig la// nga ni de yic lus gnas te// bsgrub pa gzhan lasd mchod pa len// de yis de mnyes rang gi rgyud// las kyi sgrib pa dag byed do// |360| a ‘phen par] P N S V (D and P), ‘phel bar D C b ni] D C, la P N S c yi] D C P S, yis N d las] sugg. em., la D C P N S V (D and P) . I suggest this emendation only to gramatically align with the verb len. 347 ‘di don rna bar ‘pho byed pa// gang gi dus su gnas par der// sangs rgyas bstan pa rin chen yang// gnas para rab tu bshad pa yin// |361| brgyud pa’i rim pa ‘di chad nas// sangs rgyas bstan pa nub pa ru// kun gyis rab tu shes par bya// |362| de bas khyod kyi yid gcig tu// yang dag bsdus la ma ‘ongs pa’i// gang zag47 sngonb du tshogs bskyed pa’i // skal ldan ‘ga’ la rnam par brgyud// sim byed ‘di la sbyor du chug// |363| ‘di sbyor ‘di ni yang dag pa’i// rnal ‘byor pa ru shes par bya// |364| khyod kyang zad kyi spyod pa dang// nga la cung zad ‘khrul rtogs pas// khyod kyisc tshe ‘di nyid la ni// gzugs bcas phung po rang lus ni// gnas ni yang dag mi ‘gyur te// |365| rnam par shes pa mi shigs pa// phyag rgya chen por rab tu ‘grub// |366| de bas khyodd kyis yang dag blos// sangs rgyas kun gyi ‘dus pa’i rgyud// gsang chen gsang lae ches gsang ba// gong na med pa’i lung chen po// ‘di yi rim pa dang po yi// |367| sgrub pa’i thabs dang sbyin sreg dang// gtor ma sna tshogs ‘khor lo dang// bsdus pa dang ni rnam bshad dang// dkyil ‘khor cho gaf la sogs pa// |368| mi shes mun48 bsgribsg sems can gyi// thur ma lta bur brtsam par gyis// |369| a par] P N S, pa D C b sngon] D C, mngon P N S c kyis] D C S, kyang P, kyi N d khyod] D C, chod P N S e la] P N S V (P); las D C. V (D) appears to be corrupt and reads gsal ba. f cho ga] D C P S, mchog N g bsgribs] D C V (D and P), sgrib P N S 348 de bas ma ‘ongs [16a] rnal ‘byor ches// ‘di shes bla mar rab mnyes te// yang dag ‘dod pas blangs byas nas// sems ni de nas sbyang bar bya// |370| sems ni kun rtoga nam spangs nas// de sbyorb de la rig pa skye// rig pa skyes nas rdo rje ‘dzin// dec grub pas na sangs rgyas dang// |371| pha rol phyin dang gzungs rnams dang// sa rnams thams cad bde chen po// rjes las yang dag ‘grub par ‘gyur// thams cad bde chen las byung phyir// |372| de bsgoms ci phyir de mi ‘byung// de bas rab tu ‘bad pa yis// mchog gi de nyid rab gsang gsang// smon lamd gyis kyang tha na ‘grub// bsgrub pa la ni shin tu ‘bade// |373| a la laf ho// de ltar rol pa’i gar bcas sgra brnyan lta bu rdo rje’i glu// ‘khor lor bcas pas de nyid glu bslangs bdag la bstod cing derg// nam mkha’i khams su sprin rnams med pa lta bur thim par gyurh// dge slong bla ma gnyis bcas de yang de bzhin mi snang gyuri// |374| rdo rje gdan gyi rgyabj nas rgyang grags lnga bcu song ba’i sar// parbak ta yi phug la brtenl te sems can don bya’i phyir// ‘di bsdus rab tu byed pa thams cad rtsom dang ston sogs byas// damm pas gsol ba rgya chen btab pas bdag ni shin tu brod// |375| der gnas ‘khor bcas rnams kyis yo byad gos zas nor gyi mdzod// mchod pa’i yo byad rgya chen sna tshogs ‘khor lo bya ba rnams// sa bcu‘i byang chub sems dpar gyur pa mdzod srungn gnod gnas che// a rtog] P N S V (D), rtogs D C V (P) b sbyor] D C P S, sbyong N c de] D N V (D), nga C P S, V(P) om. d lam] P N V (D and P), las D C, la S e ‘bad] D C V (D), ‘bod P N S V (P) f P N S V(P) +la g P om . h gyur] D C, ‘gyur P N S i gyur] D, ‘gyur C P N S jrgyab] D C V (D), ‘gab P N S V (P) k parba] D C V (D), ra ba S, par ba P, par pa N, spar ba V (P) l brten] C P N S V (D and P), brtan D m dam] C P N S, ngam D n srung] D C V (D), gsung P N S V (P). 349 nyin re kārshāa pa ṇa bdun brgyas rtag tu rab tu sbyor// |376| de nas bla ma chen po bā li pā da’ib drung du bgrod// bdag gis bla ma de yang mnyes bya’i phyir na sgrub pa’i thabs// cung zad bsdus pas de ru bla ma la sogs kun// mnyes par byas te sngon gnas bgrod nasc skal ldan don ‘ga’d byas// |377| de bas de ltar kun gyis gtam rgyud rgyas par shes byas te// mkhas pa’i gzu bo dam pa thabs kun gyis ni mnyes byae ste// de yi lung la rab tu byed sogs mnyanf dang bsam par bya// |378| de la rab brten dgon sogs rab tu brten byas rang gi sems// de nyid bsgoms pas ji bzhin rab tu rtogs par byas pa yis// tshe ‘di nyid [16b] la zla ba drug gis byang chub thob pa ’di ni su yis bzlog// |379| bdud rtsi mchog ‘thung ‘di ni sems can kun gyis bkur ba’i gnas// sems cheng de la rdo rje ‘chang dang bdeh gshegs kun gyis bsngags// de’i phyir log pa’i sgribi pa thams cad sa bon nyams byas te// ‘khor bar gnas kyang skyon gyis mi gos pad bzhin rtag tu gnas// |380| de ru ma thob lan grangs gzhan la yang dag myong ba yij// ‘bras bu smin byed sngags pa’i yid la yang dag ‘byung ba yi// dngos grub rgya chen thob nas gang gā’ik bye snyed dpag med kyi// ‘khor gyis bskor nas ‘jig rten khams kun thams cad rab tu bgrod// |381| de bas chu ‘khor drag po khar ltung ltar// lus dang ngag dang sems kyisl brtson ‘grus bskyed// rim par sngon bzhin rab tu bsgrub par bya// |382| dal ‘byor lus ‘di shin tu g.yo slam bas// ji ltar rlung gis mar me gsod pa ltar// skad cig tsam du mi sdod dus ‘da’ byed// de bas ‘dir ni chud gsan mi bya ste// dngos kun de nyid mchog chen rab bsgom bya// |383| de nyid spyod pa ‘di la blo nges49 pas// a kārshā] D C V (D), karsha S N V (P), ka rā P b bā li pā da’i] D C , bha li pa trī P N S c nas] P N S, gnas D C d ‘ga’] D C V (D and P), dga’ P N S e bya] D C P S, byas N f mnyan] D C, mnyam P N S g chen] P N S, can D C h bde] D C, bder P N S i sgrib] D C P S, sgribs N j yi] P S, yis D C k gang gā’i] D C, gangā’i P N S V (D and P) l kyis] D C, kyi P N S m sla] D C P S V (D and P), bla N 350 ji skad bshad bzhin rab tu ‘gyur bas na// sems ni yang dag brtena par gyis shig ces// bdag ni sems can rnams la gsol ba ‘debs// |384| gang gi phyir na ‘di nib gsang bas na// bdag gisc gsal bar ma byas ci yod pa// des na de shes bla ma yang dag par// mnyes byas de nyid ‘dod pas yang dag long// |385| ‘di ni rnal ‘byor spyod pas yang dag spyod pa la// snying po’i snying po ‘bras stsold yang dag sbyonge// mtha’ yi phar son bla chen yon tan ma lus gterf// de las ‘di ni yang dag zhal rnyed byas// |386| de bas rnal ‘byor chen po rgyud kyi don rnams ni// mi ‘tshamg par ni snang yang nyi ma’i dpes// rang dang gzhan gyi don byed nus pas na// rnal ‘byor ‘gas ni shin tu ‘bad dgos so// |387| the tshom som nyi med par dad pa yis// ‘di don grub pas yang dag blangsh byas nas// shes rab kyis ni yang dang yang du rangi// spyadj pas bdag la gnyis med ye shes che// chu gtsang nang du zla ba’i dkyil ‘khor ltar// ‘byung ‘gyur ‘di la the tshom ma byed cig// |388| ‘di ni rang bzhin grub pa’i man ngag [17a] las// brgyud pa’i bla ma yang dag rab bsten dang// rang gi bsod nams tshogs ni sngon bskyed pas// rtogs par rab tu ‘gyur ba ma gtogsk par// |389| bsod nams chung ba’i mi yis bskal ba dpag med par// ‘di ni rtogs par mi ‘gyur ‘di don ma rtogs na// rnal ‘byor chen po zhes bya de la mi bya’o// |390| de ltar rab tu shes par byas nas su// rdzogs pa chen po ye shes spyi yi gzugs// yongs su dag sku rdo rje ‘chang chen po// a brten] D C, bstan P N S b ni] D C, na P N S c gis] D C P S, gi N d stsol] sugg. em. based on V (D and P), gsol D C P N S e sbyong] P N S V (D and P), spyod D C f gter] D C rten P N S g ‘tsham] D C P S V (D and P), mtsham N h blangs] P N S V(D and P), blang D C i rang] P N S, rung D C j spyad] P N S V (D and P), dpyad D C k gtogs] sugg. em., rtogs D C P N S 351 dpal ldan kun gyi ngo bo rima gnyis ‘dib// |391| sdug bsngal lam bcas bskal pa gsum du yang// rang gi rjes mthun byang chub bla dang bcas// thob nas de bde cung tsam rab chagsc pa// rnal ‘byor ded yis ci phyir de mi bsgome// |392| dad dang brtson ‘grus ting ‘dzin shes rab dang// dran pa’i blo yis gong gi rim par ltar// yid dga’ gnas brten kun tu50 bzang po mchog// lam ‘di bsgom par bya baf kho na’o// |393| de ltar zhal gyi lung bsdus las byung dge ba dri med pa// kha ba zla ba’i ‘od ltar rab tu kdar ba de yis ni// ‘di don ma ‘ongs ‘gro ba skal ldan ‘ga’g dang phrad gyur nas// rab tu dang ba’i dad pas len cing ’di mchog rab bsgomh shog// |394| blugs dang gtor dang dag par byas pas slob dpon cher ‘gyur te// thams cad kun kyi rgyud ‘dzin gzhan rnams rgyud kun la sbyor ba’i// dang po’i rim pa rab rtogs dri ma rnams ni dag byas te// ye shes sgyu ma’i snod du rung bar rnal ‘byor de ‘gyur shog// |395| snying rje ldan pa’i bla ma’i zhabs la gus par rab ldan pas// ri bong ‘dzin pa’i gzugs kyis rang rgyud rab tu smin byas te// zhing dag byas pas chos kun sgyu sogs don du rab rtogs nas// byams pa la sogs bzhin du sems can kun gyisi ‘gro bar shog// |396| lhan cig byed pas byin brlabs rje btsun thugs rje chen po yis// dga’ ba brnyed pas chos kun dag pa’i ngo bo chen po mchog// ji bzhin gnas pa’i don la mi sluj bcu drug thig le cha// ngal gsok las thob man ngag chen po rab tu rnyed par shog// |397| de rnyed de la sems ni yang dag rab tu mgu gyur nasl// mgrin pa dma’ barm brtags nas rang rig chos kyi sku ni yang dag thob// sku dang gsung dang thugs dang dpag [17b] med sprul pa’i gzugs kyis ni// khams gsum bkang nas sems can thams cad srid las sgrol bar shog// |398| a rim] D C P S V (D and P), rims N b ‘di] sugg. em. based on V (D and C), ‘dis D C P N S c chags] P N S V(D and P), tshogs D C d de] D C V (D and P), ‘di P N S e bsgom] P N S V (D and P), sgom D C f bya ba] P N S V (D and P), bya’o D, bya’i C g ‘ga’] P N S, dga’ D C h bsgom] D C V(D and P), bsgomP N S S i gyis] D C, gyi P N S j slu] D C, bslu P N S k gso] P N S, so D C l nas] D C, bas S m bar] sugg. em. based on V (D and P), bas D C P N S 352 rima pa gnyis pa’i de kho na nyid bsgom pa zhes bya ba byang chub sems dpa’ ‘jam pa’i dbyangs kyi zhal gyis lung/ dkyil ‘khor gyi slob dpon chen pob sangs rgyas dpal gyi ye shes zhabs kyis bsdus pa/ zhal nas zhal du brgyud pa rdzogs so// rgya gar gyi mkhan po chen po ka ma la gu hya dang/ bod kyi lo tsāc ba chen po mnga’d bdage lha ye shes rgyal mtshan gyis bsgyur cing zhus de gtan la phab pa’o// a rim] D C P N V (D and P), rig S b om.] sugg. em., lha D C P N S. I suggest omitting lha here, though it is present in all recensions of the Dvitīyakrama. To call Buddhajñānapāda a deva would indeed be a very unusual epithet, and I believe it is more likely that the lha from Lha Yeshe Gyaltsen’s name was somehow added in front of Buddhajñānapāda’s name, as well, in a scribal error. c tsā] D C, tsa P N S d P N S om. e P N S om. 353 Endnotes: 1 de] C P N S, da D 2 dpal] D C P N, pa S 3 gro] C P N S, ‘go D 4 pa] C P N S, ba D 5 du] C P N S, tu D 6 de] C P N S, da D 7 ‘dzum] C P N S, ‘jum D 8 tshul] C P N S, chul D 9 byed]C P N S, byad D 10 ‘dzum] C P N S, ‘jum D 11 sems] C P N S, sams D 12 kye] C P N S, kya D 13 ‘o] C P N S, ‘a D 14 dgongs su] D C S P, dgongsu N. This appears to be khungs yig rather than a spelling mistake. 15 ‘gro ba’i] D C P N, ‘gro’i S 16 ‘khor lo] D C S P, ‘khoor N (This is presumably a form of khung yig, which I am not reporting in S, because it has quite a bit, but am reporting, where possible, in other recensions.) 17 nang] C P N S, nad D 18 byung] D C, ‘byung P N S 19 thogs] D C S P, thog N 20 krung] D C S N V (D and P) grung P 21 dwags] D, dags C P N S 22 dmigs] D P N S V (D and P), ‘migs C 23 tsam] C P N S V (D and P), cam D 24 stong] D C S P V (D and P), steng N 25 tsam] C P N S V (D and P), cam D 26 log] P N S V (D and P); ldog D C 27 steng] C P N S V (D and P), stang D 28 ste] P N S, te C, ta D 29 tsam] C P N S, cam D 30 bdud] D C S P, dud N 31 gzung] P N S V (P), bzung D C V (D) 32 log] P N S, ldog D C 33 me] D C S N, mi P 34 lha] D S S P, lta N 35 nges] D C S P, des N 36 phyi] C P N S, phya D 37 mos] D C S P, mas N 38 le] C P N S, la D 39 ster] C P N S, star D 40 mkha’] D C S P, mkhar N 41 rje] C P N S V (D and P), rja D 42 sogs] D C S N, sogso P 43 gong] D C S P, god N 44 sdom] C P N S, som D 45 sems] C P N S, sams D 46 mdo] D P N S, med C 47 zag] D P N S, za ma C 48 mun] C P N S, mar D 354 49 nges] D P N S, ngas C 50 tu] P N S V (D and P), du D C 355 Oral Instructions on Training in the Suchness of the Second Stage: A Translation of Buddhajñānapāda’s *Dvitīyakramatattvabhāvanāmukhāgama1 [1]2 In the Indian language: Dvitīya3kramatattvabhāvana-nāma-mukhāgama In the Tibetan language: Rim pa gnyis pa’i de kho na nyid sgom pa zhes bya ba’i zhal gyi lung 1 As the original Sanskrit for the Dvitīyakrama is not extant, this translation has been made on the basis of my critical edition of the Tibetan translation of Buddhajñānapāda’s Dvitīyakrama by Kamalaguhya and Lha Yeshe Gyaltsen, also included in this dissertation. I have included a critical apparatus in the notes to this English translation only when there was a variant significant enough to be reflected in the translation, or when I chose a reading from the Sukusuma, Vaidyapāda’s commentary on the Dvitīyakrama, rather than from any of the extant recensions of the Dvitīyakrama itself. For the full critical apparatus, see my critical edition. The notes to this English translation also contain a number of translations of passages from Vaidyapāda’s Sukusuma, the only extant Indic commentary on the Dvitīyakrama, but which is again only available to us in its Tibetan translation. All translated passages of the Sukusuma have been transcribed and edited based on the Derge (D) and Peking (P) editions of the Tengyur; these transcriptions and their critical apparatus are included in the notes to this translation. When no edition is specified, page numbers for the Sukusuma, are given for the Derge recension. Punctuation in the passages cited from the Sukusuma is always given according to the Derge recension; I have not recorded the many punctuation differences in the Peking recension of the commentary in my critical apparatus. 2 Page numbers given in brackets correspond to the Derge edition of the Dvitīyakrama. 3 dvitīya] sugg. em., dvi] P N S, dva D C. While the Sanskrit title as given in the Tibetan translations is rendered using the word dvikrama, the “two stages” (or as the nonsensical dvakrama in D and C), all of the Tibetan translations render this as rim pa gnyis pa (or gnyis ba), meaning the “second stage.” Indeed, while there is a brief summary of the first, or generation stage, practice in the text (twice, actually, constituting a total of just six verses out of the 399 verses in the text), the content is almost exclusively focused on instruction and practices connected to the second stage, the perfection stage. Given these facts, it seems indeed quite likely that the Sanskrit title of the work is given incorrectly in the Tibetan translations, and that the correct title of the text is

• Dvitīyakramatattvabhāvana-mukhāgama. There are many cases in the Tibetan canon where the titles of Sanskrit

works have been given incorrectly, so such a confusion of the ordinal (dvitīya) and cardinal (dvi) numbers in the Sanskrit title as given in the Tibetan translation is not terribly unusual or even unexpected. (And indeed, the appearance of the nonsensical “dvakrama” in D and C may perhaps suggest something in the direction of dvitīya, rather than just dvi, or at least that there is some confusion with the issue.) There is further evidence in Buddhajñānapāda’s and Vaidyapāda’s works that suggests that the title of the work to be the Dvitīyakrama. See, for example, verse 34 of the Dvitīyakrama, which specifies that the contents of Mañjuśrī’s oral instruction presented in this text are focused on the second stage, and verses 283 and 315 of this text, which also use the phrase “the suchness of the second stage,” a phrase I address in more detail in Chapter Three of the dissertation itself. What is more, Vaidyapāda uses the phrase “training in the suchness of the second stage” several times in his Muktitilakavyākhyāna in ways that clearly distinguish it from the first stage of practice. In one instance he writes, “Upholding, in this way, the samayas and vows, in order to [be able to] train in the reality of the second stage, [the text first] teaches, by means of example, the aspects of the first stage that are the basis for this...” de ltar dam tshig dang sdom pa la rnam par gnas pas rim pa gnyis pa’i de kho na nyid bsgom pa’i phyir de’i gzhi’i rim pa dang po rnams dpe’i sgo nas bstan pa... (Muktitilaka-vyākhyāna, D 51b.1-2; P 337b.3). The “aspects of the first stage” that Buddhajñānapāda goes on to explain at this point in the Muktitilaka are the practices of the four brahmavihāras, which are part of the preliminaries for generation stage practice in his system. At the end of the section Vaidyapāda again repeats the phrase, “Having [first] remained in the generation stage, [now] in order to teach the training in the reality of the second stage...” (da ni de ltar bskyed pa’i rim pa la gnas pas rim pa gnyis pa’i de kho na (kho na] D, P om.) nyid bsgom pa bstan pa’i phyir... (Muktitilaka-vyākhyāna, D 52b.3-4; P 339a.1-2). Modern scholars referring to the text have, up until now, consistenly given the title of this work as Dvikramatattvabhāvanā-mukhāgama following its rendering as such in the Tibetan translation. However, since the preponderance of the evidence suggests that this is based on a mistake in the Tibetan renderings of the Sanskrit title, I depart with this convention and refer to the work as the *Dvitīyakramatattvabhāvana-mukhāgama. 356 In the English language: Oral Instructions4 on Training in the Suchness5 of the Second Stage6 [Homage]7 Homage to the Omniscient One! To the Glorious One who has perfected bliss, who is endowed with the radiance Of nondual profundity and luminosity;8 To his nature, which is peace, the blazing sixteenth part,9 4 Presumably referencing the term mukhāgama, “oral instructions,” in the title, Vaidyapāda comments that the text is a “condensation of the words of the bodhisattva Mañjuśrī” (byang chub sems dpa’ ‘jam pa’i dbyang kyi lung bsdu ba) by “our great guru,” (bdag cag gi bla ma chen po) Buddhaśrījñānapāda (Sukusuma, D 87a.5-6; P 104b.5-6). Buddhajñānapāda himself, however, seems to understand himself to be representing more or less the entire content of Mañjuśrī’s direct speech (rather than a condensation of such), given that the text even includes second-person references, in which Mañjuśrī directly addresses Buddhajñanapāda as “you” (khyod). I discuss the topic of Mañjuśrī’s voice in the Dvitīykrama in Chapter Two. 5 Throughout this translation “suchness” translates the terms de kho na nyid, de kho na, de bzhin nyid, and de nyid, which appear to be used synonymously in this work. Though we can not know for certain since the Sanskrit text is not extant, if the translators followed the standard convention, the first two terms are presumably translations of the Sanskrit tattva, and the latter two of, tathatā, which are often synonymous. 6 On the “second stage” please see note 3. 7 Section headings within the text are given in brackets, as these are not part of the Dvitīyakrama, but are my own addition, included to provide more clarity and structure to the translation. Occasionally, when Vaidyapāda’s comments to a given section coincided with the heading I wanted to include for that section, the headings are translations of a brief line from his commentary; I have always indicated in the notes when this is the case. 8 Vaidyapāda indicates that the first line of this verse refers to the buddha, the second to the dharma, and the third to the saṃgha. de la dpal ldan zhes bya ba la sogs pa’i tshig gis ni mchod pa’i yul go rims bzhin du sangs rgyas dang/ de’i rang bzhin gyi (gyi] D, P om.) chos dang / de ston pa’i dgen ‘dun rin po che ston to (to] D, nyo P)// (Sukusuma, D 87a.6-7; P 104a.6-8). The term “nondual profundity and luminosity” (zab gsal gnyis med) is a central one for Buddhajñānapāda. He repeats it often, and it seems to be a centerpiece of his understanding and presentation of the nature of reality. I address this phrase in more detail in Chapter Three. In Vaidyapāda’s commentary on this verse he explains that nondual profundity and luminosity refers to the nonduality of the uncompounded and the compounded. zab gsal gnyis med ces pa ni ‘dus ma byas dang ‘dus byas gnyis su med pa ste/ (Sukusuma, D 87b.6-7; P 105a.8- 105b.1). 9 bcu drug phyed cha ’bar. I have a suspicion that the Tibetan translation may include phyed cha (literally “half”) instead of simply cha, “part,” simply in order to fill the meter of the verse. In verse 84 of this same text the term bcu drug cha is used to refer to the bodhicitta drop bestowed on the disciple during the guhyābhiṣeka, and in verse 122 bcu drug char gyur pa is used to indicate the drop of bodhicitta in the context of sexual yogic practices. I believe bcu drug cha may be a translation of ṣoḍaśakalā, understood in each of these contexts as “the sixteenth part,” referencing the sixteenth phase of the moon in the lunar month, the day when the fullness of the fifteenth phase is perfectly complete, and is here in all three verses used to indicate the bindu/bodhicitta drop. Alternatively, the term could be ṣoḍaśakala, referring more generally to the moon itself as “that which has sixteen parts.” Perhaps the term could have been in compound allowing for multiple possibilities, given that Vaidyapāda explains this verse multiple times on different levels of analysis, which he calls the general and branch meanings (spyi don and yan lag gi don), some of which fit better with reading ṣoḍaśakala and others with reading ṣoḍaśakalā. (Thanks to Harunaga Isaacson for his advice on this point.) First Vaidyapāda indicates that in the context of the first three lines of this verse referring to the three jewels, the term refers to the dharma in terms of “that which expresses it,” (bcu drug phyed cha ‘bar zhes pas ni mtshon byed kyi chos ston to//), indicating, presumably, that on this level of analysis he understands the sixteen to refer to the sixteen vowels of the Sanskrit alphabet as symbolic of the alphabet as a whole, which is the relative means by which the dharma is expressed or taught (Sukusuma, D 87b.2; P 105a.2). (See also the Samantabhadra/Caturaṅga-sādhana, verse 29, where the term ṣoḍaśakalā is used precisely to indicate the sixteen vowels of the Sanskrit alphabet.) Later, though, on a different level of analysis of the verse Vaidyapāda explains that the “blazing sixteenth part” refers to the bindu (thig le). The sixteen, he says, are the sixteen vowels, and the “part” refers to the final single bindu (bcu drug phyed cha ‘bar zhes pa ni thig le zhes pa’i don do// de yang bcu drug ni dbyangs yig rnams so// de yi phyed cha ni tha ma’i thig le gcig ces pa’i tha tshig go//) (Sukusuma, D 88a.3-4; P 105b.5-6). Here in his commentary on this verse, Vaidyapāda goes on to describe the bindu as the very one that is 357 The ultimate essence; and to the three supreme gurus10 who teach that I constantly bow with my three activities [of body, speech, and mind] equally.11 |1| [Pledge to Compose] The lamp of the three worlds, praised by all, The essence of all phenomena, the suchness of things, manipulated at the tip of the vajra during the perfection stage practices described below in Buddhajñānapāda’s text. It is the manipulation of this bindu according to the ritual, he explains, that brings about the realization of suchness: “Moreover, through practicing, by means of the agitation of the locations, the sixteen syllables appear. And these, then, become the sun and moon. Having transformed into a bindu like that, they go to the tip of the vajra. This itself, in a form which blazes with thousands of light rays, is meditated upon by the yogin in accordance with the ritual that will come below. When this happens, the suchness that has been spoken of will be realized, [and that is the] purpose [of this practice.]” de yang bsgrub pas gnas rnams dkrugs pa las yi ge rnams bcu drug par gyur/ de yang nyi zlar gyur/ de lta bu’i thig ler gyur nas rdo rje rtse mor ‘gro ba ste/ de nyid ‘od zer stong du ‘bar ba’i gzugs su rnal ‘byor pa rnams kyis ‘og nas ‘byung ba’i cho gas bsgoms nas/ ji skad du gsung pa’i de kho na nyid rtogs par ‘gyur pa’i phyir ro// (Sukusuma, D 88a.4-5; P 105b.6-8). 10 Vaidyapāda explains that these three are the causal, conditional, and sahaja ācāryas (de yang gsum ste/ rgyu dang rkyen dang/ lhan cig byed pa’i slob dpon no//) (Sukusuma, D 88a.6; P 106a.1). The sahaja ācārya is mentioned by Buddhajñānapāda himself in verse 142 of the Dvitīyakrama, and Buddhajñānapāda mentions the “three gurus” in the Muktitilaka, as well; Vaidyapāda gives the very same gloss on the identity of these three in his Muktitilakavyākhyāna. In both the Sukusuma and the Muktitilaka-vyākhyāna Vaidyapāda provides a citation of a passage about the three gurus from a work that he identifies in the Muktitilaka-vyāhkhāna as The Precious Garland (rin chen phreng ba (phreng ba] P, phrod pa D); I have been unable to identify this source). In the Sukusuma Vaidyapāda mentions that the passage was cited by Buddhajñānapāda himself on this topic (possibly in the context of oral instructions, since the citation is not found in any of Buddhajñānapāda’s surviving writings). There are some slight variations in the transmission of the verse in the Sukusuma and Muktitilaka-vyākhyāna, but in summary the verse identifies the causal ācārya as the master who gives vows and commitments and who purifies one’s mind through the stages of initiation, beginning with the water initiation; the conditional ācārya as the “great goddess” with whom one engages in play and who purifies the field of one’s mind by means of the “sixteenth part;” and sahaja ācārya as the one from whom one receives that (bindu?) and by means of whom and through whose blessing one realizes innate joy. Vaidyapāda further adds that these three are supreme because they are superior to other gurus (Muktitilaka-vyākhyāna, D 47b.5-7). The difference between the conditional and the sahaja ācāryas is difficult to understand from the passage that Vaidyapāda cites, as both seem to refer to the tantric consort. However, in his Yogasapta, Vaidyapāda states that the kalaśābhiṣeka is bestowed by the causal ācārya, the guhya initiation is bestowed by the causal and the conditional ācārya, and the prajñājñāna is bestowed by means of the causal, conditional and sahaja ācāryas (Yogasapta, D 70a.4; 70a.7; 70b.4). This suggests that the “conditional” guru may be the consort in the role as the partner of the guru for the guhya initiation, while the sahaja guru is the consort in her role as the disciple’s partner in the prajñājñāna initiation. Later in the Sukusuma, Vaidyapāda clearly states that the sahaja ācārya is the consort (shes rab, prajñā), and that uniting with her entails receiving her “blessing” (Sukusuma, D 111b.3-4; P 134a.6-7). 11 Vaidyapāda notes, in classical fashion, that this praise was composed in order to take refuge in the three jewels as a way of overcoming obstacles and ensuring that the author would be able to complete the task of composing (Sukusuma, D 87a.5-6; P 104b.6-7). 358 Reverser of the poison waters of existence,12 inside the triangle,13 blazing upon vaṃ:14 So that [beings] can realize [it] though the words of guru Mañjuśrī,15 I will explain this.16 |2| [Autobiography, Part I] 17In a town called Takṣaśilā,18 in the area of Khapir,19 in the land of Magadha, I pleased the guru Haribhadra,20 who had attained great fame. I received his instruction and studied many scriptures21 I investigated those and derived understanding.22 |3| At Śrī Nālandā,23 in response to the one of noble birth called *Guṇamitrā24 12 Vaidyapāda identifies existence as “conceptuality” (rnam par rtog pa ni srid pa) (Sukusuma, D 88b.7; P 106b.3). 13 Vaidyapāda indicates that the triangle represents the “secret lotus,” i.e. the vagina. gsum ni chos ‘byung gi phyag rgya stong ba nyid la sogs pa’i rang bzhin gru gsum dang ldan pa gsang ba’i padma’o// de’i khong pa ni nang ste/ (Sukusuma, D 89a.1; P 106b.5). This triangle is also understood as the letter e, which in conjunction with vaṃ creates the first word of the first phrase of most sūtras and tantras, including the Guhyasamāja-tantra: evam mayā śrutam. Here e is understood the be the dharmodaya, representign the vagina, while vaṃ is understood as the vajra, the penis. Such an explanation is found in many tantric commentaries; one early presentation, and almost certainly the one the Buddhajñānapāda is drawing from, is his guru Vilāsavajra’s commentary to the nidāṇa of the Guhysasamāja-tantra which reads, e ni chos kyi ‘byung gnas/ waṃ ni rdo rje ma ni rnam par snang mdzad chen po’i ye shes kyi sa bon/ (Śrīguhyasamājatantranidāṇagurūpadeśabhāsya, D 91b.4-5). 14 Vaidyapāda identifies vaṃ as the vajra and that which blazes upon it as the bindu of relative bodhicitta, that is to say semen. vam ni rdo rje de’i steng ni nor bu’i cha ste/ de na ‘od zer ‘bar ba’i kun rdzob byang chub kyi sems ji bzhin pa’i don la mi slu ba’i thig le’o// (Sukusuma, D 89a.2; P 106b.6-7). 15 Vaidyapāda notes that Buddhajñānapāda’s indicating that the teachings are the words of Mañjuśrī is meant to contradict the view that Buddhajñānapāda himself had composed the instructions. ‘jam dbyangs bla ma’i lung gi zhes te rang bzo dgag pa’o// (Sukusuma, D 89a.3; P 106b.7-8). 16 Vaidyapāda here explains the topic, purpose, connection, and essential purpose of the text (brjod bya, dgos pa,‘brel pa, and dgos pa’i yang dgos pa) (Sukusuma, D 89a.5-6; P 107a.1-3). 17 Vaidyapāda notes that it is in order to inspire faith in beings that the master himself here gives the “story of his own encounter with suchness.” da ni ‘gro ba rnams dad par bya ba’i phyir/ rje brtsun bdag nyid kyis de kho na nyid brnyes (brnyes] P; D, mnyes) pa’i lo rgyus gsungs pa/ (Sukusuma, D 89a.6-7; P 107a.3-4). 18 Rdo ‘jog is a common translation of Takṣaśilā (see C. Dalton and Szántó, forthcoming). However, this identification with the town of Taxila is also somewhat problematic. See the next note. 19 kha pir] D C S V (P), kha bir P N V (D). This may possibly be a corrupted rendering of Kaspir, i.e. Kaśmir (see C. Dalton and Szántó, forthcoming). However, since this Khapir is specified as being in Magadha, such an identification is only possible if Magadha is understood to mean the Indian subcontinent more broadly, rather than the region of Magadha, which is not near Kashmir. Moreover, Vaidyapāda describes Magadha as “in the area of Nālandā,” which again renders the identification of Khapir as Kaśmir difficult (Sukusuma, D 89a.7; P 107a.5). Any certainty about these toponyms may be difficult to ascertain. 20 Haribhadra’s name is here and in Vaidyapāda’s Sukusuma given as Bzang po seng ge, rather than the more common Seng ge bzang po. There is little doubt, however, about the identity of this guru, as Vaidyapāda explains that Buddhajñānapāda studied Prajñāpāramitā with this guru, a well known Prajñāpāramitā scholar, and Buddhajñānapāda himself wrote Prajñāpāramitā works. Later Tibetan histories also corroborate that this guru is Haribhadra. 21 Vaidyapāda mentions the Prajñāparamitā scriptures “and many others” as those studied under Haribhadra (Sukusuma, D 89b.1-2; P 107a.6). 22 rig ‘byung. I am slightly unsure about this reading. Vaidyapāda’s commentary does not address this phrase; he seems to conclude his comments on Buddhajñānapāda’s studies with a gloss of the term rnam dpyad, “I investigated.” 23 D and C read shī len nalendrar, adding an extra syllable, which fits the meter. P N and S read shrī nalāndar, which is a closer approximation of Śrī Nālandā, but which does not fit metrically. 24 Vaidyapāda refers to her as the bhikṣunī *Gunamitrā (dge slong ma yon tan bshas gnyen) who lived in the “great dharma school of logic” (rig pa’i chos sgrwa chen po) (Sukusuma, D 89b.2; P107a.7). This a valueable reference to the presence of a bhikṣunī living at Nālandā in the late eighth century. She is described as having stable faith, and being brahmin by birth. (shrī na le ndrar rig pa’i chos sgwra chen po na gnas pa’i dge slong ma yon tan bshas 359 With a [still] ignorant mind25 I composed some treatises26 joyfully, Thinking to benefit those who live there with those treatises [2a] [While] I stayed there, I composed and taught. |4| 27Then I travelled to the land of Uḍḍiyāna, the source of all positive qualities, [Where there lives] someone known as Vilāsavajra28 From him I learned much29 and investigated, as well. And also in that same place I pleased a guru called Guṇeru30 |5| gnyen zhes bya ba (ba] P; ‘ D)/ bram ze’i rigs su skyes pa dad pa brten ba zhig yod pa) (Sukusuma, D 89b.2-3; P 107a.7-8). In fact the colophon of Buddhajñānapāda’s Prajñāpāramitā work the Sañcayagāthā-pañjikā mentions Guṇāmitrā by name as the petitioner (Sañcayagāthā-pañjikā, D 189a.5). 25 blun blos. Vaidyapāda notes that this term means that although Buddhajñānapāda was engaged in the practice of the pāramitās, he had not yet realized suchness just as it is. blun blos zhes te ph rol tu phyin pa’i slos gnas pas de bzhin nyid ji lta ba bzhin (bzhin) D, nyid P) du ma rtogs pa’i phyir ro// (Sukusuma, D 89b.3; P 107a.8). 26 Vaidyapāda says this refers to a synopsis of the Prajñāpāramitā and other texts. (shes rab kyi ph rol tu phyin pa’i bsdus don la sogs pa) (Sukusuma, D 89b.3; P 107a.8-b.1). This synopsis likely refers to the *Sañcayagāthā-pañjikā, mentioned in note 24, a Prajñāpāramitā commentary which does seem likely to have been a composition that Buddhajñānapāda wrote early in his career. 27 Vaidyapāda’s commentary to this section of Buddhajñānapāda’s text has been translated in Davidson 2002, 311- 13. My reading of Vaidyapāda parts ways with Davidson’s translation in a number of places, and I have provided a full translation of Vaidyapāda’s commentary to the autobiographical sections in Chapter One. 28 ‘Jo sgeg do rje. This master has generally been identified in both traditional and modern scholarship as Vilāsavajra. Vilāsavajra is usually rendered into Tibetan as Sgeg pa’i rdo rje, but ‘jo sgeg is a synonym for sgeg pa, so the identification here seems rather certain. However, as Tribe and Szántó have noted, Vilāsavajra cites Buddhajñānapāda’s Mahāyānalakṣaṇasamuccaya in his commentary on the Nāmamantrārthāvalokinī. Szántó points out that this would be an unusual instance of a master citing a work by his disciple (Tribe 1994, 16; Szántó 2015, 541). I concur that this most likely is, in fact, the case here, since the text by Buddhajñānapāda which is cited by Vilāsavajra is a Mahāyāna text which was likely composed by Buddhajñānapāda in his youth, before he moved on to writing tantric treatises. Vaidyapāda (Sukusuma, D 89b.5; P 107b.3) mentions that this same master was also called sna tshogs gzugs, which Davidson (2002, 311) renders as *Citrarūpa. However, in the colophon to the Sanskrit text of Vilāsavajra’s commentary on the Mañjuśrīnāmasaṃgīti, it is stated that the author was also known as Śrī Viśvarūpa, and that he lived in a place called Ratnadvīpa, exactly as Vaidaypāda states (see Tribe 1994, 19; Sukusuma, D 89b.5; P 107b.3). 29 Vaidyapāda mentions that Buddhajñānapāda studied many Kriyā and Yoga tantras with Vilāsavajra (Sukusuma, D 89b.5; P 107b.4). 30 gu ne ru] S P V (D and P), gu ne nu D C N. Vaidyapāda describes Guṇeru as having received instructions on the bsam gyi mi khyab pa’i rim pa (which Davidson (2002, 311) renders as the *Acintyakramopadeśa) and as a great yoginī who had encountered suchness (rnal ‘byor ma chen mo de nyid brnyes pa) (Sukusuma, D 89b.5; P 107b.4). It is unclear whether the bsam kyis mi khyab pa’i rim pa’i man ngag is meant to refer to the title of a text or not. A text of precisely this title is extant in the Tengyur (*Acintyakramopadeśa, Bsam kyis mi khyab pa’i rim pa’i man ngag Tōh. 2228), where it is attributed to one *Kuddālīpāda (tog rtse zhabs). The same work survives in a second Tibetan translation, apparently of a slightly different recension of the Sanskrit text, in a compendium of Sakyapa works; within the Sakyapa tradition the work is understood to represent one among a series of eight subsidiary instructions connected to the Lamdre (lam ‘bras) root text (Davidson 2005, 194-95). The *Acintyakramopadeśa is also considered, in the Tibetan tradition, among a set of six Indian mahāmudrā works called the Sixfold Corpus on the Essence (Snying po skor drug) (Krug 2018, 328-9). The Sanskrit of the work, under the title Acintyādvayakramopadeśa, survives and has been edited (Samdhong and Dwivedi). I have not had the opportunity to compare this against the Tibetan translations and am unaware of any such comparison having been reported in modern scholarship. (Thanks to Harunaga Isaacson for first drawing my attention to the existence of this Sanskrit edition.) Regarding its author and period of composition, in his History of Buddhism in India Tāranātha mentions a

• Kaudālika/Mahā-koṭali (tog rtse ba che ba), who Chimpa and Chattopadhyaya have taken to refer to the same

figure as Kuddālipāda, and who Tāranātha says lived during the reign of King Gopāla, the Pāla king who reigned prior to Devapāla and Dharmapāla, who ruled when Buddhajñānapāda composed his writings (Chimpa and Chattopadhyaya, 262). However, the work as it survives at present focuses on perfection stage practices and was understood, at least by the 15th-century Tibetan scholar Ngorchen, to be based on the Sampuṭa-tantra (though 360 And received teachings from her.31 At the northern gate of that place32 I pleased a girl of sixteen years named Jātig Jālā,33 Mahālakṣmī.34 For eight months I took her instruction, and having received it, I achieved accomplishment.35 |6| Then I went to the village of Ko no dze36 in the area of Jālandhara37 And met Bālipāda,38 who had attained great renown. Having pleased him, I studied the scriptures and received many instructions. Then I went to “the place with sky trees”39 in the Koṅkana, to the south. |7| Davidson (2005, 196) notes that the connection is “only indirect” and Isaacson (personal communication) has also expressed some doubt as to the connection with the Sampuṭa) (See also Davidson 2005, 195-96; Stearns 2006, 135). It is questionable whether this text is early enough to be the referent here in Vaidyapāda’s commentary. Krug (2018, 341) identifies the work as focused on the generation and perfection stage yogas of the Yogiṇi tantras. Apart from its content, further clues to the period of the author may be found in a lineage list given in the work itself, culminating in the author’s own guru, who he styles Bhadrapāda (Krug 2018, 335-6). For now, whether or not Vaidyapāda is referencing this particular work must remain a question. It is possible, as well, that the text as written down was meant to preserve a tradition of oral instructions that had not yet been previously recorded, and that such a set of oral instructions (which of course would have been supplemented over time) could conceivably be Vaidyapāda’s intended referent here (Harunaga Isaacson, personal communication). 31 While Buddhajñānapāda does not specify the guru’s gender, and the unusual name gives no indication of gender, either, I have followed Vaidyapāda’s identification of this guru as a great yoginī (see previous note) and translated the pronoun in the feminine. Vaidyapāda (Sukusuma, D 89b.6) notes that he received instructions on Niruttara tantras (bla med rgyud) from this guru. However given that Vaidyapāda has in an earlier passage referred to Yoga tantras and in a later passage of the Sukusuma (D 108a.6-108b.1) he explicitly distinguishes between Yoga tantras (rnal ‘byur rgyud) and Yoganiruttara tantras (rnal ‘byor bla na med pa’i rgyud) (the latter of which he equates in that passage with Dākiṇī tantras (mkha’ ‘gro ma’i rgyud)), I believe it is likely that Yoganiruttara tantras is what is intended here. 32 See C. Dalton and Szántó (forthcoming) for a differing interepretation of this term where the term chab sgo is interepreted as a proper noun. I prefer to read it here as simply “gate,” especially given Vaidyapāda’s reading which includes some grammatical particles omitted for metrical reasons in the Dvitīyakrama itself. Vaidyapāda reads: u rgyan gyi gnas de yi byang phyogs kyi chab sgo na/ (Sukusuma, D 89b.7; P 107b.6). 33 Dzā (dzā] D C V (D), dza P N S, ‘dza’ V (P)) thig dzā (dzā] sugg. em. based on V (D); dza D C P N S, dzva V (P)) lā (lā] sugg em; la D C P N S). 34 Vaidyapāda notes that Jātig Jālā was sixteen-year-old outcaste girl who was actually the yoginī Mahālakṣmī, born of noble family. gdol pa’i rigs dzā (dzā] D, ‘dza’ P) thig dzā (dzā] D, dzva P) la zhes bya ba la bu mo lo bcu drug lon pa zhig yod kyis/ de ni rigs las skyes pa’i rnal ‘byor ma la kshmī chen mo zhes bya ba yin kyis (Sukusuma, D 89b.7-90a.1; P 107b.6-7). 35 Vaidyapāda notes that at this time also Buddhajñānapāda attained accomplishment of Jambhāla. Later in the Dvitīyakrama Buddhajñānapāda himself mentions receiving provisions from Jambhāla and he is also credited with composing three Jambhāla sādhanas. 36 Ko no dze] D C P N S V(P), ka no dze V(D). I have been unable to identify this location. At first glance it does seem to be a transliteration of Kannauj, and Davidson (2002, 312) has rendered it as such. However, Szántó (2015) places some doubt on this identification, since modern-day Kannauj is not near the modern-day city of Jalandhar, and C. Dalton and Szántó (forthcoming) note that at the time Kannauj was referred to as Kanyākubja, making the identification even less likely. 37 dzā lendha] D C, dzā lāndha P N S. See previous note. 38 Here D and C read bā li pā da, while P, N and S read ‘ba’ mo pa ta. Vaidyapāda’s commentary has the name translated as byis pa chung ba’i zhabs, which supports the reading from D and C (Sukusuma, D 90a.2; P 108a.1). Szántó reconstructs the name as Bālikapāda and suggests that the name may even read Bālhikapāda as reflective of a master from the area of Balkh (Szántó 2015, 542; see also C. Dalton and Szántó, forthcoming). 39 nam mkha’ shing ldan. I am here indebted to Szántó’s work, which suggests—I think convincingly—that the place mentioned here is Kadri, near Mangalore. See Szántó (2015) for the full details of this assessment. I also discuss this further in Chapter One. According to Khenpo Chodrak Tenphel the term “sky tree” (nam mkha’i shing) means mangrove (personal communication, March, 2016). 361 [There] the lord of siddhas, renowned as Pālitapāda40 Was surrounded by his disciples who could perform miraculous feats. All of them regularly received requisites, clothing, food, and wealth. I bowed at the feet of this sublime guru for nine years. |8| I listened to the great Samāja-tantra together with its commentaries for eighteen [months].41 [I said] “I have not realized it” and the great guru said the same. Thinking, “Until I realize this,42 anything else is useless,” [2b] I affixed the volume around my neck and set off to the north. |9| Behind Vajrāsana is the forest called Kuvaca Which is full of tigers and bears—a terrifying place. There I spent six months, and thus realized the suchness of phenomena. I met an emanated monk together with two gurus. |10| [Vision of Mañjuśrī and Supplication to Him] On the eighth day of of the seventh month, during [the constellation] Puṣya At the time when Mṛgaśīrṣa and Hasta are fading,43 in the early morning, right at dawn, Towards the emanated maṇḍala-cakra of Mañjuśrī44 I made a fervent supplication to understand the meaning: |11| “You are the father and the mother of all beings!45 Protect me and others from great danger! Master, lord of beings, dispel suffering! Emptier of the three realms, greatest of the great,46 you protect beings |12| 40 bā li pā dar] D C, ba li pa tar P N S. Vaidyapāda (Sukusuma, D 90a.4; P 108a.4) identifies the teacher as bsrung ba’i zhabs, which Davidson has rendered as *Rakṣapāda. However, Szántó has recently provided evidence from a Sanskrit manuscript of the Sāramañjarī, a commentary to another of Buddhajñānapāda’s works, that this teacher’s name was, in fact Pālitapāda (Szántó 2015, 542-50; see also C. Dalton and Szántó, forthcoming). 41 bar du mnyan] sugg. em., rab tu mnyan D C P N S. My suggested emendation is based upon Vaidyapāda’s commentary which reads, bco brgyad bar du mnyan ni zla ba bco brgyad kyi bar du bsgrub pa’o, which suggests that rab tu is just a textual transmission error (Sukusuma, D 90a.7; P108a.8). 42 ‘di] suggested em., ‘dir D C P N S. The suggested emendation is based on Vaidyapāda’s commentary (Sukusuma, D 90b.1; P 108b.1) which reads di ma rtogs par. 43 Puṣya is the eighth lunar mansion in Indian astrology; Mṛgaśīrṣa is the fifth; Hasta is the thirteenth. 44 ‘jam dpal dbyangs kyi (kyi] P N S, kyis D C) dkyil ‘khor ‘khor lo (lo] P N S, lor D C) sprul pa la. Here Vaidyapāda explains that this supplication took place subsequent to a question from the emanated monk about whether Buddhajñānapāda had faith in the guru who emanated the maṇḍala or in the deity within the maṇḍala. After Buddhajñānapāda answered that he placed his faith in the deity of the maṇḍala, the monk, along with the woman and dog, departed and entered a small house, and Buddhajñānapāda thsus directed his supplication directly to Mañjuśrī in the maṇḍala. 45 Vaidyapāda explains that he is the father because beings are born from Mañjuśrī’s wisdom, and the mother because beings are born from Mañjuśrī’s dharmodaya. yab ste zhes pa ni de rnams kyang de’i ye shes las ‘khrungs pa’i phyir ro// yum yang yin zhes pa ni de’i chos kyi dbyings las byung ba’i phyir ro// (Sukusuma, D 91a.1; P 109a.2-3). 46 che (che] N V, chi D C, cha S P) ba’i che. Vaidyapāda also concurs with the reading che and even explains why Mañjuśrī is to be called the “greatest of the great,” namely, because he has realized the ultimate state, rendering him greater than ordinary beings, śrāvakas, prateykabuddhas, and even bodhisattvas of the pāramitās (Sukusuma, 91a.3- 4). 362 47You are beginningless, unvoiced, lacking the upper part of the bindu, The revered, the letterless,48 producer of nectar, the empty49 bliss of great joy. In order to benefit beings, O you Great Protector, Please bestow50 bliss—51 the bliss that is great joy—52 upon all the buddhas. |13| The path to awakening, not stained by faults, Which pacifies all types of suffering, and quenches thirst Liberates from the waves of saṃsāra, and places one in bliss—53 Please teach this path, which is not fathomed [even] by those who are victorious over all things.54 |14| The peace of all peace, perfectly free from impurity, having abandoned avarice,55 Beyond cultivation, unstained like the sky, Beyond all illusion, having left all desires behind, I take refuge in you, who are like this, Lord! |15| 47 This verse is somewhat opaque and is one of the few instances where I have edited the root verse based on its citation in Vaidyapāda’s commentary in a way that differs from all available recension of the root text. The first two lines seem to be identifying Mañjuśrī in various ways with emptiness, and yet noting that he still serves beings. The following lines seem to be requesting him to bestow bliss—which Vaidyapāda identifies as the splendorous empty and luminous aggregates (presumably those of the deity)—upon the buddhas, which Vaidyapāda identifies as the aggregates. I here translate Vaidyapāda’s commentary to this verse in full, although again there are a number of opacities in his comments, as well. “Having explained [Mañjuśrī] as the source, now in the second part [he] is taught to be letterless, with [the verse] beginning You are beginningless... You are beginningless refers to the profound, which lacks beginning. That being absent, there is [also] no sound (nga ro, *svara) that [appears] in the form of the waxing moon. [Could this perhaps mean the candra of a candrabindu??] Drop means bindu, and the upper part is oṃ, and it should [also?] be understood otherwise(?). Since these do not exist [he] is letterless lacking the distinctions made by those [aspects that constitute a syllable?]. Although being nonexistent in that way, since he nourishes the realms of beings by means of the wisdom that realizes that [emptiness], he is the producer of nectar, the empty bliss of great joy. This is exactly what he teaches in order to benefit beings. Bliss is supreme joy. All the buddhas means the aggregates and so forth. The bliss that is great joy [that Mañjuśrī is supplicated to bestow] upon them are the delightful splendorous aggregates which are characterized by the profound and the genuinely luminous. O you great protector means Mañjuśrī, because he is the refuge of those who abide [in the state] of wandering. Please bestow means by means of giving [it] via methods, please bring this about in the minds of others.” de ltar ‘byung ba’i gnas su bstan nas/ da ni gnyis pa’i chas yi ge dang bral bar ston pa / thog ma zhes pa la sogs pa’o// thog ma med khyod zhes pa ni zab mo ste dang po med pa’o// de med pas zla ba tshes pa’i rnam pa lta bu’i nga ro med pa ste/ thigs pa ni thig le’o// steng cha ni oṃ ste gzhan yang rtogs par bya’o// de rnams med pas yi ge (yi ge] D, yig P) med ces te de rnams kyis khyad par du byas pa rnams med do// de ltar med kyang de rtogs pa’i ye shes kyis sems can gyis khams rnams gso bar byed pas na/ bdud rtsi rab dga’ bde stong byed/ ces so// de nyid gsung pa/ ‘gro la phan phyir zhes so// bde ba ni mchog gi dga’ ba’o// sangs rgyas rnams kun zhes pa ni phung po la sogs pa’o// de rnams la rab dgyes bde ba zhes te zab mo dang/ yang dag par gsal ba mtson pa nyams dga’ ba’i gzi brjid phung po’o// de mgon chen khyod kyis zhes te ‘khyam pa lta bur gnas pa’i skyabs su gyur pa’i phyir na ‘jam dbyangs so// gtong bar mdzod ces pa ni thabs kyis ster bas gzhan gyi rgyud la ‘gro bar mdzad cig ces pa’o// (Sukusuma, D 91a.4-91b.1; P 109a.8-109b.5). 48 yig (yig] P N S, yid D C) med 49 stong] sugg. em based on V (P and N), gtong P N S, btang D C 50 gtong] P N S, btang D C 51 bde ba] sugg em. based on V (D and P), bde ba’i P N S D C 52 rab dgyes bde ba] sugg em based on V (D and P), rab dgyes bde ba’i P N S D C 53 Vaidyapāda explains that that which places one in bliss is the great samaya, vajra-like suchness. dam tshig chen po rdo rje lta bu’i de kho na nyid do// (Sukusuma, D 91b.5; P 110a.3). 54 Vaidyapāda explains that this means is unknown to those who lack the oral instructions of the master. bla ma’i man ngag dang bral bas mi shes par ston pa (pa] D, pas P) ste/ (Sukusuma, D 91b.6; P 110a.4-5). 55 ‘jungs] P N S V (in P), ‘jum D, ‘dzum C, ‘jums V (in D) 363 Until I reach great awakening that is without pride and without fear, [Though] I may remain in saṃsāra, [in order] not to fall into the unpleasant hells of the world, I and others will always maintain The vows and uncommon samayas that you teach.56 |16| Playfully dancing the great dance57 With your various arms twisting58 and holding tight59 You open the eight soft lotus petals And insert the vajra, the cause of nondual bliss.60 |17| 61The secret suchness, undefiled, becomes clear. The moon that is born from the vajra and petals is perfectly gathered This is the supreme suchness of all phenomena born from means and wisdom. Revered master, [3a] in order to benefit me, explain what is hidden!” |18| [Mañjuśrī’s Acceptance of the Supplication] 56 Vaidyapāda describes the samayas that Mañjuśrī teaches as the things to abandon, like disparaging the guru, and the things to adopt, like consuming the five nectars and killing evil beings. He notes that the vows include avoiding (spang ba) an untrained consort and inciting the stopping of the enjoyment of objects (yul gyi nye bar longs spyod pa’ bkag bskul) (Sukusuma, D 92a.6-7; P 110b.7-8). 57 Vaidyapāda notes that the playfully dancing refers to the practice of union, which is the dance renowned in the texts of the “lotus treatises.” reaIt is unclear to which type of treatises Vaidyapāda might be referring here. gar chen zhes pa ni dbyugs ‘byin par byed pa’i spyod pa ste/ snyoms par ‘jug pa’o// de yi rnam par rol pa ni padma’i bstan bcos kyi gzhung gis grags pa’i gar te de mdzad cing zhes so// (Sukusuma, D 92.b.1; P 111a.1-2). 58 gcu] V(D), bcu D C P N S V(P). Vaidyapāda’s comments make it clear that gcu is the correct reading. “Holding tight while twisting (gcu pa), rubbing, and embracing...” gcu pa dang mnye ba dang ‘khyud pa la sogs pas bsgrims nas (nas] P, rnam D) zhes so// (Sukusuma, D 92b.2; P 111a.2). 59 bsgrims] P N S V(D and P), bskyings D C. 60 Vaidyapāda notes that “The nondual bliss is the moon-like bodhicitta and its cause is the vajra from which it emerges. [That vajra], blessed with oṃ, five-pronged and with the redish white jewel, is to be placed there [in the lotus]. Let the moon that emerges from it purify our field, making us equal those who have the wisdom of the ten bhūmis!” gnyis su med pa’i bde ba ni zla ba lta bu’i byang chub kyi sems te/ de ‘byung ba’i rgyu ni rdo rje ste oṃ gyis byin gyis brlabs pas dmar skya’i nor bu ldan pa’i rtse lnga pa ste/ de bzhag par mdzad de de las byung ba’i zla bas bdag cag gi zhing dag par byas nas sa bcu’i ye shes can dang skal ba mnyam par mdzod cig ces pa’o// (Sukusuma, D 92b.3-4; 111a.3-5). 61 Vaidyapāda’s commentary on the first three lines of this verse details some of the elements of practices involving aspects of the subtle body and, as an example of subtle-body anatomy from an author we can date with some certainty to the 9th century, is worth quoting in full. “This secret suchness is the guhyābhiṣeka. Because it is the cause for the manifestation of suchness it is undefiled. Through that the wisdom that arises from prajñā (this is likely a double entendre for the consort, also called the prajñā) becomes clear. And what is this suchness? [It is explained by the line] beginning with The moon which is... Vajra means the secret vajra. Petals are clearly mentioned since this is the unique cause, [but here] one must also understand the anthers. What is born from these two is the moon, which is the jasmine-like seed (i.e. semen). Gathering that perfectly means inciting through practice the seventy-two thousand channels and gathering [the essences/bindus in these?] into sixteen. These are then gathered into three. The suchness (de nyid) which has travelled to the lotus at the heart center is then invoked and held by means of actions and the winds. This exactly is taught [with the line] starting This is the supreme suchness... which is easy to understand.” gsang ba’i ( gsang ba’i] P, gsang ba’i dbang D) de nyid ces pa ni gsang ba’i dbang ngo// de nyid sngon du song ba’i rgyu bas ni mi mnyams (nyams] D, nyam P) pa ste/ de las shes rab las skyes pa’i ye shes gsal bar ‘gyur ces so// de nyid kyang gang zhe na/ rdo rje zhes pa la sogs pa’o// rdo rje ni gsang ba’i rdo rje’o// ‘dab ces pa ni khyad par gyi rgyus las bas (bas] D, las P) nges par mtshon pa ste/ ge sar yang shes par bya’o// de gnyis las skyes pa ni zla ba ces te kunda (kunda] D, kun da P) lta bu’i sa bon no// de yang dag par ‘du ba ni na di (na di] sugg. em., na li D P) stong phrag bdun cu rtsa gnyis bsgrubs pas bskul bas bcu drug du ‘dus/ de yang gsum du ‘dus/ de nyid snying ga’i padmar phyin pa bskul nas bya ba dang rlung gis gzung ba’o// de nyid gsung pa/ thabs shes las byung zhes pa la sogs pa’o// de ne go sla’o// (Sukusuma, D 92b.4-7 P 111a.5-8). 364 Then, the great bodhisattva Mañjuśrī62 Looked upon me with a smiling face and said, “Excellent” three times. With this vajra song, like an echo,63 he taught to me64 The playful dance and the suchness of all phenomena. |19| [Mañjuśrī’s Condensed Teaching/Pledge to Teach] A65 vi yaṃ raṃ vaṃ laṃ hūṃ66 a la la la ho!67 The great compassionate ones, Who have realized this, Those vajra holders of the past, present, and future |20| Who obtained the excellence of the sugatas, Have taught, teach, and will teach [this truth] To [only] some worthy [disciples]. In order that they may realize the genuine meaning, |21| I will teach this to you— Concentrate your mind and listen! |22| [Mañjuśrī’s Teaching] [The Nature of Phenomena is Nondual Wisdom] 62 Vaidyapāda is keen here to indicate that Buddhajñānapāda’s teacher, the “bodhisattva” Mañjuśrī is fully awakened, not just a bodhisattva on the path. He states, “He is called a bodhisattva because he is integrated with awakening (bodhi), not because awakening is his goal.” de nyid byang chub dang ‘dres pa’i phyir byang chub sems dpa’ ste/ byang chub la dmigs pa ni ma yin no// (Sukusuma, D 93a.2; P 111b.3-4). 63 Vaidyapāda’s commentary confirms that sgra bsnyan here means echo rather than lute. sgra brnyan lta bur zhes pa ni brag ca lta bu ste grag (grag] D, grags P) kyang ma grub ces pa’i don to// (Sukusuma, D 93a.4; P 111b.6-7). 64 Vaidyapāda notes, “With the words he taught [this] to me, the great guru makes others feel confident.” de lta bus bdag la bstan zhes bla ma chen pos gzhan yid brtan par mdzad pa yin no// (Sukusuma, D 93a.5; P 111b.7). 65 According to Vaidyapāda, this is the beginning of Mañjuśrī’s direct speech. It is worth noting that the very beginning of Mañjuśrī’s direct speech in the Dvitīyakrama as a series of syllables is evocative of the very beginning of Mañjuśrī’s direct speech in the Mañjuśrīnāmasaṃgīti, which is likewise a series of syllables, starting with a. There, however, the syllables in question are the first twelve vowels of the Sanskrit alphabet (see Tribe 1997, 118). Commenting on the syllables here in the Dvitīyakrama Vaidyapāda states, “a begins the words of Mañjuśrī himself.” da ni a zhes ba ba la sogs pas ‘jam dbyangs kyi lung nyid gsungs te/ (Sukusuma, D 93a.5; P 111b.7-8). Vaidyapāda’s comments on the first two syllables of Mañjuśrī’s speech are also worth quoting in full here, as they correspond closely with the gnostic cosmogony expressed later in the Dvitīyakrama itself. “A is the nature of all things because they are unarisen. It is said that “A is the gateway to all pheonomena.” If we examine that statement, [we can understand that a is] the gateway through which all [phenomena] emerge. Moreover it should be known as the nature of the nonduality of the profound and the luminous which is like the maṇḍala of space, not arisen from any sort of conceptual imputations, [but] primordially and spontaneously present. That which appeared from its essence as mere knowing is vi, the first named syllable which is called ‘awareness.’” de la a zhes pa ni dngos po thams cad kyi rang bzhin te/ ma skyes pa’i phyir a ni chos thams cad kyi sgo’o zhes pa’i gsung la dpyad na/ thams cad byung ba’i sgo ste/ de yang gsal bzab gnyis su med pa’i rang bzhin nam mkha’ dkyil ‘khor lta bu brtags pa thams cad kyis ma skyes pa dang po nas lhun gyis grub pa nyid du shes par bya ba’o// de yi ngo bo las shes pa tsam lta bur snang ba ni bi ste/ rig pa zhes pa’i ming gi yi ge dang po’o// (Sukusuma, D 93a.5-7; P 111b.7-112a.2). 66 Here in the context of the Dvitīyakrama, Vaidyapāda homologizes the syllables subsequent to a vi (which were addressed in the previous note)—yaṃ, raṃ, vaṃ, and laṃ, with the four elements in the same way that they are commonly used in the visualization of the gradual emergene and stacking of elements that support the celestial palace in the practice of tantric sādhana (Sukusuma, 93b). I discuss this further in Chapter Three. 67 Vaidyapāda suggests that the rest of Mañjuśrī’s quotation is simply a clarification or an unpacking of these syllables. da ni ‘jam dbyangs kyi gsung gis de nyid gsal por bstan pa’i phyir/ rje btsun zhes pa la sogs pa’o// (Sukusuma, D 94a.4; P113a.1). 365 68 The nature of phenomena, From form and the rest up to omniscience, Is the perfectly pure wisdom of the nonduality of the profound and the luminous69 Which is like the center of space. 70 |23| That71 is not a thing; it is not a meditation object. Free from all entities,72 Not encompassed by the elements nor the sense-sources,73 It is naturally luminous, |24| Primordially pure, like space. Lacking existence, phenomena are free from characteristics. Since there are neither phenomena nor their nature, Entityless, it is similar to space. |25| Free from all words and letters. This is the essence of all time, Directions, and phenomena. It is not body, not speech, nor mind; |26| Not the realm of desire Nor the form nor formless realms.74 It is not the four great elements. Therefore, because it does not reside anywhere— |27| 68 Vaidyapāda states that at this point Mañjuśrī begins to teach the nondual wisdom that is the nature of all phenomena. da ni chos thams cad kyi rang bzhin gnyis su med pa’i ye shes gsung pa (Sukusuma, D 94b.1; 113a.7). 69 Zab gsal gnyis med ye shes. The term zab gsal gnyis med ye shes could certainly be translated more concisely as “profound, luminous nondual wisdom.” Yet that phrase, in English, suggests the terms “profound” and “luminous” to be adjectives describing nondual wisdom, which I do not believe to be Buddhajñānapāda’s intent. I have therefore opted for the more lengthy and awkward translation, “the wisdom of the nonduality of the profound and the luminous,” because I feel it reflects the nuance of Buddhajñānapāda’s understanding of the terms zab and gsal nominally rather than adjectivally. First, Vaidyapāda’s gloss of the phrase parses it exactly in the way I have translated it above: “the wisdom of the nonduality of the profound and the luminous” (zab mo dang gsal ba gnyis su med pa’i ye shes) (Sukusuma, D 94b.2-3; P 113a.8-b.1). And, indeed, Buddhajñānapāda’s use of the term zab gsal gnyis med (he does not always include ye shes) in a number of other instances, as well, indicates that he is speaking of the nonduality of the profound (emptiness) and the luminous (the apparent aspect), rather than using the terms “profound” and “luminous” as adjectives describing nonduality. 70 Vaidyapāda explains that this first verse is the general explanation and that what follows is an elaboration (Sukusuma, D 94b.3; 113a.8-b.1). 71 Vaidyapāda specifies that “that” refers to “that nature.” de ni zhes pa ni rang bzhin no// (Sukusuma, D 94b.3; P 113b.1). 72 Vaidyapāda specifies these as entities such as the manu and mānava (shed can dang shed bu) imputed by heretics (Sukusuma, D 94b.4; P 113b.2-3). 73 Vaidyapāda explains that those elements and sense sources are “understood by the śrāvakas as being subject and object (gzung ‘dzin), and even these do not encompass that nature.” nyan thos kyis de gzung ‘dzin du rtog (rtog] P, rtogs D) par byed pa des kyang rang bzhin de la ma zin zhes zo// (Sukusuma, D 94b.5; P 113b.4). 74 Vaidyapāda notes that these references to the various realms all refer to the sentient beings who live in them; what follows refers to the outer world—the production of the elements up until the generation of the deity’s celestial palace. de rnams kyang ma yin zhes bcud kyi sems can ma yin par bstan to// ‘byung chen bzhi yang ma yin te zhes rten (rten] D, P om.) ‘brel gyi stobs las rlung skye ba la sogs pa’i rim par lha’i gzhal yas khang grub pa yan chad kyi bar du snod kyi ‘jig rten grub pa ste de yang ma yin pa’o// (Sukusuma, D 95a.5; P 114a.6-7). 366 It is equanimity. The great Vajradhara who is like this Is the supreme nature75 of all phenomena, That which is to be accomplished through the method.76 |28| [It] is totally free from all constructs The cause of excellence, difficult to fathom,77 [And yet] appears as the mahāmudrā,78 Whose light rays ripen oneself and others. 79 |29| That is the supreme nondual nature, The great body which cannot be described Even by the great vajra-holder, That itself80 is the Victors, their offspring, and their fields of influence,81 |30| The three existences, together with the animate and the inanmiate.82 That, the identity of everything,83 Is the very essence of one’s mind, [3b] Supreme among all things. |31| When that is realized, this is the awakening of buddhahood.84 75 chos nyid 76 Vaidyapāda identifies this method as being comprised of the two stages and the four branches. rim pa gnyis yan lag bzhis rnam par bsdus pa (Sukusuma, D 95b.2; P 114b.2-3) . 77 Vaidyapāda identifies these first two lines as referring to the aspect of the profound, and second two lines of the verse as referring to the aspect of the luminous (Sukusuma, D 95b.3-5). 78 Vaidyapāda clearly identifies the mahāmudrā here as the form of the deity “with a face, hands, and so forth,” the usual use of this term in the 8th and early 9th centuries. phyag rgya chen por (por] P, po D) rab snang ba/ zhes (zhes] D, ces P) te/ zhal dang phyag la sogs pa dang ldan pa ni phyag rgya chen po’o// (Sukusuma, D 95b.6; P 114b.7-8). 79 zer gyis rang gzhan (rang gzhan] D C V(D), rang nyid S P V(P)) smin byed pa (pa] P N S V(P and D), pa’i D C). Vaidyapāda’s comments on the verse (rang phul du byung zhing gzhan rgyud dang par byed pa’o) also make it clear that he was reading rang gzhan, despite the fact that the citation of the verse in the Peking edition of his commentary reads rang bzhin in the citation of the verse (Sukusuma, D 95b.7; P 115a.2). 80 Vaidyapāda identifies this as nondual wisdom. de nyid cas pa ni gnyis su med pa’i ye shes so// (Sukusuma, D 96a.2; P 115a.4). 81 Vaidyapāda identifies the zhing as those to be tamed. de rnams kyi zhing ni gdul bya rnams te/ zhing dang chos mthun pa’i phyir ro// (Sukusuma, D 96a.2; P 115a.5). 82 rgyu] sugg. em., rgyur] D C P N S. Vaidyapāda’s commentary also supports this reading, providing the classical gloss that “the animate and inanimate refer to the world and its contents.” rgyu ba dang mi rgyu ba ni snod dang bcud do// (Sukusuma, D 96a.3; P 115a.6) . 83 Vaidyapāda identifies this as the fundamental ground that pervades everything. de thams cad kyi gyi bdag nyid de zhes don gi gzhi (gzhi] P; bzhi D) pos kun la khyab par bstan pa’o// (Sukusuma, D 96a.4; P 115a.6). 84 Here Vaidyapāda seems to advocate a sort of subitist position on awakening connected specifically to the realization of the nature of the mind. He says, “That which is supreme among all things is the nature of the mind. Why is that? Because When one realizes it... Those who first blaze with dillgence, [striving] in the methods through which one realizes that [nature] come, at some point, to experience it directly. When that happens, their lack of knowledge is abandoned, and a buddha endowed with the twofold accumulation is nothing other than that; it is like a person waking from sleep, or like a water lily(?).” dngos po kun gyi gtso bo ni/ rang gi sems kyi ngo bo nyid/ ci’i phyir zhes na/ de rtogs na zhes te/ de nyid dang por rtogs par bya ba’i thabs la brtson ‘grus ‘bar ba dang ldan pa rnams kyis nam zhig mngon du gyur na mi shes pa spangs shing tshogs gnyis kyi gang ba’i sangs rgyas kyang de nyid de/ skye bu gnyid sad paam/ ku mu da (da] P, ta D) bzhin no// (Sukusuma, D 96a.4-5; P 115a.7- 8). 367 The three worlds also are exactly this. The great elements also are exactly this.85 Why is that? Because all phenomena |32| Abide in the mind. This, as well, Abides in space.86 Space itself Abides nowhere;87 it is luminous. It is primordially completely pure and empty. |33| The sphere88 of the buddhas’ nirvāṇa89 The unborn vajra, manifest awakening, The supreme essence of all sugatas, This great nondual nonconceptual reality90 85 Vaidyapāda here notes that when the Dvitīyakrama says the three worlds and the four elements are “exactly this,” it means that the sentient beings who are the contents of the world are that mind nature and the outer vessel-like world itself is also that mind nature, respectively. de bzhin du (du] D, P om.) ‘jig rten gsum ste/ bcud kyi sems can rnams kyang sems de nyid do// ‘byung chen bzhi ste snod kyi ‘jig rten rnams kyang sems de nyid do zhes so// (Sukusuma, D 96a.5-6; P 115b.1-2). 86 Vaidyapāda clarifies that abiding in space means abiding in emptiness. nam mkha la gnas zhes ste stong pa nyid la gnas so zhes so// (Sukusuma, D 96b.1; P 115b.5). 87 Padmasambhava’s Man ngag lta ba’i ‘phreng ba cites a parallel passage, which is attributed in the commentarial literature on the Man ngag lta ba’i ‘phreng ba to the Guhyasamāja-tantra, though Karmay (2007, 158) notes that the passage in question is not found in the Guhyasamāja. The passage from the Man ngag lta ba’i ‘phreng ba reads: “All phenomena abide in the mind/ The mind abides in space/ And space abides nowhere.” (chos rnams thams cad ni sems la gnas so// sems ni nam mkha’ la gnas so// nam mkha’ ni ci la yang mi gnas so// (Man ngag lta ba’i ‘phreng ba, 5b.2-3). While Karmay is correct that no such passage is to be found in the Guhyasamāja-tantra, I have identified a passage with very similar content at the end of Chapter 15 of the Guhyasamāja-tantra. The passage from Matsunaga’s edition reads: atha te sarvatathāgatāḥ sarvatathāgatakāyavākcittavajraṃ tathāgatam evam āhuḥ/ sarvatathāgatadharmā bhagavan kutra stitāḥ kva vaā sambhūtāḥ/ vajrasattva āha/ svakāyavākcitta saṃstitāḥ svakāyavākcitta sambhūtāḥ/ bhagavantaḥ sarvatathāgatā āhuḥ/ svakāyavākcittavajraṃ kutra stitham/ ākāśasthitam/ ākāśaṃ kutra stitham/ na kvacit/ (Matsunaga 1980, 85). Fremantle’s edition and translation, however, give a version of the passage from the Guhyasamāja-tantra that is even closer to the quotation from the Dvitīyakrama and Garland of Views. Her edition reads: atha te sarvatathāgatāḥ sarvatathāgatakāyavākcittavajram evam āhuḥ/ sarvatathāgatadharmā bhagavan kutra stitāḥ kva vā saṃbhūtāḥ/ vajrasattva āha/ svakāyavākcittasaṃstitāḥ svakāyavakcittasaṃbhūtāḥ/ bhagavantaḥ sarvatathāgatā āhuḥ/ cittaṃ kutra sthitam/ ākāśastitam/ ākāśaṃ kutra sthitam/ na kvacit/ (Fremantle 1970, 348; see also Fremantle 1970, 349 for the Tibetan edition of the passage). Fremantle’s English translation of the passage reads: “Then all the Tathāgatas said to the Tathāgata, Vajra Body, Speech and Mind of all Tathāgatas: O Blessed One, where do the dharmas of all the Tathāgatas exist and where do they come from? Vajrasattva said: they exist in your body, speech, and mind, and they come from your body, speech and mind. The Blessed Tathāgatas said: where does mind exist? He answered: it exists in space. They asked: where does space exist? He answered: nowhere.” (Fremantle 1970, 110). While it does seem that this parallel passage in the Dvitiyakrama and in the Garland of Views is related to, or perhaps based on, the passage I have cited here from Chapter 15 of the Guhyasamāja-tantra, it is not drawn directly from that tantra, and yet it appears in both the Dvitīyakrama and the Garland of Views. I discuss the relationship between Buddhajñānapāda’s writings and the early literature of the Great Perfection in Chapter Four. 88 khams. 89 Vaidyapāda specifies that it is non-abiding nirvāṇa that is intended here. srid pa dang zhi ba la mi gnas pa’i mya ngan las ‘das pa ste/ (Sukusuma, D 96b.4; P 115b.8-116a.1). 90 don. Throughout this translation “reality” is used to translate don. 368 Is explained as91 the second stage.92 |34| [How Saṃsāra Arises out of Nondual Wisdom]93 The reality which is like that Is present pervading all entities. Yet,94 from beginningless time, even from this There was arising in the manner of the great thought.95 |35| And from that also the great elements [arose]: The maṇḍala of wind arose, And from that also, the great element of fire Arose and spread. |36| From that, the great element of water also Arose and spread, and from that also earth. From the essence of the gathering of the four [elements] Mountains, and so forth, and all sentient beings also |37| In all their variety, subtle and gross: Men, women, and hermaphrodites, The young and old, Gods and nāgas and yakṣas, |38| Evil spirits, planets, Yāma, The Lord of Water, Indra,96 hell beings, 91 rim pa gnyis par. This could be translated equally as either “in the second stage” or “as the second stage.” It seems that it is possible also to understand it in both ways. It could be that this truth is explained in the teachings on the second stage, or that the truth itself is the second stage, as it is the “perfection” stage, in which the truth itself is perfected or manifest. 92 Vaidyapāda gives three synonyms for the second stage: the spontaneously arisen stage (lhan cig skyes pa’i rim pa), the perfection stage (rdzogs pa’i rim pa), and the stage of [things] just as they are (ji bzhin pa’i rim pa). He then gives a brief description of the generation stage: “as for the generation stage it is for the purpose of reversing the coarse delusions of the world and its contents. This yoga that involves engaging with the conceptual mind is the first [stage].” rim pa gnyis par rab tu bzhad zhes pa ni/ lhan cig skyes pa’i rim pa am/ rdzogs pa’i rim pa am/ ji bzhin pa’i rim pa (rim pa] P; om. D) rnam grangs so// bskyed pa’i rim pa ni snod bcud rags par ‘khrul pa bzlog pa’i phyir ro// blos rnam par gzhag pa’i rnal ‘byor pa ste dang po’o// (Sukusuma, 96b.6-7; P 116a.5). 93 Vaidyapāda gives this heading for the section: da ni gnyis su med pa’i ye shes las ‘khor ba ji ltar ‘byung ba (Sukusuma, D 97a.2-3; P 116b.1-2). 94 Vaidyapāda explains further, “Although that kind of nonduality pervades and remains [as the nature of] all entities, the reason that this is not apparent is explained with the lines beginning, Yet, from beginningless time...” de yang gnyis su med pa de lta bus dngos po kun rnam par khyab ste gnas kyang de mi gsal ba’i rgyu ni thog med dus nas zhes te/ (Sukusuma, D 97a.4; P 116b.2). 95 Vaidyapāda elaborates, “From that time, just as clouds arise within space, the great thought, the mind alone, arose in a manner [in which it appeared] as if it were endowed with conceptuality.” de’i dus nas nam mkha’ la sprin ‘byung pa bzhin du rnam rtog chen po zhes te sems tsam rtog (rtog] P, rtogs D) pa dang ‘brel pa lta bu’i tshul du byung ste zhes bya ‘o// (Sukusuma, D 97a.4; P 116b.3-4). See Muktitilaka verse 101, which is parallel (Muktitilaka, 50b.4-5). This seems to refer to the cosmogonic moment where conceptuality (seemingly) arises out of a reality that is nonconceptual. 96 Vaidyapāda omits references to the Lord of Water and Indra. This is unusual since he otherwise follows the root verses exactly here in this section (Sukusuma, D 97b.1). 369 Pretas, animals, and those who abandon all of this,97 Beings who rely upon consciousness alone,98 Such beings abide, spread far and wide.99 |39| Therefore, the nondual nonconceptuality That is higher than that is completely obscured.100 Because of not realizing it, all beings Cycle around in saṃsāra. |40| Due to their delusion they bring about The great suffering of saṃsāra, which is like poisoned water. [And yet] it is just like a snake, which is “seen” when looking at a rope But, in fact, is held not to exist.101 |41| Thus, this great nondual reality Is the supreme great maṇḍala of self-awareness102 Which abides continuously, and in many forms—103 97 Vaidyapāda identifies these as the śrāvakas and so forth. de kun spangs pa ni nyan thos la sogs pa’o// (Sukusuma, D 97b.1; P 117a.1-2). 98 Vaidyapāda identifies these as those beings of the realm of limitless space, and so forth, since they have abandoned form. nam mkha’ tha’ yas la sogs pa ste/ gzugs spangs ba’i phyir ro// (Sukusuma, D 97b.1-2; P 117a.2). 99 Vaidyapāda comments that the statement that these beings live far and wide means that, “having been produced by conceptuality, they appear in the ten directions.” de kun rgyas par gnas zhes pa ni rtog pas bzo byas nas phyogs bcu kun du snang ba’i phyir ro// (Sukusuma, D 97b.2; P 117a.2-3) This cosmogony, where beings are said to be produced by conceptuality or thought, is echoed in the perfection stage ritual practices outlined later in the Dvitīyakrama; see verse 165. 100 bsgribs] P N S V(D and P), bsgrims D C. 101 In using the classical Yogācāra metaphor of the “rope-snake,” Buddhajñānapāda here displays the strong Yogācāra bent found throughout his works, which is echoed in Vaidyapāda’s writings, as well. The same metaphor is also used in the Muktitilaka (D 47a.7). Nonetheless, when providing a doxography of philosophical systems later in this work Buddhajñānapāda places Madhyamaka above Yogācāra. I discuss Buddhajñānapāda’s philosophical views briefly in Chapter Three. 102 Vaidyapāda notes that, “In the maṇḍala-cakra there have never been any saṃsāric phenomena, so despite [their] being produced by conceptuality, they are not [actually] there.” dkyil ‘khor gyi ‘khor lo la ‘khor ba’i chos ye mes pa la rtog pas bzo byas kyang de de la med pa yin no// (Sukusuma, D 97b.5; P 117a.6-7). At this point in the commentary he gives a supportive citation which he attributes to the Vajrajñānasamuccaya-tantra (Sukusuma, D 97b.6-7; P 117a.7-8). Vaidyapāda gives the very same citation from this tantra in his Ātmārthasiddhikaranāmayogakrama (D 87a-b) and in his commentary to the Caturaṅgasādhana (Caturaṅgasādhana-ṭīkā, D 165b). The Vajrajñānasamuccaya-tantra is one of the so-called “Ārya-School” explanatory tantras, which sets out, among other things, the six parameters and four modes that characterize the Ārya School’s unique interpretive framework for the Guhyasamāja-tantra, so it would be interesting to see it cited by Vaidyapāda, probably the earliest commentator on Buddhajñānapāda, in what is clearly a Jñānapāda-School commentary. However, the citation that Vaidyapāda gives—a teaching from the Lord of Secrets (Vajrapāṇi) on the fact that the apprehension of form, which is a delusion, is due to habit just like the apprehension of a rope-snake—is not found anywhere in the Vajrajñānasamuccaya-tantra. I have thus far been unable to locate the citation in any other work, either. In the Muktitilaka-vyākhyāna Vaidyapāda likewise makes reference to the Vajrajñānasamuccaya (or rather the Jñānavajrasammucaya—the title is rendered there as ye shes rdo rje kun las btus pa—but presumably refers to the same text), but there his reference does not include a citation (Muktitilaka-vyākhyāna, D 51a.5) 103 rgyun dang rnam grangs gyis. Vaidyapāda’s explanation here seems to suggest that he understands this nature to be, in some sense, permanent: “Abides continuously and in many forms means the profound, which is permanent in that its continuity is permanent. The wisdom that realizes this is to be seen as permanent due to the permanence of its many forms.” rgyun dang rnam grangs kyis gnas pa zhes pa ni zab mo ni rgyun gyi (gyi] P, gyis D) rtag pas na rgag pa’o// de chub pa’i ye shes ni rnam grangs kyi (kyi] P, kyis D) rtag pas na rtag par blta’o// (Sukusuma, D 370 One must realize its essence. |42| [The Individual [i.e. Disciple] Who Undertakes the Path to Realize This]104 Whoever aspires towards realizing105 that, And genuinely holds in mind an aspiration, Who has previously generated the power of merit, And relies as his foundation,106 upon an authentic being, |43| Who has joy, respect, honor, And proper intention, who trains, Is stable, and has completely abandoned doubts, Who is compliant107 and is generous—he will understand it. [4a] |44| [The Guru Who is to be Pleased by Him]108 He should please a guru who is genuine and venerable, Who possesses the lineage of supreme oral instructions,109 Who is intent upon the conduct and training Of the Mahāyāna path, |45| Who knows the secret, great secret, and exceptional secret110 Of the ten suchnesses,111 and so forth, 98a.3; P 117b.5). Khenchen Chodrak Tenphel explains that rgyun here refers to the fact that the nature is continously present in the context of ground, path, and fruition, and that rnam grangs refers to the various enumerations and elaborations that we use to describe this nature, such as the five kāyas or the five wisdoms, and so forth (Khechen Chodrak Tenphel, personal communication, February 2016). 104 Vaidypāda heads this section in the commentary this way. da ni rtogs par byed pa’i lam la ‘jug pa’i skyes bu gsung pa/ (Sukusuma, D 98a.4; P 117b.7-8). 105 rtogs] P N S C, rtags D 106 gzhi] sugg. em. based on V (D and P), bzhi D C P N S 107 rjes mthun, Skt. *anurūpa? Vaidyapāda takes rjes mthun here to refer to dbang gi rjes su mthun pa, indicating offerings that correspond with the specific initiations for which they are given, which he notes will be explained below (Sukusuma, D 98b.2; P 118a.6). 108 Vaidyapāda gives this headline: des mnyes par bya ba’i bla ma (Sukusuma, D 98b.3; P 118a.7). 109 Vaidyapāda elaborates: “[A guru] who possesses the lineage of supreme oral instructions means [a guru] who has ascertained in his own mind the essence of Vajradhara [and] who knows that which has been passed from ear to ear, just exactly as it is.” mchog gi gdams ngag rgyud la ldan zhes te/ rdo rje ‘chang chen po’i ngo bo rang gi thugs la mnga’ ba rna ba nas rna bar ‘pho ba ji lta ba bzhin du shes pa’o// (Sukusuma, D 98b.3-4; P 118a.8). 110 sang ba, rab gsang, and shin tu gsang. Vaidyapāda identifies these as the generation stage, (bskyed pa’i rim pa), illusory samādhi (sgyu ma lta bu’i ting nge ‘dzin) and the suchness of the *ādhideva and all phenomena (lhag pa’i lha dang dngos po thams cad kyi de bzhin nyid do/), respectively (Sukusuma, D 98b.7-99a.1; P 118b. 6-7). The latter two appear to correspond to different levels of perfection stage practice. The first of those mentioned here has the same name, māyopama-samādhi, that we find in later Ārya School terminology referring to the fourth of the five stages of Ārya School perfection stage practice. The term is used in verse 19 of Vaidyapāda’s Yogasāpta, in the context of the practices connected with the guhyābhiṣeka (Yogasapta, D 70b; P 84a). 111 See Klein-Schwind’s dissertation on the daśatattva (Klein-Schwind, 2012). Vaidyapāda lists the ten suchnesses here as: de nyid bcu zhes pa ni rgyud thams cad kyi de nyid bcu ni dkyil ‘khor dang sbyin sreg dang/ ting nge ‘dzin dang/ phyag rgya dang/ stang stabs dang/ (D adds: la ni lhur len ba zhes pa ni lam la mos/) ‘dug stabs dang/ bzlas brjod dang/ mchod pa dang/ las la sbyor ba dang /slar bsdus pa’o// (Sukusuma, D 98b.5-6; P 118b.3-4). The infelicitious addition of the line in the middle of the list of the ten suchnesses in the Derge recension appears simply to be an instance of dittography from the previous line. This list of ten is not consistent, however, with the (admittedly variable) list of ten suchnesses found in other later sources (see Klein-Schwind 2012, 47-8 for a comparative table from several sources) and indeed not even with those among the ten tattvas that Vaidyapāda identifies in his Guhyasamājamaṇḍalopāyikā-ṭīkā. 371 And who will teach Those to whom such reality is concealed. |46| [Offerings to the Guru] One’s lands,112 houses, mansions, horses, Elephants and the like, beds, Beautiful wife and delightful sons, Daughters, sisters, and nieces,113 |47| Also, gold114 and silver, Beautiful things made of copper, Iron, and the like, Strings of pearls,115 rubies, |48| Amber,116 sapphire, Emerald, turquoise, and others, these many offerings— The skillful person fills this realm with such things117 And offers them to such a venerable teacher.118 |49| [Descriptions of the Types of Tantric Consort] 112 The offering of land to the guru mentioned here is an important indicator of the changing socio-political climate of Indian religious practice in the late 8th and early 9th centuries. Royal clientele began giving land grants to religious institutions, and this was part of the major shift in political structures that took place in the early medieval period (see, e.g. Thapar 2002, 451). In fact, there are accounts of Buddhajñānapāda’s having been patronized by the Pālā kings, mostly in the later Tibetan sources, but such an alliance is also briefly reported by Atīśa (Szántó 2015, 539; the Tibetan accounts include those by Chögyal Phagpa and Tāranātha). Vaidyapāda also, however, refers very briefly to several events from Buddhajñānapāda’s life that appear serve as the basis for the accounts that were expanded in the Tibetan histories to describe royal patronage (Sukusuma, D 135b). These expanded accounts were presumably based on oral history, as well as on written histories to which we no longer have access. I discuss some of these features of the socio-political context in which Buddhajñānapāda lived, as well as the accounts of his life, in Chapter One. 113 Vaidyapāda notes that these first substances are what is to be offered for the kalaśābhiṣeka (bum pa’i dbang) because they accord with that initiation since it “is the foundation” (gzhir gnas pa’i phyir ro) (Sukusuma, D 99a.2; P 119a.1). 114 Vaidyapāda notes that from here the offerings are for the guhyābhiṣeka (Sukusuma, D 99a.3; P 119a.2). 115 Vaidyapāda notes that from here the offerings are for the prajñājñānābhiṣeka (Sukusuma, D 99a.3; P 119a.2). He only mentions offerings connected with three initiations here, suggesting that here he only understands there to be three. However, Vaidyapāda also composed the Yogasapta-nāma-caturabhiṣeka-prakaraṇa (Tōh. 1875), in which, as evidenced by its title, he acknowledges a “fourth,” as well, though he shies away from calling it a “fourth initiation.” I address the topic of “the fourth” briefly in Chapter Three and in more detail in Chapter Seven. 116 pu shel] sugg. em., pu shar D C P N S. This suggested emendation is based on the oral commentary of Khenchen Chodrak Tenphel who suggests that the text should read pu shel, “amber,” rather than pu shar, which does not yield any sense (personal communication, February 2016). Vaidyapāda’s Sukusuma does not mention the term, but in his Yogasapta Vaidyapāda includes this substance in a list of initiation offerings which he has obviously taken from the Dvitīyakrama (Yogasapta, D 70b.4). There the term is given as pur sha, which also lacks sense. 117 Vaidyapāda offers the possibility of filling all of space with mentally produced offerings here “if one is unable,” to offer in the way just described (Sukusuma, D 99a.3-4; 119a.3). To substantiate this, he cites Chapter 8, verse 22 of the Guhyasamāja-tantra, in which it is not possible to offer the offerings described physically, so they must be understood as mentally created offerings. This eighth chapter of the Guhyasamāja-tantra has been understood in the commentarial tradition to be referring to the context of initiation, which is exactly the context that Vaidyapāda understands here—offerings made for receiving initiation (Sukusuma, D 99a.4; P 119a.4). 118 Vaidyapāda notes that this was the ordinary way of pleasing the guru (mnyes byed thun mong pa) (Sukusuma, 99a.6). 372 That which is luminous and joyful, equal to space—119 One will not know120 it any other way.121 Thus, a woman,122 the illusory mudrā, Is superior among all illusions.123 |50| This illusion here in this world, Because of124 having Locanā and so forth as her pure forms,125 Is of four types. 119 Vaidyapāda specifies that this refers to nondual wisdom (gnyis su med pa’i ye shes) (Sukusuma, D 99a.7; P 119a.7). 120 rig] S P V(P), rigs D C N V(D). 121 Vaidyapāda notes that it cannot be recognized without the seven yogas, a system of yogas connected to the practices of “the fourth,” addressed by Vaidyapāda in his own Yogasapta-nāma-caturabhiṣeka-prakaraṇa. Buddhajñānapāda himself references, but does not list, the seven yogas in his Muktitilaka, and Vaidyapāda in his commentary on that text mentions to the seven in reference to their connection with the three joys, but does not list them individually. In Vaidyapāda’s Yogasapta, however, the seven yogas are explained in much greater detail as seven states or experiences that are part of the result of awakening, and which somehow constitute “the fourth;” the seven are perfect example-less bliss (dpe med bde rdzogs), nonduality (gnyis su med pa), great bliss (bde ba chen po), lacking nature (rang bzhin med pa), unfolding compassion (thugs rjes rgyas pa), unbroken continuity (rgyun mi chad pa), and non-cessation (‘gog pa med pa). The fact that Vaidyapāda discusses, in that text, “the fourth” despite the fact that Buddhajñānapāda’s tradition is known for including just three initiations, is an issue I have taken up briefly in Chapter Three, and in more detail in the discussion of initiation in Chapter Seven. Regarding these seven yogas, here the Sukusuma Vaidyapāda writes: “That [nondual wisdom] comes about through training. Regarding its realization, [the text says] in any other way, meaning that one will not know that [by means of relying upon] any other thing besides the prajñāpāramitā-mudrā (i.e. a female consort). Even though [nondual wisdom] pervades all things, since other [things] lack the seven yogas [one can not know it through them].” de ni bsgom (bsgom] P, bsgoms D) pa las byung la/ rtogs pa de yang gzhan du zhes te shes rab kyi pha rol tu phyin pa’i phyag rgya las gzhan pa’i chos can gyi dngos po gang du yang de rig (rig] P, rigs D) par mi gyur te/ dngos po thams cad la khyab kyang gzhan ni sbyor ba bdun dang bral bas na’o// (Sukusuma, D 99a.7-b.2; P 119b.1-3). The same seven factors (with the second called “union” rather than “non-duality”) are addressed in Vāgīśvavarakīrti’s later Saptaṅga and his Tattvaratnāvaloka and its auto-commentary, where they are called the seven aṅgas of mahāmudrā, with reference to which see Isaacson (2010b, 271, 271n27) and, with a bit more detail, Isaacson and Sferra (2014, 271), where they are mentioned with reference to a citation from the Saptaṅga in Rāmapāla’s Sekanirdeśapañjikā. The seven aṅgas are listed in Vāgīśvarakīrti’s work as sambhoga, sampuṭa, mahāsukha, niḥsvabhāva, kāruṇyanirbhara, nirantara, anirodaḥ. I discuss the seven yogas in more detail in Chapter Seven. 122 Vaidyapāda helpfully defines a woman as someone with a lotus (padma, i.e. vagina) and breasts (dkar ‘chang) (see note 132 for a discussion of the latter term) (Sukusuma, 99b.2). 123 bud med sgyu ma’i phyag rgya ni/ sgyu ma kun las khyad par ‘phags. These two lines have strong parallels with the first two lines of the Sarvabuddhasamāyoga-tantra 1.4, which read, in Sanskrit, sarvāsām eva māyānāṃ strīmāyā praviśiṣyate |, and in Tibetan translation, sgyu ma dag ni thams cad pas/ bud med sgyu ma khyed par che/ (Sarvabuddhasamāyoga, D 151a.3). The Sarvabuddhasamāyoga also mentions the woman as a mudrā in the last two lines of the immediately preceding verse: sarvastrīmāya mudreyam advayaṃ yānam uttamam |; bud med kun gyi sgyu ma’i rgya/ ‘di ni gnyis med theg pa’i mchog (Sarvabuddhasamāyoga, D 151a.2). Thanks to Ryan Damron for bringing these Sarvabuddhasamāyoga parallels to my attention and to Péter Szántó for sharing with me his draft edition of the Sarvabuddhasamāyoga-tantra. The two lines from the Dvitīyakrama are also strongly paralleled in Śākyamitra’s Anuttarasandhi, included as the second stage in Nāgārjuna’s Pañcakrama, which reads: sarvāsām eva māyānāṃ strī-māyaiva viśiṣyate/ (Mimaki and Tomabechi 20); sgyu ma dag ni thams cad las/ bud med sgyu ma khyad par ‘phags (Pañcakrama, D 49a.7; Mimaki and Tomabechi 20). Tomabechi (2006, 132n128) has already noticed all of these parallels and additionally notes that a passage identical to that in the Pañcakrama is found in the Vajramaṇḍālaṃkāra. 124 pas] P N S, par D C. 125 Here Buddhajñānapāda is referring to the viśuddhi, or “pure forms” of the consorts, who are identified as the four female buddhas from the Guhyasamāja-tantra. Vaidyapāda further relates this to the fact that the four female buddhas are identified in the tantras as the pure forms of the four elements (Sukusuma, 100a.1). 373 Their names and characteristics will be explained.126 |51| [They are called] kamalī127 and śaṅkhīnī 126 Buddhajñānapāda’s mention of the four consort types, which correspond to the classic four-fold typology of women in Indian kāmaśāstra, is an early one in Buddhist literature, and indeed in Indian literature on the whole. In terms of Buddhist sources, it is the earliest mention of such a four-fold classification—either in the scriptural or commentarial literature—that I am familiar with. The very same four types of consort mentioned here in Buddhajñānapāda’s text are found, however, in later Buddhist tantric literature, including Chapter 18, verse 1 of the Samvarodaya-tantra (except there the first type, instead of kamalī, is called padminī, which is in fact the much more common name for this particular type; the two categories obviously correspond, though, given that both are derived for words meaning “lotus.” The Samvarodaya passage on the four types appears likely to have been influenced by the Dvitīyakrama, as there are a number of parallels. For the Samvarodaya passage see Tsuda 1994, 155-57 and 324-35.). The Samvarodaya, however, is quite a bit later than the Dvitīyakrama; Isaacson has suggested it may be of Nepalese origin and date to as late as the 12th century (English 2002, xxi, 384n2). With regards to non-Buddhist Indian literature, after Vātsyāyana’s famed 3rd-4th century Kāmasūtra, which does not mention such a four-fold categorization of women (Vātsyāyana instead has a six-fold schema), Kokkoka’s Ratirahasya, and the Buddhist author Padmaśrī’s Nāgarasarvasva are considered to be the earliest of the “later” kāmaśāstra works (Ali 2011, 43). The Ratirahasya is reported to be the first text to describe women in these classical four types (Datta 1988, 1203; Ali 2011, 45). That text is, however, difficult to date, and its dates are given by some scholars as 9-10th century, by others as the 10-12th century, and by others as late as the 13th century (See Ali 2011, 44 and 44n14; Datta 1988, 1203; and Hopkins 1992, 35 and 35n4). The 9th-10th century dates posited by some scholars, seem to be based on erronous and/or ambiguous references to the Ratirahasya in the works of the 10th century author Somadevasūri, and all that may be said with certainty is that the Ratirahasya is cited by commentators beginning only in the 13th century (Ali 2011, 44n14 and 44). In any case, even with the earliest dates posited by scholars, it seems that the Ratirahasya is certainly later than Buddhajñānapāda’s late 8th/early 9th-century Dvitīyakrama, which lists precisely these four categories (though as in the Samvarodaya the first category according to Kokkoka is padminī rather than Buddhajñānapāda’s unusual kamalī). There is one other possibly early but difficult-to-date source for this four-fold categorization: *Surūpa’s Kāmaśāstra (Tōh. 2500), which, like the Dvitīyakrama, is not extant in Sanskrit but is preserved in Tibetan translation (and may therefore not have been considered by Indologists discussing early kāmaśāstric literature). We know nothing of the author of this treatise, though the homage and one of the concluding verses of the work suggest he was a Buddhist, and he tells us that his work was composed on the basis of Nāgārjuna’s treatise on erotics. Vogel, who has edited and translated the Kāmaśāstra, suggests that this must be the “tantric” Nāgārjuna, who he dates to the 6th century (Vogel 1965, 5). However, if it is indeed a work by this Ārya School author Nāgārjuna on which *Surūpa’s work is based (and this point is itself not entirely certain; Vogel notes that “the work admittedly used by Surūpa as his source does not appear to be identical with any of the several known treatises entitled Ratiśāstra that go by the name of Nāgārjuna and differ widely from each other (Vogel 1965, 5)— this author, we now know, was likely writing slightly later than Buddhajñānapāda, in the 9th, not the 6th, century. (A Nāgārjuna, by the way, is also listed as a source in Kokkoka’s Ratirahasya (Ali 2011, 60).) *Surūpa’s work, then, also appears to be later than Buddhajñānapāda’s, making the Dvitīyakrama the earliest known locus, Buddhist or non-Buddhist, of this important four-fold kāmaśāstric categorization of women. The (much) later Tibetan commentaries on the Guhyagarbha such as Longchenpa’s 14th-century Phyog bcu mun gsel commentary to the Guhyagarbha likewise have a four-fold typology, but it is different from the classical kāmaśāstra typology (See Dorje 1987, 902). The Guhyagarbha-tantra itself, which we can date to the 8th century, only says “discriminating between devīs, nāginīs, and female mudrās of inferior species, or else without discrimination,” at the point where Longchenpa gives his extensive commentary on the four types of consorts (Dorje 1987, 883). Other Buddhist tantras have different schemas of classification of consorts, or of women in general, like the tantras pertaining to the Cakrasaṃvara tradition, including the Cakrasaṃvara-tantra itself and the Abhidhanottara-tantra, which have a seven-fold classification (See Gray 2007, 227-29 and Kalff 1979, 237-38, respectively); and the Sampuṭa-tantra and Caṇḍamahāroṣaṇa-tantra, which each have a five-fold one corresponding to the five buddha families (See Sampuṭa, 1.1.42-1.1.45 and Caṇḍamahāroṣaṇa, 8.15-8.17). I briefly discuss the relationship between Buddhajñānapāda’s writings and kāmaśāstra in Chapter Six. 127 All versions of the text read kamalī, which is an unusual and unexpected form. Kamalinī would be the expected feminine form that would correspond with the other names given in the verse. It is possible that the Tibetan translators may have simply shortened the form for metrical reasons. Vaidyapāda’s Sukusuma also reads kamalī, but this may again simply be because the translators of the commentary were referencing the Tibetan translation of the 374 Citriṇī and hastinī. The first is of the class of nāga women. The second is of the class of tigers and lions. |52| The third is [of the class] of the wild black antelope,128 and so forth.129 The fourth is of the class of elephants. Here, as for kamalī I will explain her shape and characteristics: |53| She is a girl who is redolent with the scent of lotus Her face is round, the tip of her nose like a mustard seed, Her nails are red and her back is bent [out of respect].130 The soles of her feet rest flat upon the earth. |54| Her body hairs coil and she is golden.131 Her breasts132 are like the fruit of the mustard plant. She has three wrinkles at her waist.133 Her chest is lovely, and she has the [leisurely] gait of an elephant. |55| The taste of her blood is sour134 Her skin is reddish. The pure form of this goddess is Māmakī. I will explain to you the characteristics |56| And shape of śaṅkhinī: The girl has the scent of sulphur.135 Her hair is long, and her nose is long. root text. I have not taken the liberty of changing the text in my edition, however, as this would render unmetrical all verses in which it occurs. I am grateful to Mattia Salvini for a helpful conversation on this topic. 128 ri dwags kṛṣṇa 129 sogs] D C V (P and D), tshogs P N S 130 rgyab sgur (Skt. *kubjā?). The term normally means hunchbacked, but Khenchen Chodrak Tenphel explains that here it is meant to indicate a respectful body posture. (Khenchen Chodrak Tenphel, personal communication, January 2016). 131 ser] sugg. em. based on V (D and P), sen D C, se P N S. Vaidyapāda makes it clear that this refers to her coloring. Below Buddhajñānapāda states that her skin color is reddish. In the Cakrasaṃvara-tantra the type of woman who corresponds to kamalī is described as “reddish-golden” (Gray 2007, 236). 132 dkar ‘chang. Vaidyapāda makes it clear that this term refers to breasts: “White means milk. That which holds this are breasts.” dkar ba ni ‘o ma’o// de ‘chang ba ni nu ma ste/ (Sukusuma, D 100a.7; P 120b.6). I believe that this is likely a translation of one of the Sanskrit terms for breasts payodharā—literally “that which holds milk.” The Tibetan dkar ‘chang could also be understood to mean this, given that the term dkar is often used for dairy products, in general, thus including (at least cow’s) milk. However, I have not been able to find any other uses of the term dkar ‘chang in Tibetan. 133 Vaidyapāda comments: “Below her navel [she has three wrinkles] that look like a triśūla.” (lte ba’i ‘og tu tri shū la (shū la] D shu la P) lta bu zhes pa’o// (Sukusuma, D 100b.1; P 120b.7). This is a classical mark of beauty in Indian literature. 134 skyur] C P N S V(D and P), skar D 135 tsha’i] D C V(D and P), tshī S P, tshwi(?) N. Reading the word as ba tsha Khenchen Chodrak Tenphel explains that this refers to the scent, somewhere between pleasant and unpleasant, that comes from natural hot springs. I understood this to be the scent of sulfur. 375 Her breasts are like oranges.136 |57| Like milk mixed well with yogurt, The taste of her blood is sweet. Her color is whitish yellow and Her pure form is Pāṇḍarāvasiṇī. |58| Likewise, I will explain the shape And characteristics of137 citriṇī138 The girl is redolent of the scent of fresh meat. Her body is small, and her thighs are exquisite. |59| Her breasts are like bel fruits139 She is not shy, and she is keen on anger. She always likes to quarrel. Her calves are like the legs140 of a crow. |60| Her lower lip protrudes downwards, and she sleeps on her back. Her speech sounds like a pigeon and her skin is dark.141 The taste of her blood is salty. Her pure form is the goddess Tārā. |61| Then, I will explain the characteristics And shape of hastīnī.142 The girl smells like beer. Her calves are thick and her nose is slightly crooked. |62| She smells foul and her body is thick Her conduct is crude and her color is a dusty gray. Her breasts curve to the right and left.143 Her skin is purple and Her blood tastes unpleasant.144 136 nā ga ra] sugg. em., na ra ga D C P N V(P), ma ra ga S, na ga ra V(D). I have translated this as “orange” based on my suggested emendation of the text to nāraga, which I derive from the reading in the Derge recension of Vaidyapāda’s commentary, nagara, which I then emend to nāgara, one of the meanings of which is “an orange.” Neither hell (naraga (= naraka?)) nor a village (nagara) seem to fit well as a description of breasts, and Vaidyapāda anyway clarifies that this is a type of fruit. However, Vaidyapāda's description of the particular oranges intended here is not exactly what one might expect: "Like a fruit whose top part is large, but whose main part is slack." shing tog rtse mo che la rtsa ba zhum pa ste de ‘dra ba’o// (Sukusuma, D 100b.2; P 121a.1). 137 yi] P N S, ya D C 138 tsi tri nī] sugg em., tsi tri ni D C, tsi tra ni P N S 139 pa la. In Vaidyapāda’s commentary this is rendered as dpal. He specifies that “they are like the bel fruit means [that they are] small and round.” dpal gyi ‘bras bu ‘dra (‘dra] D, P om.) zhes bya ste zlum shing chung ba’o// (Sukusuma, D100b.4; P 121a.3-4). 140 rkang] D C, dang P N S 141 ngo bsangs. 142 hastī nī] sugg. em. hastī ni D C, hasti ni P N S 143 dkar ‘chang g.yas g.yon du dgye ba. I am unsure about this phrase. 144 bsngal. I am unsure about the meaning of this term. 376 The pure form of this type of girl is Buddhalocanā. Now I will teach about the The acceptable and unacceptable conduct for all of them. |64| [Unacceptable Conduct for the Consort] At the time of sporting she turns her back, She is inwardly lustful145 and overly talkative, Breaks the commands that she has been given, And wipes her mouth when it has been kissed, |65| Even though she knows the guru’s qualities She relates them only a little to others.146 She pretends that she has not seen the guru, And though coming [before him] does not prostrate to him. |66| She feasts with other people who Do not get along with the revered [yogin]. These, and so forth, are completely inappropriate [behaviors]. The intelligent [yogin] avoids147 someone who is like this.148 |67| [Acceptable Conduct for the Consort] The intelligent [yogin] searches for the someone appropriate: Radiant eyed,149 with an alert gaze,150 She repeatedly sets aside passing thoughts, Or momentary excited distractions, |68| Thinks [first], and [only then] either speaks or smiles.151 When seeing the venerable guru, She looks joyfully and beautifully at him and smiles. She listens to his commands with interest. |69| She is charmingly respectful. Moreover, she speaks sweetly to him. When she sees him she embraces and serves her companion. She discreetly embraces him, and massages his finger joints. |70| 145 khong du dga’. I here follow Vaidyapāda who explains this to mean desirious (Sukusuma, D 101a.2; P 121b.2). 146 gzhan du de la cung zad bsnyad. Vaidyapāda clarifies that this means that she accuses him of having faults that he does not have (cung zad bsnyad (bsnyad] D, snyad P) ces pa ni med pa’i skyon brjod pa’o) (Sukusuma, D 101a.2; P 121b.2). 147 spang 148 Vaidyapāda explains why such a partner is to be avoided: “A disharmonius consort does not bring about the occasion for one’s physical or mental happiness.” (rje su mi mthun pa’i phyag rgyas lus sems bde bar byed pa’i skabs med de/) (Sukusuma, D 101a.3; P121b.3-4). 149 Vaidyapāda comments that this means there is a clear distinction between the white and black parts of her eyes (dkar nag phyed pa’o/) (Sukusuma, D 101a.4; P 121b.5). 150 mig rtsa rgod. 151 I am not completely certain about the meaning of these few lines. slar zhing skad cig bsam pa dang// skad cig rgod pa’i ‘phro bzhag (bzhag] D C S P, gzhag N) nas// bsam zhing yang na smra ‘am ‘dzum// 377 She loosens her hair and binds it up again. She mends his clothes, and periodically smiles. She pretends not to see the guru,152 And mends his clothes153 and massages his body. |71| She bows at his feet and is inquisitive She gives milk154 to the mouth of her infant; She shows her breasts and loosens her belt. When not seeing the revered [yogin], |72| She sings and makes him visible.155 Wherever he is She goes there, making up a pointless excuse Lauging vigorously, she flirts |73| And pretends to clear her throat and yawn— [5a] One should rely upon such a human woman. |74| [Characteristics of the Consort’s Secret Place] Now I will authentically explain The shape of156 her secret place, And its appropriate and inappropriate characteristics. Listen one-pointedly. |75| If it is like pot shards, or facing downwards,157 Very deep, or very ugly, Or if it is always dry, Or always damp and dripping, |76| Or rough like the back of a toad, This [type of secret] place is to be avoided; it is unacceptable. Elevated like a turtle’s back, Its upper part symmetrical, and very smooth, |77| Like the nape of an ox’s neck, Its upper part is even and long;158 152 Presumably this time coquettishly, unlike in verse 66 above. 153 This is likely dittography from the previous line, but it is preserved in all of the Tibetan recensions. 154 While ‘o byed usually means “to kiss,” I believe that based on the context here it refers to giving milk to an infant. 155 This is, perhaps, an allusion to the section in many generation stage sādhanas, including Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana where the goddesses sing to make the main deity appear from a bindu of light as part of the generation process; the songs found in Guhyasamāja sādhanas are taken directly from the root tantra itself. 156 kyi] D C, kyis P N S 157 kha sbub. Literally, “upside down.” Vaidyapāda notes that this means it is located too (?) low. kha sbub ni ‘og na gnas pa ltar ro// (Sukusuma, D 101a.5; P121b.7). 158 Vaidyapāda comments, “Like the nape of an ox’s neck means that the two sides and the anthers (i.e. clitoris) are like the nape of the neck and the shoulders. Moreover, its upper part is even and long means that most of it is [located] [towards] the upper [area].” ba glang (glang] D, lang P) gi ni ltag pa ltar/ zhes pa ni ‘gram gnyis dang 378 Round like the center of a lotus; Not dripping at all; protruding slightly159 |78| And with little hair160—such an illusory mudrā Is skillfully and exuberantly searched for by the intelligent [yogin].161 |79| Even if she does not have these characteristics, If she has caste,162 beauty, and youth, This delightful girl, beautified by ornaments, Endowed with [these] three enjoyments, should be taken up. |80| Devoted towards the profound dharma, And not entangled with karmic relations163— A companion, as described above, who has swayed [the yogin’s] mind—164 Such a human woman165 should be obtained by the intelligent [yogin]. |81| In that way, by means of the illusory great mudrā Of that type of female, The so-called *adhideva,166 ze’u ‘bru ste ltag pa dang dpung mgo ltar gnas te/ de yang steng mnyam zhing thog gi dbyibs kyang ring zhes pa ni thog tu phel che bar gnas pa’o// (Sukusuma, D 101b.1; P 122a.2-3). 159 Vaidyapāda comments, “Not dripping means [it does not drip] when [she is] not practicing. When practicing it should slightly protrude.” ‘dzag min zhes pa ni nyams su ma blangs na zhes so// nyams su blangs na cung zad ‘byung ba ste (Sukusuma, D 101b.3; P 122a.4-5). 160 Vaidyapāda comments, “With little hair means with the hair removed to make it appealing.” spu nyung zhes pa ni spu langs pa ste yid du ‘ong bar byed pa’i phyir ro// (Sukusuma, D 101b.4; P 122a.5). 161 Vaidyapāda comments that the yogin’s exuberance in procuring a partner should not, however, involve unseemly conduct. thabs kyis shin tu btsal zhes pa ni kha na ma mtho ba med pa (Sukusuma, D 101b.3; P 122a.5-6). 162 Vaidyapāda, however, notes that her caste is indeterminiate (mi nges pa) (Sukusuma, 101b.6). 163 las ‘brel gyis ‘khyud ma byas dang. I am unsure of the meaning of this line. Vaidyapāda seems to interpret it to mean that the woman is not held on to too closely by her relatives (her mother, father, brother etc.) such that they would be resentful of the yogin’s relationship with her (Sukusuma, D 102a.3-4). It could also mean a girl who has not been sexually involved (‘khyud pa) with a person with whom the yogin has a close karmic relationship, like a fellow student, etc. 164 sngon gsungs grogs kyis shes bslus pa’i//. I am unsure about my reading of this line as well as of Vaidyapāda’s comments here. “Such a one who has swayed the mind means [the following]. Having reversed [his] first thought by means of [her] pleasant speech [is what is meant by] swayed. Then, [he] is placed in the state of [having] trust [in her].” sngon gsungs grogs zhes pa ni gong du smos pa’i rjes su mthun pa’i phyag rgya gang zhig gis/ shes pa bslus pa zhes te rjes su mthun pa’i gdam gyis dang po’i shes pa rnam par zlog pa ni slu ba ste/ dad pa’i sa la ‘god pa’o// (Sukusuma, D 102a.5-6; P 124a.3-4). 165 Vaidyapāda’s comments suggest that Buddhajñānapāda’s text included the word “et cetera” here, thus reading “human woman etc.” He writes “The word et cetera here is meant to include a devī, nāginī, yakṣinī, ḍākiṇī, and so forth.” sogs kyi sgras bsdus pa lha mo dang klu mo dang gnod sbyin mo dang mkha’ ‘gro ma la sogs pa rnams so// (Sukusuma, D 102b.1; P 123a.6). 166 lhag pa’i lha. Buddhajñānapāda uses this term several times in the Dvitīyakrama. In his usage, the term seems to refer to the final result of tantric practice. See especially verse 314 where the *adhideva is equated with the thirteenth bhūmi. (For further details this point see my notes to verse 314, as well as Chapter Three, in which I discuss this term and its function in Buddhajñānapāda’s work in more detail.) Vaidyapāda’s comments on the term here suggest is he understands this term to refer to Mahāvajradhara. He explains that Mahāvajradhara is be present in all beings, presumably as their basic nature, but can only be accomplished using the “higher methods which seal by means of wisdom.” ci’i phyir lhag pa’i lha zhes bya zhe na/ lha rnams las mchog tu gyur pa ni kha na ma tho ba med pa’o// de las mchog tu gyur pa ni byang chub sems dpa’o// de las mchog tu gyur pa ni sangs rgyas rnams so// de rnams kyi phul du gyur pa ni rdo rje ‘chang chen po sbyor ba bdun dang ldan pa’o// de lta bu sems can thams 379 So difficult to encounter in the three realms, will be accomplished. |82| [Initiation Rituals: The Second Initiation (guhyābhiṣeka)]167 Additionally, together with the ordinary [offerings],168 Perform the gaṇapūjā169 Then, having searched for a girl [who fits the description] that has been taught, She must be offered to the guru. |83| cad la mi slu (slu] D, bslu P) ba’i tshul du gnas kyang ye shes kyis rgyas btab pa’i thabs gong ma dang bral na mi ‘grub pas/ de rnams dang lhan cig tu gyur pa’i lhag pa’i lha zhes bya ba sgrub ces bya’o// (Sukusuma, D 102b.3-5; P 123b.2-5). 167 Buddhajñānapāda never uses the word “initiation” (abhiṣeka, dbang) in the Dvitīyakrama. However, he does describe the ritual procedures for the second (guhya) and third (prajñājñāna) initiations. 168 I suggest that the line “together with the ordinary” should perhaps be read as meaning “together with the ordinary offerings.” This makes sense when the two lines de yang thun mong dang bcas pa// tshogs kyi mchod pa yang byas te// are read together. The noun mchod pa, “offerings/pūjā,” is supplied in the second line and was simply omitted by Buddhajñānapāda and/or his translators in the earlier pāda due to metrical considerations. Vaidyapāda, however, uses the opportunity presented by this line, in which the noun to which “the ordinary” refers is not clearly specified, to add in a brief description of the kalaśābhiṣeka. I believe that the fact that Buddhajñānapāda himself does not make any mention to the kalaśābhiṣeka ritual in his text, as well as the fact that Vaidyapāda’s commentary on the kalaśābhiṣeka ends with the concluding initiatory rites like the vyākaraṇa, anujñā, and āśvāsa, indicates that in their tradition the kalaśābhiṣeka was likely given on a separate ritual occasion than the second and third initiations which are described here in the Dvitīyakrama in the subsequent verses. Presumably, the Dvitīyakrama does not mention the kalaśābhiṣeka because, given that the text is concerned primarily with perfection stage practices, the Dvitīyakrama assumes a student who has already received the kalaśābhiṣeka, and is now ready for the guhya and prajñājñāna initiations. In order to avail himself of the opportunity to present a summary of the kalaśābhiṣeka, in the present verse Vaidyapāda reads “ordinary” (thun mong) to refer to “that which is attained through [the vows of a] bodhisattva, and so forth, and through the ordinary vows, that is to say the vidyābhiṣeka” (thun mong zhes pa ni byang chub sems dpa’ la sogs pa dang thun mong du gyur pa’i sdom pas thob pa ste/ rig (rig] D, rigs P) pa’i dbang ngo//) (Sukusuma, D 102b.6; P 123b.5-6). The term vidyābhiṣeka was normally used to refer to the consecrations, from water to name, that correspond with the five buddha families in Yoga tantra (Mori n.d, 100). Vaidyapāda then interprets “together with that” ([de] dang bcas pa) to mean “the extraordinary,” and says that it means “that which is obtained by means of the extraordinary vows—the irreversible ācārya initiation.” (de dang bcas pa ni thun mong ma yin pa’i sdom pas thob pa ste/ rdo rje slob dpon phyir mi ldog pa’i dbang ngo//) (Sukusuma, D 102b.6; P 123b.6-7). He then proceeds to give a rather extensive explanation of the ritual of purifying the land in preparation for making an initiatory maṇḍala (sa dag par bya ba’i cho ga), followed by a rather detailed description of the various rituals of the kalaśābhiṣeka itself—though he himself calls it just a brief description and notes that a more detailed version of the ritual should be found elsewhere (Sukusuma, D 104b.5; P 126a.3-4). He finally notes that the subsequent guhyābhiṣeka is to be given to “students who [already] have those [earlier initiations]” de dag dang ldan pa’i slob ma la gsang dbang bskur bar gsungs pa/ (Sukusuma, D 104b.6; P 126a.4). Despite all of this, Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, which is generally known as the generation stage sādhana in his Guhyasamāja practice system includes several practices that appear to pertain in some way to the perfection stage (see Chapter Five for more details on this point). While generally I do believe, as I have just argued here, that the Dvitīyakrama and Vaidyapāda’s Sukusuma both suggest the likelihood of the kalaśābhiṣeka being given in a separate ritual context from the two later initiations, the fact that the Caturaṅga/Samantabhadra-sādhana contains practies that seem likely to have been permitted only for a practitioner who had obtained the later initiations does at least call this position into question. It may not be possible to resolve this point with complete certainty. I discuss this issue in Chapter Seven. 169 Vaidyapāda identifies this as the beginning of the guhyābhiṣeka. Moreover, in his commentary on the Samājottara Vaidyapāda uses the same term, gaṇapūjā, to refer to the pūjā to be performed at the outset of the guhyābhiṣeka. In that text Vaidyapāda describes the ritual as follows: “Having offered the wide-eyed one together with the gaṇapūjā to the guru, he bestows the initiation, as [will be] described, upon the disciple.” de yang ji skad du gsungs pa’i mig yangs tshogs kyi mchod pa dang bcas te bla ma la phul nas des ji skad du gsung pa’i dbang bskur ba slob ma la sbyin par bya’o// (Samyagvidyākara, D 103a.3). The ‘wide-eyed one’ is a term used to refer to the consort in Chapter Eight of the root Guhyasamāja-tantra, as well. That chapter is interpreted by some commentators as concerning initiatory practices. 380 Then, when the guru is pleased, He engages in union with her Due to which the sugatas melt and become the sixteenth part;170 This is dropped171 in the mouth of the disciple,172 |84| And having descended, it enters the lotus at his heart. Through this the field is purified173 And the twelve [experiences]—[perceiving] all phenomena as illusions, and so forth— Are realized in actuality. 174 |85| [The Third Initiation (prajñājñānābhiṣeka)]175 176And then, in order to bring about the realization 170 See note 9 regarding “the sixteenth part.” Vaidyapāda explains the melting of the sugatas following to the guru’s union with the offered girl as follows: “Having become fully impassioned, [the guru] enters into union [with her]. The sugatas, who have been invoked by the seed [syllable], enter into the mouth and one should think that having melted as the moon they become the sixteenth part, that is, the essence of the bindu.” rjes su chags pa’i mtha’ la thug pas snyoms par zhugs te/ sa bon gyis bskul ba’i bde bar gshegs pa rnams zhal du zhugs te zla bar zhu bar (zhu bar] D, P om.) gyur nas bcu drug cha zhes te thig le’i ngo bor gyur bar bsam mo// (Sukusuma, D 105a.5-6; P 126b.4-5). 171 gtung 172 Vaidyapāda comments: “That [drop] itself is dropped into the disciples’s mouth means that from the bindu comes a syllable and from that arises a maṇḍala and that [maṇḍala] itself, which has melted due to the heat of great passion, is [then] given into the mouth of the disciple.” de nyid slob ma’i kha ru ltung/ zhes te thig le las yi ge/ de las dkyil ‘khor de nyid ‘dod chags chen po’i mes bzhus pa slob ma’i khar sbyin pa’o// (Sukusuma, D 105a.6; P 126b.5-6). 173 Vaidyapāda explains that ‘the field’ means the disciple’s aggregates and so forth. de yis zhing dag byas te/ zhes te zhing ni de’i phung po la sogs pa’o// (Sukusuma, D 105a.7; P 126b.7-8). 174 Vaidyapāda clarifies that these are the twelve examples that show phenomena as being illusory: an illusion, a mirage, an echo, a spinning firebrand, a delusion, a dream, a city of gandarvas, a bubble on water, a flash of lightening, an emanation, a rainbow, and a cloud (Sukusuma, D 105b.1-2; P 126b.8-127a.2). This, moreover, concludes the second initiation, according to Vaidyapāda. He notes that having received this initiation such that his “field” is purified, the disciple “attains equal fortune to bodhisattvas such as Maitreya, and will thus travel from buddhafield to buddhafield. Thus, having entered this path he will swiftly attain accomplishment.” de byams pa la sogs pa’i byang chub sems dpa’ rnams dang skal pa mnyam pas na sangs rgyas kyi zhing nas sang rgyas kyi zhing du ‘gro ste/ de’i lam la ‘jug pa myur bar thob par ‘gyur ro// (Sukusuma, D 105b.2-3; P 127a.2-3). 175 See note 167 on the use of the term abhiṣeka in the Dvitīyakrama. 176 According to Vaidyapāda, this verse begins the explanation of the prajñājñānābhiṣeka, which he explains has the purpose of bringing the disciple who has had his “field” purified by the second initiation to a realization of the dharmakāya (da ni zhing dag par byas pa’i slob ma la chos kyi sku rtogs par bya ba’i phir shes rab ye shes kyi dbang bskur ba gsungs pa/) (Sukusuma, D 105b.3; P 127a.3). This section of the Dvitīyakrama contains quite a number of verses that are parallel with at least fourteen later tantric texts, both scriptural and authored, several of which survive in Sanskrit. I have referenced the parallels of which I am aware (I imagine there are more) in the notes to the translation, and done my best to take the readings from the parallel verses into consideration in my edition and translation of the verses from the Dvitīyakrama, but a fuller study of all of the parallels would likely further improve some of the readings here. Many of the verses that follow here correspond with a sequence of verses studied by Wedemeyer in his analysis of the relationship between the prajñājñānābhiṣeka and the vidyāvrata ritual, presented at a conference at UC Berkeley in 2014. (On that relationship, see note 186.) In a number of places, the verses from this section of the Dvitīyakrama are identical with those cited by Wedemeyer from Vaidyapāda’s Maṇḍalavidhiṭīkā, Vagīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, Advayavajra’s Samkṣiptābhiṣekaprakriyī (Tōh. 2244), Kṛṣṇācārya’s Guhyasamājamaṇḍalopāyikā (Tōh. 1819), Kuladatta’s Kriyāsamgraha, Prajñāgupta’s Abhiṣekaratnāloka (Tōh. 1333), and Prajñāśrī’s Abhiṣekavidhi (Tōh 1269), all of which are, however, later than the Dvitīyakrama. Buddhajñānapāda’s text here appears, then, to be the source of these verses. (It is worth noting here that, although Wedemeyer did not mention this in his presentation, the verses as preserved in the Tibetan translation 381 Of the self-arisen dharmakāya, great joy That is equal to space, called the *adhideva, The girl is given to him [i.e. the disciple].177 |86| “This goddess is suitable for you.178 Great being,179 all of the buddhas have given180 This delightful girl to you to enjoy181 By means of your desire |87| Through the ritual for the maṇḍala-cakra.182 of Vaidyapāda’s maṇḍalavidhi commentary are given in extremely garbled transliteration of the Sanskrit (Guhyasamājamaṇḍalopāyikāṭīkā, D 211a.4-6), not in Tibetan translation. They are so garbled that without having the Sanskrit from a later text to compare them to, I doubt that it would have been possible to reconstruct any meaning whatsoever from them. But when one compares with the later versions of the verses in Sanskrit, it is clear that they are exactly those same verses. In this regard we are quite lucky that Wedemeyer was engaging in comparative work and was thus able to see the parallels of these garbled lines of Sanskrit transliteration with the call-and-response verses in other later texts!) In any case, the earliest version of these verses now known to us appears to be the one here in the Dvitīyakrama. It is unclear whether Mañjuśrī/Buddhajñānapāda himself composed the verses or incorporated them from an earlier source that is unknown to us. In either case, the use of these verses from his writings by many later authors is yet another indication of Buddhajñānapāda’s influence on the later tantric tradition. In this section I have italicized the lines or verses that are parallel with the later sources in order to make it easier to see what has been incorporated from the Dvitīyakrama into the later tradition. 177 de la. Vaidyapāda clarifies that in response to the disciple’s supplications the guru gives the girl “that he has blessed” to “the disciple’s right hand” and recites the subsequent verses. De nas (nas] D, P om.) yang me tog gis snyim pa bkang ste/ gsol ba ‘debs pa la bla mas byin gyis brlabs pa’i bu mo slob ma’i lag pa g.yas par sbyin par byas nas lung ‘di bsgo’o// (Sukusuma, 105b.6-7; P 127a.7-8). Several scholars have noted this practice of handing over of the consort into the right hand of the disciple (in later liturgies this often serves as part of the ritual for bestowing the vidyāvrata) resembles an Indian marriage ceremony, in which the joining of the couples’ hands is an important part of the ritual. Schwind (2012, 291n 1062) cites Isaacson on this point, (noting that pāṇigrahana, “taking by the hand” is a common term for marriage); on this point see also Wedemeyer (unpublished 2014 and forthcoming), who likewise remarks on the pāṇigrahana element, and Onians (2003, 176), who refers to this ritual of the vidyāvrata as a sort of “sacred marriage.” 178 Vaidyapāda explains that “suitable” here means that, “Since she has what you desire, she is suitable for you.” khyod kyi bzhed pa ‘di la yod pa’i phyir na khyod dang mthun zhes so// (Sukusuma, D 105b.7; P 107a.8-107b.1). 179 sems chen] P N S V(P), sems can D C V(D); Vaidyapāda’s commentary also suggests that sems chen is the better reading: “Great being means someone who has the intention to liberate sentient beings.” sems chen (chen] P, can D) zhes pa ni sems can bsgral ba’i sems gang la yod pa’o// (Sukusuma, D 105b.7-106a.1; P 107b.1). 180 gnang] D C V (D and P), snang P N S; The Peking edition of Vaidyapāda’s commentary cites the line from the verse with snang, but then in the explanation of the verse uses the correct spelling, gnang. 181 This line could also be understood as “to practice with.” Vaidyapāda explains: “Given by the buddhas to enjoy means that the unsurpassed buddhas give [a woman] to some suitable disciples to enjoy.” sangs rgyas kun gyis spyod du gnang (gnang] D, snang P) / zhes pa’i bla na med pa’i sangs rgyas rnams kyis snod du rung ba’i gang zag ‘ga’ la spyod du gnang ba’o// (Sukusuma, D 106a.1; P 127b.1-2). 182 Vaidyapāda specifies that this means the ādiyoga-samādhi—the first of the three samādhis that are often connected to generation stage practice—“and so forth.” The Sukusuma reads, “By means of the ritual of the maṇḍala-cakra means the ritual of the ādiyoga[-samādhi] and so forth. Thus, by means of reversing the ordinary, one attains liberation in one life.” dkyil ‘khor ‘khor lo’i cho ga yis/ zhes te/ dang pa’i rnal ‘byor (D adds pa) la sogs pa’i cho ga ste/ tha mal pa bzlog pas tshe gcig gis grol ba’o// (Sukusuma, D 106a.1-2; P 107a.2-3). While a first glance at this statement might suggest that Vaidyapāda takes this ritual to refer specifically to generation stage practices, as I discuss in Chapter Five, it is in the third of these three samādhis from the Yoga tantras that we find perfection stage practices being incorporated into Buddhajñānapāda’s system. Vaidyapāda’s reference to the structure of practice using the three samādhis here is also, I believe, an indication of the sexual yogas incorporated into the sequence of generation stage practices, rather than being separated out from that structure, as became the case in many (but not all) later tradition. In any case the distinction between generation and perfection stage practices was newly being made at this point, so some overlap is to be expected. 382 In order to accomplish great awakening You must experience great bliss [5a] [With] the girl who liberates and gives joy.183 |88| Nothing else can bring about buddhahood This girl is the genuine supreme184 Thus, throughout endless saṃsāra You must never separate from her.”185 |89| Then that great being Should take up the delightful girl.186 183 87b-88d ~C.f. Vajrāvalī (Mori 2009 Vol. 2, 444) and Daśatattva V.14 (Klein-Schwind 2012, 209). 87b-88c ~C.f. Guhyasamājamaṇḍalavidhi, vv. 365c-366b, and Saṃkṣiptāvhiṣekavidhi (Sakurai, 417). 184 These two pādas are nearly identical with Samājottara 125 c and d. The verse in the Samājottra, however, uses the term vidyā rather than “girl” (*kanyā?). This suggests that Buddhajñānapāda’s verses may be earlier. In an earlier conference paper (C. Dalton 2014) I have argued in some detail that the verse on the two stages of tantric practice in the Samājottara is likely modeled on Buddhajñānapāda’s verse in the Muktitilaka, rather than vice versa. In that instance, it appears that the term “buddhas” from Buddhajñānapāda’s earlier verse in his Muktitilaka was transformed into “vajra holders” in the Samājottara. Just like in this verse with the use of the term vidyā in the Samājottara rather than “girl” (*kanyā?) in the Dvitīyakrama, a move towards increased tantrification is much more likely than the reverse. In this case, moreover, the second two pādas of this verse in the Dvitīyakrama are also found in the Samājottara, though with an intervening two pādas about the nondual nature of reality. Again, the fact that there are two intervening pādas in the Samāmjottara’s version suggests that if one text is based upon the other (i.e. if they are not both drawing from some separate earlier source) the Samājottara’s is likely later than Buddhajñānapāda’s verse, as it would be unlikely that Buddhajñānapāda would cite from a scriptural source—even unattributed—and not provide the complete citation. I discuss the relationship of Buddhajñānapāda’s writings and the Samājottara in Chapter Eight. 185 Vaidyapāda reports that, “You must never separate from her means that since she is the seal of the perfection of wisdom you must examine the actual and the example wisdom together with her in order that the continuity of wisdom is not severed.” khyod kyis ‘di dang ‘bral mi bya/ zhes pa ni shes rab kyi pha rol tu phyin pa’i phyag rgya bas na de dang lhan cig tu mngon sum dang dpe’i ye shes brtag par bya ba ste/ ye shes rgyun mi ‘chad pa’i phyir ro// (Sukusuma, D 106a. 3; P 107b.4-5). 186 In Wedemeyer’s 2014 conference paper on the vidyāvrata, he argues that in the Samājottara the vidyāvrata is conceived as a separate procedure that follows the prajñājñānābhiṣeka (Wedemeyer unpublished; an updated version of this paper is forthcoming). Indeed, they are two separate questions among those set out in the beginning of that tantra: “How is initiation bestowed? And likewise, what of the vidyāvrata?” (abhiṣekaṃ kathaṃ deyaṃ kathaṃ vidyāvrataṃ tathā). Wedemeyer proposes that this is one of two ritual paradigms regarding the vidyāvrata: one in which it is separate from the initiations, and one in which it is joined with, or even identical to the prajñājñānābhiṣeka. Vaidyapāda, in his Guhyasamājamaṇḍalavidhiṭīkā, Wedemeyer reports, follows the second paradigm. Likewise, here in the Dvitīyakrama, the giving of the consort appears to be identical with the ritual sequence of the prajñājñānābhiṣeka. Vaidyapāda likewise explains, “Should take this delightful girl, means that by means of that the prajñājñānābhiṣeka is recieved.” (yid ‘ong bu mo blang bya ste/ zhes pa ni des shes rab ye shes kyi dbang blang zhes (zhes] D, shes P) pa’o//) (Sukusuma, D 106a.3-4; P 107b.5). However, in his commentary on the Samājottara Vaidyapāda identifies the ritual described in verses 124-125 of that text—which corresponds with the giving over of the girl to the disciple here in verses 89-90b of the Dvitīyakrama—as the “vidyāvrata initiation” (rig pa’i brtul zhugs kyi dbang) (Samyagvidyākara, 192a.6). This is not unexpected given that the Samājottara itself uses the term vidyāvrata here, but it is interesting to note that Vaidyapāda seems to refer to more or less the same ritual as the prajñājñānābhiṣeka in his commentary on the Dvitīyakrama and as the “vidyāvrata initiation” in his commentary on the Samājottara. For Dvitīyakrama verse 89 cf. also the Vajrāvalī (Mori 2009 Vol. 2, 449). Here we can see that the Vajrāvalī verse is modeled on Buddhajñānapāda’s rather than the Samājottra’s (i.e. it lacks the two pādas on nonduality), but in a slightly “updated” version, as are several of Buddhajñānapāda’s verses that appear in this section of the Vajrāvalī. Unlike Abhayākaragupta, Vaidyapāda cites the version from the Samājottara (thus including the two pādas on nonduality) rather than from the Dvitīyakrama in his commentary on Dīpaṃkarabhadra’s Maṇḍalavidhi. Vaidyapāda does this, as well—i.e. cites the Samājottara’s verse over Buddhajñānapāda’s in an instance of parallel verses between the two—when he cites the verse on the two 383 She187 speaks to him of the sublime samayas and vows, With these words: |90| “Speak up, darling, [O] vajra holder Can you eat flesh, blood, semen, Feces and urine, and the rest? Can you kiss [my] bhaga without a second thought?188 |91| Can you delight in me, [your] consort?”189 He replies laughingly,190 “O Goddess, how could I not be delighted? I will eat feces, urine, and the rest! |92| Goddess, you require respect: I have no second thoughts about kissing [your] bhaga!” 191 Then the girl throws off her lower garment, She shows her lotus clearly and speaks these words of praise: |93| stages (found in both Buddhajñānapāda’s Muktitilaka and in the Samājottara) here in the Sukusuma, despite the fact that Vaidyapāda also composed a commentary on the Muktitilaka and thus certainly knew Buddhajñānapāda’s version of that verse. The fact that Buddhajñānapāda shows no knowledge of the Samājottara and yet several components from his writings appear to be included in it (rather than vice versa), in addition to the fact that his main commentator Vaidyapāda, who was likely a direct disciple, cites the Samājottara regularly in preference over Buddhajñānapāda’s versions of the parallel passages, seems to narrow down quite considerably the period of the earliest circulation of the Samājottara, to very soon after (or possibly even during the latter part of?) Buddhajñānapāda’s lifetime. Vaidyapāda also wrote a full commentary, the Samyagvidyākara, on just the Samājottara (i.e. not including the root Guhyasamāja-tantra). I discuss the relationship of Buddhajñānapāda’s and Vaidyapāda’s writings with the Samājottara in Chapter Eight. 187 de yis. Vaidyapāda clarifies that it is the girl who speaks (de yis zhes te bu mos so//) (Sukusuma, D 106a.4; P 107b.5-6). 188 mi rtog. Literally “without thought.” I have translated the term more colloquially as “without a second thought,” which I think very much represents the question being asked of the yogin here. Nontheless, the idea of nonconceptuality, and of the transgressive acts described in the verse as evocative of a nonconceptual state, is also very much at play here. 189 Vaidyapāda reports that the consort’s words here are “easy to understand” (de rnams go sla ‘o//) (Sukusuma, D 106a.5). 190 Vaidyapāda clarifies that the yogin is laughing as a sign of his joy to practice the vow, which presumably is a reference to the vidyāvrata committment (Sukusuma, D 106a.5; P 107b.7). 191 Quite a number of tantric texts, including at least one tantra, have variants of Verses 91a-93b. C.f. Caṇḍamahāroṣaṇa-tantra 3.26-27; Vaidyapāda’s Guhyasamājamaṇḍalopāyikāṭīkā, where similar verses are given in garbled Sanskrit transliteration (D 211a.4-5); Ratnākaraśanti’s Piṇḍīkṛasādhanopāyikāvṛtti-ratnāvalī-nāma (D 91b.6-7); Nāgabodhi’s Pañcakramaṭīkā-maṇimālā-nāma (D 130b.5-7); Advayavajra’s Samkṣiptābhiṣekaprakriyā (D 131b.5-7); Kṛṣṇācārya’s Śrīguhyasamājamaṇḍalopāyikā (258a.4-6); Prajñāgupta’s Abhiṣekaratnāloka (D 299a.7- b.2); Prajñāśrī’s Abhiṣekavidhi (D 48b.4-5), and Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi (Sakurai 1996, 418). Kṣitigarbha’s Daśatattvasaṃgraha (V.17-20) incorporates vv. 91a-94d (see Klein-Schwind 2012, 210). Wedemeyer, in a handout from his 2014 lecture at Berkeley gives these call and response verses in proper ungarbled Sanskrit, but does not make it clear which source he draws the Sanskrit from. (I suspect they are from Vāgiśvarakīrti’s and/or Kuladatta’s works. Kuladatta’s is not available to me at the moment.) I have Wedemeyer’s work to thank for pointing out that these verses are found in Vaidyapāda’s Guhyasamājamaṇḍalopāyikā-ṭīkā, Vagīśvarakīrti’s Saṃkṣiptābhiṣekavidhi, Advayavajra’s Samkṣiptābhiṣekaprakriyā (Tōh 2244), Kṛṣṇācārya’s Guhyasamājamaṇḍalopāyikā (Tōh 1819) (minus the kuru padme verse), Kuladatta’s Kriyāsamgraha, Prajñāgupta’s Abhiṣekaratnāloka (Tōh 1333), and Prajñāśrī’s Abhiṣekavidhi (Tōh 1269). Prajñāgupta’s and Prajñāśrī’s texts have the variant reading of “suck” (*cūṣaṇa?) rather than “kiss” (cumbana), which may have influenced the reading in the Caṇḍamahāroṣaṇa-tantra. 384 “Ah! This lotus of mine Is endowed with all bliss! [I]192 will always remain before He who enjoys it according to the ritual.193 |94| This sublime lotus that brings about [one’s] aim,194 Is the place venerated by all the buddhas.195 Self-arisen great bliss Always abides here.”196 |95| 192 Who or what remains before the yogin is not specified in Buddhajñānapāda’s text. Vaidyapāda identifies what remains before that yogin as wisdom (ye shes): “[I] always remain before him means that that which always remains directly in the presence of the one who knows the stages of the ritual is wisdom.” de yi mdun na rtag tu gnas/ zhes pa ni cho ga’i rim pa shes pa (pa] D, P om.) de’i mngon sum tu rtag tu gnas pa ste ye shes so// (Sukusuma, D 106b.2; P 108a.4-5). However, in the later verses from the Vajrāvalī, the word “I” (aham) is stated clearly—“I remain before him...” The verse, in the Vajrāvalī, reads: aho madīyaṃ padmaṃ sarvasukhasamanvitaṃ/ yaḥ sevati vidhānena tasyāham agrataḥ sthitā (Mori 2009 Vol. 2, 445. I have emended Mori’s sevayati to sevati, following the advice of Harunaga Isaacson; sevati is indeed reported as the reading in one of Mori’s manuscripts). There is no difficulty in reading Vaidyapāda’s commentary in this way, as well, as the term “wisdom” can easily be understood as a gloss of or reference to the consort (though Vaidyapāda does use the term ye shes (jñāna), rather than shes rab (prajñā), which is the usual term referring to the consort). The referent for who or what is remains before such a yogin in the Tibetan translation of the Dvitīyakrama may simply have been left out due to metrical considerations. 193 vv. 91a-94d ~c.f. Daśatattvasaṃgraha, V.17-20 (Klein-Schwind 2012, 210). Vaidyapāda notes that “according to the ritual means the actions of body, speech, and mind, together with the pith instructions on [manipulating] the winds that will be explained below.” de la cho ga’i rim pa ni ‘og nas ‘byung ba’i lus ngag yid gsum gyi bya ba rlung gi man ngag dang bcas pas so// (Sukusuma, D 106b.2; P 108a.4). 194 Don byed padma dam pa ‘di. This line in the Dvitīyakrama is a bit awkward, particularly the description of the lotus as don byed—performing a function, or “bringing about [one’s] aim,” as I have translated it based on Vaidyapāda’s commentary. The pāda has been modified in the Vajrāvalī where it reads, “Perform in the lotus that which is to be done” (kuru padme yathā kāryaṃ) (Mori 2009 Vol. 2, 445). It would be possible to read the line in the Dvitīyakrama in more or less same way as the Vajrāvalī’s reading by making just two small modifications to the Dvitīyakrama (rendering the verb in the imperative byos insteady of byed and adding a locative (emending ‘di to ‘dir)). However, it is unlikely that this is what the Dvitīyakrama intends, given that all of the readings in the Dvitīyakrama and the Sukusuma, as well as the verse as it is cited in Vaidyapāda’s maṇḍalavidhi commentary (see Guhyasamājamaṇḍalopāyikāṭīka, D 211a.6) are consistent. Moreover, in the Sukusuma Vaidyapāda explains: “Thus because it is the cause of the natural, excellent aim, [the text says] “For bringing about [one’s] aim the lotus is sublime.” de bas na rang bzhin gi don phun sum tshogs pa’i rgyur gyur pas na (na] D, P om.) don byed padma dam pa’o// (Sukusuma, D 106b.2-3; P 128a.5). It therefore seems that the verse as found in the Vajrāvalī may have undergone some minor changes from what was found in Buddhajñānapāda’s work. 195 Vaidyapāda explains that, “In order [that one might] see that performance as proper (brtsun pa) [the text says] the place that is venerated by all the buddhas.” byed pa la btsun par blta bas na/ sangs rgys kun gyis bkur ba’i gnas so// (Sukusuma, D 106b.3; P 128a.5-6). The version of this line in the Vajrāvalī differs here, describing the place as one “where the buddhas are venerated,” but the third case indicating that the buddhas are the ones doing the venerating is clear here in all recensions of the Dvitīyakrama, as well as in both editions of Vaidyapāda’s commentary (P and D) that I have consulted, and also in the verse as rendered in Vaidyapāda’s maṇḍalavidhi commentary (Guhyasamājamaṇḍalopāyikāṭīka, D 211a.6). Again, it seems that the verse as included in the Vajrāvalī may have undergone some minor changes from the earlier version from Buddhajñānapāda’s work. 196 Verses 94-95 are likewise found in all of the sources mentioned above in note 191. That is: Caṇḍamahāroṣaṇatantra 3.28-89; Vaidyapāda’s Guhyasamājamaṇḍalopāyikāṭīkā, where they, as opposed to the previous call and response verses which were rendered in garbled Sanskrit transliteration, are translated into Tibetan (D 211a.5-6); Ratnākaraśanti’s Piṇḍīkṛtasādhanopāyikāvṛtti-ratnāvalī-nāma (D 92a.1-2); Nāgabodhi’s Pañcakramaṭīkāmaṇimālā- nāma (D 131a.7-132b.1); Advayavajra’s Samkṣiptābhiṣekaprakriyā (D 131b.7-132a.1); Kṛṣṇācārya’s Guhyasamājamaṇḍalopāyikā (D 258a.6-7) (here the verses are abbreviated); Prajñāgupta’s Abhiṣekaratnāloka (299b.2); and Prajñāśrī’s Abhiṣekavidhi (D 48b.5) (the verses are also abbreviated here). They are also found in 385 Bhaja197 mokṣa hoḥ [Mentally Cultivating Passion]198 Then from one’s own bīja199 Dense rays of light illuminate [one’s own] interior200 Causing the maṇḍala deities201 to become impassioned And thus [they]202 emerge from the vajra |96| And enter into the lotus. Thus the consort’s maṇḍala deities Become intensely impassioned And emerge from [her] mouth into [one’s own] mouth203 |97| The maṇḍala deities [again] become impassioned and emerge from the vajra They enter [the lotus] again, as before, and so forth. Abhayākaragupta’s Vajrāvalī (Mori 2009 Vol. 2, 445); The last two pādas of verse 94 in Buddhajñānapāda’s and Vaidyapāda’s readings are a bit confusing, and it appears that a later author like Abhayākaragupta, or perhaps some other intervening author, cleaned the verse up slightly to make it more understandable, thus also slightly changing the meaning. See notes 194 and 195 for two such changes. 197 bhaja] sugg. em. based on the parallel verses in the Vajrāvalī and Daśatattvasaṃgraha, bhanydza] D C V(D), bhaṃdza P N S, bhaga V(P). Vaidyapāda clarifies, “Then the disciple should say ‘Enjoy! Liberate! hoḥ!’” (bhaga mokṣa hoḥ), due to his delight [in] the celestial palace of great nirvāṇa.” de nas slob mas kyang bhaga (bha ga] P., bhanydza D) mokṣa ho zhes brjod par bya ste/ mya ngan las ‘das pa chen po’i gzhal yas khang dgyes pa’i phyir ro// (Sukusuma, D 106b.3-4; P 128a.6-7). The Vajrāvalī and Daśatattvasaṃgraha here read bhaja mokṣa hoḥ here (Mori 2009 Vol. 2, 445; Klein-Schwind 2012, 210), and while this mantra is generally problematic in all of the readings with which I am familiar, I find bhaja, found there, more plausible than bhañja, which seems to be the reading from all recensions of the Dvitīyakrama itself. In either case reading mokṣa as an imperative as I have done here is not, of course, in accordance with standard Sanskrit grammar. Thanks to Harunaga Isaacson for his advice on choosing a reading here. 198 This section represents, according to Vaidyapāda, the guru teaching the physical, verbal, and mental practices that are meant to bring about the experience of wisdom in the disciple (Sukusuma, D 106b. 4; P 128a.7). He explains that the first set of practices described here are meant to bring a state of mental arousal to oneself, the girl, and the maṇḍala deities. bdag dang bu mo lha’i dkyil ‘khor yid kyis shin tu chags pa’o zhes so// (Sukusuma, D 106b.7-107a.1; P 128b.4). It seems that mental arousal is specified here since the subsequent practices are said to bring about a state of arousal through verbal, and finally, physical means. Given that it is not until later that the partners join in the yogic coitus that constitutes the main part of the third initiation, and that what is described here is followed by the partners speaking to each other in arousing ways, it seems that the practices of “mentally cultivating passion” are visualized, rather than entailing actual sexual union. 199 Tib. sa bon; Skt. bīja. Vaidyapāda identifies this as the “seed [syllable] of the wisdom being.” rang gi sa pon zhes pa ni ye shes sems dpa’i sa pon las so// (Sukusuma, D 106b. 4-5; P 128a.8). 200 Vaidyapāda specifies that it is one’s own interior that is so illuminated: de las ‘od zer byung (‘byung] P, byung D) bas rang gi nang gsal te/ (Sukusuma, D 106b.5; P 128a.8). 201 Vaidyapāda specifies that these are the maṇḍala deities of one’s own body maṇḍala. de las ‘od zer ‘byung (‘byung] P, byung D) bas rang gi nang gsal te/ des lha’i ‘khor lo zhes pa ni des lus kyi dkyil ‘khor du gtogs pa’i lha rnams so// (Sukusuma, D 106b.5; P 128a.8-b.1). 202 It is slightly unclear in both the root text and the commentary whether it is the deities themselves that emerge from the vajra path or simply the light rays. Vaidyapāda reads ‘od zer des rdo rje’i lam nas phyir byung ste/ (Sukusuma, D 106b.5; P 128b.1). 203 Vaidyapāda makes this clear: “From the mouth means [from] the girl’s [mouth]. Enter into the mouth means my own [mouth].” zhal nas zhes pa ni bu mo’i ‘o// zhal du zhugs zhes pa bdag gi ‘o// (Sukusuma, D 106b.6; P 128b.2). 386 Through repeating this again and again,204 the maṇḍala deities Become intensely mentally impassioned. |98| [Verbally Cultivating Passion]205 Then, one recites these illusory words of desire And thus the girl becomes filled with passion: “Nondual supreme great bliss, Goddess, you are the illusory mudrā! |99| Sweet-faced one,206 come play207 with me And [we] will have an experience that is like the sky!”208 Then [she] also supplicates With illusory words of desire, |100| And one becomes impassioned oneself And searches for the cakra.209 |101| [She says,] “Vast Supreme Bliss,210 pay heed to me! [9a] Inconceivable great bliss that is vocalized211 is unshakeable!212 204 Vaidyapāda clarifies, “Again and again means four times with a steady mind.” yang nas yang du zhes pa ni brtan pa’i sems lan bzhi’i bar du… (Sukusuma, D 106b.7; P 128b.3-4). 205 Vaidyapāda describes this as the “ritual for verbally cultivating passion” ngag gis chags par byas pa’i cho ga (Sukusuma, D 107a.1; P 128b.4). 206 Vaidyapāda notes that “[He says] sweet-faced one because she is charming and so forth, and not scowling.” zhal bzang khyod ni zhes pa ni steg pa la sogs pa dang ldan zhing ‘dzum gnag pa ma yin pas (Sukusuma, D107a.2; P 128b.5-6). 207 rtsen] P N S V(P), brtson] D C V(D). Vaidyapāda’s commentary, which glosses the term as rnam par rol pa also supports the reading of rtsen (Sukusuma, D 107a.2; P 128b.6). 208 Vaidyapāda explains that “that which is like the sky is innate wisdom (*sahaja-jñāna?), which is also called ‘luminosity.’” mkha’ mnyam ni ye shes lhan cig skye pa ste ‘od gsal ba zhes rnam grangs so// (Sukusuma, D 107a.2; P 128b.6). 209 Vaidyapāda mentions that this is to be done “subsequent to the physical activity” (lus kyi spyod pa’i rjes la), presumably indicating that the yogin should wait until after completing the other preliminaries to sexual union to perform this search (Sukusuma, D 107a.3; P 128b.6-7). Indeed such a reference to using his fingers to search for the cakra is found again in the Dvitīyakrama in verses 115-18, which occur after the description of a number of preliminary sexual acts (described in verses 105-113) that Vaidyapāda indicates as the methods by which the yogic couple are to “phyisically cultivate passion” (lus kyi spyod pas chags par bya ba’i thabs) (Sukusuma, D 107b.1; P 129a.5-6). It is clear that many of the sexual acts described in vv.105-113 do not involve actual coitus, and if we follow Vaidyapāda’s explanation of the five positions described in verse 105 as positions from which the couple are to gaze upon one another, then none of these acts involve coitus. It is not until after the description of these acts that the yogin is instructed to “search for the cakra,” (vv. 115a-118b), and only after that to “embrace” his partner in sexual union (v. 118c). 210 bde mchog rgya chen, *vistaraśaṃvara?. Vaidyapāda clarifies that this is, in this instance, a reference to the practitioner who is endowed with relative bliss. bde mchog rgya chen po ni kun rdzob kyi bde ba ste de dang ldan pa’i sgrub pa po bdag la dgongs su gsol zhes pa’o// (Sukusuma, D 107a.3; P 128a.7-8). 211 ngag (ngag] sugg. em. based on V (D and C); dag D C P N S) dang ldan. Not only do both the Derge and Peking editions of Vaidyapāda’s commentary here read ngag, his comments also make it clear that this is his reading: bsam mi khyab pa’i bde chen te/ zhes te thams cad la khyab bdag tu gnas pa’i bde ba chen po ni mi slu bas mtshon par byed pas na ngag dang ldan zhes bya ‘o// (Sukusuma, D 107a.3-4; P 128b.8). 212 I am not entirely clear what the referent of this line is in its relation to the rest of the verse. Vaidyapāda comments on the line but does not clearly indicate how it relates gramatically to the rest of the verse. What is clear is that, in his reading at least, the two terms ‘inconceivable great bliss that is vocalized’ and ‘immovable’ refer to the same thing—the great bliss that abides in and pervades all things. bsam mi khyab pa’i bde chen te/ zhes te thams 387 With a beautiful melody beyond words I ask you to sport with me—rouse yourself! |102| This supreme path leads to the essence of awakening And is shown213 by means of marvelous play: [Even] without relying upon any of the authentic tantras214 I, delightful great bliss, will liberate!215 |103| hoḥ hoḥ hoḥ!216 Sport, sport, sport with me! Rouse your desire, O you who plead217 for bliss! Have no doubts that you will realize That which is not realized by other [means]! |104| A la la la ho!” [Physically Cultivating Passion]218 Then, with great passion Engage in physical practice with her; Practicing this play in an isolated place You should examine bliss.219 |105| cad la khyab bdag tu gnas pa’i bde ba chen po ni mi slu (slu] D, bslu P) bas mtson par byed pas ngag dang ldan zhes bya ‘o// de nyid gzhan gyis mi bskyod pas na mi gyo ba’o// (Sukusuma, D 107a.3-4; P 128b.8-129a.1). 213 snang bar mdzad. Khenchen Chodrak Tenphel explains that this means it is shown to the initiated (Khenchen Chodrak Tenphel, personal commmunication, February 2016). 214 Vaidyapāda explains: “Without relying upon any of the authentic tantras means without relying upon any of the Yoga or Mahāyoga tantras; those are teachings that are like a raft (i.e. to be left behind upon reaching the destination). Since this is liberation that comes about through the power of knowing, even if someone has various karmic obscurations, he nonetheless enters [into it] instantaneously.” rgyud (rgyud] P, rgyu de D) rnam yang dang gang la mi rten (rten] P; brten D) par/ zhes pa ni rnal ‘byor dang (rnal ‘byor dang] D, P om.) rnal ‘byor chen po’i rgyud gang la mi rten (rten] P; brten D) te/ de rnams gzings dang ‘dra ba’i chos so// ‘di ni ye shes kyi stobs kyis grol ba’i phyir na las kyi sgrib pa sna tshogs dang ldan yang dus gcig par chud pa’o// (Sukusuma, D 107a.6-7; P 129a.3-4). 215 de chen yid ‘ong nga yis grol bar gyis// The grammar of this line is somewhat unclear. Vaidyapāda comments, “Therefore, may you be liberated together with delightful great bliss!” de bas na bde ba chen po yid du ‘ong ba dang lhan cig tu grol bar gyur cig ces pa’i don to// (Sukusuma, D 107a.7; P 129a.4-5). 216 Vaidyapāda explains that the three syllables here are expressions of joy at the knowing of the three wisdoms that arise from that practice. ho ho ho zhes pa ni de las skyes pa’i ye shes gsum rig par ‘gyur ba la dgyes pa’o// (Sukusuma, D 107a.7-b.1; P 129a.5-6). 217 gsol] D C, gsal P N S 218 Vaidyapāda notes that from here on the text explains the methods for cultivating passion through physical actions. de nas lus kyi spyod pas chags par bya ba’i thabs gsungs pa. (Sukusuma, D 107b.1; P 129a.5-6)). 219 Vaidyapāda explains that this bliss is “*sahajānanda, which is composed of three [aspects]” (lhen cig skyes pa’i dga’ ba ste gsum gyis bsdus pa’o) (Sukusuma, D 107b.2; P 129a.7). I discuss the system of blisses in Buddhajñānapāda’s system in Chapter Six. 388 220 First, coming together221 Then the [posture] characterized by the elbows222 Additionally, the one [characterized by] extending [the legs]223 And likewise, the [posture] characterized by lifting up And then the complete extending [of the legs]—these are the five. |106| Completely raising the shoulders Elbows bent, embrace [her] around the neck Holding tight, embracing firmly, [with] the right and the left hands He holds [her] hair [and] head unmoving and looks [at her]. |107| Then looking between224 her two thighs Singing śīt like a225 bee 220 Vaidyapāda comments, “How should one look? Teaching the five principal [ways] as taught in “skyod (skyod] P, skyed D) byed (byed] D, phyed? P) thams cad kyi gtsug lag (*sarvasaṃcalaśāstra?), the text says First…” This is a tantalizing reference from Vaidyapāda, as he appears to actually tells us Buddhajñānapāda’s source for some of these sexual practices, but unfortunately, I am unable to understand clearly what he means! ji ltar blta bar bya zhe na/ skyod (skyod] P, skyed D) byed (byed] D, phyed? P) thams cad kyi gtsug lag las bshad pa’i gtso bo lnga gsungs pas/ dang por zhes pa la sogs pa’o// (Sukusuma, D 107b.2-3; P 129a.7-9). 221 Vaidyapāda describes the characteristics of this first posture as follows: “The woman firmly embraces [him] around neck/ And the man’s forearms/ Are placed against her elbows (even)/ It is is also said that her calves should be brought together (i.e. with her legs around his body).’ This is [how] to perform the position.” de’i mtshan nyid kyang ji skad du/ bud med mgul par (par] D, pa P) dam ‘khyud de// pho yi dung pa gnyi ga yis// gru mo gnyis la bzhag pa yag// bud med rje ngar ‘dus par bshad// ces te/ bsdam pa’i bya ba’o// (Sukusuma, D 107b. 3-4; P 129a.8- 129b-1). Regarding the term “position” (bsdam pa) the Sanskrit term bandha, which is likely what bsdam pa is translating here, is used in kāmaśāstra to refer to sexual positions. Vaidyapāda may here and in his subsequent comments be citing a kāmaśāstric source (perhaps the aforementioned Skyod byed thams cad kyi gtsug lag?). As noted above there seem not to be many (any!?) extant such texts from the period between the 3rd-4th-century Kāmāsūtra itself and the later kāmaśāstra texts starting with the work of Kokkoka and Padmaśrī, both dated to no earlier than the 9th century, and quite possibly as late as the 12th. Vaidyapāda’s citations here may thus provide a window into a kāmāśāstric source from this intermediary period. However, there is a reference in one of the passages that Vaidyāpāda cites here to a sequence of blisses, which is either a noteworthy reference to stages of bliss in a kāmaśāstric work, or an indication that the text Vaidyapāda cites here is actually not a kāmaśāstric text, but rather a Buddhist one. See also notes 126 and 222. I briefly discuss the relationship between Buddhajñānapāda’s writings and kāmaśāstra in Chapter Six. 222 Vaidyapāda explains that this line refers to the action of the arms and legs and describes the posture with the following verse: “The woman’s bent knees/ are to be placed on the man’s elbows/ This [posture] is called “knees on elbows”/ These [postures] are asserted to be (to produce?) the stages of bliss.’ This is the act of looking closely.” de nas gru mo mtshan nyid ces pa ni rkang lag gi bya ba ste/ ji skad du/ bud med pus mo bkug pa ni// pho yi gru mor bzhag par bya// bus mo gru mor bshad pa ste// de dag dga’ ba’i rim par ‘dod// ces te/ rnam par lta bar bya ba’o// (Sukusuma, D 107b.4-5; P 129b.1-2). The reference to the “stages of bliss” in this verse, which, if it is indeed from a kāmaśāstric source rather than a Buddhist one, might suggest that perhaps the progression of the “blisses” in tantric texts was developed on the basis of their being such a progression already in the literature on kāma, which was then adapted to a soteriological context. However, it is also possible that the reference to the stages of blisses indicates, rather, that this is a Buddhist source rather than a kāmaśāstric one. I have been unable to determine at this point which is more likely to be the case. I briefly discuss the relationship between Buddhajñānapāda’s writings and kāmaśāstra in Chapter Six. 223 Vaidyapāda’s explanation of these positions via quotations from the aformentioned scripture continues. As this and the subsequent descriptions are difficult to understand, I have not translated the remainder of them here. I will note, however, that as Vaidyapāda describes them none of these positions seem to involve actual coitus, but rather to be positions from which the partners are to visually regard one another in order to stimulate passion. 224 bar] D C S, par P N; Although Vaidyapāda’s commentary also reads par (in both D and P), his comment suggests bar: brla gnyis par bltas te zhes pa ni ‘og gi padma la bltas na/ (Sukusuma, D 108a.1-2; P 129b.7). 225 zid sgra. A sound used in works on Indian erotics to indicate arousal and pleasure. 389 He should play using his lips.226 |108| Embracing her in the majestic posture, With his left hand holding the hair on the crown of her head And his right hand supporting her throat, He should suck the honey of her lower lip. |109| 227 While sucking and making the sound śīt, He plays with her breasts, the tips of her fingers. Her throat, lower lip, cheeks, and earlobes,228 Her eyes, the crown of her head, and her secret place229— Kissing these with his mouth. |110| At her two ears and her armpits The two [sides] of her throat, and the place where the three meet230 He should make marks with his fingernails.231 |111| Her two breasts, and two armpits, Her two main [places],232 and her two cheeks Her two palms and the soles of her two feet By rubbing these places, he creates great affection.233 |112| With his left hand he should massage the lotus maṇḍala And stir it with his tongue.234 226 Vaidyapāda explains: “Just as a bee at a flower maṇḍala sings and sucks honey, just like that the yogin, as well, should sing with the sound ṣīt and issue forth a long “hūṃ” as he uses his two lips to play in the lotus maṇḍala.” ji ltar bung ba me tog gi dkyil ‘khor la glu len cing rtsi ‘jibs pa ltar/ rnal ‘byor pas kyang zid (zid] D, zing P) sgra lta bu’i glu len (len] D, P om.) cing (cing] D, P om.) hūṃ ring po blangs nas mchu gnyis kyis padma’i dkyil ‘khor du rtse (rtse] D, brtse P) bar bya zhes so// (Sukusuma, D 108a. 2-3; P 129b.8-130a.1). 227 Vaidyapāda explains that the practices described in this verse are done in order to “invoke the places that are the sources of bodhicitta.” de nas byang chub kyi sems ‘byung ba’i gnas rnams bskul ba’i phyir (P + ro) / nu ma lag rtse zhes pa la sogs pa ‘o// (Sukusuma, D 108a.3; P 130a.1-2). 228 rna ba’i rtsa. 229 Verse 110b-d is mostly parallel with a verse cited in the only surviving manuscript of Kalyāṇavarman’s Catuṣpīṭhapañjikā, where it appears combined together with verses 171-174 from the Dvitīyakrama (which deal with the practice of placing sixteen syllables on specific places in the body during the practice of the bindu yoga) and some other related verses. The passage from the Catuṣpīṭhapañjikā is said to come from the Aṣṭāṣṭaka, which perhaps may be the title of a text, though this is not certain. Thanks to Péter Szántó for sharing his diplomatic transcript of these verses with me. The parallel verse here reads: pīnastane karāgre ca grīvāyāṃ adhare tathā | gaṇḍākṣikarṇṇamūle ca mūrdhni sarvāṅgam eva ca | (I have edited the text slightly following Harunaga Isaacson’s suggestions, for which I am grateful.) 230 Khenchen Chodrak Tenphel explains that this refers to the secret place (personal communication, February 2016). 231 Khenchen Chodrak Tenphel explains that these are all places where there are particular channels running through her body (personal communication, February 2016). 232 gtso. I am unsure about the meaning of this term here. 233 Vaidyapāda clarifies that “Accomplishes great affection means that in all three ways (physically, verbally, and mentally(?)) she [experiences] great passion towards oneself.” mdza’ ba chen po grub ces pa ni rnam pa gsum po (po] D, pa P) des bdag la lhag par chags pa zhes so// (Sukusuma, D 108a.3-4; P 130a.2). 234 At this point, Vaidyapāda clarifies, he may only use his vajra to stir it ever so slightly. ‘og sgo lces bskyod par bya zhes te/ rdo rje ni thabs cha tsam gyis bskyod par shes par bya’o// (Sukusuma, D 108a.5; P 130a.3). See also 390 Looking, moreover, both above and below, His mind becomes passionate about her. |113| [Searching For The Cakra] And then that delighted girl Shows her lotus and recites these words: “The king of natural great bliss Abides in this lotus |114| Because it is realized by means of the channels and winds235 You should search for the cakra.”236 And then with his fingers237 [He238 should search for] the great cakra, which abides239 inside. |115| Having ascertained the anthers, stamen, And the eight-petaled-one ornamented by the five essences,240 That abide in the lotus241 He should search for the āli [and] kāli;242 [the] mantra,243 |116| Daśatattvasaṃgraha (V.23), which likewise describes the stimulation of the yogin’s partner with his fingers and tongue immediately before the yogic partners’ union during the prajñājñānābhiṣeka (Klein-Schwind 2012, 210). 235 Here Vaidyapāda clarifies this with reference to the Yoga tantras, which he says focus on winds, and the Yoganiruttara tantras, which he says focus on channels. “It is realized by the oral instructions of the Yoganiruttara tantras, that is to say the Dākiṇī tantras, which focus principally on the channels. With regards to the Yoga tantras which emphasize method, since the suchness of the winds is primary [there], one is brought to realization through that. That being the case, since bodhicitta is attained [through?] the abiding and resting of the winds in the pathways of the channels, you should understand that both are necessary.” rnal ‘byor bla na med pa’i rgyud rnams kyi man ngag ste mkha’ ‘gro ma’i rgyud rnams rtsa rtso bor byed pa’i rgyud de/ des rtogs pa’o// rnal ‘byor thabs gtso bor byed pa’i rgyud ni rlung gi de (de] P, D adds kho na) nyid gtso bos des (des] P, de D) rtogs par byed pa’o// de lta na yang byang chub sems rtsa’i lam na rlung gi gnas dang ngal gso (gso] P, so D) bas (bas] sugg. em, ba D, P) thob pas na gnyis ka (gnyis ka] D, gnyi ga P) la yang dgos par shes par bya’o zhes so//) (Sukusuma, D108a.6-108b.1; P 130a.6-7). 236 See notes 209, 237, and 248. 237 Vaidyapāda comments that he should do this “using the three fingers of his left hand drawn together.” sor mo yis zhes pa ni lag pa g.yon pa’i sor mo gsum ‘dus byas pas so// (Sukusuma, D 108b.1; P 130a.8). See also notes 209 and 248. 238 yis] sugg. em. based on V (D and P), yi D C P N S. 239 gnas] sugg. em based on V (D and P), nas D C P N S. Vaidyapāda reads de nas nang na gnas pa’i rtsa’i ‘khor lo chen po btsal bar bya’o// (Sukusuma, D 108b.1; P 130a.8). 240 snying po. Vaidyapāda here specifies and names five channels in the body—the central, left, right, front, and back—which he says correspond to the five buddhas and the five elements (Sukusuma, D 108b.2-3; P 130b.1-3). 241 116a-c ~c.f. Daśatattvasaṃgraha, V.22a-d (Klein-Schwind 2012, 210). 242 While the āli and kāli traditionally refer to the vowels and consonants of the Sanskrit alphabet, Vaidyapāda clarifies that here āli and kāli refer to “the left and right [channels? respectively].” de ne ā li kā li zhes pa ni g.yon pa dang g.yas pa’o// (Sukusuma, D 108b.4; P 130b.5). 243 Vaidyapāda explains that here “Another name for [what is here referred to as] mantra is the lalana (‘phyang ma = rkyang ma?), which is the place where the moon descends.” mantra zhes pa ni rnam grangs ghzan du ‘phyang ma (rkyang ma?) zhes pa ste zla ba ‘bab pa’i gnas so// (Sukusuma, D 108b.5; P 130b.5) 391 The kūrmaka,244 and the śaśāṅka245— These three nāḍīs.246 The vajradhatvīśvarī nāḍī, free from subject and object, [Abides] in the center of the bhaga, |117| By means of the oral instructions from the guru One must find247 this using his fingers.248 Then, endowed with the ten bhūmis,249 244 kur] D C, kun P N S V (D and P). Above in his list of the five channels Vaidyapāda notes that the right-hand channel is called the rus sbal, which is probably a translation of kūrmaka. Vaidyapāda explains that here “Another name for [what is here referred to as] kūrmaka is the rasana (ro ldan ma), the place where rakta descends.” kur (kur] sugg. em based on root text in D and C, kun D P) ma ka zhes pa ni rnam grangs ghzan du ro ldan ma zhes te/ rakta ‘bab pa’i gnas so//. (Sukusuma, D 108b.5; P130b.5-6). 245 sha (sha] D C, shang P N S V (D and P)) shāng (shāng] sugg. em based on V (D and P) which read shang (I suggest adding the long ā), sha D C P N S) ka. This Sanskrit term used here, the “hare-marked [one]” is usually a term for the moon, but it is also attested in the Pradīpoddyotana, for example, as a term for the central channel, which is clearly what it refers to here. Thanks to Harunaga Isaacson for both pointing out the correct Sanskrit term and its attestation in the Pradīpoddyotana. Vaidyapāda explains here that, “The śaśāṅka is that [channel] which is located in the center between those two, and which is also otherwise known as the *mūrdhanī (spyi gtsug ma). It is the place where wisdom desends.” Again, Vaidyapāda’s commentary leaves no doubt that this is a reference to the central channel, which is more commonly termed the avadhūti. shang shang ka zhes pa ni de gnyis kyi dbus te gzhan du spyi (P +bo) gtsug (gtsug] D, gtsugs P) ma zhes kyang grags/ de ye shes ‘bab pa’i gnas so// (Sukusuma, D 108b.5-6; P 130b.6). 246 na ḍi] sugg. em., na li D C P N S. 247 go ba. Literally “understand.” 248 ~C.f. Vagīśvarakīrti’s Saṃkṣiptābhiṣekavidhi: so ‘pi vāmāmbhoruhavāmapārśvasthitāṃ vajradhātvīśvarīnāḍīṃ gurūpadeśavalād upalabhya vihitotphullabhramarījālādikaraṇasaṃhāre/ (Sakurai 1996, 418). The procedure of “searching for the cakra” appears to involve the yogin’s seeking out, with the fingers, the so-called vajradhatvīśvarī nāḍī in his partner’s body. Vaidyapāda elaborates, “Moreover, the vajradhatvīśvarī channel which is beyond subject-object [duality] is located in the center of the bhaga, like the string of a lute. You must find this with your fingers in reliance upon the instructions of a compassionate guru.” de yang rdo rje dbyings kyi dbang phyug ma’i rtsa gang gi phyir gzung ‘dzin (‘dzin] P, ‘don D) dang bral ba bha (bha] P, bu D) ga’i dbus na gnas pa’i pi bang (bang] P, wang D) gi rgyud ltar gnas pa ste/ bla ma thugs rje dang ldan pa’i man ngag gis (gis] P, gi D) sor mos go bar bya dgos so zhes te/ (Sukusuma, D 108b.6-7; P 130b.7-8). This same procedure is described in the Piṇḍikṛta commentary, the Maṇimālā, attributed to Nāgabodhi. Here, just after the call and response between the yogin and consort, and before she recites the verse of praise to the bhaga, the girl is to show the yogin the naḍī-cakra inside of her lotus: “Then that devī holds the two sides of her lotus with her hands, and thus pulling on the lotus she should show him the nāḍī-cakra. Regarding this, she should show him the nāḍī-cakra inside her lotus in this way: “Hey, son of noble family! [Here] in the center is a nāḍī which, because it is covered by the pleasure nāḍī that is similar to a person’s nose, corresponds with the man’s liṅgam. This is the central channel, called vajradhatviśvarī. It is also called samantabhadrī, and from among the thirty-two channels described in the Vajrāmṛta, it is the main one where blood and semen are brought together. This itself is that from which the three realms arise, and they also dissolve [back] into this. Since this itself is the essence of the Tathāgata Akṣobhya, it is the prajñāpāramitā, the nature of the dharmadhatu wisdom, that which produces beings, and which gathers them back...” de nas lha mo des rang gi lag pa dag gis padma’i ngos gnyis nas bzung ste padma brgyangs (brgyangs] D, brgyad P) nas rtsa’i ‘khor lo bstan par bya’o// de la padma’i nang du rtsa’i ‘khor lo ‘di ltar bstan par bya ste/ kye’o (kye’o] D, kye’i P) rigs kyi bu dbus na gnas pa’i mi’i sna lta bu’i rtsa ra mas g.yogs par gyur pas skyes bu’i linga dang mtshungs pa’i rtsa de ni rtsa dbum ma chos kyi dbyings kyi dbang phyug ma zhes bya ste/ rnam grangs gzhan yang kun tu bzang mo yin zhing dpal rdo rje bdud rtsi las gsungs pa’i rtsa sum cu rtsa gnyis kyi nang nas gtso bor gyur pa rakta dang shu kra dag sdud par gyur pa ste/ de nyid las khams gsum pa skye bar ‘gyur zhing de nyid du thim par ‘gyur ro// de nyid ni de bzhin gshegs pa mi bskyod pa’i ngo bo yin pa’i phyir shes rab kyi pha rol tu phyin ma (ma] D, pa P) dang/ chos kyi dbyings kyi ye shes kyi rang bzhin dang sems can thams cad skyed bar byed pa dang sdud par byed pa yin no// (Maṇimālā, D 130b.6-131a.3; P 122b.4-7). See also notes 209 and 237. 249 See verses 299-312 of the Dvitīyakrama below where ten stages of sexual union are described in terms of their correspondence with the bodhisattva bhūmis. 392 He should embrace that goddess. |118| [Union] 250 When the vajra touches the lotus This is explained to actually be sevā.251 The vajra entering the lotus Is actually upasādhana. |119| Then, through moving and stirring a bit, The heart quivers and attentiveness wanes The hair on the crown falls loose and garments are cast off Sweat252 covers the body and it takes on a reddish hue,253 |120| And with reddened eyes [she]254 looks at one. Moving255 repeatedly brings about sādhana. Thus, without concern256 Moving that which is bow-shaped,257 the vow-holder |121| Causes the blazing of the triangular wisdom fire.258 Thereby the elements melt and the sixteenth part, 250 Vaidyapāda mentions that these verses describe the way in which the master teaches the fourth tattva, which according to the list of the ten tattvas that Vaidyapāda gave before in the Sukusuma, corresponds with the tattva of the mudrā (Sukusuma, D 109a.1-2). In his commentary on Dīpaṃkarabhadra’s Maṇḍalavidhi Vaidyapāda similarly notes that presentation of the fourth tattva corresponds with the presentation of the prajñājñānābhiṣeka (Guhyasamājamaṇḍalopāyikāṭīkā, 211a.2-3). See note 111 for the list of the ten according to the Sukusuma. 251 bsnyen pa’i de nyid. I understand this to be an instance where Buddhajñānapāda is homologizing terms from nonsexual tantric practices (the four aṅgas of sevā and so forth, which are usually associated—even in Buddhajñānapāda’s own writings—with phases of generation stage sādhana practice) with perfection stage sexual practices, and thereby asserting sexual practice to be the actual identity or suchness (de nyid) of those terms and practices. Vaidyapāda says the term suchness is used in each of these instances in order to indicate that it represents the “unchanging” sevā and so forth, because it unfailingly brings about non-abiding nirvāṇa (bsnyen pa zhes pa ni rdo rje dang padma zhes sngon gyi tha (P om.) tshig go// de nyid ces pa ni mi ‘gyur ba’i ste/ ‘dis mi gnas pa’i mya ngan las ‘das pa la mi bslu bar byed pa’i phyir ro// phyi ma la yang de bzhin du sbyar ro// (Sukusuma D 109a.2; P 131a.3-4). 252 rngul] sugg. em., rdul D C P N S V (D and P). This emendation is based on the line from the parallel verse in Vaidyapāda’s Yogasapta which reads rngul chu thigs pas lus kun khyab// (Yogasapta, D 71a.5; P 84b.7) 253 Vaidyapāda links the preceding five “signs” that come about through moving and stirring with the five wisdoms (Sukusuma, D 109a.3-4; P 131a.5-6). 254 Khenchen Chodrak Tenphel explains that this actually refers to both partners looking at one another (personal communication, February 2016). 255 bsgul] sugg. em. based on V, bskul D C P N S. Buddhajñānapāda’s text here reads bskul but given the fact that earlier the text read bsgul ba, as well as the fact that this is glossed in Vaidyapāda’s commentary as yang dang yang du bskyod pa suggests that it is bsgul that is is meant. This passage has been translated at least twice, in J. Dalton (2004, 13) and Roberts (2010, 486). I part ways from both of their interpretations of the passage in several places. 256 Vaidyapāda seems to suggest that this means something like “effortlessly.” He writes, “Without concern means without having to search for it. Since the causes have already come about, have no doubt that the fourth tattva will arise.” sems khral med pa ru zhes pa ni btsal (btsal] D, brtsal P) dgos pa med de/ rgyu sngon du song ba’i phyir te de kho na nyid bzhi pa skye ba la the tshom mi bya’o// (Sukusuma, D 109a.5; P 131a.7-8). 257 This is a reference to the wind element, the “maṇḍala” of which is represented in the traditional sādhana visualizations as a bow-shape. Vaidyapāda makes it clear that this refers to the wind maṇḍala. (Sukusuma, D 109a.5; P 131a.8). 258 Vaidyapāda explains that this is “the fire which has the nature of the triangle.” sum mdo’i rang bzhin gyi me (Sukusuma, D 109a.6; P 131b.1). 393 Which is like a jasmine flower, Should be offered by unifying the winds.259 |122| Naturally perfectly pacified The suchness that is the pacification of all phenomena, That bliss itself, dwells at the jewel [for] an instant.260 Free from recollection, [it] is made to move—261 This itself is mahāsādhana.262 |123| 259 Vaidyapāda elaborates “The sixteenth part which is like a jasmine flower means the bindu of bodhicitta. Through joining and uniting, this is made to enter the lotus saṃputa at the heart center.” bcu drug char gyur pa me tog kunda (kunda] D, kun da P) dang ‘dra zhes te byang chub kyi sems kyi thig ler gyur pa sdus cing (P adds sdus pa’i) sbyor bas snying ga’i padma kha sbyar du gzhug pa’o// (Sukusuma, D 109a.6-7; P 131b.2). Vaidyapāda explains that after causing the bindu to enter the heart, the downward clearing wind is then used to send it to the tip of the vajra, and at this point the disciple should come to know the seven yogas through the oral instructions of the master (Sukusuma, D 109a.7; P 131b.2-3). On these seven yogas see also note 121. Dakpo Tashi Namgyal cites this passage in the context of the experience of the third initiation performed with a visualized wisdom consort, but later references the possibility of a karmamudrā (Roberts 2010, 486). 260 Vaidyapāda comments that, “Regarding its being made to remain, it remains there for one or two ‘instants of a completed action.’” gnas par byed pa ste/ der yang bya ba rdzogs pa’i skad cig ma gcig dang/ gnyis la sogs pa’i bar du gnas nas/ (Sukusuma, D 109b.2; P 131b.5-6). This is one of two types of “instant” found in Buddhist literature and refers to, as an “instant,” the time it takes to complete a given action. 261 Vaidyapāda explains, “[One is] free from recollection because of having reached the natureless yoga. The text says [it] is made to move because one has completely abandoned all conceptuality.” dran med ces te/ rang bzhin med pa’i sbyor bas zin pa’i phyir ro// g.yo bar byed pa (byed pa] D, P om.) zhes pa ni rnam par rtog pa thams cad spangs pa’i phyir ro// (Sukusuma, D 109b.2-3; P; 131b.6-7). While in both earlier translations of this passage, J. Dalton (2004, 13) and Roberts (2010, 486) seem to take the subject of the movement to refer to consciousness, I am inclined to read this as referring to the bindu. Thus, following Vaidyapāda’s earlier explanations, the bindu has been brought to the tip of the vajra, remains there for an instant or two, and then the practitioner enters a state free from conceptuality and “it [i.e. the bindu of bodhicitta] is made to move.” This would then refer to the moment of emission. This reading also fits better with the statement two verses later where Buddhajñānapāda notes that “The intelligent one will take up the liquid nectar that is in the lotus with his mouth and drink it.” Without having emitted it there, there would be no nectar (i.e. semen), for the yogin to receive from his partner’s lotus and drink. Indeed, emission is a standard feature of the prajñajñānābhiṣeka and other manuals also instruct the yogin to drink the fluids that result from the sexual union as part of the ritual for this initiation (see e.g. Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi (ed. Sakurai 1996, 419)). This verse appears, then, to associate mahāsādhana both with the holding of the bindu at the tip of the vajra, as well as with its emission, or perhaps even with the precise moment in which the bindu begins to be emitted. Later commentators writing about the prajñājñānābhiṣeka did get into detail about the precise location of the bindu of bodhicitta at the moment when suchness is experienced (see, for example, the Sekanirdeśapañjikā (Isaacson and Sferra 2014, 104-5). 262 Vaidyapāda elaborates, “This itself is mahāsādhana because it is the essence of the accomplishment of the mahāmudrā.” de nyid la sgrub pa chen po zhes te phyag rgya chen po dngos grub kyi ngo bo nyid kyi phyir ro// (Sukusuma, D 109b.3; P 131b.7). Here Vaidyapāda seems to be further supporting Buddhajñānapāda’s claim that these sexual yogic practices are in fact identical to these processes of sevā etc., which are generally used to describe the stages of generation stage practice. That is, he appears to be saying that the wisdom experienced through the bliss that occurs at the climactic moment of sexual yogic practice is called mahāsādhana, because it is indeed the essence of the result of deity yoga practice, of mahāmudrā. The term mahāmudrā, as I noted above, is in the 8th and 9th centuries used to refer to the form of the deity; Vaidyapāda clearly uses the term in this way at multiple places in the Sukusuma. 394 263 From the uniting of the realm of space and the vajra, great bliss that has genuine vision arises,264 which brings about genuine bliss. 263 It seems that the text of verses 124 and 125 may be corrupt in several places. My translation of these verses relies on Vaidyapāda’s commentary, as well as the parallel verses that survive in Sanskrit from Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavedhi (on which see below), and in Tibetan in Vaidyapāda’s Yogasapta. These two verses are also metrically unusual in the Tibetan translation. Departing from the seven syllables per pāda in the preceding and subsequent verses, verse 124 has four pādas with nineteen syllables each, and verse 125 has two pādas with seventeen syllables each, two with thirteen syllables each, and a final pāda again with nineteen syllables. Verses 124a-125a are found in Vaidyapāda’s Yogasapta where they appear with slight variation from the Dvitīyakrama, which, on the basis of a comparison with the Sanskrit from the Saṃkṣiptābhiṣekavedhi, suggests that the Yogasapta’s verses are also somewhat corrupt (Yogasapta, D 71a.3-4; P 84b.4-6). As just noted, 124a-d and 125b-d are mostly parallel with Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi vv. 10-14 on the prajñājñānābhiṣeka, which fortunately survive in Sanskrit. However, while the extant Sanskrit of these verses is very helpful in providing a more clear reading of the Tibetan translation of some parts of these verses from the Dvitīyakrama (for example mahādbhutaṃ/ sukham utpadyate yat tat paramānandadāyakaṃ// viramānandayor mmadhye is extremely helpful in clarifying the confusing bde chen ‘byung byar ‘gyur// gang gang yang dag dga’ byed chags bral dga’ gnyis bar du, with which it appears to be precisely parallel, but which would otherwise not be naturally read that way just on the basis of the Tibetan), there are also several places where the wording as found in the Dvitīyakrama and the Saṃkṣiptābhiṣekavidhi clearly differ, and where Vaidyapāda’s Sukusuma supports the reading in the Dvitīyakrama, rather than that in the Saṃkṣiptābhiṣekavidhi (such as mchan can versus samspaśāt). Thus the verses as found in Vāgīśvarakīrti’s work appear to have undergone some transformation from the way they appeared earlier in the Dvitīyakrama. Vāgīśvarakīrti’s verses (following Sakurai’s edition) read: khadhātuvajrasaṃyogāt samsparśāc ca mahādbhutaṃ/ sukham utpadyate yat tat paramānandadāyakaṃ// 10// viramānandayor mmadhye lakṣyam vīkṣya dṛḍhīkuru/ kamalākāśe maṇivaratakayoḥ pīḍanasthāne // 11// vajraparyaṅkataś cittaṃ maṇyantargatam īkṣayan/ yat tad utpadyate jñānaṃ jñā[na]n tadrūpam ity alam// 12// na rāgo na virāgaś ca madhyamā nopalabhyate/ jñānadṛṣṭir yadā yogī sukhaṇ tiṣṭhet kṣaren na ca// 13 // praharam vātha vaikāhaṃ pakṣaṃ māsañ ca vatsaraṃ/ kalpaṃ kalpasahasra[ñ] ca tiṣṭhet jñānābhiyogataḥ // 14// (Sakurai 1996, 418-19). Thanks to Harunaga Isaacson for drawing my attention to this parallel. I have also emended Sakurai’s edition very slightly based on Harunaga Isaacson’s suggestions. 264 Vaidyapāda reads “that which has genuine vision” and “great bliss” separately, suggesting that the latter arises out of the former, but the grammar of the verse, at least as it has been rendered into Tibetan, does not allow for this reading, so I have not followed it here. 395 Between the cessation of bliss265 and bliss,266 an absence267 is seen and should be stabilized. 268 265 Chags ‘bral. The Sanskrit from Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi reads viramānandayor, and it seems that this is indeed probably what the Tibeatan translators were reading here. They seem to have understood virama in the compound to mean viramānanda (thus they the two members of the compound would be viram[ānanda] and ānanda), and thus chags ‘bral is here a translation of viramānanda. The more common Tibetan translation of viramānanda, however, would be dga’ bral rather than chags bral. 266 This line seems to contain a reference to at least two of the three blisses that are mentioned in verse 241 and listed in verses 290 and 291 of the Dvitīyakrama, and which Vaidyapāda also refers to in his Muktitilaka-vyākhyāna. In both the Dvitīyakrama verses 290-91 and in the Muktitilaka-vyākhyāna the three blisses are listed as bliss (dga’ ba), middling bliss (dga’ ba bar ma), and bliss of cessation (dga’ bral). Vaidyapāda correlates them with the seven yogas (Muktitilaka-vyākhyāna, D 58a.1). A similar set of categories also appears to be referenced in verse 125 of the Dvitīyakrama as “passion, freedom from passion, and something in between” (‘dod chags, chags bral, bar ma). Those lines from verse 125 are, however, excerpted from the Sarvabuddhasamāyoga-tantra, where the context does not seem to be the series of blisses arising from sexual yogic practice. On this line from the Sarvabuddhasamāyogatantra and its relation to the three blisses see note 276. On the seven yogas see note 121. 267 dben (dben] sugg. em. based on V (D and P), bden D C P N S) nyid (nyid] sugg. em. based on parallel verse in Vaidyapāda’s Yogasapta (see Yogasapta, D 71a.3; P84b.4), gnyis D C P N S). Here Vaidyapāda states, “The absence of the two blisses should be seen by means of the oral instructions, and [the text] is stating that one should stabilize that.” dga’ ba gnyis gyis dben pa de man ngag gis mthong ba de la blo brtan par gyis shig ces gdams pa’o// (Sukusuma, D 109b.5; P 132a.2). I have emended the Dvitīyakrama here in accordance with Vaidyapāda’s commentary to read “an absence” (dben nyid), rather than the implausible “the two truths” (dben gnyis), found in all recensions of the Dvitīyakrama itself. However, Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi (as well as the other citations of this line in later sources, on which see note 268 below) here reads “the goal” (lakṣya). I will note here that it was very tempting to make an even more serious emendation of the Dvitīyakrama from bden gnyis not just to dben nyid following Vaidyapāda, but to ‘ben nyid, “the goal,” to match the Sanskrit (lakṣya) of all of the later sources of which I am aware that cite this passage. However, I have resisted doing so because, while this reading works quite well with the Dvitīyakrama itself, it is more difficult to coherently make this emendation in Vaidyapāda’s commentary on the relevant passages. It is not absolutely impossible to make some sense of Vaidyapāda’s commentary with the reading of ‘ben nyid—in the passage cited above in this note it is not so difficult, but a passage cited below in note 275 is significantly more difficult to comfortably emend in that way—but the Sukusuma reads much more smoothly and naturally without this emendation, and Vaidyapāda’s commentary on the Dvitīyakrama is several centuries earlier than any of the Sanskrit sources that include this line, so my guess for the moment is that the Dvitīyakrama and the Sukusuma represent an earlier reading of the line that read “an absence” (dben nyid; I unfortunately cannot guess what the Sanskrit may have been), and which was later emended to read “the goal” (lakṣya). Moreover, the line as cited in Abhayākaragupta’s Āmnāyamañjarī, which reflects the later reading translates lakṣya into Tibetan in a way that is more expected, not as ‘ben but as mtshon bya (Āmnāyamañjarī, D 67a.1). In the end, though, this point remains something of a question. 268 Apart from its inclusion, along with most of the rest of verses 124-125 from the Dvitīyakrama in the Saṃkṣiptābhiṣekavidhi (see note 265 above), this particular line is cited—with some variant readings—in a number of later sources, including the Caturmudrānvaya (attributed, by some authors at least, to Nāgārjuna; p 32), the Abhiṣekanirukti (fol. 43 r) Kumāracandra’s Ratnāvalī (p. 102), Rāmapāla’s Sekanirdeśapañjikā (see Isaacson and Sferra 2014, 275), and the Kriyāsaṅgrahapañjikā (chapter 6, prajñājñānāvhiṣekavidhiḥ, st. 13ab). (This list of citations is provided in Isaacson and Sferra (2014, 275 n 120) in the notes to their translation of the Sekanirdeśapañjikā, which cites the passage twice. The page numbers that I give here are those provided in their citation, and include sources I, myself, have not looked at. The interested reader is therefore directed to Isaacson and Sferra’s bibliography for further details). Isaacson and Sferra note that the original source of this line is unknown (they do not reference its occurrence here in the Dvitīyakrama), and suggest that it may derive from a lost tantra, since the sources that cite it tend to give it the reverence normally attributed to scripture (ibid., 98-99). They note, however, that Abhayākaragupta (who, it should be noted, holds a position on the sequence of blisses that seems to be contradicted by this statement) in his Āmnāyamañjarī, casts doubt on its scriptural authority (ibid.). Although Abhayākaragupta does, indeed, as Isaacson and Sferra have noted, cast doubt on the scriptural authority of the line, he does still provide (“in the case that it is scriptural….” lung yin na de’i cha/….) a way of interpreting the line that does not undermine his position on the sequence of blisses (Āmnāyamañjarī, D 67a.1). While Isaacson and Sferra may be correct that this line is a from a lost tantra, it is also possible that the issue of its scriptural authority or lack thereof may be with reference to its presence here in the Dvitīyakrama, in a mukhāgama, a work that lies precisely on the borderline of scripture and authored commentary. It should be noted, however, that while Abhayākaragupta questions the scriptural authorituy of this line, he cites as scriptural—he attributes it to the Paramādya-tantra—a 396 In the space of the lotus, the jewel of the vajra and the heart of the lotus join,269 in vajra posture270 The mind271 is observed within the jewel.272 The bliss that arises is ascertained—and that itself is wisdom.273 |124| This is explained by all of the genuine supreme gurus to be the perfection stage.274 verse that is parallel to the very next two lines (124 cd) of the Dvitīyakrama (Āmyāyamañjarī, D 68a.1-2). Lastly, the line as it is cited in the Sekanirdeśapañjikā does include a variant. While in the Dvitīyakrama and Saṃkṣiptābhiṣekavidhi the line reads: viramānandayor madhye… /, the line in the Sekanirdeśapañjikā (and apparently also the Caturmudrānvaya, from which it appears to cite the line) reads: paramaviramayor madhye lakṣyaṃ vīkśya dṛdhīkuru. In both cases the first compound refers to two among the series of blisses that arise in the context of the prajñājñānābhiṣeka. In Buddhajñānapāda’s system, in which there are just three blisses, the compound seems to refer to viramānanda and ānanda, the third and first of the blisses, respectively. In the later systems under discussion in the Sekanirdeśapañjikā, there are four blisses, among which the compound appears to reference paramānanda (the second) and viramānanda (the fourth, in the system upheld by Rāmapāla, who cites the passage). This, however, does not seem to be such a significant difference, given that the issue is in what lies between the two, and in Buddhajñāṇapāda’s system there is no division into the first two blisses (ānanda and paramānanda, respectively) of the later system. So, the middle place between the two corresponds in Buddhajñānapāda’s system to *madhyamānanda, and in Rāmapāla’s system to sahajānanda, which may indeed be understood as parallel here—the absence (dben nyid) or goal (lakṣyam) which is to be marked by the practitioner during the initiation. I discuss the blisses according to Buddhajñānapāda’s system in Chapter Six. This line is also transmitted differently in Vaidyapāda’s Yogasapta (see note 265 above), which includes a parallel to Dvitīyakrama 124a-125a. The remainder of the pādas in this segment of the Yogasapta are the same content-wise as those in the Dvitīyakrama. There therefore seems to have been some error of transmission with respect to this particular line. 269 In the later versions of this verse which survive in Sanskrit, what we find here in the Tibetan translation of the Dvitīyakrama as sbyor, “unite,” is instead the slightly more forceful pīḍana, “pressing” or “squeezing.” My inclination is that the Sanskrit verse in the Dvitīyakrama also likely read pīḍana, and that the choice of sbyor here was simply a choice made by the Tibetan translators. I have, however, translated it into English in accord with the text as it reads here in the Dvitīyakrama as it survives in Tibetan translation, rather following the Sanskrit of the later parallel citations, since I cannot be absolutely certain about this point. The full Sanskrit line from Vāgīśvarakīrti’s Saṃkṣiptābhiṣekavidhi reads kamalākāśe maṇivaratakayoḥ pīḍanasthāne (Sakurai, 1996, 418). Note that the Tibetan translation of the Sekanirdeśapañjikā, where this same verse is cited as coming from the Paramādya-tantra (though Isaacson and Sferra (2014, 297 n 239) have noted that it is not found in any surviving recension of that work), translates the whole verse somewhat differently (though giving the same sense), and translates pīḍana, specifically, with the more semantically accurate Tibetan term mnan (ibid., 229). 270 Vaidyapāda clarifies that vajra posture here means union (Sukusuma, D 109b.6, P 132a.3). 271 Sems here refers to byang chub kyi sems, the bindu of bodhicitta. 272 ~Cf. Guhyasamājamaṇḍalavidhi, verse 366c-d. vajraparyaṅkataś cittaṃ maṇyantargatam īkṣyan. Cf. also the Vajrāvalī (Mori 2009 Vol. 2, 444), which incorporates these two pādas immediately after the incorporation of Dvitīyakrama 88a-c; and Daśatattva V.15, which follows the Vajrāvalī in incorporating these pādas after the incorporation of Dvitīyakrama 88a-c. 273 Apart from their inclusion, along with most of the rest of verses 124-125 from the Dvitīyakrama, in the Saṃkṣiptābhiṣekavidhi (see note 265 above), these two lines (124cd) are cited—with some variant readings—in several other later sources, at least two of which attribute the verse to the Paramādya-tantra. The verse is cited in the Abhayapaddhati (MS A fol. 15v2), the Sekanirdeśapañjikā (ed. Isaacson and Sferra 2014, 173), the Kriyāsaṅgrahapañjikā (ed. Sakurai 1996, 514) and the Yamāritantramaṇḍalopāyikā (fol. 24r.3), and the Āmnāyamañjarī (D 68b.1-2) (this list of sources is given by Isaacson and Sferra (2014, 297 n 239) in reference to the verse’s citation in the Sekanirdeśapañjikā; the page numbers I have given here are those provided in Isaacson and Sferra’s citation, and include sources that I, myself, have not looked at. The interested reader is therefore directed to Isaacson and Sferra’s bibliography for further details). Both the Sekanirdeśapañjikā and the Āmnāyamañjarī attribute the verse to the Paramādya-tantra, though Isaacson and Sferra note that it is not found in any of the surviving recensions of that tantra (ibid.). 274 Vaidyapāda explains that it is called the “perfection stage” (perhaps better translated here as “perfected stage”) “because it is naturally accomplished, and is not something posited by the mind” rnam grangs gzhan du na rdzogs pa’i rim pa yin par bla ma mchog rnams kyis kyang bshad de/ rang bzhin gyis grub pa ste/ blos gzhag (gzhag] D, bzhag P) pa ma yin pa’i phyir ro// (Sukusuma, D 109b.7; P 132a.4-5). 397 Neither passion, dispassion, nor something in between275 is perceived;276 the wisdom deity is seen there in a single instant. For eight hours, one day, one month, One year, one aeon, up to a thousand aeons one should experience that wisdom. The intelligent one will take up the liquid nectar that abides in the lotus with his mouth and drink it.277 |125| [A Doxography of Philosophical Views] Thus, the final identity of all things Is profundity and luminosity.278 [But] since beginningless time, ordinary beings Have fixated upon it as “me” and “mine;” Thus, without examining, they grasp to the self.279 |126| 275 Vaidyapāda writes, “How is it that there is the absence of two blisses? Neither passion, dispassion, nor something in between are observed means that there is no conceptualization in terms of these three.” (dga’ ba gnyis kyis ji ltar dben zhe na/ ‘dod chags chags bral bar mi dmigs zhes te ‘di gsum gyi (gyi] P, gyis d) rtog pa med pa’o// (Sukusuma, D 109b.7-110a.1; P 132a.5-6). 276 C.f. Sarvabuddhasamāyoga-tantra 1.3a-b. na rāgo na virāgaś ca madhyamā nopalabhyate|. Thanks to Péter Szántó for sharing with me his draft Sanskrit edition of the Sarvabuddhasāmayoga-tantra. While the context of this line in the first chapter of the Sarvabuddhasamāyoga-tantra, from which it is drawn, is not one of sexual yoga, given the strong parallels between this verse and the names of the three blisses that are given in Buddhajñānapāda’s system—they are listed in verses 290-91 of the Dvitīykrama as bliss (dga’ ba), middling bliss (dga’ ba bar ma), and bliss of cessation (dga’ ba dang bral ba)—along with its incorporation into the Dvitīyakrama precisely in the context of sexual yogic initiatory practice in which the three blisses are experienced—I wonder if the line from the Sarvabuddhasamāyoga may have served as a scriptural source or inspiration for the classification of the blisses as three-fold, as well as for the names ascribed to them in Buddhajñānapāda’s system. Certainly, Vaidyapāda does not take the line that way, however; as we will see below, he understands it to be a three-fold description of the goal itself. Nonetheless, my suspicion remains. Ronald Davidson (2002, 62) has suggested that the source of these three blisses may have been the oral tradition, and specifically the teachings of Buddhajñānapāda’s guru Pālitapāda (Davidson erroneously refers to this guru as *Bālipāda; he did not, at the time of writing his article, have access to Sanskrit sources that we now have). My speculation here with regards to a scriptural inspiration for the three-fold system of blisses does not necessarily contradict Davidson’s suggestion that this system may have been passed down to Buddhajñānapāda by means of an oral tradition. In any case, it seems that the Dvitīyakrama represents a very early example of the classification of the blisses that arise in the context of tantric practice (Davidson 2002b, 60-1); later systems appear to be based on Buddhajñānapāda’s, but with the addition of a fourth bliss, and a change in the name of one of the blisses. The reference to the “stages of bliss” the text cited in the Sukusuma’s commentary to verse 105, if it is indeed a kāmaśāstric source rather than a Buddhist one, might suggests that perhaps the progression of the “blisses” in tantric texts was developed on the basis of their being such a progression already in the literature on kāma, which was then adapted to a soteriological context. See note 224. 277 Vaidyapāda follows up this section by adding the concluding elements of the initiation ritual, including the vyākaraṇa, aśvāsa, and anujñā. After the final initiation Vaidyapāda says that the oaths should be taken and a pacifying homa performed, the maṇḍala reabsorbed and so forth (Sukusuma, D 110a; P 132b). 278 Given that they follow immediately after the descriptions of the rituals for the guhya- and prajñājñāna-abhiṣekas, these two lines are actually quite similar to Vaidyapāda’s interpretation of “the fourth” that would follow (or take place within the context of) these initiations: that is, a verbal description of suchness, or reality. However, Vaidyapāda explains “the fourth” as consisting of a much more elablorate presentation on the seven yogas. Could the presence of these two lines at this point in the text, though, perhaps serve as an allusion to the basic character of such an oral instruction on suchness immediately following the prajñājñānābhiṣeka? 279 Vaidyapāda explains that the incorrect opinions expressed in this section of the text are those of ordinary beings who hold to certain philosophical systems but who have not investigated by means of genuine valid cognition, and thus perceive things mistakenly because they follow untrustworthy scriptures, inferences, and mistaken forms of samādhi. thog ma dang tha ma (dang tha ma] D, P om.) med pa’i dus na so so’i skye bo grub pa’i mtha’ ‘dzin pa rnams kyis nga dang nga yir ‘dzin pas yang dag pa’i tshad mas ma brtags par ni bdag tu bzung zhes te yid mi ches 398 Arisen from time, arbitrary, Appointed by the heavens, inferior 280 Emanated by the grasper, Emerged from the all realms, |127| Iśvāra, issued forth, Created by time, and by prakṛti Not the creator, 281 Yoga, valid cognition, |128| Pure and impure,282 A self that is inexpressible and dwells within, The individual, the pervasive self, Life force, and individual, 283 consciousness, |129| The universal basis, and the knower, 284 The seer, and subject and object, Knowing and the known285 Man, and those born from him, 286 |130| Life force, sustanance,287 and so forth— The nonbuddhists hold these ideas. |131| [Asserting] space and the two types of cessation [To be] completely uncompounded, And perfectly stable; all compounded things As momentary, and without an owner, |132| Made up of subtle particles And not of aspects of the mind— The Kaśmīri Vaibhāṣikas understand [things in this way]. |133| pa’i lung dang/ rjes su dpag par snang ba dang/ log par ‘dzin pa’i ting nge ‘dzin gyis so// (Sukusuma, D 110a.7- 110b.1; P 132b.7-8) 280 Vaidyapāda notes that four terms beginning with “arisen from time” (dus las byung ba) are assertions of the Sāṃkhyas (grang can pa) (Sukusuma, D 110b.1; P 132b.8-133a.1). 281 Vaidyapāda notes that the eight terms beginning with “emanated by the grasper” (‘dzin pa pos sprul pa) are the assertions of the Vaiśeṣikas (bye brag pa), although I have found it difficult to find precisely eight terms in this list; the subsequent list begins with “yoga” (rnal ‘byor). (Sukusuma, D 110b.1-2; P 133a.1a). 282 The four terms starting with “yoga” (rnal ‘byor) are presented by Vaidyapāda as assertions of the followers of Kapila (ser skya pa) (Sukusuma, D 110b.2) 283 The five terms starting with “a self that is inexpressible and dwells within” (nang gnas brjod du med pa’i bdag) are presented by Vaidyapāda as assertions of the Jains (nam mkha’ gos can pa) (Sukusuma D 110b.2). 284 The three terms starting with “consciousness” (rnam shes) are presented by Vaidyapāda as secret assertions of the Vedantins (rig byed kyi mtha’ gsang bar smra wa) (Sukusuma, D 110b.2). 285 The five terms starting with “the seer” (mthong ba po) are presented by Vaidyapāda as assertions of the Carakas (tsa ra ka) (Sukusuma, D 110b.2-3). 286 These two terms, ‘men’ (shed bu) and ‘born from men’ (shed las skyes pa) are presented by Vaidyapāda as assertions of “the kings and so forth” (rgyal po la sogs pa) (Sukusuma, D 110b.3). 287 This last point is presented by Vaidyapāda as an assertion of the “red-robed ones” (gos dmar pa). 399 Brought into being [by] one’s awareness, The experienced object which is seen is an aspect [of the mind]; The three unconditioned [things] Are nonexistent, like the son of a barren woman; |134| All conditioned things are material There is no transformation in the three times;288 [There exist] the smallest particles [and] unobstructed form— This is what is understood by the Sautrāntikas. |135| That which has parts is not the ultimate; This is the case even for subtle particles. One cannot observe them individually; They do not appear, but are just like a dream. |136| The wisdom289 that is free from subject and object Is the ultimate, pure like a crystal— This is what the Yogācārins understand. |137| All of these different traditions Are not the ultimate, because [All things] are beyond the nature of being singular or multiple, Just like a lotus in the sky. |138| Peace [beyond] non-duality or non-non-duality Completely stainless like space— The intelligent Mādhyamikas understand [reality to be] thus.290 |139| [Though] reality abides as suchness, [Beings] conceptualize it distinctly In these and countless other [ways]. Therefore, all of these [perspectives] |140| Are not the genuine; they can be surpassed. The perspective of the higher yogins Is superior to that of the lower.291 288 This line is strange given that it appears to be the exact opposite of the assertion of the Sautrāntikas, who precisely do assert the three times to be changing. Khenchen Chodrak Tenphel suggests reading this as something that should be negated, i.e. dus gsum rgyu ba ma yin [pa med], “The three times are not unchanging, that is to say, they are impermanent” (personal communication, February 2016). 289 Vaidyapāda’s commentary preserves a different reading of this line. Instead of reading ye shes te “that wisdom,” it reads rnam shes che, “that great consciousness” (Sukusuma, D 111a.4; P 133b.5). 290 Buddhajñānapāda here clearly asserts the superiority of Madhyamaka to Yogācāra, and asserts Yogācāra to hold an idealist position, exactly like the critique of the system that is found in many later Tibetan sources. Vaidyapāda bears this out even more clearly in his analysis of the subsequent verses. However, as we will see, the Madhyamaka position is also here clearly asserted to be lesser than the tantric one. 291 Vaidyapāda notes that compared to what is held by the non-Buddhists, the Vaibhāṣikas improve by asserting these things to be impermanent. Compared to that, the Sautrāntikas improve by asserting those impermanent things 400 The lower view is refuted |141| By the wisdom of the higher one. Therefore, by means of the higher stage The sahaja master Performs the genuine blessing.292 |142| Luminous and perfectly joyful like the sky The self-arisen great *adhideva Is realized through spontaneously arisen wisdom In reliance on the words of the guru.293 |143| Thus, in the vessel294 possessed of samayas and vows, Through water-like one-pointedness The yogin should examine The reflection-like wisdom. |144| Once he has achieved this, the yogin While abiding in cyclic existence Will not be stained by those evils. Just as someone possessed of mantras and medicines |145| Enacts the slaying of snakes, Likewise when the great lord of yogins, Seals [them] with the medicine of wisdom What can the afflictive emotions do? |146| What can the rain do To someone with an umbrella in his hand? Likewise, when carrying the umbrella Of nondual wisdom295 |147| as imputations. Better yet, the Yogācāras assert that these imputations are merely mind, while the Yogācāra- Mādhyamikas improve this further by asserting the idea of things as merely the mind to be just the relative level of things. Even better, the Annuttara-ists(!) (bla na med pa’i gzhung pa) improve further by asserting that even the relative level of things is nothing but wisdom. (Sukusuma, D 111b.2-4; P 134a.4-5). 292 Here Vaidyapāda identifies the higher stage as bestowing initiation, the sahaja ācārya as one’s consort (Tib. shes rab; Skt. prajñā) and her blessing as uniting with her. (Sukusuma D 111b.3-4; P134a.6-7). See also note 10 with regard to the three types of ācārya, including the sahaja ācārya. 293 Here it seems Buddhajñānapāda is indicating that the wisdom that the disciple gains comes from the words of the master. Vaidyapāda makes this even more explicit. He writes, “From the words of the guru means, from what is transferred from the words of the great causal master, one directly experiences the bliss which is to be realized.” bla ma’i kha las zhes te/ de (de] P, de’i D) rgyu’i (rgyu’i ] P, rgyud D) slob dpon chen poi’i kha las rnam par ‘pho ba las mngon sum du bde ba rang la (la] D, las P) ‘byung ba rtogs par bya’o zhes so// (Sukusuma, D 111b.4-5; P 134a.8). The “causal master” is explained in a verse cited by Vaidyapāda earlier in the Sukusuma, as well as in the Muktitilaka-vyākhyāna, to refer to the guru who bestows initiation upon the disciple. See also note 10 with regards to the “causal master.” 294 Here the common metaphor of the worthy student described as being a “proper vessel” for the teachings is being played upon. 295 Vaidyapāda identifies the umbrella of nondual wisdom as the second stage of tantric practice (Sukusuma, 112a.2). 401 Even if a rain of concepts should fall How could they do any harm? That kind of perfect supreme wisdom How could it be known by an ordinary being? |148| It is not known by the śrāvakas Nor by the pratyekabuddhas The Yogācāras, Mādhyamikas, And bodhisattvas do not know it. |149| Even all of the non-superior buddhas296 Do not know this at all. [But] by pleasing the future vajra-holders, Who know this reality, |150| Due to the power of one’s great merit It will be transferred [even] without words.297 Thus the maṇḍala, homa, Bali, recitation, the counting rosary, |151| Sitting cross-legged, maintaining postures,298 and so forth299— Are in contradiction300 to the unelaborate, [Thus] they should not be [exclusively] taken up; but neither should they be [wholly] rejected Since they are emanated by the *adhideva. |152| The yogin who holds actions To be the great path Is like a wild animal chasing a mirage— [The goal] continually appears but can never be grasped. |153| When infected by the great sickness of actions, The one who heals [himself] with the great medicine Of unwavering wisdom is a sublime being. |154| Therefore, maintain the supreme three vows Of body, speech, and mind301 296 Tib. bla bcas sangs rgyas. Literally “those buddhas who are surpassed by something else.” Vaidyapāda identifies these as the buddhas of the Kriyā, Caryā, and Yoga tantras (Sukusuma, 112a.4-5). 297 This is one of a number of references in Buddhajñānapāda’s writings to the transference of wisdom directly from the teacher to disciple. 298 Buddhajñānapāda, however, composed an entire treatise on postures, the *Gativyuya. 299 In the Muktitilaka Buddhajñānapāda mentions the very same list of activities but makes an even stronger statement about them: they are “meant to fool beginners” (dang po pa rnams ‘drid phyir ro) (Muktitilaka, D 47b.7). 300 Tib. rnam par slu ba, Skt. *visaṃvāda? 301 Vaidyapāda explains these vows. He says that the vow of body is not to have contempt for the many forms of male and female bodies since they are arisen from Vajrasattva; the vow of speech is to speak coarse words of desire and the like, and not to hold them back, since they have the nature of vajra speech; and the vow of mind is not to reject bad or good thoughts since these have the nature of Vajrasattva. (Sukusuma, D 112b.6-7). His commentary 402 And having generated the majesty of the thought “I am [this]!”302 Practice the second stage. |155| [Instructions on the Second Stage, Beginning with the Framework of the Utpattikrama] 303 In an isolated place or on the edge of town Having completed all the required tasks, As appropriate Sit down on a comfortable seat. |156| Then bring to mind all sentient beings By means of the four great brahmacaryas.304 [With] these and the rest Purify the karmic obscurations in one’s mind-stream. |157| Looking at [it] as mind alone here is clearly taken from the Muktitilaka where Buddhajñānapāda makes almost identical statements about the three samayas (Muktitilaka, D 48a.4-5). 302 This seems to be an acknowledgement of one’s own body, speech, and mind, as having a vajra-nature, or even, as in Vaidyapāda’s commentary, a Vajrasattva nature. It likely alludes both to the acknowlegdement of the innate nature of the three gates, as well as the practice of generation stage deity yoga that is preliminary to and provides the framework for the perfection stage practice that is subsequently described. Vaidyapāda writes: “When [the text] says, Generate the majesty of thinking, “I am [this]!” [it means] to draw one’s own body, speech, and mind together as wisdom through uniting both means and wisdom, and to hold the pride of thinking that [this] is me. Mentally generate majesty in that way.” nga’o snyam pas gzi bskyed la/ zhes pa ni thabs dang shes rab gnyis kyi sgo nas mnyam par sbyar bas bdag gi lus ngag yid gsum ye shes su rab tu bsdus nas de lta bu’i rlom pa’i nga’o snyam pas sems kyi gzi bskyed pa’o// (Sukusuma, D 112b.7-113a.1; P 105b.7-8). Buddhajñānapāda’s Muktitilaka also has a parallel passage (Muktitilaka, D 48a.3-4), in which it is made clear that this process is to be engaged in while in union (“uniting the two organs”), which can likewise be understood from the reference to “uniting means and wisdom” in Vaidyapāda’s commentary to the Dvitīyakrama passage, translated above. Vaidyapāda’s commentary on the two parallel passages is also nearly identical (See Muktitilaka-vyākhyāna, D 51a.1-2). 303 These verses are a very condensed version of the first two samādhis of generation stage sādhana. Vaidyapāda fills out the details significantly. 304 The term tshangs spyod normally translates the Sanskrit term brahmacarya. However, given the context— clearly the beginning of the generation stage, in which the four brahmavihāras are commonly practiced—and Vaidyapāda’s commentary which defines them as “the four—compassion and so forth,” it is clear that what is intended here are what are more commonly referred to as the brahmavihāras. Buddhajñānapāda also uses the term mahābrahmacarya to refer to the brahmavihāras in the Muktitilaka (D 49a.1). Vaidyapāda explains here in the Sukusuma that these are called brahmacarya because, “The four [practices] of compassion and so forth are the conduct (carya) of the Bhagavan Buddha Great Brahmā.” (ston pa sangs rgyas bcom ldan ‘das tshangs pa chen po ste de’i spyod pa ni snying rje la sogs pa bzhi’o// (Sukusuma, D 113a.4-5; P 136a.5). He goes on to describe their contemplation in some detail, in a rather usnusual matter that is connected with tantric sādhana practice: “Emanate from the seed syllable light rays that place the sentient beings who are illuminated by them on the path of seeing, and imagine that this liberates them from suffering and its causes. This is the contemplation of compassion. Having been freed from that [suffering] they achieve worldly and transcendent bliss. This is the contemplation of love. One might wonder how it is possible to have the capacity to benefit beings who are incorrigible and difficult to cure. Think with joy that it is possible to do this by appearing in the body of one's own [yidam] deity by means of the seed-syllable that is like a wish-fulfilling tree. This is the contemplation of joy. Mentally engaging with unsurpassed awakening in these [three] ways and not thinking at all about praise and the like is the contemplation of equanimity.” sa pon gyi ‘od zer gyis snang ba’i sems can rnams mthong ba’i lam du byas la/ de rnams sdug bsngal dang de’i rgyu las grol bar bya’o snyams pa ni/ snying rjes dmigs pa’o// de las grol nas ‘jig rte dang ‘jig rten las ‘das pa’i bde ba sgrub po snyams pa ni byams pas dmigs pa’o// dmu rgod ltar bcos dka’ ba’i sems can rnams la don de ji ltar nus snyams pa la/ dpag bsam gyi shing lta bu’i sa bon las rang gi lha’i (lha’i] D, om. P) skur snang na sgrub nus so snyam pa la rangs pa ni dga’ bas dmigs pa’o// de rnams kyis bla na med pa’i byang chub yid la byed pas bstod pa la sogs pa la mi sems pa ni btang snyoms kyis dmigs pa’o// (Sukusuma, D 113a.5-7; P 136a.5-8). 403 The outer world is seen to be empty of nature305 Seeing306 mind alone, as well, to be empty Remain in self-awareness alone. |158| That awareness, as well, is imagined As a moon, and so forth, upon a seat, which when struck With the pen of the syllable307 becomes a characteristic implement.308 From that generate yourself as the deity. |159| And while possessing divine pride Seal with the four mudrās309 And emanate the maṇḍala-cakra, [Then] please them. Accustom oneself to this through training. |160| [First Bindu Yoga: The Indestructible Bindu] Then, having cast away the outer body One should train in the ultimate suchness310 That is the buddhas’ supreme sphere of experience. At the heart center of311 the samayamudrā312 at one’s own heart,313 |161| Abiding within the symbolic implement there 305 Tib. stong bar bya; literally “made empty.” I have rendered the term less literally here, as the meaning is that the yogin is to see or perceives the world as empty, which is indeed its fundamental nature. Vaidyapāda comments that this means not to mentally engage with the appearance of the world as appearing separately (Sukusuma, 113b.4). Khenchen Chodrak Tenphel explains that the term “made empty” is used here to indicate that the practitioner is to “make” his perception of the world accord with the way the world actually is (personal commmunication, February 2016). 306 Again, the same wording of “making empty” is used here. 307 ge yi] sugg. em based on V (D and C) which read yi ge’i, ge’o D C P N S. 308 This is a tenative rendering of these two lines based on emending the text (see previous note) following Vaidyapāda’s commentary. Vaidyapāda writes, “Strike with the pen of the syllable means upon that seat, having completed the first ritual, striking with the pen of one’s own awareness, it is imagined as a syllable.” (yi ge’i smyu gus bsnun byas zhes pa ni gdan de la cho ga dang po rdzogs pas rang rig pa’i smyu gus bsnun nas yi ger brtags pa’o// (Sukusuma, D 114a.1; P 138a.3-4). 309 Vaidyapāda writes “The second point is to uphold the pride of oneself as the deity. The eyes and so forth are consecrated and then Kāyavajra and the rest are sealed with the four mudrās and the initiation is bestowed.” (bdag nyid rang gi lha’i nga rgyal ‘chang ba ni gnyis pa’o// de’i mig la sogs pa byin gyis brlabs nas de la sku rdo rje la sogs pa phyag rgya bzhis rgyas btab ste dbang bskur blang ba’o// (Sukusuma, D 114a.4 ; P 137a.7-8 ). 310 Vaidyapāda identifies the previous pāda as concluding the part of the text describing the practice of the generation stage, which forms the foundation for perfection stage practices. This line, he contends, begins the articulation of those perfection stage practices (Sukusuma, D 114b.2-3; P 137b.7-8). This section of the text corresponds with the very same perfection stage practices described—in a significantly abbreviated form—in the Muktibindu (D 49a.2-49b.7). 311 yi] sugg. em. following on V (D and P), yis D C P N S 312 Vaidyapāda glosses this samayamudrā as the jñānasattva, and indeed what is described here in Buddhajñānapāda’s text is exactly what is usually referred to as the jñānasattva (Sukusuma, D 114b.4; P 138a.1-2). Buddhajñānapāda himself uses the term jñānasattva to refer to the same in a later passage in the Dvitīyakrama (see Dvitīyakrama verses 248, 252, 261, and 262). 313 Vaidyapāda makes clear that this is all taking place within the framework of oneself visualized as the deity. “The samayamudrā at one’s own heart means the samayamudrā, that is the jñānasattva together with his seat, at the heart center of one’s own deity.” rang snying dam tshig phyag rgya’i zhes te rang gi lha’i snying gar dam tshig gi phyag rgya ye shes sems dpa’ gnas dang bcas pa’o// (Sukusuma, D 114b.4; P 138a.1-2). 404 Is the supreme wisdom of all tathāgatas. Which, for those practicing the first stage314 Appears clearly in the form of a syllable, |162| The seed syllable of one’s deity, indestructible, Blazing with five[-colored] light. One should emanate the five-colored light of that [syllable] 315 Out of the right [side] |163| Of the upper opening [i.e. the right nostril].316 [Emerging] from the tips of these [light rays]317 The tathāgatas, together with the maṇḍala-cakra, Fill the world and its ten directions. |164| Generate all sentient beings, who have emerged from the Breath of conceptuality,318 In319 the forms of buddhas: They are made to melt as the moon, dissolve, and are purified.320 |165| Then, bring them, as the essence of wisdom, Into the left nostril.321 Bring them into the seed syllable That abides within the center of the symbolic implement of [the samayamudrā].322 |166| 314 Vaidyapāda identifies the first stage here as the generation stage (Sukusuma, (D 114b.5; P 138a.3). 315 yi] sugg. em. following V (D and P), yis, D C P N S. 316 Vaidyapāda identifies the upper opening on the right side as the right nostril of the deity (Sukusuma, D 114b.6; P 138a.4). 317 rtse las] sugg. em following V (D and P), rtsa la D C P N S. Likewise, Vaidyapāda’s commentary on the Muktitilaka also makes it clear that the Tathāgatas are emanated from the tips of the light rays (Muktitilakavyākhyāna D 52b.5-6). 318 Vaidyapāda clarifies this term, “Sentient beings who emerge from the breath of conceptuality means all sentient beings who have been produced by inauthentic conceptuality riding on the horse of the winds.” des dbugs kyi kun rtog las byung ba’i sems can zhes pa ni yang dag pa ma yin pa’i kun du rtog pa rlung gi rta la zhon pas sems can thams cad bsgrub pa ste/ (Sukusuma, D 114b.7; P 138a.5-6). 319 skur] sugg. em. following V (D and P), sku D C P N S. 320 Vaidyapāda elaborates on this process: “They are generated in the form of buddhas. How is this done? They are made to melt as the moon, dissolve, and are purified, which means that the emanated maṇḍala, which has melted like the moon, dissolves into sentient beings and they become as above. That itself, as well, enters into oneself as the essence of wisdom means that they are gathered as the essence of wisdom which is pure like water and ushered into one’s left nostril.” de rnams sangs gyas kyi skur bskyed pa’o// de gang gis she na/ zla bar zhu byas te/ thim pas rnam par dag byas nas/ zhes (zhes] D, shes P) pa ni ‘phros pa’i dkyil ‘khor zla ba lta bur zhu ba sems can thams cad la thim pas de rnams gong ma lta bur gyur pa’o// de nyid kyang ye shes ngo bo ru (bo ru] P, bor D) rang la zhes te chu ltar dang (dang] P, dangs D) ba’i ye shes kyi ngo bor ‘dus zhing rang gi g.yon pa’i ha sar zhugs par bya zhes so// (Sukusuma, D 114b.7-115a.2; P 138a.6-8). A similar (but not identical) process involving the purification of sentient beings is found in the process of the generation of the causal deities in Buddhajñānapāda’s generation stage sādhana, the Caturaṅga/Samantabhadra-sādhana. 321 ha sa. It is clear from the earlier and later context in the commentary that this is meant to be a nostril. However, the Sanskrit word hasa normally means either laughter or is used to indicate the consonants of the Sanskrit alphabet. Khenchen Chodrak Tenphel suggests that here this term may be used to refer to the name of a particular channel in the subtle body, presumably one culminating in the nostrils (personal communication, March 2016). 322 The procedure described in verses 161-166 is also described (in a less elaborated form) in the Caturaṅga/Samantabhadra-sādhana in a single verse—verse 109—at the outset of the karmarājāgrī-samādhi. 405 This itself [becomes]323 the precious jewel That produces the qualities of all buddhas The self that pervades all things The great indestructible bindu. |167| Full of five-colored light, It is about the size of a chickpea. Within this [bindu], within one’s own mind, All phenomena are gathered324—contemplate thus. |168| Its light, bit by bit, Fills its own area. Then spills outward Illuminating the interior of the samayamudrā |169| This light then illuminates the outer body [of the samayamudrā] Which illuminates the maṇḍala and its area. Spilling outward [it illuminates] the sixteen forms of bodhicitta In one’s own interior: |170| At the base of the big toe[s], the syllable a; On the two calves, likewise, [the syllable] ā; On the two thighs, the form of I; On the secret place the form of ī; |171| The form of u is at the navel; Likewise on the abdomen, the form of ū; On the two breasts rests the form of ṛ; Likewise on the hand[s] the form of ṝ; |172| On the throat the form of ḷ; Likewise on the lower lip the letter ḹ; On the two cheeks the form of e; On the two eyes, as well, ai; |173| At the base of the ears the form of o; 323 Vaidyapāda clarifies that the seed syllable becomes the indestructible bindu. This also simply makes sense, since first there is a syllable there and then the practice is done with a bindu, rather than a syllable, so some sort of transformation from syllable to bindu must take place at some point. Vaidyapāda reads: “Once it has been made to enter inside in that way, it is brought into the seed syllable that abides in the center of the symbolic implement mentioned above; this sets forth the source and locus of the practice. By means of that [process], what does it become? This is expressed in the verse beginning, This itself…” nang gi la yang de bzhin du zhugs nas gong gi mtshan ma’i dbus na gnas pa’i sa bon la zhugs par bya ste ‘grub pa’i rgyu (rgyu] D, rgyud P) dang gnas bstan (bstan] P, brtan D) pa’o// des cir ‘gyur zhes na/ (des cir ‘gyur zhes na/] D, P om.) de nyid ces pa la sogs pa’o// (Sukusuma, D 115a.2; P 138a.8-b.1). 324 Vaidyapāda writes that both the mind and all things are gathered into the bindu. de (de] P, da D) nyid ‘od zer lnga dang ldan pa’i sran tsa na ka tsam du rang gi sems bsdus te der chos thams cad kyang bsdus nas rnam par bsam pa ni de’i tshad bstan pa’o// (Sukusuma, D 115a.3-4; P 138b.2). 406 On the crown there is the form of au; The forms of aṃ and aḥ on all the joints— These are completed at the time of the sixteenth.325 326 |174| 325 Verses 171-174 from the Dvitīyakrama are attested in Sanskrit in a citation (?) given in the single surviving manuscript of Kalyāṇavarman’s Catuṣpīṭhapañjikā. The verses appear as part of a larger passage, in which they are preceded by part of verse 111 from the Dvitīyakrama, and followed by some other verses that are not found in the Dvitīyakrama. The verses are attributed in the Catuṣpīṭhapañjikā to the Aṣṭāṣṭaka, which presumably is the title of a work, though this is not entirely clear. Thanks to Péter Szántó for pointing the parallel of verses 171-174 out to me and for sharing me with his diplomatic transcript of the verses vrom the Catuṣpīṭhapañjikā. akārāṅguṣṭhamūle ca ākārāñ caiva jaṅghayoḥ | ikāram ūruyugale | īkāraṃ guhyam āśṛtaṃ | ukāraṃ nābhimūle tu ūkāram udare tathā | ṛkāraṃ stanamūle tu ṝkāraṃ tu kare sthitaṃ | ḷkāraṃ tu gale caiva ḹkāra mūrdhayas tathā | ekāraṃ gaṇḍadeśe tu aikārañ caiva cakṣuṣī | okāraṃ karṇṇamūle tu aukāra mūrdhni saṃsthitaṃ | aṃ aḥ sarvāṅgike kāya | vidhānāni prayojayet |. 326 Vaidyapāda explains, “Also, one should know that this is with regard to the stages of the first day [of the month] and so forth. They are completed at the time of the sixteenth, means that at the time when the outer moon comes to fullness, these are also perfected. One must understand that this is then reversed. Regarding being perfected at the time of the sixteenth, the sixteen places that are stirred up through practice also become “the sixteen.” These then [become] the bindu and this becomes like the moon, which produces the consciousness of joy. The previous light rays hook, means that they hook the sixteen syllables and draw them into the bindu. By slightly holding one’s mind, like the first wisdom, there for a moment, what happens? [The text then says] Meditate with determination/ On the great [maṇḍala]-cakra of deities together with its support.” de yang tshes gcig la sogs pa’i rim par shes par bya’o// bcu drug dus su rdzogs gyur pa zhes pa ni phyi’i zla ba rdzogs par’i dus su de yang rdzogs pa’o// de ne bzlog ste shes par bya’o// bcu drug dus su rdzogs par ‘gyur ba ni sgrub (sgrub] P, bsgrub D) pas dkrugs pa bcu drug/ de (de] D, ste P) yang bcu drug par ‘gyur/ de yang thig le/ de yang zla ba lta bur song nas/ dga’i ba’i shes pa ‘byung ba’o// gong gi ‘od kyis rnam pa bkug ste zhes pa ni yi ge bcu drug po rnam par bkug nas thig le’i nang du bcug la der rang gi sems dang po’i ye shes ltar bag zhad bzung bas cir ‘gyur zhe na/ lha’i ‘khor lo che/ rten dang bcas pa mos pas bsgom/ (Sukusuma, 115a.6-115b.1; P 138b.6-139a.2). This is very similar to Vaidyapāda’s comments earlier in the Sukusuma: “Moreover, through practicing, by means of the agitation of the locations, the sixteen syllables appear, and these, then, become the sun and moon. Having transformed into a bindu like that, they go to the tip of the vajra. This itself, in a form which blazes with thousands of light rays, is meditated upon by the yogin in accordance with the ritual that will come below. When this happens, the suchness that has been spoken of will be realized, [and that is the] purpose [of this practice.]” de yang bsgrub pas gnas rnams dkrugs pa las yi ge rnams bcu drug par gyur/ de yang nyi zlar gyur/ de lta bu’i thig ler gyur nas rdo rje rtse mor ‘gro ba ste/ de nyid ‘od zer stong du ‘bar ba’i gzugs su rnal ‘byor pa rnams kyis ‘og nas ‘byung ba’i cho gas bsgoms nas/ ji skad du gsung pa’i de kho na nyid rtogs par ‘gyur pa’i phyir ro// (Sukusuma, D 88a.4-5; P 105b.6-8). Tāranātha, who reports having received initiation into and teachings on the Jñānapāda lineage from his master Buddhaguptanātha, in his much later instruction manual on the perfection stage rituals of the Jñānapāda School, gives an ever-so-slightly more clear presentation of this practice that does not seem substantially different from what is already here in Buddhajñānapāda’s text, with Vaidyapāda’s clarifications. He writes, “For the second part, the light from the bindu illuminates the jñānasattva, and from that light radiates forth and illuminates the interior of the foundational body. Like holding up a lamp in darkness, one sees clearly the sixteen bindus, which are the white substance.... [He lists here the syllables at all of the locations on the body, exactly as they are described in the Dvitīyakrama]...All of these are white and radiate white light. Think of them as being of the nature of bliss. The light from the heart center, either in stages or all at once, as one prefers, dissolves those syllables into the indestructible bindu at the heart center, and [it] then blazes with light and causes a strong increase in the essence of bliss. Contemplate thus.” (gnyis pa ni thig le’i ‘od kyis ye shes sems dpa’i sku gang / de las ‘od ‘phros gzhi lus kyi nang gsal zhing gang bar byas/ mun khung du sgron me bteg pa ltar dkar cha thig le bcu drug po rnams gsal bar mthong ba ni/ ...... thams cad kyang kha dog dkar po ‘od zer dkar po ‘phro ba/ bde ba’i rang bzhin can du bsam/ snying ga’i ‘od kyi yi ge de rnams rim pas sam cig car gang mos kyis snying ga’i mi shigs pa’i thig ler bstims pas/ ‘od zer ‘bar zhing/ bde ba’i ngo bo lhag par rgyas par bsam mo// (Dpal grol ba’i thig la’i khrid yig, 243-4). In both Vaidyapāda’s and Tāranātha’s descriptions, what is located at the sixteen places is sometimes described as syllables and other times as bindus. Likewise, in the Dvitīyakrama itself the bindu is first described as a syllable and then as a bindu without anything explicit describing its transformation from one state into the other. Tāranātha does not include any reference to the line “They become complete at the time of the sixteenth,” which is unfortunate because Vaidyapāda’s commentary here is still rather cryptic. Vaidyapāda’s comments about the phase of the moon in relation to these syllables in different parts of the body can, I believe, be understood more clearly with reference to the later kāmaśāstric doctrine of candrakalā in 407 The previous light rays hook them [i.e. the syllables] And draw them into the bindu. By holding the mind briefly there, Meditate with determination |175| Upon the great [maṇḍala]-cakra of the deities together with its support. At the heart center of the lord in the center of that [bindu]327 By means of the stages given above Meditate upon the great indestructible bindu |176| which Kāmadeva was understood to dwell in different parts of the body at different points in the moon’s phases (See Ali 2011, 47). As described in Kokkoka’s Ratirahasya and in Padmaśrī’s Nāgarasarvasva, this involves Kāma moving gradually through the left side of the body in the moon’s waxing phase, pervading the entire body for two days during the moon’s fullness, and traveling down the right side of the body during the waning phase (ibid.) A man is meant to stimulate these specific location on his lover’s body at particular days in the lunar calendar in order to please her, and the texts even prescribe the visualization of the vowels of the Sanskrit alphabet (i.e. precisely the syllables listed in the Dvitīyakrama), along with a candrabindu, at these various places on the body, on the appropriate dates (ibid., 47-48). While the doctrine of candrakalā described in these works is several centuries later than the Dvitīyakrama—the Ratirahasya is likely not earlier than the 10th century, and perhaps even later, and the Nāgarasarvasva dates to the12th century—and pertains to the genre of erotics rather than tantric practice, Ali has shown clearly that kāmaśāstra authors from this period, including Padmaśrī (who perhaps not incidentally was a Buddhist), were drawing on tantric Buddhist ideas in their writings (Ali 2011, esp. pp. 53-54). A similar practice is described also in *Surūpa’s Kāmaśāstra, which may be earlier than the Ratirahasya and the Nāgarasarvasva, but as I noted above, that work is difficult to date (See Vogel 1965, 24). In any case, we saw earlier that Buddhajñānapāda and Vaidyapāda both seem to show some familiarity with kāmaśāstra. The specific association with Kāmadeva is unlikely to be relevant here in the Dvitīykrama— While the practices described in the Dvitīyakrama with regard to these syllables are also presumably to be done in union with a partner, the syllables and bindus that are visualized, agitated, and drawn in with light rays are specified in the Dvitīyakrama as being present within the yogin’s own body, whereas the practice of candrakalā in a kāmaśāstric context seems always to pertain specifically to a woman’s body (see Desmond 2011, 26). However, the idea of syllables or bindus in the practitioner’s body becoming fully “perfected” at the time of the full moon (and perhaps otherwise individually “perfected” on the waxing or waning days of the moon) may be related to a more widely shared conception of specific areas of the body being associated with the progression of the lunar calendar. Indeed, such a conception appears to be a more broadly Indic idea, as it is also found in Indian medical traditions, where both the life force and the pulse are also said to travel through the body on specific days of the lunar calendar, and are likewise associated with vowels located at the same places on the body described in the Dvitīyakrama (Somānanda Dharmanātha, personal communication). Moreover, a passage on the syllables at the sixteen places found in Kalyāṇavarman’s Catuṣpīṭhapañjikā that is parallel with Dvitīyakrama verses 171-174 (see previous note for details) makes reference to the waxing and waning phases of the moon, and its context is clearly one of sexual yogic practice. While this work is also several centuries later than the Dvitīyakrama, it further confirms the connection between the syllables and locations described in the Dvitīyakrama with the practice of candrakalā in a sexual (and in this case also a yogic) context. 327 Vaidyapāda makes explicit that this is taking place within the lord at the center of the bindu, suggesting that one is to meditate upon the maṇḍala within the bindu and focus on the “great” indestructible bindu at the heart center of the lord of this maṇḍala. This is thus a further “nesting” of deities inside of deities—the bindu at the heart center of the samayamudrā (=jñānasattva) now holds its own maṇḍala, with yet another indestructible bindu within the symbolic implement within the samayamudrā at his heart. Here we may also note that the first bindu was described as the size of a chickpea while the second one is the size of a mustard seed. Vaidyapāda writes, “At the heart center of the lord in the center of that [bindu] means that in the center of the bindu is the samayasattva. Meditate with special determination on the cakra of the jñānasattva at his heart center following the stages given. Meditate on the great indestructible bindu in the symbolic implement at his heart center. How big is it? The size of a mustard seed.” de dbud bdag po thugs ka ru/ zhes pa ni thig le’i dbus su dam tshig sems dpa’i thugs kar ye shes sems dpa’i ‘khor lo gong ma’i rim pas lhag par mos pas bsgoms la/ de’i thugs kar mtshan ma la yang mi shigs pa’i thig le chen po bsgom par byas nas zhes’o// ci tsam zhes na/ yungs kar tsham/ (Sukusuma, D 115b.2-3; P 139a.2-3). 408 About the size of a mustard seed from which [light rays] emanate Illuminating its own area and the interior of the [samaya]mudrā [And its] maṇḍala.328 [Then those rays], illuminate its area and the interior of the lord.329 |177| This illuminates the outer body And the maṇḍala together with its basis.330 The light that resides in the sixteen places Spills outward |178| Illuminating the outer body.331 This [light] illuminates the maṇḍala-cakra Together with its basis.332 It then goes before all of the tathāgatas |179| Who reside in the ten-directional world. It melts into nectar and enters their mouths And takes up nectar from the bindu [at their heart centers], Which then emerges from the vajra path. |180| This comes like a stream of nectar [toward oneself] And the wisdom of method enters into [one’s] right [nostril]. Likewise, the other enters into the left [nostril]. 333 Then it dissolves into the wisdom bindu 328 Vaidyapāda identifies “its area” as “the subtle symbolic implement;” the mudrā as the “samayamudrā in which that [symbolic implement] abides;” and the interior which is illuminated as “the interior of its [the samayamudrā’s] body.” He explains that the maṇḍala is the “maṇḍala of that [samayamudrā],” i.e. a small maṇḍala within the larger of the two bindus, in which the samayamudrā is the central deity (Sukusuma, D 115b.3; P 139a.3-4). 329 Vaidyapāda identifies “its area” as the symbolic implement. This presumably is the larger symbolic implement which houses the larger of the two bindus. He explains that ‘the interior of the lord’ means “the other mudrā of the samaysasattva.” Presumably this is the larger samayasattva/mudrā at the center of oneself as deity (Sukusuma, D 115b.3-4; P 139a.4-5). 330 Vaidyapāda identifies the ‘outer body’ as the body “of the lord;” the ‘maṇḍala’ as “the eighteen,” presumably referring to the other 18 deities of the 19-deity maṇḍala; and ‘its basis’ as “the celestial palace and the dharmodaya” (Sukusuma, D 115b.4; P 139a-5-6). These two lines must be a brief description of what is described again in more detail in the subsequent lines, that is, the light spilling out of the body of oneself as the deity and onto the other eighteen deities of the maṇḍala-cakra as well as the “support maṇḍala” of the celestial palace, dharmodaya, etc. They have to be describing the same process as the subsequent lines, because otherwise we end up with more maṇḍalas than have been visualized—three instead of just two. 331 Vaidyapāda explains that this is the body of the “outer samayasattva,” i.e. oneself-as-deity (Sukusuma, D 115b.5; P 139a.6-7). 332 Vaidyapāda explains that this is the maṇḍala, meaning the other eighteen deities, of oneself-as-deity, and the basis is the support maṇḍala which consists of the palace, dharmodaya, and protection circle (Sukusuma, D 115b.5; P 139a.5-6). Since he did not mention the protection circle above in the description of the basis of the smaller maṇḍala inside the bindu, it seems that there is no second smaller protection circle visualized there. 333 Vaidyapāda explains that that which enters into the right nostril is the nectar emerging from Akṣobhya, and so forth, the male deities, while “the other” is that which originates from the Locanā and so forth, the female deities (Sukusuma, D 115b.7-116a.1; P 139b.1-2). Tāranātha likewise describes this process in the same way, but is more explicit, explaining that the substance emerges from the vajra of the male deities and the lotus of the female deities. (Dpal grol ba’i thig la’i khrid yig, 244). 409 Via the apparent bindu.334 |181| That [bindu] has the nature of dripping. Blazing with five[-colored] light, It is brilliant white, with a reddish tinge, Hold the mind stably within this.335 |182| From here,336 in accordance with the above stages, Correctly perform the emanation, Absorption, and holding.337 Through this eventually one will come to encounter338 |183| The mind as the vajra of cessation.339 334 Vaidyapāda explains: “The wisdom bindu is the first bindu. The apparent bindu is the one that appeared from that, which is suchness.” ye shes kyi thig le zhes pa ni dang po’i thig le’o// snang ba’i thigs pa (thigs pa] P, thig le D) zhes pa ni de las snang ba’i thig le ste chos nyid do// (Sukusuma, D 116a.1; P 139b.2-3). The wording of these two lines are a bit strange, suggesting that the transmission may be corrupted. I am tempted to emend as follows: de nas ye shes thig le las// snang ba’i’ thigs par (or thig ler) thim par ‘gyur//. It seems that this emendation could be made following Vaidyapāda’s commentary, but it’s not clear if Vaidyapāda is actually glossing the root text there or merely explaining. The translation, following both emendations, would be: “Then, it dissolves into the bindu that has appeared from the wisdom bindu.” For now, I will simply leave this here as an alternative way to read the line. Tāranātha, for what it’s worth, does not mention anything about two bindus here, but he also left out the nesting sequence above in which the second bindu was visualized. He simply states that the two flows of nectar dissolve into the “root bindu” (rtsa ba’i thigs las) (Dpal grol ba’i thig la’i khrid yig, 244). 335 Vaidyapāda clarifies that this is to be done “one-pointedly, and without [thought] proliferation, via the practice of the entering of the winds.” rtse gcig pa ma ‘phros pa dugs ‘jug pa’i spyod pas so// (Sukusuma, D 116a.3; P 139b.5). 336 las] sugg. em. following V (D and P), la D C P N S 337 Vaidyapāda explains that “From here, in accordance with the above stages” means starting from the final point (of the previous practice), without setting the practice aside for a moment, but continuing directly into the emanation (along with exhalation), which then draws up the nectar (from the hearts of the buddhas) as before. The yogin should then engage in drawing the nectar in (along with inhalation), and then holding the mind in the bindu as before (Sukusuma, D 116a.3-4; P 139b.5-7). 338 Tib. reg, literally “touch.” 339These lines are part of a verse adapted from the Guhyasamāja-tantra, VI, 41 (see note 341 below) and are also included in the Muktitilaka (D 49a.4-5). Vaidyapāda explains: “The mind will become the vajra of the cessation of all entities, and the signs of stability [in that] arise. That is to say, Eventually, one will come to encounter means at some time [one] will encounter the unchanging bindu by means of the path of [practicing with?] the goddess (i.e. the consort?), and at that time.... [the signs will authentically arise] of having become the glorious wish-fulfilling gem/ That contains all the great buddhas... ” ” sems dgnos po kun las ‘gog pa’i rdo rjer gyur nas brtan pa’i rtags skyes pa ste/ nam zhig de la reg gyur pa/ zhes pa ni dus nam zhig na mi ‘gyur ba’i thig le la lha mo (mo] P, mo’i D) lam gyis reg par gyur pa de’i tshe/ sangs rgyas kun gyi mchog ‘dzin pa // yid bzhin (bzhin] sugg. em based on Dvitīyakrama D C, and Guhyasamāja-tantra, VI, 41), sbyin V (D and C)) dpal dang ‘dra bar ‘gyur ba’ ‘o// (Sukusuma, D 116a. 4; P 139b.6-7). Vaidyapāda’s commentary on the parallel lines from the Muktitilaka reads: “The mind as the vajra of cessation indicates that [this takes place] by means of the mind [which engages in] the action of inhalation into(?) that bindu. Whoever comes to eventually encounter that, means one should not have any doubts that by means of encountering [it?] with that mind the signs will arise.” ‘gog pa’i rdo rje sems nyid du// zhes pa ni thig le de la(?) (la?] sugg. em., las D P) dbugs ‘jug pa’i spyod pa’i sems kyis so// nam zhig de la sus reg pa/ zhes pa ni sems des reg pas rtags rnams skye bar ‘gyur ba la som nyi mi bya’o// (Muktitilakavyākhyāna, 53a.4- 5). In any case, this seems to be the result of the first bindu yoga, the yoga of the indestructible bindu which is meditated upon at the heart. A different result—the arresting of the breath—is described below with bindu yoga performed with the “secret bindu” located at the tip of the vajra. 410 Thus signs will authentically arise340 Of the glorious wish-fulfilling gem, The great receptacle of all the buddhas.341 |184| Because through this ritual The deities of the aggregates, elements, and sense sources Are at the outset gathered into the heart by means of wisdom fire This is explained as the first.342 |185| [Second Bindu Yoga: The Secret Bindu] 343 There, through the power344 of practice,345 [The bindu] travels to the tip of the nose [i.e. the vajra].346 It is explained that one should meditate upon the secret bindu Immediately following the [meditation on] the indestructible bindu. |186| From the above-mentioned wisdom bindu347 Light rays in the form of hooks Emanate to the ten directions Hooking all the sugatas together with their maṇḍala-cakras. |187| In the form of the essence of wisdom They are drawn into one’s own heart center. Those348 melt and enter into the bindu. 340 The signs mentioned by Vaidyapāda are the same signs mentioned by Buddhajñānapāda himself later in the text as indications of the effectiveness of perfection stage practices: laughter, yawning, and trembling. (Sukusuma, D 116a.5-6; P 139b.1-2). 341 C.f. Guhyasamāja-tantra, VI, 41. nirodhavajragataṃcittaṃ yadā tasya prajāyate/ sa bhavec cintāmaṇiḥ śrīmān sarvabuddhāgradhārakaḥ/ (I have emended Matsunaga’s edition of the line from the tantra to follow the variant - dhārakaḥ found in two of his manuscripts, since that is the reading found here in the Dvitīyakrama (it is also, incidentally, the reading found in the Tibetan translation of the tantra). It is fortunate that these two lines are based upon a verse from the Guhyasamāja-tantra for which we have Sanskrit—which was very helpful in making sense of this verse from the Dvitīyakrama—as there appears to be some problem in the transmission of the lines in the Tibetan translations of Buddhajñānapāda’s works. The verse is found also in the Muktitilaka, but it is transmitted differently in each of the four places we find it—the Dvitīyakrama, Vaidypāda’s Sukusuma, the Muktitilaka, and Vaidyapāda’s Muktitilaka-vyākhyāna! 342 Vaidyapāda indicates that this refers to the ritual of the indestructible bindu that has just been explained. Because it has the function of drawing the deities of the aggregates and so forth into the heart, it is explained first (Sukusuma, D 116a.6-7; P 139b.3). 343 Vaidyapāda explains that this is the ritual for the training on the “secret bindu which has the nature of the wisdom of the intermediate joy” (Sukusuma, D 116a.7-116b.1; P140a.3-4). 344 Tib. byin rlabs, Skt. adhiṣṭhāna. 345 Vaidyapāda specifies that this is performed by the “moving wind that has the form of hūṃ phaṭ.” hūṃ phaṭ rnam pa dang ldan par gyo ba’i rlung gis so// (Sukusuma, D 116b.1; P 140a.4-5). 346 Vaidyapāda makes this clear: “It goes to the tip of the nose means the tip of one’s vajra, and it is held there.” sna yi rtse mor ‘gro bas na/ zhes pa ni rang gi rdo rje kha ste der ‘dzin pa’o// (Sukusuma, D 116a.1-2; P 140a.5). 347 Vaiyapāda specifies that “The wisdom bindu is the bindu at the center of the symbolic implement of the jñānasattva.” ye shes thig le zhes pa ni ye shes sems dpa’i mtshan ma’i dbus kyi thig le... (Sukusuma, D 116b.2; P 140a.6). 348 Vaidyapāda specifies that “Those melt means the locuses and so forth melt.” de rnams zhu nas zhes (zhes] P, zhen D) pa ni gnas la sogs pa zhu nas/ (Sukusuma, D 116b.3; P 140a.7-8). 411 Due to this, in the form of rajas, tamas, and sattva349 |188| The bindu also Descends from350 the lotus at the heart center And then abides at the center of the Jewel at one’s own vajra. |189| The supreme form of the five elements, The identity of the five wisdoms, Blazing with five[-colored] light— Visualize the form of the subtle symbolic implement of one’s own deity [thus].351 |190| In its center352 In the middle of a support353 the size of a mustard seed Meditate intently upon the great [maṇḍala-]cakra, together with a bindu.354 If while [regarding] that |191| One’s mind becomes dull or weary And it emerges from the vajra Make it remain at the tip of the nose355— And examine by means of the bliss of cessation.356 |192| 349 Vaidyapāda notes that the aspect of the bindus that “travels down the right [channel] is rajas, that which travels down the left [channel] is tamas, and that which travels down the central [channel] is sattva.” g.yas nas ‘gro ba rdul dang g.yon nas ‘gro ba mun pa dang/ dus nas ‘gro ba snying stobs kyi tshul du babs.... (Sukusuma, D 116b.3-4; P 140a.8). This use of the three guṇas from the Sāṃkhya system to describe the constituent aspects of the bindu is unusual, and is another example of Buddhajñānapāda’s use of non-Buddhist terminology, suggesting that he was indeed operating in an eclectic milieu. 350 las] sugg. em., la D C P N S. The reading of las also seems to have been transmitted in Tsongkhapa’s commentary on the five stages (see its translation in Kilty 2013,174-75). The passage makes much more sense this way. 351 It seems that the descriptions in this verse pertain to the visualized subtle symbolic implement that is to be visualized; it is unclear in both the root text and the commentary, however, precisely what the relationship is between this implement and the bindu described in the previous verse. 352 Tib. ‘bum pa 353 Vaidyapāda identifies this as the “support maṇḍala” (rten gyi dkyil ‘khor) (Sukusuma, D 116b.6; P 140b.3 354 thig le] sugg. em. following V (D and P), thig ler D C P N S 355 While looking only at the root verse, one might presume that the “nose” referred to here is (as is often the case) the tip of the yogin’s penis. Vaidyapāda, however, specifies that one is to “make it remain at the tip of the nose of the goddess’ lotus.” lha mo’i padma’i sna rtser rnam par gnas par byas nas/ (Sukusuma, D 116b.6-7; P 140b.4). A similar passage in the Samantabhadra/Caturaṅga-sādhana is described in Vaidyapāda’s and Samantabhadra’s commentaries as indicating a process by means of which the yogin appears to be instructed to draw the bindu which was previously emitted into the lotus of the consort out onto the “nose tip” of her lotus by means of transforming the “prong” of his vajra into hook-like light rays that hook the bindu and draw it out to this location. (See Samantabhadra/Caturaṅga-sādhana verse 130 (D 34b.3), Vaidyapāda’s Samantabhadrī-ṭīkā (D168a.2-5), and Samantabhadra’s Sāramañjarī (D 38a.1-3; Szántó unpublished 125). 356 These lines describing the procedure for when the drop emerges from the vajra appear to be something of an aside, as they occur right in the middle of the description of the yoga itself, which continues below with the instructions on emanating, absorbing, and holding, which is performed in the same way with the bindu at the tip of the vajra as it was performed with the bindu at the heart. The less detailed instructions in the Muktitilaka lack the instruction on what to do if the mind becomes tired and the bindu emerges from the vajra. This verse in the Dvitīyakrama is one of several points in Buddhajñānapāda’s system where we see an association of the “bliss of cessation” with emission. I address this point briefly in Chapter Six. 412 In this bindu yoga Following the stages described above, the five[-colored] light Emerges from the upper door And again draws in the nectar and |193| Likewise brings it into the foundation357 Stopping the breath—[this is] the branch of emptying.358 Then [within] the very subtle Great secret bindu, just like before |194| 357 Vaidyapāda clarifies that this foundation is the bindu (Sukusuma, D 117a.2). 358 There appears to be some confusion regarding the name of this branch in Buddhajñānapāda’s system. Vaidyapāda’s Sukusuma reads “the branch of casting out” (gtong pa’i yan lag) (as does Vaidyapāda’s Samantabhadrī-ṭīkā, which also mentions the practice), but Tāranātha’s later text follows Buddhajñānapāda’s root text in reading stong pa’i yan lag for this practice—actually he calls it the “branch of emptying which stops the breath” (dbugs dgag stong pa’i yan lag), which clearly follows the line in the Dvitīyakrama itself that mentions this branch (Dpal grol ba’i thig le’i khrid yig, 247). Vaidyapāda here writes, “The bindu is the foundation. Abiding there is called stopping the breath, the branch of casting away (or “emptying” if one uses the term from the root text). This is because movement, abiding, emanating, and absorbing are stopped and cast aside, and thus [the winds?] are united.” thig le ni gzhi ste der bzhugs pa ni dbugs dgag gtong gi yan lag go zhes te/ rgyu dang gnas dang spro bsdu dgag pa dang gtong bas bsdus pa’i phyir ro// (Sukusuma, D 117a.2-3; P140b.8-141a.1). This is the first of three branches (yan lag, aṅga) that are mentioned in this section of the Dvitīyakrama, the other two of which are clearly two among the classical tantric Buddhist version of the ṣadaṅga (praytāhāra, dhyāna, prāṇāyāma, dhāraṇā, anusmṛti, and samādhi), the Buddhist locus classicus of which seems to be the Samājottara, verse 141. The fact that Buddhajñānapāda refers to some of these six aṅgas but not all of them and also includes a branch that is at least not referred to by the name that is used for that branch in the classical sixfold list—is further suggestive of the fact that he did not know the Samājottara. Even in the later presentation by Tāranātha of these practices only these same three branches—named just as they are in the Dvitīyakrama—are mentioned, as Tāranātha appears to be presenting Buddhajñānapāda’s system quite faithfully. However, while he keeps to the name used in Buddhajñānapāda’s text, Tāranātha does correlate this branch of emptiness with the branch of prāṇayāma, which is one of the classical six branches (Dpal grol ba’i thig le’i khrid yig, 247). Indeed, the connection of this gtong pa’i yan lag or dgag pa’i yan lag with prāṇayāma, one among the classical six branches, is also suggested by Vaidyapāda’s Samantabhadrī-ṭīkā, his commentary on Buddhajñānapāda’s Caturaṅga-sādhana, where he mentions that the practice of the sūkṣma yoga according to that sādhana constitutes “the branch of emptying, or cessation, the third” (dgag pa dang gtong ba’i yan lag gsum pa’o) (Samantabhadrī-ṭīkā, D 168a.6). Prāṇayāma is the third of the six yogas in the traditional list as given in the Samājottara. In any case, Tāranātha’s later description of the practice here is clarifying, “Just like before, meditate upon the bindu at the jewel. The light from the bindu, beginning with the handle of the sword, illuminates the inner and outer maṇḍala of the bindu in two stages and the foundational body together with its outer maṇḍala. Together with the drawing back in of the light the outer and inner maṇḍalas gradually dissapear, and in the end one holds the mind only on the bindu. One should bring this to a halt with the forceful vase breath. This is a sublime method for bringing about the cessation of impure perceptions and gathering the ordinary mind and winds into the central channel. Regarding [the practice] in the aspect of nonconceptuality,* by practicing in union with a karmamudrā, through holding the bindu unmoving at the jewel, the earth, wind, and so forth are forcefully brought to a halt. Up until here is the branch of prāṇayāma.” snga ma bzhin du nor bur thig le bsgoms/ thig le’i ‘od kyis ral gri’i yu ba nas brtsams te/ nor bu’i thig le’i phyi nang gi dkyil ‘khor rim pa gnyis dang/ gzhi lus phyi’i dkyil ‘khor dang bcas pa gsal bar byas/ ‘od zer tshur bsdu ba dang lhan cig phyi nang gi dkyil ‘khor rnams rim gyis mtha’ nas yal te/ mthar thig le gcig pu la sems gzung/ rlung bum pa can drag tu dgag par byas ste/ ‘di ma dag pa’i snang ba ‘gag cing tha mal gyi rlung sems rnams dbu mar sdud pa’i thabs dam pa’o// rtog med kyi phyogs la ni las rgya dang mnyam par sbyor bas nor bur thig le g.yo med du bcings pas sa rlung sogs ches ‘gags pa’o// de yan chad srog rtsol gyi yan lag go// (Dpal grol ba’i thig le’i khrid yig, 247). *Tāranātha specifies each of the three branches in terms of a conceptual and a nonconceptual practice. Presumably for the “conceptual” aspect of the practice of the branch of emptiness one would not be practicing with an actual partner, since the practice done in union with a karmamudrā is specified as the nonconceptual aspect. This seems to be the case for the other branches in which such a distinction is made, as well. 413 Hold your mind very gently The state of entitylessness ensues.359 |195| Through training in retention,360 as well, The mistaken earth is withdrawn.361 This experience of [things] being like a mirage— Know that this is the first sign. 362 |196| Likewise, when water is withdrawn There is the experience of something which appears like smoke— Know that this is the second sign. Due to the withdrawing of fire |197| One experiences something like a bright sky—the third [sign].363 When the wind likewise is withdrawn There is an appearance like a lamp— Know this to be the fourth sign. |198| Likewise, when the mistaken consciousness Is withdrawn, something similar364 to the profound, luminous Nondual state [appears], clear like the cloudless sky— This is the fifth sign.365 |199| 359 Vaidyapāda explains that this does not mean that entities are empty because they are destroyed or overcome. Rather, by means of the yoga of lacking nature (one of the seven yogas mentioned by Vaidyapāda in his Yogasapta treatise on the fourth initiation) one turns away from other mental states, and since one therefore remains only in suchness, the state of entitylessness ensues. (Sukusuma, D 117a. 4). On the seven yogas see note 121. 360 gzung ba. Vaidyapāda specifies that this refers to the “branch of retention” (gzung ba’i yan lag), which is one of the six-branch yogas, dhāraṇā (although this branch is normally rendered into Tibetan as ‘dzin pa). Two of these six yogas are mentioned in this section of the Dvitīyakrama, along with a third “branch”—the branch of “emptying,” which appears to be an alternate name for what is usually termed prāṇayāma, the third among the classical set of six (on this branch see note 358), whose Buddhist locus classicus is the Samājottara. See also verse 210. 361 Here Vaidyapāda explains that earth and the rest of the elements have two aspects—they can either appear in a mistaken way or appear to wisdom. He seems to understand the ‘withdrawal’ of these elements here as referring to the reversal of their mistaken perception (Sukusuma, D 117a.5). 362 Here Vaidyapāda explains that this refers to “the light rays of that bindu [appearing] clearly and [then] unclearly, like the example of a mirage. In this context one should understand it like this. One should then think, ‘This mistaken experience of earth in my mindstream which ought to be withdrawn—I know [this to be] the sign [of that withdrawal taking place],’ and in this way feel a sense of encouragement and conviction. Apply this [attitude] to all of the other [signs] also.” thig le de’i ‘od zer gsal ba dang/ mi gsal ba’i dpe smig rgyu dang ‘dra ba ‘byung ste/ de’i skabs su ‘di ltar shes bar bya’o// bdag gi rgyud la gnas (gnas] D, gnang P) pa’i phyin ci log tu snang ba’i sa ‘di ni rnam par ldog tu rung ba ste/ bdag gis (gis] D, gi P) ni rtags shes so snyam du gzengs bstod la yang de la zhen par bya’o// gzhan rnams la yang de bzhin tdu sbyar bar by’o// (Sukusuma, D 117a.7-118a.1; P 141a.7-141b.1). 363 Here Vaidyapāda’s commentary describes this “bright sky” as referring to the more commonly mentioned third sign—fireflies (Sukusuma, D 117b.1-2). 364 ‘dra] sugg. em. based on V (D and C), dran D C P N S 365 Vaidyapāda notes that these signs are visible not only to oneself, but to others. “The previously mentioned light rays themselves appear in the form of the deity or of signs, and so forth, and are perceptible by my sense faculties and within the sphere of experience of others, as well. They are not [exclusively experienced] from one’s own perspective.” gong gi ‘od zer nyid lha ‘am mtshan ma la sogs pa’i gzugs su snang zhing bdag gi dbang po dang gzhan gyi yang spyod yul du snang bar ‘gyur ba’o// rang dngos ni ma yin no// (Sukusuma, D 117b.3-4; P 141b.4-5). 414 Since by means of these five signs One will attain non-abiding nirvāṇa,366 A yogin should strive in this— This is the branch of retention.367 |200| In this way having made the five signs appear To make this reality368 pervasive, Emanate [it] from the vajra path Into the realm of space. |201| From those emanations, those appearances Which were previously369 seen should be generated. The recollection of the buddha And the authentic recollection of the dharma |202| The recollection of vajra, The genuine recollection of family, The recollection of the wrathful ones, Vairocana, Amitabha, |203| Akṣobhya, Ratnasambhava, and the rest, Yamāntaka and the others—the maṇḍala forms370 Of the wrathful ones, as many as there are— 366 Vaidyapāda here correlates several of the signs with the stages of practice—the second through the fourth of the signs mentioned in the text are, he says, signs of attainment for the yogin practicing sevā, and the “first” and “second” levels of sādhana—presumably sādhana and mahāsādhana (Sukusuma, D 117b.5). 367 Retention is one among the ṣaḍaṅga yoga, dhāraṇā (normally rendered in Tibetan as ‘dzin pa, though this does not present so much of a problem given that gzung ba and ‘dzin pa are simply different tenses—future and present, respectively—of the same verb). But below in verse 210 Buddhajñānapāda specifically mentions only three branches: emptying (stong pa), retention (gzung ba), and recollection (‘dran pa). With respect to this particular branch of retention Vaidyapāda says “The text says This is the branch of retention, because this is the branch where one attains the signs that arise due to holding (“retaining”) one’s life force like a precious gem.” gzung ba yi (ba yi] D, ba’i P) ni yan lag go// zhes pa ni rang gi srog rin po che lta bu gzung bas rtags rnyed pa’i yan lag tu ‘gyur ba’i phyir ro// (Sukusuma, D 117b.6; P 141b.8). Tāranātha retains the spelling of gzung ba’i yan lag—he calls this branch “the branch of retention where the signs appear” (rtags snang gzung ba’i yan lag)—when listing the branches from the Jñānapāda School practices, but later refers to the same practice using the more common Tibetan translation ‘dzin pa’i yan lag. (Dpal grol ba’i thig le’i khrid yig, 247-48). Tāranātha further clarifies that it is through the practice of the previous branch that the five signs begin to appear and slowly stabilize, but when they have become stable and one focuses one-pointedly on their appearance, this is the branch of retention (Dpal grol ba’i thig le’i khrid yig, 247-48). 368 Vaidyapāda identifies “this reality” (don de) as the deity, symbolic implements, and the rest (lha dang phyag mtshan la sogs pa) (Sukusuma, D 117b.7; P 142a.1). Likewise, earlier he wrote of “light rays appearing in the form of the deity, signs, and so forth,” (‘od zer nyid lha am mtshan ma la sogs pa’i gzugs su snang) (Sukusuma, D 117b.3-4; P 141b.4). 369 sngon] sugg. em. based on V (D and P), mngon D C P N S. 370 Tib. sku 415 Visualize the activity371 of their emanation and absorption. 372 |204| Likewise, regarding the recollection of the samayas,373 They are passion and the four-fold ritual— First visualize the activity of emanation and absorption. Regarding the recollection of the maṇḍala |205| Clearly perform the activity of emanating and absorbing The maṇḍala of the second ritual and the rest As for the recollection of body, Speech, and mind, |206| And the recollection of sentient beings, Having emanated and absorbed the three vajras, Which have been cultivated as awakened body, speech, and mind Aspire to374 and recall these. |207| Likewise, regarding the recollection of the General form of all mantras: By means of the five wisdoms Generate Vajrasattva, |208| And encounter the clarity 375 Of your own deity endowed with the four enjoyments, [through] the activity of emantion and absorption. Regarding the recollection of samaya: Perform the tasting of nectar, and so on, |209| [And] the activity of emanation and absorption clearly. As for the recollection of the prajñāpāramitā 371 I am not completely sure of the way the term las is being used in this section. I am translating it now as “activity,” understood as an appositive of “emanating and absorbing.” Vaidyapāda does seem to read it this way, as well, but this is not completely clear. It does seem clear that he does not take las as a third member of the list, however. The Tibetan translators or redactors of the Tibetan canon also seem to have had some issue with the term, as a number of places in several of the various recensions of the canon read la rather than las in some but not all instances. The majority of instances across the root text and the commentary, however, read las, so I have kept it in all instances. 372 Vaidyapāda identifies recollection of the buddha as the emanation, absorption, and activity of Vairocana; the recollection of the dharma as the emanation, absorption, and activity of Amitabha; the recollection of vajra as that of Akṣobhya; the recollection of family as that of Ratnasambhava, Amoghasiddhi, Locanā and the rest; and the recollection of the wrathful ones as that of Yamāntaka and the other three (Sukusuma, D 118a.2-3; P 142a.4-6). 373 For this and each of the subsequent recollections, Vaidyapāda introduces the section by writing, “Regarding the recollection of samaya, what should one recall with respect to the reality that was previously seen?” dam tshig rjes su dran pa ni smgon mthong ba’i don la gang dran zhe na/ (Sukusuma, D 118a.4; P 142a.7). And later Vaidyapāda clarifies that “it is called the branch of recollection because of recalling [something] with respect to the reality that one has previously seen.” de rnams ni sngon mthong ba’i don la rjes su dran pas na rjes su dran pa’i yan lag go// (Sukusuma, D 118b.6; P 143a.3-4). This suggests that what is meant in this section on recollection is to bring to mind different aspects of the reality that one has previously experienced, which suggests that the various deities of the maṇḍala and their qualities are understood here as constituting different expressions of suchness itself. 374 smon] P N V (D and P), smin D C S 375 Tib. rjes su gsal ba; Skt. *anuspaṣṭa(?) = noticed, clearly perceived. 416 And non-arising: Those emanated maṇḍalas, as well, |210| Do not exist in any way— Recalling their nondual essence Perform the activity of emanation and absorption. Regarding the yoga of recollecting |211| The pūjā of the family of anger and so forth, The supreme girl of one’s own [buddha-] family Purified by the intermediate four enjoyments376 Should be pleased by means of passion— |212| This activity of emanation and absorption should be performed; This is the branch of recollection.377 In this way, by means of these three branches, Meditate upon the secret bindu. |213| [Third Bindu Yoga: The Emanated Bindu (=Vajrajapa)] 378Even when it emerges from the jewel379 It is made to pervade the three realms— This is called meditation upon the emanated bindu.380 376 The “four enjoyments” here are the four stages of sevā etc. that constitute the ritual procedures for the generation stage practice. The “intermediate ones” are those four stages in Buddhajñānapāda’s generation stage sādhana that relate to the consort. The “lesser ones” are the four stages as performed in relation to oneself as the deity, and the “greater” ones are the same four stages performed for the maṇḍala deities. Tāranātha here notes that this practice could be performed either with an actual or a wisdom (i.e. visualized) partner (Dpal grol ba’i thig le’i khrid yig, 249). 377 This is one among the ṣaḍanga yoga, anusmṛti. As noted above, Vaidyapāda clarifies that “it is called the branch of recollection because of recalling [something] with respect to the reality that one has previously seen.” de rnams ni sngon mthong ba’i don la rjes su dran pas na rjes su dran pa’i yan lag go// (Sukusuma, D 118b.6; P143a.3-4). Tāranātha calls this “the branch of recollection in which the deity appears” (lha snang rjes dran gyi yan lag) and gives a rather detailed description of the practice that elaborates somewhat significantly on Buddhajñānapāda’s text (Dpal grol ba’i thig le’i khrid yig, 247-51). In particular, in the practice of recollection, Tāranātha often gives instruction to follow the procedures described in “the sādhana,” which clearly refers to passages in Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana, which is generally understood to constitute generation stage practice. This is perhaps yet another an indication of the fact of the generation and perfection stage practices not being completely separate in Buddhajñānapāda’s system. These practices of recollection, Tāranātha notes, constitute “the perfection stage being sealed with the generation stage,” which is an unusual reversal of the more common phrase “the generation stages sealed with the perfection stage” at least within the later Tibetan tradition of sādhana practice within the Nyingma School. Tāranātha further clarifies that when these appearances of the deity actually appear, they exclusively constitute the appearance of the perfection stage deity (Dpal grol ba’i thig le’i khrid yig, 249). 378 Vaidyapāda specifies that the emanated bindu has the nature of the wisdom of the “bliss of cessation.” da ni dga’ bral gyi ye shes kyi rang bzhin sprul pa’i thig le gsungs pa/ (Sukusuma, D 118b.6-7; P 143a.4) 379 i.e. the head of the penis 380 Vaidyapāda clarifies that it is what follows in the Dvitīyakrama that is the procedure for meditating on the “emanated bindu” (Sukusuma, D 119a.2; P 143a.7-8). He also seems to link this practice both to the context of initiation and to post-initiatory practice. He writes: “Then, having performed the meditation on the secret bindu, now, in order to indicate the purpose of the meditation on the emanated bindu [the text] states, Even when it emerges from the jewel/ It is made to pervade the three realms. Thus, after the conclusion of the initiation, due to abiding in the branch of increase, [there is] the emanated [maṇḍala]-cakra, the meditation upon the nirmāṇakāya; 417 At the center of the crown of your head |214| At the heart center of Kāyavajra381 Resting in a symbolic implement is a wind maṇḍala. [The color] of smoke. Here, upon a moon disc, Imagine a white syllable oṃ that symbolizes coming. |215| Likewise at the center of one’s throat At the heart center of Vacvajra382 At the center of the symbolic implement Is a white water maṇḍala. At the moon disc at its center |216| Meditate upon a red syllable āḥ with the nature of abiding. At one’s heart center, in the heart center of Cittavajra, Resting in the center of the symbolic implement Is a red fire maṇḍala. On this is a moon disc |217| At its center meditate upon the black syllable hūṃ Which has the nature of going. Also between the two breasts At the heart center of the samayamudrā |218| On the symbolic implement rests the Yellow maṇḍala of Iśvāra.383 On a moon disc there Is the great seed of your own deity. Think that this is the seed of liberation From arising, engaging, and abiding.384 |219| 385 In this way, having joined the four great mudrās Together with the [four] identities,386 By means of the [following] procedure or, alternatively, because of emanating the four bindus and the rest [it is called the practice of] the emanated [bindu]. This meditation is [now] explained.” De gsang ba’i thig le bsgom par byas nas/ sprul pa’i thig le bsgom pa’i dgos pa gsungs pa/ nor bu las ni byung nas kyang// khyams kun khyab par byed pas ni// zhes te dbang gi mthar thugs rjes rgyas pa’i yan lag tu gnas pas sprul pa’i ‘khor lo ste (‘khor lo ste] D, P om.) sprul pa’i sku sgom par byed pa’am yang na thig le bzhi la sogs par spros pa’i phyir na sprul pa ste de bsgom pa (bsgom pa] P, bsgoms D) bshad do zhes so// (Sukusuma, D 118b.7-119a.2; P 143a.8-143b.2). 381 According to Vaidyapāda, Kāyavajra (another name for Vairocana) has been generated from the seed syllable placed in this part of the body during the generation stage practice, which was first transformed into a symbolic implement before being transformed into the form of Kāyavajra. (sku yi rdo rje thugs ka ru/ zhes pa ni bskyed pa’i rim pa’i skabs su gnas bskyed nas/ sa bon mtshan ma phra mo las gyur pa’i sku’i rdo rje bsams te de’i thugs ka ru’o// (Sukusuma, D 119a.2-3; P 143a.8-143b.1). 382 i.e. Amitabha 383 i.e. the earth maṇḍala 384 After this section Vaidyapāda notes that “The four maṇḍalas which arise in this way are [the practice of?] the perfection [stage?] yogin.” de ltar ‘byung ba’i dkyil ‘khor bzhi ni rdzogs pa’i rnal ‘byor pa’o (Sukusuma, D 119a.4- 5; P 143b.3). 385 The next few verses are translated into French in Tomabechi (2006, 66-7). 386 Vaidyapāda omits the mention of mudrās and simply writes, “Having thus joined together with the four identities...” de bas na bdag nyid bzhi dang ‘brel par byas nas... (Sukusuma, D 119a.b 5; P 143b.4). 418 Perform the essence of the indestructible vajra recitation: |220| From the right [nostril]387 emerges the great maṇḍala Of the element of wind, Smoky-colored and with the activity of moving. By means of the essence of hūṃ, emergence,388 |221| It is made to pervade all the phenomena of appearance and existence. It purifies all phenomena and transforms them into nectar. By means of the essence of oṃ it is made to return And cleanses the habitual patterns in one’s own mindstream, |222| Purifying them, and then enters into that [oṃ].389 By means of the essence of āḥ it is made to remain. Then, the suchness of the deity, Is gently, gently held. |223| This brings about the state of entitylessness. [Then] from the left [nostril] emerges the great maṇḍala Of the element of water White in color and perfectly remaining |224| From that the hūṃ of emergence, The oṃ whose nature is gathering, The āḥ whose nature is remaining. And one holds suchness, just as before. |225| [Then] from both [nostrils] emerges forcefuly The maṇḍala of fire itself It is red in color. From its essence Hūṃ emanates and oṃ draws in |226| Āḥ causes [it] to remain, and suchness is held— Understand that it is just as before. Likewise, from both [nostrils] slowly emerges The causal maṇḍala of Maheśvara,390 |227| 387 Vaidyapāda clarifies that this is the right nostril. rang gi steng sgo’i bug pa gnyis kyi g.yas so// (Sukusuma, D 109a.6; P 143b.6). 388 The last two lines of this verse are a bit problematic. Reading them in a natural way following Tibetan grammar, the las in the penultimate line of the verse seems to be an ablative. However, Vaidyapāda—though his comments are also slightly unclear—seems to read this as a noun, karma, rather than a particle, providing a reading of the passage that makes more sense, though it is at odds with a more natural grammatical reading of the Tibetan translation (Sukusuma, D 110a.7; P 143b.7). The final line also appears to be preserved in a different version in the Derge recension of Vaidyapāda’s commentary, where it reads ‘byung ba rlung gi ngo bo yis// rather than ‘byung ba hūṃ gi ngo bo yis// (Sukusuma, D.110a.7). The Peking recension of Vaidyapāda’s commentary accords with the root text (Sukusuma, P 143b.7). 389 Vaidyapāda makes this clear: oṃ de la zhugs pa (Sukusuma, D 109b.3; P 144a.2). Tāranātha says it enters back into the wind maṇḍala (Dpal grol ba’i thig le’i khrid yig, 253). 390 i.e. the earth maṇḍala 419 Gold colored and perfectly apparent. Then, by means of hūm it emerges Oṃ draws it back in, and āḥ makes it abide, Then suchness is held there. |228| Have no doubts or hesitations That you will encounter a wisdom That is free from drawing in, abiding, and letting go. |229| This recitation combined with the four mudrās391 [Is done] two hundred and twenty-five times That, when multiplied [by four], is nine-hundred [recitations] In one day, because of the twenty-four— |230| The great lord of yogins Always [performs] 21,600 recitations392 Day and night. |231| By means of this, he will know all appearing phenomena To be [like] an illusion, a mirage, An echo, a spinning firebrand, A delusion, a city of gandharvas, |232| Bubbles in the water, an optical illusion, A reflection, the moon in the water, and so on, And will share the fortune Of the lords of the tenth bhūmi.393 |233| Therefore, the yogin394 should put effort Towards this natural recitation. Although this has been the genuine nature, Since beginningless time, [even] if one constantly recites395 |234| 391 Vaidyapāda explains, “Regarding the four mudrās, the [recitation] together with the wind support is the karmamudrā; with the water support it is the dharmamudrā; with the fire support it is the mahāmudrā; with the earth support it is the samayamudrā.” phyag rgya bzhi ni rlung rten dang bas pa ni las kyi phyag rgya’o// chu rten dang bcas pa nichos kyi phyag rgya’o// me rten dang bcas pa ni phyag rgya chen po’o// sa rten dang bcas pa ni dam tshig gi phyag rgya’o// (Sukusuma, D 110a.2; P 144b.3-4). 392 This passage on the number of recitations connected with the four mudrās is parallel to Muktitilaka, D 49b. 393 Here we see an instance of Buddhajñānapāda homologizing the results of the tantric path with those of the sūtric one, in a move that seems directed towards legitimizing these tantric practices. 394 Interestingly, given that this seems to be part of the perfection stage sādhana practice, Vaidyapāda specifies that this is something that “beginners, and so forth” should train in (Sukusuma, 120b.5). 395 Here I follow Vaidyapāda’s commentary with respect to the phrase rtag zlo. He says, “Because since beginningless time all sentient beings have arisen together with wind, they remain in the vajra recitation. But without being accepted by a teacher one will not realize this.” thog ma med pa’i dus nas sems can thams cad kyang rlung dang lhan cig tu byung bas na rdo rje bzlas pa la gnas kyang bla mas ma zin pas rtogs par mi ‘gyur ro// (Sukusuma, D 120b.5-6; P 145b.2-3). 420 Without relying upon a genuine teacher One will not realize this truth.396 Knowing this correctly Abandon the obstacle to meditation, External recitation. |235| The great supreme essence of recitation Is the inexpresssible awakened body of Vajradhara Which completely transcends thought and expression— How could this be recited by speech? |236| Thus, having found an elephant What need does one have of his footprints? |237| [Yet] for a great yogin who abides in that reality It is not a contradiction for him to rely upon outer [recitation]. This is like the [effortless] aquisition of firewood (i.e. elephant dung) Due to the [mere] power of the elephant[’s presence].397 |238| Therefore, due to striving in the supreme vajra recitation, When the sleep398 that is Blessed by the tathāgatas occurs The signs of accomplishing the mahāmudrā Will appear again and again.399 |239| Since it is doubtless and certain that Remaining in the [practice of] the ritual that brings that about For [just] a single moment will transfer Great onmicience into400 one’s mindstream, Strive in the practice of this ritual. |240| 396 don 397 The sense behind these lines on their own is rather unclear, but it seems, based on Vaidyapāda’s commentary together with Khenchen Chodrak Tenphel’s explanation of it, that the meaning is that the excrement of the excellent elephant is able to turn ordinary gold into the superior Jambu-river-gold, when that ordinary gold is placed in the elephant’s excrement and the excrement is then burned in a fire, i.e. used for firewood. And once one has found the elephant, the firewood one seeks—i.e. his excrement—is also naturally found. Like that, even when one knows the essence of the vajra recitation, doing exernal recitation still perfects the accumulation of merit, and therefore remains powerful (Sukusuma, D 120b.7-121a.1; P 145b.5-6.; Khenchen Chodrak Tenphel, personal communication, February 2016). 398 gnyid] sugg. em. based on V (D and P), nyid D C P N S. While all recensions of the Tibetan translations of Buddhajñānapāda’s text here read nyid, Vaidyapāda’s commentary makes it very clear that the Sanskrit text read a word which would, in Tibetan, be rendered as gnyid, sleep. “Of course, the yogin who strives in the vajra recitiation abides within spontaneity in which there is no lying down, resting, repose, or sleep. However, when, a little bit of sleep that is blessed by the sugatas takes place, at that time [the signs of accomplishing] the mahāmudrā...” rdo rje bzlas pa la ‘bad pa’i rnal ‘byor pa lhun gyis grub pa nyid du gnas pa nyal ba dang snyes pa dang ‘phres pa dang gnyid la sogs pa med mos kyi/ bde bar gshegs pas byin gyis brlabs pa’i gnyid cung zad ‘ong (‘ong] , ‘od D) pas na de’i dus su phyag rgya chen po zhes te/ (Sukusuma, D 121a.2-3; P 145b.7-8). 399 Vaidyapāda specifies that these signs refer to auspicious dreams of the deity, and so forth (bkra shis pa’i lha’i rmi (rmi] P, smon D) lam la sogs pa) (Sukusuma, D 121a.3; P 146a.1). 400 la] sugg. em. based on V (D and P), las D C P N S 421 This was the authentic teaching of the ritual Of meditating on the three bindus That correspond with the three joys. |241| 401 Although nondual wisdom itself Takes on a relative form402 Even when the inanimate, and so forth, along with the animate, Brahmā and the others, the gods, asuras, and the rest |242| Completely disappear, That bindu will not cease;403 Everything animate and inanimate404 Will again be made to emerge from it. |243| However, because they do not realize What is genuine, Beings are confused, [believing everything] to be arisen from Brahmā’s egg. But that [bindu], which cannot be moved by any phenomena, |244| Which cannot be destroyed by anything at all, As long as it remains embodied Brings about [engagement in virtuous] activity and non[-virtuous] acts.405 Therefore406 the meditation on the indestructible bindu, Stable and beyond destruction, is explained. |245| [Perfection Stage Practices with Detail on Dissolution Process] The four brahmacaryas,407 Seeing408 the outer [world] as empty, and so forth,409 And having generated oneself as the deity, One should also seal with the four mudrās— |246| Understand these procedures, just as [explained] before, And [then] by means of410 the [subsequent] procedures, Hold one’s mind within the bindu. 401 This passage is quoted at length in Vaidyapāda’s commentary on the Muktitilaka (Muktitilaka-vyākhyāna, D 47a). 402 Vaidyapāda specifies that it remains in the conventional form of the five-colored bindu the size of a chickpea (Sukusuma, D 121b.2; P 146a.7-8). He explains that the phenomena described below in fact emerge from that conventional form of nondual wisdom (Sukusuma, D 121b.2-3; P 146a.8-b.2). 403 This seems to indicate that the indestructable bindu remains at the end of the aeon, which Vaidyapāda states even more clearly in his commentary (Sukusuma, D 121b.4; P 146b.3). 404 rgyu] sugg. em. based on V (D and P), rgyur D C P N S 405 Here I rely upon Vaidyapāda’s commentary, which notes that abandoning killing is an example of “acts” and killing is an example of “non-acts” (Sukusuma, D 122a.2-3; P 147a.2-3). 406 pas] sugg. em. based on V (D and P), pa D C P N S 407 Here brahmacarya clearly refers to what are usually called the brahmavihāras. See also verse 157 and note 304. 408 Literally “making,” but it what is meant is making oneself see it as such. 409 This is a concise description of the beginning steps of the generation stage practice, and Vaidyapāda says as much (Sukusuma, D 122a. 4-5; P 147a.5-6). 410 yis] P N S V (D and P), yi] D C 422 From the seed that rests in the center of the symbolic implement |247| Of one’s jñānasattva411 Five [-colored] light rays radiate outward. At their tips countless maṇḍala-cakras are emanated; These fill all realms. |248| [Thus] the concepts that ride on the horse of the breath Are perfectly cleared away. These, as well Become the maṇḍala-cakra. Second,412 the maṇḍala is gathered in |249| [As] the essence of wisdom, which enters into the left nostril. And is made to abide in the center of the symbolic implement. This then dissolves into the seed, Which then illuminates |250| The indestructible bindu, blazing with five [-colored] light, Thus authentically producing All the qualities ascribed to the buddhas Who abide in the [ten] directions and [three] times. |251| From that, as well, light radiates forth Illuminating its own space and the interior of the jñānasattva. That illuminates its exterior, and the maṇḍala and its area. The [light] which emerges from this draws in413 the |252| Bindus in the forms of the syllables āḥ and so forth that abide in The channels of the big toes, the calves, The two thighs, the secret place, The navel, the belly, |253| The two breasts, the tips of the fingers, The throat, the two lips, The two cheeks, the two eyes, The channels of the ears, the crown of the head, |254| And all the joints, And dissolves them into the indestructible bindu. 411 Interestingly here in the Dvitīyakrama the term jñānasattva is used for what was termed the samayamudrā in the previous section of the very same text. (Vaidyapāda’s commentary clarified in the earlier usage that by samayamudrā what was meant was the jñānasattva.) This raises the question of whether this summary section of the Dvitīyakrama was added later by a different author using the more updated vocabulary that is present in the commentarial literature or whether Mañjuśrī/Buddhajñānapāda himself just used two different terms to refer to the same thing in these different sections of the Dvitīyakrama. 412 gnyis pa] sugg. em. based on V (D and P), gnyis pa’i D C P N S 413 bkug] sugg em. based on V (D and P), bsgrub D C, sgrub P N S. This is also what happens in the more elaborate description of this practice earlier in the Dvitīyakrama. 423 Gently414 holding your mind there, Visualize415 the support and supported maṇḍala,416 |255| And, in its center, meditate Upon the indistructible bindu, wisdom, Which has an essence of cognizance.417 The light rays from this illuminate its area and onwards. |256| The light rays of the wisdom bindu That have been placed to illuminate the interior418 Go before those [tathāgatas] who reside in the world of the ten directions Melting into nectar |257| They enter their mouths And travel to the indestructible bindu at their hearts Nectar,419 in the form of kṣuṃ is taken up It emerges from the vajra path |258| Like a stream of milk And, coming from the ten directions, Draws all sentient beings and buddhas along with the inanimate Together420 and brings them into the [vajra-]pañjara. |259| Likewise they are drawn into the dharmadhātu mudrā421 And the support maṇḍala. These themselves are drawn in And brought into the [supported] maṇḍala |260| This is then drawn into oneself Then oneself, as well, becomes no longer apparent. Focus on the abode of the jñānasattva. This also [is drawn into] the [maṇḍala-]cakra |261| And the [maṇḍala-]cakra is made to enter the jñānasattva. The jñanasattva becomes no longer apparent And one focuses only on the symbolic implement. The symbolic implement itself is drawn inwards, |262| 414 cung zad 415 mos 416 i.e. the maṇḍala and the deities of the maṇḍala 417 snang ba. Vaidyapāda notes that “the essence of cognizance” refers to the bindu which is endowed with a subjective aspect (snang ba’i ngo bo zhes pa de ni de’i yul can dang ldan pa’i thig le) (Sukusuma, D 123a.4; P 148a.7-8). 418 Here I follow Vaidyapāda’s commentary to interpret the otherwise unclear term “placed” (bzhag). He explains that the “light rays which have been placed” refers to those which have been placed at the sixteen locations in the interior of the body (Sukusuma, D 123a. 5-6; p 148b.2). 419 bdud rtsi] sugg. em. based on V (D and P), bdud rtsir D C P N S 420 Vaidyapāda specifies that they are drawn together as the essence of bodhicitta (Sukusuma, D 123b.1; P 148b.6). 421 Vaidyapāda explains that this is the dharmodaya (Sukusuma, D 123b.2; P 148b.6). 424 Then, hold the mind within the indestructible bindu. From that as well, the inner gathering is all drawn in. Genuinely hold the mind within Just that great self-appearing bindu. |263| Meditating [on this] as long as one is able Bring the mind again and again into this. And then, when it emerges from there422 Focusing on its own area and one’s own body, |264| The maṇḍala-cakra And its support and the three realms—all are illuminated. And then again, just as before, Gradually they dissolve into one another. |265| And the mind should be placed upon the bindu. When one’s faculty423 is held there, The earth maṇḍala enters into water, That water likewise enters into fire, |266| The fire then enters into wind, And the wind enters into mind.424 As an indication that the mind has to some degree Entered nondual wisdom |267| There are five signs that will appear: Appreances like a mirage, like smoke, Like a lamp, like a bright sky, And like a cloudless sky, |268| Because one has entered into Vajrasattva. Holding the mind within the bindu When the yogin experiences Yawning, laughing,425 trembling, and so on |269| The bindu should be genuinely emanated By means of the higher stage, making it pervade everything. When one has genuinely trained in this One attains great non-abiding nirvāṇa, |270| The supreme attainment of every method, 422 Vaidyapāda says that its emergence means, “casting [it] out immediately after [its/the?] genuine perfection.” (yang dag par rdzogs pa’i de ma thag du gtong ba) (Sukusuma, D 123b.7; P 149a.5). 423 Vaidyapāda explains that this is the life force (Sukusuma, D 124a.1; P 149a.7124a.5). 424 Vaidyapāda clarifies that these five signs as explained in the tantra unfold only appear when one practices with the “secret bindu” located at the tip of the vajra (Sukusuma, D 124a.2; P 149a.8-149b.1). 425 dgod] D C V (D), rgod P N S V (P) 425 [The state of] great Vajradhara. Innate wisdom alone Brings control over this. The method for training in the second stage Is the meditation upon the indestructible bindu. |271| [The Names of Suchness] In this way, having explained the ritual for training In that which is genuinely brought about By focusing on the suchness of all phenomena, The excellent immeasurable sublime— |272| Its various names will be set forth: Suchness, authentic limit, The inconceivable dhātu, Dharmatā, stainless dharma, |273| Emptiness, signlessness, Wishlessness, as well, That which throws off the great load of the negative emotions, The unborn, the luminous, |274| Manifest awakening, That which brings about knowledge of others’ minds, Bestower of the ear of the gods, Bestower of the eye of the gods, |275| The great emmanator of countless miracles, The perfection of entities,426 The ultimate truth, The perfection stage,427 |276| The completely pure body, The reliance of all, Completely pure like space, Unsoiled by adventitious stains, |277| Primordially luminous, Indestructible by any means, Entitylessness itself, That which brings about cause, and so forth—the twelve,428 |278| Perfectly pure wisdom of the great glorious ones 426 dngos po 427 Vaidyapāda explains “It is called the perfection stage because it is not posited by the intellectual mind.” blos gzhag pa ma yin pas na rdzogs pa’i rim pa’o. (Sukusuma, D 124b.7; P 150a.8-150b.1). 428 This seems to be the twelve links of dependent origination. Khenchen Chodrak Tenphel suggests that it refers to their pure aspect, that is, the twelve links in reverse order (personal communication, February 2016). 426 The perfectly pure great429 bindu, The great secret of all buddhas, Space, [and]430 the object of experience [like] space, |279| Non-meditation itself, The great pith instructions of the revered master, Transferred from ear431 to ear, Not known by the śrāvakas, |280| Not known by the pratyekabuddhas and others, The letterless itself, Wordless, inexpressible, and so on. In the sūtras and tantras |281| It has been expressed, and will be again, With these countless names and others. There is nothing at all taught there Besides this suchness. |282| Therefore, with a mind that has already [generated] faith, Genuinely maintain432 the nature of all phenomena, The profound, luminous, nondual great reality, The suchness of the second stage, |283| Which has been taught by the guru. Maintaining this, by means of the previously-described procedures, [Results/Benefits of the Training] The individual who constantly habituates himself to it Based on this [practice] will give rise to the signs |284| As if leaping from bhūmi to bhūmi! The capacity of his intellect433 will increase By means of emanating as a vidyādhara, and so forth He will connect others with this truth.434 |285| The yogin who has become stable by means of [practicing] this Will possess the identity of accomplishment The one who possesses that identity should engage In the supreme practice just as it is taught. |286| A devī, nāginī, yakṣinī 429 Vaidyapāda’s commentary suggests that he was reading “stable bindu” (thig le brtan po) here (Sukusuma, 125a.3). 430 Vaidyapāda’s commentary appears to read simply “the object of experience [like] space” (Sukusuma, 125a.5). 431 rna] P N S V (D and P), sna D C 432 gzung 433 blo] P N S V(D and P), de D C 434 In the Muktitilaka, as well, Buddhajñānapāda writes of the yogin taking different forms to benefit beings. 427 A human female, a kiṃnārī, Dākīṇīs, and others One should [summon them] with one’s power and practice [with them]. |287| Put forth effort for six months, and so on In the observance of a madman (unmatta-vrata) and others.435 By means of that [one will attain]436 the revered, the letterless,437 The essence of all the glorious buddhas, |288| The state of all vajradharas, The profound suchness of all phenomena, The supreme attainment of all buddhas. Just as the yogin of the higher stage,438 |289| Having put forth tireless effort, Remained there,439 for a moment, In the manner of the example,440 In bliss, middling bliss, |290| And the bliss of cessation, [Likewise,] in time, he will attain just as has been taught, The three blisses just as they are. |291| Then until saṃsāra’s end He will remain, free from torment, 435 The observance of a madman (unmatta-vrata) is a ritual observance in which the practitioner takes a vow to act like a madman in order to test the stability of and further his practice. Vaidyapāda’s commentary here instructs that the yogin should engage in the observance of a madman for six months, and then he will experience the signs of the main part of practice, and through the goddesses’ initiation the realm will be purified. Then he should practice with a consort for six months and the results should become manifest. But if that does not bring success, he must to do it for another six months. If that does not bring success, he must to do a peaceful summoning (bskul ba) ritual for seven days. If that still does not bring success, he should practice for another seven days following each of the various divisions (dbye ba’i sgo nas), which perhaps means using another one of the divisions of the four activities, increasing and then magnetizing. If that does not bring success, he should do a wrathful ritual for summoning the buddhas for seven days. He should use a kilaya and so forth. Then he will obtain the result (Sukusuma, D 126b.4- 7). 436 There is no clear verb here, but Vaidyapāda’s commentary indicates that these phrases describe the result that is attained by the previously described practices (Sukusuma, D 126b.7). 437 yig] P N S, yid D C. This is also strongly supported by Vaidyapāda’s commentary, which provides a gloss of yi ge med pa (Sukusuma, D 126b.7; P 150b.8). 438 Vaidyapāda identifies these here as “the six-month period and so forth,” suggesting that he is referring to the practice of the unmatta-vrata and so forth, which was just mentioned (Sukusuma, D 127a.1-2). 439 Vaidyapāda specifies that this means in the lotus of the consort (Sukusuma, D 127a.2). 440 In his comments on this passage, Vaidyapāda links the three blisses not only to the three kāyas, but also to the seven yogas, and notes that the yogin remains in these “in the manner of an example” during the prajñājñānābhiṣeka, and that they will “in their own time” be fully attained as explained in the tantras (Sukusuma, D 128a.1-4; P 153a.1-6). He specifically links bliss with two of the seven yogas, middling bliss with two, and the cessation of bliss with one. He does not specify which yogas are linked to the different blisses. Vaidyapāda makes this same link of the yogas to the blisses in his Muktitilaka-vyākhyāna (D 58a.1), but there additionally notes that the remaining two of the seven yogas pertain to all three blisses. In that statement he also does not identify which of the yogas pertain to which of the blisses. I imagine that this point is clarified in the Yogasapta, but did not yet have the opportunity to check this. 428 Cool, singular, Blissful, stainlessness, |292| Joyful, and mentally joyful— These are the eight [signs] of having tasted great bliss. The lord,441 supreme Vajradhara, The yogin of the perfection stage |293| He who has performed the actions, completed the activities, The great lord who has cast off the great load, Unagitated by thorns, omniscient, The hero of beings, knower of all, |294| The great bull-among-men,442 the tamed one, The one who has gone to the far shore of saṃsāra, The great yogin for whom The ultimate and relative truths are nondual, |295| He who has abandoned misdeeds,443 completed all paths, Source of all qualities, Samantabhadra, The suchness that encompasses everything, Filling all the realms with kāyas and so forth, |296| The authentic state, above which nothing is higher— It is known in this way, and by many other names. The names that indicate nondual wisdom444 In all the sūtras and tantras Are limitless— The intelligent ones must realize this. |297| [Stages of Sexual Practice Homologized with Ten Bhūmis] 445 A vajra holder, Who abides in the thirteen virtues446 Genuinely realizes the stainless In a single moment through [relying upon] The previously-mentioned goddess,447 While endowed with the ten bhūmis. |298| 441 Tib. bdag po 442 Tib. glang po chen po; Skt. *gopati. 443 tha ba] P N S V (D and P), thab D C. 444 The names given here include some that seem to refer to a state, and others that seem to refer to an individual who abides within that state, as if these are somehow indistinguishable. 445 Vaidyapāda explains that these verses are meant to teach about the substantial cause of such realization who is the goddess, together with the ten bhūmis (Sukusuma, D 128a.1). 446 Vaidyapāda indicates that this is refers to the thirteen bhūmis (sa bcu gsum). The term “the thirteenth bhūmi” is used below in verse 314 to refer to the final result of practice, and this verse does explain that the vajra-holder reaches attainment in a single instant by means of practicing with a consort. 447 Vaidyapāda indicates that this refers to the consort types, like kamalī and the others who were mentioned before (Sukusuma, D 128a.1). 429 448[Adorned] with garlands, necklaces, anklets,449 and more, [Beholding] her complexion, breasts,450 and the rest, Knowing the bliss of examining the lotus— This should be known as the first. |299| Praising [with] melodious tones Like the ṣaḍja, ṛṣabha, niṣāda451 and others, And delighting with the sweet sound śīṭ —452 This should be known as the second. |300| At the time of anointing the body With sandalwood and other scents, The genuine bliss which is so produced— This should be known as the third. |301| Having sucked the honey from [her] lower lip The bodhicitta that abides in the head melts, Tasting it brings pleasure,453 thus delighting oneself— This should be known as the fourth. |302| Through anointing the body and a variety of acts At the time of playing Genuine bliss is brought about through touch— This is known to be the fifth. |303| 448 What follows is a revisioning of the classical ten bodhisattva bhūmis in terms of tantric sexual practice. The ten verses correspond with first the six sensory experiences— visual, auditory, olfactory, gustatory, tactile, and mental—and then with the four elements—earth, water, fire, and wind—respectively. Thanks to Harunaga Isaacson for pointing out these correspondences, which were obviously intented in the text. 449 Tib. ha ra nu pur. This seems to be a Tibetan transliteration of hāranūpura, necklaces and anklets. Thanks to Harunaga Isaacson for his assistance with this point. 450 ku tsa. This may be a Sanskrit transliteration of kuca, breasts. Thanks to Harunaga Isaacson for this suggestion. 451 ṣa dzdza ṛī ṣa ni ṣā] sugg. em. following V (D), ṣa dzdze rī ni ṣā na D, ṣa dzdza rī ni ṣā na C, sha rdzas gri tra gri na P N S), sha rdza gri ta ghri na V (P). These terms seem to have caused some confusion for the translators and scribes of the Tibetan canon, as they are rendered in four different ways in the five extant recensions of the root text and two further ways in the two recensions of Vaidyapāda’s commentary that I consulted. Vaidyapāda indicates that they refer to the singing of erotic songs from the *devīśāstras (lha mo’i bstan bcos) (Sukusuma, D 128a.3). Following Harunaga Isaacson’s suggestion, for which I am grateful, I believe that these garbled Tibetan translations of Sanskrit terms (although the transliteration from Vaidyapāda’s Sukusuma in the Derge Tengyur—the one I have chosen to use in my edition of the Dvitīyakrama—is pretty close) are meant to read ṣaḍja, ṛṣabha, and niṣāda, which are the Sanskrit names of the first two and the seventh, respectively, among the seven tones of the classical Indian musical scale: ṣaḍja, ṛṣabha, gāndhāra, madhyama, pañcama, daivata, and niṣāda. It seems, then, that the intent of the passage is simply to indicate that the praises are to be rendered musically. Thanks also to Grant Damron for his clarifications on the nature of the Indian scale. 452 sid sgra] P V (D), sing sgra D C S N V (P). 453 I am not entirely sure abou this line, but this seems to be the meaning. Vaidyapāda writes, “The bodhicitta that abides in the head melts means that regarding the path of the bodhicitta that resides in the head, it is by means of that path, that one drinks this elixir.” mgor gnas byang chub sems ‘ju bas/ zhes pa ni mgor gnas pa’i byang chub kyi sems kyi lam ni des te des ro ‘thung ba’o// (Sukusuma, D 128a.3-4; P 154a.7-8). 430 By means of this the three wisdoms454 Are known, and one’s mind Is made to experience great bliss— This should be known as the sixth. |304| By means of the hardness that results From one’s relying on her body 455 Genuine bliss is produced— This should be known as the seventh. |305| The dew from her lotus and The wetness of bodhicitta Bring about great bliss in the mind— This is known as the eighth. |306| Due to heat—the warmth and so forth of the secret place— One’s mind is brought to the supreme, Genuine bliss— This should be known as the ninth. |307| Then, through stirring, the wisdom fire Burns the aggregates, elements, and the rest Through this the mind becomes genuinely blissful— This should be known as the tenth. |308| By means of these ten The first and the later supreme result Are attained, just as explained above. But for those disciples |309| Who are unable to authentically engage in this great reality The tathāgatas have taught it in terms of characteristics Like “Perfect Joy” and the rest.456 Through engaging in this truth, and by means of [its practice] |310| They gain realization—though there is still something higher.457 That itself458 has been taught,459 454 Vaidyapāda explains that the three wisdoms are the wisdoms of “sevā and the rest.” (Sukusuma, D 128a.4). 455 de yi lus ni bdag gi ni// rten du gnas pa sras pa yis// I am unsure about the translation of these two lines. 456 Rab tu dga’ ba, “Perfect Joy,” is the name of the first bodhisattva bhūmi. 457 It is not completely clear here whether Buddhajñānapāda is asserting the path or the result of tantra to be higher. It seems, however, like he is talking about the result. Vaidyapāda reads it this way: he says that via the bodhisattva bhūmis they attain the final result that is not the highest, whereas the unsurpassable result is attained by means of the unique path (Sukusuma, D 128b.1-2). This is not the only instance in which Buddhajñānapāda seems to say that the result of tantra is superior to that of sūtric practice. In his Ātmasādhanāvatāra he also states that only by means of deity yoga can one attain the result of perfect awakening. I discuss Buddhajñānapāda’s position on the superiority of tantra in Chapter Three 458 Vaidyapāda explains that this refers to the ten bhūmis (Sukusuma, D 128b.2) 459 bstan] P N S V (P), brtan D C V (D) 431 To the yogins of the first [stage] 460 [As] the support and supported maṇḍala-cakra. |311| Engaging in and relying up on that one may gain realization, But those who do not know this truth Are not genuine buddhas. |312| This is the self-arisen bhagavan The sole supreme deity (adhidaivata), 461 Called the *adhideva [And] explained as the “thirteenth bhūmi.” |313| In this way, as for the suchness Of the second stage,462 Whichever yogin drinks this supreme nectar463 Together with the method |314| [The Greatness of This Kind of Practitioner] Certainly becomes a son of the buddhas, A companion of the bodhisattvas, A leader of the vidyādharas, The husband of the dākiṇīs. |315| The main guide, Leader of the śrāvakas and pratyekabuddhas,464 The revered master of ordinary beings. To him [i.e. the yogin] the buddhas, |316| Bodhisattvas, wisdom deities, wrathful [deities], and others Who abide in the realms of the ten directions, Together with offerings of flowers, Worship him and sing his praises |317| From the sky First in the morning at dawn, then in the warmth of mid-day, And as evening comes on,465 And [then] depart to their own realms. |318| 460 Vaidyapāda clarifies that this refers to yogins who are at the generation stage level of practice (Sukusuma, D 128b.2-3). 461 These two lines correspond to the first two pādas of Sarvabuddhasamāyoga-tantra, I.2. asau svayaṃbhūr bhagavān eka evādhidaivataḥ. Thanks to Harunaga Isaacson for bringing this parallel to my attention and to Péter Szántó for sharing with me his draft Sanskrit edition of the Sarvabuddhasamāyoga-tantra. 462 Here Vaidyapāda describes this as “the perfection stage of the perfection stage,” following Buddhajñānapāda’s four-fold classification of the generation and perfection stages in the Muktitilaka (Sukusuma, D 128b.6-7). 463 Vaidyapāda seems to understand the nectar spoken of here to be wisdom, as he writes that it is to be received again and again from the guru’s mouth (Sukusuma, D 129a.1). I am inclined to read the line more literally. 464 rang ‘dren 465 chal chil mtshams su. Vaidyapāda also clarifies that worship at the three times of the day is being referred to here (Sukusuma, D 129a.4). 432 Thus if the pure deities Worship him thus, Why would the impure deities Not do so, as well? |319| Other ordinary sentient beings, as well, Filling their cupped hands with flowers With their necks [bent] low, bow at his feet And constantly respect him—this is absolutely appropriate. |320| He is the lord among the two-legged, The one set forth by the omniscient,466 The one who throws off the mistaken great load, The future vajra holder. |321| [Faults of Depricating Such a Practitioner] Anyone who deprecates him, That practitioner who is like a bull[-among-men], Because they [also] depricate me I will always abandon them.467 |322| On the other hand, since I468 abide in [his] body, By praising and worshiping [him] [Their] physical obscurations will be cleared away. |323| Here in the second stage, the practitioner Practices one-pointed retention. Maintaining that, By means of the vratavidhi,469 |324| [And] with the goddess acting as the condition— 466 Vaidyapāda explains that this means he is sealed by omniscient wisdom, and therefore is set forth by the omniscient ones, like bodhisattvas held back from awakening by only one life time who reside in Tuṣita (Sukusuma, D 129a.6-7). 467 This verse is rather perplexing. Khenchen Chodrak Tenphel suggests emending the final line to nga ni des kun dus kun spong. “They also constantly abandon me [i.e. Mañjuśrī].” It does seem strange for Mañjuśrī, a wellknown bodhisattva, to promise to forever abandon someone who depricates him, but I have not taken the liberty of making such a significant emendation in the text. Jamgön Kongtrül, who cites this passage in his Torch of Certainty, does seem to take the passage to intend exactly what it says with regard to abandoning. He uses the citation to argue that one should view the guru as the Buddha and takes it to mean that Mañjuśrī dwells in the body of the guru receiving offerings, and so forth (Kongtrül 1994, 126). 468 Vaidyapāda is very clear that I means, I, Mañjuśrī (Sukusuma, D 129b.1). 469 Given that the “goddess acting as the condition” is mentioned in the subsequent line, it is presumably the vidyāvrata and its associated sexual yogas that is intended here. However, the unmattavrata was also specifically mentioned earlier in the text (in verse 288), so perhaps this vrata and its associated practices are also being referenced. 433 Have no doubt that the mahāmudrā470 Will be transfered to his mindstream in this very life. |325| [The Yoga of Utkrānti] 471 Now for the stage of svādhiṣṭhāna472 This will be explained To a few yogins Who are fortunate due to their actions. |326| Someone who has pleased the guru And received the vase [initiation] and the others Together with the samayas and vows given by him And has thus obtained the suchness473 |327| That is found through the guru’s words,474 And has realized the secret and supreme secret [But] is not able to genuinely train by means of the activities In the way explained [above]— |328| He should train in this stage 470 Here Vaidyapāda glosses mahāmudrā: “It is [called mahāmudrā ] because it is a great (mahā) accomplishment that is generated through the mudrā.” (phyag rgya las skyes pa’i dngos grub chen po bas na de skad de/ (Sukusuma, D 129b.5-6; P 156a.5). 471 Vaidyapāda states, “Having in this way taught the stages [of practice] for attaining nirvāṇa in this life, now he teaches the stages [of practice] for attaining nirvāṇa in the intermediate state with the verse beginning ‘Now...’.” de ltar mthong ba’i chos la mya ngan las ‘da’ ba’i rim pa bstan nas/ da ni bar ma dor mya ngan las ‘da’ ba’i rim pa gsungs pa/ da ni zhes pa la sogs pa’o (Sukusuma, D 129b.6-7; P 156a.7-8). 472 The term svādhiṣṭhāna is used in several works of the Guhyasamāja system, as well as in the corpi of later tantras, to refer to a number of different practices, but its use here to refer to the yoga of utkrānti appears unique. Svādhiṣṭhāna is mentioned in the Samājottara (verse 77), in reference to what appears to be a practice within the context of the generation stage, and the term is also used within the literature of the Ārya School to describe the third of the five stages of that tradition’s perfection stage practices, called the svādhiṣṭhānakrama, and also termed the practice of the the illusory samādhi (māyopama-samādhi), or of the illusory body (māyādeha) (see Wedemeyer 2007, 68 and Tomabechi 2006, 79-81). The Hevajra-tantra uses the term svādhiṣṭhāna in what has been interpreted by commentators as just a reference to utpannakrama practice more generally (see Isaacson and Sferra 2014, 267 n 74). None of these usages of the term relates to utkrānti. However, the practice of svādhiṣṭhāna according to the Ārya School is the method by which the yogin produces the body or form of an awakened buddha (ibid.), and indeed, as we shall see below, the utkrānti instructions given here in the Dvitīyakrama seem to serve precisely this same function of generating a saṃbhogakāya form, which is done here by means of first bringing the mind into the dharmakāya at the time of ejecting the consciousness in the moment of death. Once the saṃbhogakāya form is achieved, the Dvitīyakrama contends, one will naturally take birth in the next life in a nirmāṇakāya form (see Dvitīyakrama verses 351-353). 473 Vaidyapāda specifies that this refers to having received the instructions on suchness together with the sādhana for accomplishing suchness via the seven yogas (Sukusuma, D 130a.3; P 156b.3-4). On the seven yogas see note 121, and Chapter Seven. 474 This is one of a number of instances in which Buddhajñānapāda writes about the practice of the guru directly showing the state of suchness to students, here by means of words. Such a “showing” of suchness by means of words seems, in his system to have been a practice that was connected to or immediately followed the third initiation, and indeed such a verbal communication of suchness is the dominant one among the various positions recorded in the literature about what is meant by the so-called “fourth initiation,” that later became part of the standard sequence of initiations (see e.g. Isaacson 2010b, 270-1). I discuss this topic briefly in Chapter Three and in more detail in Chapter Seven. 434 Of suchness, just as it is.475 At some time in the future One will see the signs of death. |329| When the time of death has arrived And one is not completely overcome by illness Engage in the yoga of utkrānti. |330| The forehead and the navel, The crown and the eyes, The ears and nose, The urethra, the anus, |331| And the mouth— Know the signs of wisdom traveling through these places. Know that [wisdom departing from] the forehead is a sign Of being born in the form realm, |332| [From] the navel, as a god in the desire realm— If the sign appears, birth there is certain. [Emergence from] the crown of the head is a sign That one will be reborn in the formless realm. |333| If the wisdom is transferred out from the two nostrils476 One will be reborn in the abode of yakṣas. If from the two ears,477 one will certainly go To the abode of the vidyādharas. |334| If from the two eyes, this is a sign That one will certainly be born as a king among men. If the wisdom leaves from the mouth This should be known as a sign of [birth among] the pretas. |335| [If from] the urethra this should be known As a sign of [birth among] the animals. If the widom exits from the anus 475Precisely what practice background is necessary for taking up the yoga of utkrānti is not made entirely clear. The Dvitīyakrama appears simply to suggest that this practice is for someone who received initiation and “obtained suchness” from the guru, “and has realized the secret and the supreme secret” but was unable to train (fully?) in the practices “described above.” Vaidyapāda, explains that this refers to a disciple who has received suchness from the guru by means of the seven yogas, but who has been unable to genuinely train in it, meaning that he has begun with the generation stage, but been unable to train in accordance with both stages (Sukusuma, D 130a.3-4; P 156b.4-6). Vaidyapāda’s subsequent explanation gives many options for the type and frequency of practice that a yogin who wishes to perform utkrānti may have engaged in before undertaking this final practice, but the very fact that he includes such a list seems to indicate that he understood some type of training in suchness by means of the generation and perfection stages as a necessary prerequisite for performing the yoga of utkrānti at the moment of death (Sukusuma, D 130a.4-7; P 156b.6-157a.1). 476 sna] P N S V (D and P), rna, D C 477 rna] P N S V (D and P), sna D C 435 This should be known as a sign of [birth] as a hell being. |336| Knowing in this way the aspects Of wisdom being transferred One should block the seven higher doors With the syllable proclaimed by [all] five [buddhas].478 |337| Block the urethra with suṃ Block the anus with kṣuṃ. Having thus blocked the nine doors By means of this procedure, |338| Search for the abode of your mind By doing this, it will certainly enter into space itself. By meditating on the aggregates, Elements, and sense sources |339| As they are explained in the Yoga tantras,479 Through wisdom480 you will realize the unsurpassed state. Just as explained, One transforms oneself into the body of the deity,481 |340| And482 the dharmadhatu and [one’s] consciousness Remain as the identity of the body of the buddha.483 Imagine a supreme nine-pronged vajra Perfectly adorned with the five colors |341| Of a variety of gems Above the crown of one’s head. There, imagine one’s own mind As a five-pronged white vajra one-tenth of the size [of the previously visualized vajra] 484 |342| And examine it. The five upper prongs represent the five methods 478 Vaidyapāda specifies that this refers to hūṃ, since it is is proclaimed by the five buddhas (Sukusuma, D131a.1; P 157b.4). 479 Vaidyapāda provides an unattributed quotation here to illustrate how to visualize these—the aggregates as the buddhas, the sense sources as the maṇḍala bodhisattvas, and the elements as the buddha consorts (Sukusuma, D 131a. 5-6; P 158a.2-4). 480 gyis] sugg. em. based on V (D and P), so D C P N S. 481 Vaidyapāda explains that this means to generate oneself in the form of the deity following the four-branch ritual of the generation stage (Sukusuma, D 131a.7-131b.1; P 158a.4-5). 482 la] P N S, pa D C. Vaidyapāda supports this reading: lha’i lus su gyur par byas la/ (Sukusuma, D 131a.7; P 158a.5). 483 Here I am reading the verse following Vaidyapāda’s commentary which indicates that it is these two things—the dharmadhātu and one’s consciousness—that have the identity of the body of a buddha. The third genitive particle in the Tibetan of the root verse might suggest reading the “identity” as a third member of the list, but Vaidyapāda’s reading is gramatically possible even with the Tibetan translation of the verse as it is, and also seems to me a more plausible reading, so I have followed it here (Sukusuma, D 131b.1; P 158a.6). 484 bcu tshal tsam. I am unsure of the meaning of this term. 436 Likewise the ones facing downwards are the five wisdoms Imagining that at485 its center is the rabbit-holder486 |343| There, meditate upon great wisdom Yellow, and like bodhicitta, With the nature of dripping.487 This bindu is the size |344| Of five chickpeas joined together. Then imagine that all phenomena Dissappear,488 and focus [only] on oneself. Then oneself dissolves into oneself, |345| And there is only mind. This, [shoots up] like an arrow, entering into the opening At the lower end of the vajra above one’s crown and dissolves into the bindu. Know that [bindu] |346| To be the natural abode Of the tathāgatas [and] the goddesses. That is the primordially accomplished form. Having placed one’s consciousness there |347| Again and again hold the mind there. When the mind becomes dissapated from that It emerges from the opening at the top Of the nine-pronged vajra and, |348| On a moondisc on top of a multi-colored lotus, It transforms into the body of Vajrasattva.489 485 bar] sugg. em. based on Vaidyapāda’s commentary, ba D C P N S. Vaidyapāda reads dbus kyi bum pa ni de’i dbus kyi bum par ro// (Sukusuma, D 131b.4; P 158b.2-3) 486 i.e. the moon. Here, according to Vaidyapāda, in the form of the reflection of the moon (Sukusuma, D 131b.5; P 158b.3). 487 Vaidyapāda explains the reason for each of the characteristics described. He says it has the nature of dripping because when one places one’s attention upon it a definitive mental state does not arise. de yid la byed pa la nges pa’i shes pa mi skye ba’i phyir ‘dzag pa’i ngang tshul dang ldan pa’o// (Sukusuma, D 131b.6; P 158b.4-5). 488 Vaidyapāda explains that this means “do not direct your attention to the appearances” of these things (Sukusuma, D 131b.7). 489 In Buddhajñānapāda’s generation stage writings, like many other sādhanas and sādhana commentaries from this period, vajrasattva (rdo rje sems dpa’) does not mean the specific deity/buddha Vajrasattva, but is an epithet for the causal deity from which the main deity of the sādhana is generated, and is, in that context, interchangeable with vajradhara (or at least some of the commentaries on his Caturaṅga/Samantabhadra-sādhana also use the term vajradhara to refer to this causal deity). As such, the form of this vajrasattva in the generation stage is indeterminate; he can have different colors or attributes, as determined by the particular sādhana, and does not have the specific form of Vajrasattva as the primordial buddha (who is usually white in color and holding a bell and vajra). I am unsure how to read the term vajrasattva here in the Dvitīyakrama, but my inclination is to read it as referring to the primordial buddha Vajrasattva, in part since no details of his appearance are given and they therefore appear to be assumed, which would only be possible if they were standard rather than indeterminate. (Though of course it is possible that this Vajrasattva should be visualized in the form of the progenitor deity vajrasattva described in Buddhajñānapāda’s Caturaṅga/Samantabhadra-sādhana) Presumably there is anyway a relationship 437 He is adorned with all the major and minor marks Fully ornamented, but without clothing, |349| Visualize him very clearly. Then imagine the emanation And absorption of the great [maṇḍala-]cakra Arisen from the blessings of nondual union. |350| In this way engage in this supreme meditation Again and again, for as long as one is able. When one’s mind enters into the [dharma]dhatu Due to this contemplation |351| One realizes that which is luminous and perfectly joyful,490 like the sky.491 Then, possessed of miraculous power,492 One accomplishes the form of a five-year-old child At that time, one realizes unparalleled perfect bliss. |352| When one is transferred493 from that To another rebirth One will genuinely realize the nirmāṇakāya. |353| Therefore, it is by means of the various Attitudes of sentient beings That the human mind is purified. Such [cause] and such [result] [is explained] in the sūtras Like [the resulting appearance of] different types of gems.494 |354| The dharmakāya, perfect joy equal to the sky, between the use of the term vajrasattva as an epithet of the causal deity in sādhana practice and the primordial buddha Vajrasattva, as the causal deity is the deity from which the iṣṭadeva for a given sādhana is produced, and, as such, it makes sense for the primordial buddha to function as the source/progenitor of any and all other deities. 490 dga’] P N S V (D and P), dag D C 491 Vaidyapāda explains that this describes the attainment of the dharmakāya, but that this moment is referred to as the “death state” (shi ba’i srid pa) by “followers of karma who do not know the nature of mind” (sem kyi rang bzhin ma shes pa’i las su smra ba) (Sukusuma, D 132b.2; P 159a.1). According to Vaidyapāda’s commentary, this and the subsequent two verses are about attaining the three kāyas. At the time when ordinary beings experience the moment of death (shi ba’i srid pa) the practitioner realizes dharmakāya; at the time when ordinary beings would be in the intermediate state (for seven days etc) the practitioner realizes the sambhogakāya (here this is the practitioner taking on the form of a five-year-old child as described in the next verse); and at the time when they would be born into another body the practitioner realizes the nirmāṇakaya. This amounts to a description of the death process for the practitioner who has realization of suchness, in which the death processes of an ordinary being instead become the stages for the practitioner’s realization of the three kāyas. 492 ldan pa] D C V (D and P), ldan pa’i P N S 493 ‘phen par] P N S V (D and P), ‘phel bar D C 494 Vaidyapāda explains that this is an example given to show that the statement above, about appearing in the form of the emanated body, is not contradictory. Following his explanation, the text generally seems to be saying that in the same way, due to giving rise to different thoughts (Vaidyapāda gives the example of pleasure, pain, and the seed of liberation) the human mind is variously trained, likewise, as explained in the sūtras, because of these differences, different forms—like those of different jewels—appear (Sukusuma, D 132b.4-5). This passage, which is cited in Tsongkhapa’s Lamp to Illuminate the Five Stages is translated rather differently in Kilty (2016, 433-4). 438 Is experienced for just an instant At death, when fainting, falling asleep, When yawning, and during intercourse. Therefore, by training in this, embodied beings purify their minds.495 |355| With this ritual even someone Who has committed one of the acts of immediate retribution,496 A deluded being, or a brahmin-slayer— None of these are precluded from accomplishment. |356| Therefore recieve the instructions and transmission And maintain the vows and samayas— Then there is no doubt that one will attain the three kāyas! If one does not attain the three kāyas497 |357| He will become the leader of the vidyādharas And gradually will transform into the mahāmudrā.498 In that way, it is explained that meditating By means of these three [procedures499 brings about] accomplishment. |358| Anyone who, without having realized this, Speaks about these secrets, I and the tathāgatas Will never join with him And give him blessing. |359| I abide in the bodies Of a few [individuals] who possess this understanding,500 Receiving offerings from501 other practitioners.502 Through pleasing [those individuals], the karmic obscurations 495 Vaidyapāda explains that because these experiences are so short, without the instructions of a guru one will be unable to gain realization. However, when one does have those instructions and trains in the suchness that is experienced, then the mind becomes purified, as explained in the example above (Sukusuma, D 132b.6-7; P 159b.8- 160a.1). 496 Mtsham med pa. A set of five particularly heinous evil deeds: killing one’s father, one’s mother, or an arhat, creating a rift in the saṅgha, and maliciously drawing blood from a tathāgata. 497 Vaidyapāda specifies that this means attining the three kāyas in the manner that was just described above in the text (Sukusuma, D 133a.1-2). 498 This passage is cited by Dakpo Tashi Namgyal as indicating the situation of the lesser individual who attains siddhi in another lifetime (Roberts 2010, 614). 499 Vaidyapāda clarifies that this refers to the three meditations on the bindu (Sukusuma, D 133a.2). 500 ‘di ni don ldan. Vaidyapāda is very clear that this is still Mañjuśrī speaking in the first person: “The lines I abide in the bodies... is the opposite [of the situation previously described]. The text says in a few because there are not many suitable individuals. I Mañjuśrī abide in their bodies.” ‘di ni don ldan zhes pa ni bzlog pa’o// ‘ga’ zhig la zhes pa ni snod du rung ba’i gang zag mang ba ma yin pas so// ‘jam dyangs nga (nga] P, D om.) ni de’i lus la gnas so// (Sukusuma, D 133a.5-6; P 160a8-160b.1). 501 las] sugg. em., la D C P N S V (D and P). I suggest this emendation to gramatically align with the verb len. 502 I have interpreted this line following Vaidyapāda’s commentary: “If someone asked, ‘Why do you abide there?’ It is [in order] to receive offerings from other practitioners.” khyod ci’i phyir gnas she (she] D, zhe P) na/ sgrub pa (P om.) po gzhan la mchod pa len zhes so// (Sukusuma, D 133a.5; P 160b.1). The grammar of this passage remains problematic, however. See previous and next note. 439 In the mindstreams of [those practitioners] are purified.503 |360| [Essential Nature of These Instructions] For as long as those who pass along this teaching Into the ears [of disciples] remain, For that long it is said That the Buddha’s precious teaching will remain. |361| When this lineage is broken This should be known to everyone As the [time of the] dissappearance of the Buddha’s teaching. |362| Therefore you should one-pointedly Compile [these instructions] And bring some fortunate future individuals Who have previously generated the accumulations Into connection with this lineage transmission. |363| Those who connect with this [lineage] Should be known as authentic yogins. |364| [Mañjuśrī’s Prediction and Command for Buddhajñānapāda] However, because of [your] conduct regarding food, And holding a slight delusion with respect to me You will not, in this very life, Bring about a complete transformation of the state of Your body—the aggregates including form. |365| However, you will accomplish consciousness, Which is indestructible, as the mahāmudrā.504 |366| 503 Khenchen Chodrak Tenphel explains that this verse refers to the practice of generating oneself as the deity in generation stage practice, and making offerings to oneself (in the form of offering goddesses) as a way of purifying obscurations (personal communication, February 2016). The grammar of the verse is problematic, however, given that the verb len does not go with the particle “to” (la), and we find this problematic reading in all versions of the text. I have suggested emending la to las, in order to make the grammar work better, but with this reading it is difficult to understand the verse as Khenchen Chodrak Tenphel has suggested. 504 This verse seems, in Vaidyapāda’s commentary—and definitely is in its interpretation by later commentators and historians—to be understood to refer to the fact that Buddhajñānapāda gave rise to doubt with respect to the monk, himself an emanation of Mañjuśrī, who emanated the maṇḍala of Mañjuśrī from which Buddhajñānapāda received these instructions. Vaidyapāda reports that the “conduct regarding food” refers to Buddhajñānapāda’s refusal of “the cooked rice and yogurt vomited by the female dog and the cooked fish.” Later Tibetan historians elaborate that these were foods served to Buddhajñānapāda by the female companion of the “emanated monk” who eventually emanated Mañjuśrī’s maṇḍala for Buddhajñānapāda, prior to that visionary experience. The “slight delusion with respect to me,” according to Vaidyapāda, refers to Buddhajñānapāda’s lack of faith toward the “vajra holder” (Vaidyapāda’s term for the emanated monk) based on his confusion (Sukusuma, D 133b.1-2). Vaidyapāda’s comments on the subsequent lines of the verse are difficult to understand, but the later historians interepret the whole episode to mean that these mistakes made it impossible for Buddhajñānapāda to attain the final fruition in this body (i.e. not bringing about a complete transformation of his body as referred to in the verse itself), so he had to wait for full awakening in the intermediate state at the time of death. Various versions of these accounts are found in Chogyal Phagpa’s Gsang ‘dus ye shes zhabs kyi rnam thar dang rgyud pa’i rim pa, Gö Lotsāwa’s Deb ther sngon po, Tāranātha’s Rgya gar chos ‘byung and his Bka’ babs bdun, Amnye Zhab’s Gsang ‘dus chos ‘byung, and Dudjom Rinpoche’s History of 440 Therefore you should compose with a genuine intention A sādhana, homa, Bali, gaṇacakra,505 Summary, commentary, Maṇḍala-vidhi, and so forth506 |367| For the first stage Of the tantra that is the gathering all the buddhas, Which is greatly secret,507 secret and, supremely secret—508 This great scripture, surpassed by none— |368| [To be] like a scalpel509 for sentient beings who are Obscured by the darkness of ignorance. |369| the Nyingma School. Certainly this is not an exhaustive list of Tibetan accounts of Buddhajñānapāda’s life. I am grateful also to Khenchen Chodrak Tenphel for his explanations clarifying the details of Buddhajñānapāda’s life story and the intent of this verse (Khenchen Chodrak Tenphel, personal communications, January and February 2016). I discuss several accounts of Buddhajñānapāda’s life in Chapter One. 505 sna tshogs ‘khor lo. 506 Vaidyapāda explains that each of these refer to texts that Mañjuśrī is commanding Buddhajñānapāda to compse, and he identifies many of these texts, making this section of his commentary the earliest list we have of texts purportedly composed by Buddhajñānapāda. The sādhana, Vaidyapāda says, is the “three Samantabhadrīs” (kun tu bzang mo gsum); the homa is [for?] the generation stage, and he notes that there are two such homa rituals; the bali ritual is that for the unfaltering Tārā (mi nub pa’i sgrol ma); the gaṇacakra text is Mahāgaṇacakra (though it is unclear if this is meant to be the name of a text or simply stating that it is a ritual for the practice of the mahāgaṇacakra; the summary is the Blazing Gem (rin po che ‘bar ba), the commentary “he did not compose.” As for the maṇḍalavidhi, Vaidyapāda notes, “Regarding the maṇḍalavidhi, the one in two hundred and fifty verses is said to have been taken off to Kaśmir and I, myself, have not seen it.” (dkyil ‘khor cho ga ni shlo (shlo] D, shlō P) ka nyis brgya lnga bcu pa de kha che’i yul du khyer zhes grags te/ bdag cag gis ma mthong ngo//. (Sukusuma D 134a.1-2; P 161a.7) Vaidyapāda then goes on to explain that the “and so forth” includes the Great Root Wisdom (rtsa ba’i ye shes chen po) and the Treasure of Verses (tshigs su bcad pa’i mdzod), the Muktitilaka (grol ba’i thig le), and the Ātmasādhanāvatāra (bdag nyid grub par ‘byung ba; usually rendered as bdag nyid grub pa la ‘jug pa, but presumably it refers to the same text here), the *Bodhicittabindu (byang chub sems kyi thig le), the Great Commentary on Glorious Auspiciousness (dpal bkra shis kyi rnam par bshad pa chen po), The Method for Engaging in the Fourth (bzhi pa la ‘jug pa’i thabs), and three Jambhala sādhanas. Vaidyapāda then notes that these fourteen teachings were composed in accordance with Mañjuśrī’s prediction (Sukusuma, D 133b.7-134a.3). Some of these texts we know and others we do not. The only way I have been able to make this list total fourteen is by counting each of the texts listed in the root text as one (7; ignoring the fact that Vaidyapāda says that the sādhana actually refers to three texts, and the homa to two), subtracting the commentary that Vaidyapāda says was not composed (-1), and adding the texts Vaidyapāda lists in as part of the etc (+8; again ignoring the fact that the “three Jambhala sādhanas” counts only as one of the eight). (I discovered only later that Gö Lotsāwa had previously engaged in a similar mathematical endeavor regarding this list of fourteen! (Deb ther sngon po, Vol I, 550).) The texts in this list that are identifiable among Buddhajñānapāda’s surviving works are: the Samantabhadra-sādhana (It is interesting that Vaidyapāda mentions “three Samantabhadrīs,” as there are two translations of the sādhana into Tibetan under two different names, the Samantabhadra-sādhana and the Caturaṅga-sādhanopāyikā-samantabhadrī-nāma, as well as a third “Samantabhadra” text, the Kun tu bzang po bsdus don, which is listed in the Peking Tengyur catalogue, but the text itself is strangely absent from the place where it should be in that Tengyur, and is not mentioned even in the catalogues of the other Tengyurs); the Muktitilaka, the Ātmasādhanāvatāra, and the three Jambhala sādhanas (Bhaṭṭārakāryajambhalajalendra-sādhana, Guhyajambhalasādhana, and Vistarajambhalasādhana). I discuss Buddhajñānapāda’s extant compositions in Chapter One. 507 gsang chen 508 ches gsang ba 509 thur ma, śalākā. According to Vaidyapāda’s commentary such an instrument is used “to clear away cateracts” (ling tog rnam bsal) (Sukusuma, D 134a.4; P 161b.2). 441 Therefore the great yogins of the future, Should please a guru who knows this, And, having received it with genuine desire, Should train their mind in this. |370| The mind, when it has fully abandoned conceptuality, Joins with that and awareness arises there. When awareness has arisen, one accomplishes The [state of] a vajra holder. |371| Due to that they will then510 genuinely accomplish Buddhahood,511 the pāramitās, dhārāṇī, All the bhūmis, and great bliss. Since everything arises from great bliss, |372| If one trains in this, why would this [result] not occur? Therefore, with great effort, At least make aspirations512 Towards this supreme suchness, the perfectly secret secret, [Or513] endeavor feverently towards its accomplishment! |373| A la la ho!” [Conclusion of the Visionary Encounter] In this way with the vajra song like an echo, together with the playful dance And the [maṇḍala-]cakra, right then514 he sang and praised me515 Then, right there, he disappeared like a cloud into the sky516 And the monk and two gurus also likewise disappeared. |374| [Autobiography: Part II] In a place fifty krośas behind Vajrāsana I lived in the Parvata cave. In order to benefit beings I compiled this [text, the Dvitīyakrama], composed and taught all of the treatises,517 and so forth. 510 Vaidyapāda clarifies that this does not mean subsequently, but rather that these other attainments arise from the power of the accomplishment of the state of being a vajra holder (Sukusuma, D 134b.2). 511 It is worth noting here that Buddhajñānapāda states that after having become a vajra holder then subsequently the practitioner will accomplish buddhahood, the pāramitās, dhāraṇī, etc, as if buddhahood itself were somehow of a lower status than being a vajra holder. Vaidyapāda specifies that this is the “buddhahood of the lesser stages” (sa ‘og ma’i sangs rgyas) (Sukusuma, D 134b.1; P 161b.8). 512 smon lam] P N V (D and P), smon las D C, smon la S 513 Presumably what is meant here is, “Or, better yet…” 514 de nyid. I am following Vaidyapāda in interpreting that as referring to the immediate moment (Sukusuma, D134b.6). 515 Vaidyapāda seems to suggest that the song of praise is from, or perhaps located in, the previously mentioned

• Gathākoṣa (Sukusuma, D 134b.6).

516 Unusually, this short section of Vaidyapāda’s commentary appears to be commenting on a line or lines of the root text that are not extant in our version of the Dvitīyakrama (Sukusuma, D 134b.7-135a.2). 517 Vaidyapāda explains that this refers to the treatises mentioned above, that Mañjuśrī commanded Buddhajñānapāda to compose (Sukusuma, D 135a.3-4). 442 Since excellent beings made extensive supplications, I was delighted [to do so]. |375| Living there together, my retinue518 and I [received] necessities, Clothing, food, a treasury of jewels, and various vast offering substances for gaṇacakra. [From] the tenth-ground bodhisattva the treasure guard,519 great Jambhala Each day we regularly received seven hundred kārṣāpaṇa. |376| Then I traveled to meet the great guru Pālitapāda520 In order to please that guru, I compiled521 some short sādhanas522 And the guru and all the others there were pleased. I returned to the place I had come from and523 joyfully performed the benefit of some524 fortunate [individuals]. |377| [Buddhajñānapāda’s Advice and Injunction to Practice] Thus in this way everyone, having come to know the detailed accounts [of my life],525 Should, using all methods, please the sublime and sincere learned one,526 And listen to and contemplate his teachings, compositions, and so forth. |378| Through relying upon that, remaining in isolated places and the rest, Training one’s mind in suchness, and genuinely realizing the way things are, 518 Vaidyapāda notes that this includes eighteen disciples who functioned as his regents, among whom there were four disciples who attained nirvāṇa in this lifetime: Dīpaṃkarabhadra (mar me mdzad bzang po) *Praśāntamitra (rab tu zhi ba’i bshes gnyan), *Rahulabhadra (sgra gcan ‘dzin bzang po) and *Vajramahāsukha (rdo rje bde ba chen po) (Sukusuma, D 135a.5-6; P 162b.8-163a.1). 519 srung] D C V (D), gsung P N S V (P). 520 bā li pā da’i] D C, bha li pa trī P N S. Vaidyapāda’s commentary, however, reads bsrung ba’i zhabs. For this and a number of other reasons cited in his article, I follow Szántó in identifying this guru as Pālitapāda (Szántó 2015, 542). In the edition, however, I have left the rendering from the Derge and Cone Tengyurs because to “correctly” phoneticize the teacher’s name would render the line unmetrical. 521 It is worth noting that Buddhajñānapāda here uses the word “compile” (bsdus) rather than “compose.” He uses the term compile to describe the compilation of the Dvitīyakrama, but that is presumably because it is in fact Mañjuśrī’s teaching, which he is only compiling within the framework of his own narrative. “Compiling” rather than “composing” these sādhanas may hint to a process more revelatory than compositional, or it may simply be an acknowledgement that sections of the sādhana were compiled from other sources, most prominently the Guhyasamāja-tantra itself. 522 Vaidyapāda seems to take this to refer to more than one sādhana, as he notes that it refers to “those” that were mentioned above (Sukusuma, D 135b.1). 523 nas] P N S, gnas D C. 524 ‘ga’] D C V (D and P), dga’ P N S 525 Vaidyapāda here refers to several accounts of Buddhajñānapāda’s life as if they are already well-known stories that will be understood by anyone reading his text. “Having come to know all of the detailed accounts means having engendered even more faith in the fortunate by means of the detailed accounts of the great master: the taming of Nālandā, the offerings made at Vajrāsana, the consecration, and so forth.” (gtam rgyud rgyas par shes byas nas/ zhes pa ni bla ma chen po’i gtam rgyud (P om.) rgyas pa ni na landa (landa] P, lendra D) ‘dul ba dang/ rdo rje gdan gyi mchod pa byas pa dang/ rab tu gnas pa byas pa la sogs pa’i lo rgyus kyis skal pa dang lcdan pa cher dad par byas nas/ (Sukusuma, D 135b. 3-4; P 163a.7-8). These same accounts are described in a number of later Tibetan histories in much more detail, though unfortunately not to my knowledge in any earlier Indian sources, so it seems they survived only in the oral tradition until their recording centuries later by Tibetan historians. See note 504 for a list of Tibetan accounts of Buddhjñānapāda’s life. I also discuss some of these accounts in Chapter One. 526 Vaidyapāda says the “learned one” refers to the bla ma himself, which is the term Vaidyapāda uses throughout the commentary to refer to Buddhajñānapāda (Sukusuma, D 135b.4). 443 [One can] attain awakening in this very life, or [even] in [just] six months, and so forth—who could refute this!? |379| The one who drinks this supreme nectar is the object of respect of all sentient beings That great527 being is praised by Vajradhara and all the sugatas Since the seeds of all of mistaken obscurations are exhausted, And although he may remain in saṃsāra he remains always unstained by faults, like a lotus. |380| If he does not attain [awakening] now, then in the future528 The result of genuine experience will be brought to ripening. In the mantrin’s mind, the genuinely arisen vast accomplishment will be obtained And he will travel to all worlds, with a retinue as numerous as sand grains from the [River] Ganga. |381| Therefore like [someone who has] fallen down into a whirlpool, One should bring forth exertion in body, speech, and mind, And, stage-by-stage, as [explained] before, accomplish this [result]. |382| This body [endowed with] freedoms and riches so easily wavers Just like a flame blown out by a gust of wind It lasts barely a moment—time passes by. Therefore don’t let this go to waste— Train in the great supreme suchness of all things! |383| Having become certain about this practice [of] suchness, One will become just as explained above. Therefore, make the mind stable! I supplicate [thus] to sentient beings. |384| Because this [teaching] is secret Those who desire [to realize] suchness should Please a genuine guru who understands this and ask of him Whatever [points] I have not explained clearly. |385| Engage genuinely with yogic conduct in this and Train529 well in that which bestows the result of the essential essence. This should be authentically received from the words Of a great guru who has gone beyond, who is a treasure530 of limitless qualities. |386| Thus the meaning of the Mahāyoga tantras 527 chen] P N S, can D C 528 Vaidyapāda explains that de ru means “this year” and lan drang gzhan means some time after this; after two, three, or an indeterminate number of years (Sukusuma, D 136a.4). 529 sbyong] P N S V (D and P), spyod D C 530 gter] D C rten P N S 444 Though it may appear unpalatable,531 like the example of the sun,532 Is capable of benefiting oneself and others Therefore, certain yogins must genuinely endeavor towards it. |387| With faith, free from doubts and hesitations, Having acomplished this purpose, and taken up what is genuine, Train oneself again and again with wisdom. In this way the great nondual wisdom, like the circle of the moon in clear water, Will arise within oneself—of this have no doubts! |388| With the naturally accomplished pith instructions, Through relying upon a genuine lineage teacher, And one’s own previously gathered accumulation of merit— One will come to realize this. |389| Apart from533 these [circumstances], those with little merit Even in countless aeons will not realize this If one has not realized this reality534 he is not called a great yogin. |390| Having come to fully understand this, [One knows] the universal form of the wisdom of the great perfection,535 The perfectly pure body, Great Vajradhara, The essence of all the great glorious ones, this second stage.536 |391| Even following the path with suffering for three aeons The concordant awakening that one attains can be surpassed. Why would a yogin who is so attached537 to the very limited bliss of that [result] Not train in this [path instead]?538 |392| Residing in a delightful place With a mind [full of] faith, diligence, concentration, wisdom, and attention, 531 mi ‘tsham par 532 I am still unclear on this example and its explanation in Vaidyapāda’s commentary (Sukusuma, D 137a.4). 533 gtogs] sugg. em., rtogs D C P N S 534 Vaidyapāda notes that this remains the case even if one has attained the hightest realization of a bodhisattva, and so forth (Sukusuma, D 137b.3-4). 535 rdzogs pa chen po. This line is often cited by Tibetan Nyingma scholars as evidence of the practice of the “great perfection” in Indian Buddhism. Vaidyapāda glosses the term as “the second part of the second stage” (rim pa gnyis pa’i rim pa gnyis pa), thus associating it with the four-fold schema of generation and perfection stages found in Buddhajñānapāda’s Muktitilaka (Sukusuma, D 137b.6). I discuss this and another use of the term rdzogs pa chen po in the translations of Buddhajñānapāda’s writings in Chapter Four. 536 rim gnyis ‘di (‘di] sugg. em. based on V (D and P), ‘dis D C P N S). Vaidyapāda clearly indicates that rim gnyis ‘di is to be understood as the “second stage,” the perfection stage (or, according to P, the “perfection stage of the perfection stage”) only, rather than to the “two stages.” (Sukusuma, D 137b.7-138a.1; P 166a.5). I have translated in accordance with his comments, somewhat (but not completely unfeasibly; rim gnyis could very easily be an abbreviation of rim pa gnyis pa made for metrical reasons) against the grain of the Tibetan translation of the root text, which would be more easily translated as the “two stages.” Moreover, given that the topic of the verse is wisdom, “the second stage” really seems to be the better reading. 537 chags] P N S V(D and P), tshogs D C. 538 Dudjom Rinpoche cites this passage of Buddhajñānapāda, as well as Vaidyapāda’s commentary, in his Nyingma School (Dorje and Kapstein, 313). 445 One should train only in this ever-excellent supreme path In accordance with its previously-described stages! |393| [Dedication and Aspiration] By the stainless virtue, perfectly white like the light of the snow[-white] moon, That has arisen from compiling these oral instructions May some539 fortunate beings of the future meet with this truth And with sincere faith may they take it up and train in this supreme [truth]! |394| When one has been cleansed and sprinkled and made pure, and thus become a great ācārya540 Who holds all of the tantras, and brings others to connect with all tantras,541 And having perfectly realized the first stage and purified all stains, May the yogin become a suitable vessel for illusory wisdom!542 |395| Through respectfully [serving at] the feet of a compassionate guru And by means of that which has the rabbit-holder’s form,543 may one’s mindstream be perfectly ripened So that the field is purified,544 and one perfectly realizes the reality of phenomena to be illusory and the like: In this way may all beings, like Maitreya and others, arrive [in that state].545 |396| 546 Through the blessings of the sahaja [guru]547 [and] the great compassionate revered master,548 [One] encounters bliss, through which one [realizes] the undeciving truth, just as it is, The supreme, great pure essence of all things, the drop which is the sixteenth part,549 539 ‘ga’] P N S, dga’ D C 540 Vaidyapāda explains that this is by means of receiving the ācārya initiation (Sukusuma, 128b.2). This line and Vaidyapāda’s commentary on it are cited in Jamgön Kongtrül’s Systems of Buddhist Tantra (Kongtrül, 205). 541 Vaidyapāda here specifies that all the tantras refers to “the Kriyā, Cārya, Yoga and Yoganiruttara mantras and tantras” (bya ba dang / spyod pa dang/ rnal ‘byor dang/ rnal ‘byor bla na med pa’i sngags dang rgyud) (Sukusuma, D 138b.3; P 167a.2-3) 542 Here Vaidyapāda associates the vessel mentioned here with the vase (kalaśa) initiation, and identifies this passage as referring to the receiving of the kalaśābhiṣeka (Sukusuma, D 138b.4). The next two verses refer to the guhya- and prajñājñānābhiṣekas, respectively. 543 This is a reference to the moon, and therefore a metaphor for the bindu of bodhicitta. 544 I believe this is yet another instance in which Buddhajñānapāda takes a Mahāyāna concept—here the concept of zhing sbyang ba, the “cultivation/purification of the [buddha]field,” and reenvisions it according to a tantric paradigm. Vaidyapāda explains the zhing here as referring to the aggregates of the yogin himself (Sukusuma, D 138b.6-7). Thus the field that is purified here is indeed the body of the yogin himself. This is an internalization of the concept of the purification of the field, directing it towards the locus of the yogin’s body—the macrocosm having become microcosm. This supports Jacob Dalton’s (2004) analysis of the interiorization of ritual during precisely this period. 545 This line is a bit gramatically unclear. Khenchen Chodrak Tenphel explains that the basic sense is the aspiration for all beings to follow in Maitreya’s footsteps (Khenchen Chodrak Tenphel, personal communication, March 2016). 546 The grammar of this verse is somewhat unclear, but the general sense seems clear enough. This verse is a reference to the guhyābhiṣeka. Much of the language used in this verse reflects the language in the verses above on that initiation. 547 Vaidyapāda identifies the sahaja as the “sahaja guru,” which in this system refers to the consort, and incicates that her “blessing” refers to union with her (Sukusuma, D 139a.1). While the previous two verses refer to the kalaśābhiṣeka and guhyābhiṣeka, respectively, this one refers to the prajñājñānābhiṣeka. 548 Vaidyapāda explains that this refers to the causal guru, which in this system is the guru from whom one receives initiation and instruction (Sukusuma, D 139a.1-2). 549 Vaidyapāda explains that this is bodhicitta (Sukusuma, D 139a.3). 446 Achieved through resting,550 the great instruction551 —may you come to encounter this! |397| Having encountered this, with a mind filled with deep respect, With neck held low,552 and through investigating,553 the self-aware dharmakāya is attained. [Thus] may the three realms be filled with awakened body, speech, and mind And uncountable emanations, liberating all beings from existence! |398| [Colophon] This completes [the treatise] called Training in the Suchness of the Second Stage, the oral instructions of the bodhisattva Mañjuśrī, a lineage passed from mouth to mouth, which were compiled by the great maṇḍalācārya,554 Buddhaśrījñānapāda. It was translated, edited, and finalized by the great Indian scholar555 Kamalaguhya and the great Tibetan lotsāwa Ngadak556 Lha Yeshe Gyaltsen. 550 Vaidyapāda clarifies that it is “attained through resting” since it is encountered through the winds resting in the central channel (Sukusuma, D 139a.3-4). 551 Vaidyapāda notes that this is called a “great instruction” because it is encountered by means of method (Sukusuma, D 139a.4). 552 Vaidaypāda explains this refers to keeping undistracted focus on the generation stage practice (Sukusuma, 139a.5). 553 Vaidyapāda explains that this means examining the innate nature via the second stage of practice (Sukusuma, 139a.5). 554 om.] sugg. em., lha D C P N S. I suggest omitting lha here, though it is present in all recensions of the Dvitīyakrama. To call Buddhajñānapāda a deva would indeed be a very unusual epithet, and I believe it is more likely that the lha from the translator Lha Yeshe Gyaltsen’s name was somehow added in front of Buddhajñānapāda’s name, as well, in a scribal error. 555 mkhan po 556 P N S om. 447 Bibliography Primary Sources Indic Sources in Sanskrit or Tibetan1 Scriptures Akṣayamatinirdeśa-sūtra. • ‘Phags pa blo gros mi zad pas bstan pa zhes bya ba theg pa chen po’i mdo. Tōh. 175. Sde dge bka’ ‘gyur, vol. 60 (mdo sde, ma), ff. 79a.1-174b.7. Nīlāmbaradharavajrapāṇi-tantra. • Bcom ldan ‘das phyag na rdo rje gos sngon po can gyi rgyud ces bya ba. Tōh. 498. Sde dge bstan ‘gyur, vol. 87 (rgyud ‘bum, da), ff. 158a-167a. • 2013-16. The Tantra entitled “The Blue-Clad Blessed Vajrapāṇi.” Translation by Dharmachakra Translation Committee for 84000: Translating the Words of the Buddha. http://read.84000.co/translation/toh498.html Caṇḍamahāroṣaṇa-tantra. • 2016. The Glorious Canḍamahāroṣaṇa Tantra “The Sole Hero.” Translation and Sanskrit Edition by Dharmachakra Translation Committee for 84000: Translating the Words of the Buddha. http://read.84000.co/translation/toh431.html • See George 1974. Caturdevīparipṛcchā-tantra. • Lha mo bzhis yongs su zhus pa. Tōh. 446. Sde dge bka’ ‘gyur, Vol. 81 (rgyud, ca), ff. 277b.3-281b.7. • Pe king bstan ‘gyur, rgyud, ca, ff. 294b.5-351a.5. Daśabhūmika-sūtra. • Sa bcu pa’i mdo. Tōh. 44. Sde dge bka’ ‘gyur, Vol. 36 (‘Phal chen, kha). Gaṇḍavyūha-sūtra. • Gaṇḍavyūhasūtram. 1960. edited by P.L Vaidya. Darbhanga: The Mithila Institute. Guhyagarbha-tantra. 1 Unless otherwise noted references to page numbers in the dissertation, translation, and notes are always from the Derge edition of the Kangyur and Tengyur. All passages of canonical texts that I have translated in the dissertation or the notes have been critically edited from at least both the Derge and the Peking editions (as is obvious in my citations when there are divergences in the two editions, but not obvious when there are not), but the page and line numbers for many such passages are still given only for the Derge edition. 448 • Dpal gsang ba’i snying po de kho na nyid nges pa. Tōh. 832. Sde dge bka’ ‘gyur, Rnying rgyud, kha, ff. 110b.1-132a.7. • Rnying ma rgyud ‘bum. Sde dge par ma, Vol. 9, ff. 1a-26b. Guhyasamāja-tantra. • The Guhyasamāja Tantra: A New Critical Edition. 1978. edited by Yukei Matsunaga. Osaka: Toho Shuppan, Inc. • Dpal gsang ba ‘dus pa. Tōh. 442. Sde dge bka’ ‘gyur, Vol. 81 (rgyud, ca), ff. 90a.1- 148a.6. • Pe king bstan ‘gyur, rgyud, ca, ff. 95b.5-157b.5. • Dpal gsang ba ‘dus pa. Rnying ma rgyud ‘bum. Sde dge par ma, Vol. 12, ff. 89a-157a. • See Fremantle 1971. Vajramālā-tantra. • Rnal ‘byor chen po’i rgyud dpal rdo rje phreng ba mngon par brjod pa rgyud thams cad kyi snying po gsang ba rnam par phye ba zhes bya ba. Tōh. 445. Sde dge bka’ ‘gyur, Vol. 81 (rgyud, ca), ff. 208a.1-277b.3. Samājottara. • The Guhyasamāja Tantra: A New Critical Edition. 1978. edited by Yukei Matsunaga. Osaka: Toho Shuppan, Inc. • Rgyud phyi ma. Tōh. 443. Sde dge bka’ ‘gyur, Vol. 81 (rgyud, ca), ff. 148a.6-157b.7. • Pe king bstan ‘gyur, rgyud, ca, ff. 148a.6-157b.7. • In Dpal gsang ba ‘dus pa. Rnying ma rgyud ‘bum. Sde dge par ma, Vol. 12, ff. 89a-157a. Saṃdhivyākaraṇa-tantra. • Dgongs pa lung bstan pa zhes bya ba’i rgyud. Tōh. 444. Sde dge bka’ ‘gyur, Vol. 81 (rgyud, ca), ff. 158a.1-207b.7. Sampuṭa-tantra. • Forthcoming. Dharmachakra Translation Committee for 84000: Translating the Words of the Buddha. http://read.84000.co/section/O1JC114941JC21412.html2 Sarvabuddhasamāyoga-tantra. • Szántó, Péter-Dániel Szántó. Sarvabuddhasamāyogaḍākinījālaśaṃvara. Unpublished draft Sanskrit edition. • Dpal sangs rgyas thams cad dang mnyam par sbyor ba mkha’ ‘gro ma sgyu ma bde ba’I mchog ces bya ba’i rgyud phyi ma. Tōh. 366. Sde dge bka’ ‘gyur, Vol. 77 (rgyud, ka), ff. 193a.6-212a.7). Sarvatathāgatatattvasaṃgraha-tantra. 2 I am grateful to James Gentry and Wiesiek Mical of the Dharmachakra Translation Committee for sharing their draft translation of the Sampuṭa-tantra with me. 449 • Sarva-tathāgata-tattva-saṅgraha-nāma-mahāyāna-sūtra: A Critical edition based on a Sanskrit manuscript and Chinese and Tibetan translations. 1981. edited by Isshi Yamada. Delhi: Jayyed Press. • De bzhin gshegs pa thams cad kyi de kho nan yid bsdus pa zhes by aba theg pa chen po’I mdo. Tōh. 479. Sde dge bka’ ‘gyur, Vol. 84 (rgyud, nya), ff. 1b.1-142a.7. Vajrajñānasamuccaya-tantra. • Ye shes rdo rje kun las btus pa zhes bya ba’i rgyud. Tōh. 447. Sde dge bka’ ‘gyur, Vol 81 (rgyud, ca), ff. 282a.1-286a.6. Indic Commentaries in Sanskrit or Tibetan Abhisamālaṃkārāloka of Haribhadra. • See Sparham 1989. • ‘Phags pa shes rab kyi pha rol tu phyin pa brgyad stong pa’i bshad pa mngon par rtogs pa’i rgyi gyi snang ba. Tōh. 3791. Sde dge bstan ‘gyur. Vol. 85 (shes phyin, cha), ff. 1b.1-341a.7. Abhiṣekaratnāloka of Prajñāgupta. • Dbang rin chen snang ba. Tōh. 1333. Sde dge btsan ‘gyur, rgyud, ta, ff. 294b.2-303a.5.

• Abhiṣekavidhi of Prajñāśrī.

• Dbang bskur ba’i cho ga. Tōh. 1269. Sde dge bstan ‘gyur, rgyud, ta, ff. 34b.7-49a.7. Acintyādvayakramopadeśa of Kuddālapāda. • In Samdhong Rinpoche and Vrajvallabh Dwivedi, ed. 1987. Guhyāda-Aṣṭasiddhisaṇgraha. pp. 193-208. Sarnath: Central Institute of Higher Tibetan Studies, Rare Buddhist Texts Project. Ātmasādhanāvatāra of Buddhajñānapāda. • Bdag sgrub pa la 'jug pa zhes bya ba. Tōh. 1860. Sde dge bstan ‘gyur, rgyud, di, ff. 52a.7-62a.7. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 62b.1-75a.6. Ātmārthasiddhikara of Vaidyapāda. Ātmārthasiddhikara-nāma-yogakriyākrama. • Bdag gi don grub par byed pa zhes bya ba rnal ‘byor gyi bya ba’i rim pa. Tōh. 1878. Sde dge bstan ‘gyur, rgyud, pi, ff. 84b.2-94b.2. • Pe king bstan ‘gyur, rgyud ‘grel, thi, 100a.7-112a.8. Bhaṭṭārakāryajambhalajalendra-sādhana of Buddhajñānapāda. • Rje btsun 'phags pa gnod 'dzin chu dbang gi sgrub pa'i thabs. Tōh. 1861. Sde dge bstan ‘gyur, rgyud, di, ff. 62a.7-64b.1. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 75a.6-77b.5. 450 Bodhipathapradīpapañjikā of Atīśa. • Byang chub lam gyi sgron ma’i dka’ ‘grel. Tōh. 3948. Sde dge bstan ‘byur, vol. khi, ff. 241a4-293a4. Caturaṅga-sādhana of Buddhajñānapāda. Caturaṅgasādhanopāyikā-samantabhadrī-nāma. • = Samantabhadra-sādhana of Buddhajñānapāda; See Samantabhadra-sādhana of Buddhajñānapāda. Dpal gsang ba 'dus pa'i rgyud kyi man ngag gi rgya mtsho thigs pa attr. to Viśvamitra.3 • Dpal gsang ba 'dus pa'i rgyud kyi man ngag gi rgya mtsho thigs pa. Tōh. 1844. Sd edge bka’ ‘gyur, rgyud, ji, ff. 53b.7-161b.1. Dvitīyakrama of Buddhajñānapāda. Dvitīyakrama-tattvabhāvanā-mukhāgama. • Rim pa gnyis pa'i de kho na nyid sgom pa zhes bya ba'i zhal gyi lung. Tōh. 1853. Sde dge bstan ‘byur, rgyud, di, ff. 1a-17b. • Co ne bstan ‘gyur, rgyud ‘grel, di, ff. 1a-17b. • De pe bsdur ma bstan ‘gyur, vol. 21, pp. 852-894. • Snar thang bstan ‘gyur, rgyud ‘grel, ti, ff. 1a-20a. • Gser bris ma bstan ‘gyur, rgyud ‘grel, ti, ff. 1a-22b. • Peking bstan ‘gyur, rgyud ‘grel, ti, ff. 1a-21b.

• Gativyūha of Buddhajñānapāda.

• Stang stabs kyi bkod pa. Tōh. 1864. Sde dge bstan ‘byur, rgyud, di, ff. 66b.2-69a.4. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 80a.4-83a.6. Guhyajambhalasādhana of Buddhajñānapāda. • Gsang ba'i dzam bha la'i sgrub thabs. Tōh. 1862. Sde dge bstan ‘gyur, gyud, di, ff. 64b.1- 65a.7. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 77b.5-78b.6. Guhyasamājamaṇḍalavidhi of Dīpaṃkarabhadra. • In Dhīḥ Journal of Rare Buddhist Texts Research Unit, vol. 42. 2006. Saranath, pp. 109- 154. • online edition input by Sabine Klein-Schwind and proofread and revised by Harunaga Isaacson. http://www.tantric-studies.uni-hamburg.de/e-texts/bauddha/GuSaMaVi.txt • See Bahulkar 2010. • Dpal gsang ba ‘dus pa’i dkyil ‘khor gyi cho ga. Tōh. 1865. Sde dge bka’ ‘gyur, rgyud, di, ff. 69a.4-87a.3. Guhyasamājamaṇḍalopāyikā-ṭīkā of Vaidyapāda. • Dpal gsang ba ‘dus pa’i dkyil ‘khor gyi sgrub pa’i thabs rnam par bshad pa. Tōh. 1873. Sde dge bstan ‘gyur, rgyud, ni, ff. 179a.1-218a.7 3 I have included this text here in the section on Indic texts because it is found in the Tengyur, despite the fact that no Sanskrit title is provided in the Tibetan translation and my conclusion, based on this and a number of other factors, that the text is in fact a Tibetan composition rather than a translation of an Indic text. See Chapter 1, note 108. 451 • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 498a.6-549a.8. Jinamārgāvatāra of Kaśmiri Buddhaśrījñāna. • Rgyal ba’i lam la ‘jug pa. Tōh. 3964. Sde dge bstan ‘gyur, dbu ma, gi, ff. 201b.6-235b.1. Jinamārgāvatārodbhavapraṇidhāna of Kaśmiri Buddhaśrījñāna. • Rgyal ba'i lam la 'jug pa las byung ba'i smon lam. Tōh. 4391. Sde dge bstan gyur, sna tshogs, nyo, ff. 319b.4-320b.5. • Pe king bstan ‘gyur, ngo mtshar bstan bcos, sna tshogs, mo, ff. 307a.7-308b.4. Kośalālaṃkāra of Śākyamitra. Kośalāṃkāra-tattvasaṃgrahaṭīka. • De kho na nyid bsdus pa’i rgya cher bshad pa ko sa la’i rgyan. Tōh. 2503. Sde dge bstan ‘gyur, rgyud, yi, ff. 1b.1-245a.7, rgyud, ri, 1b.1-202a.5. Kṛṣṇayāmarisādhana of Śrīdhara. • Gshin rje gshed nag po’i sgrub thabs. Tōh. 1923. Sde dge bstan ‘gyur, rgyud, mi, ff. 1b.3-7b.4. Kusumāñjali of Ratnākaraśānti. Kusumāñjali-guhyasamājanibandha-nāma. • Gsang ba 'dus pa'i bshad sbyar snyim pa'i me tog ces bya ba. Tōh. 1851. Sde dge bstan ‘gyur, rgyud, ti, ff. 202b.1-325a.7. • Pe king bstan ‘gyur, rgyud ‘grel, ji, ff. 233b.8-383a.8. Laṅkāvatāra-vṛtti of Jñānavajra. Āryalaṅkāvatāra-nāma-mahāyānasūtravṛttitathāgatahṛdālaṃkāra- nāma. • 'Phags pa lang kar gshegs pa zhes bya ba theg pa chen po'i mdo'i 'grel pa de bzhin gshegs pa'i snying po'i rgyan zhes bya ba. Tōh. 4019. Sde dge bstan ‘gyur, mdo ‘grel, pi, ff. 1b.3-310a.7. Madhyamakaratnapradīpa of Bhāviveka. • Dbu ma rin po che’i sgron ma. Tōh. 3854. Sde dge bstan ‘gyur, dbu ma, tsha, ff. 259b.3- 289a.7. Mahābalividhi of Vaidyapāda. • Gtor ma chen po’i cho ga. Sde dge bstan ‘gyur, rgyud, pi, ff. 75b.4-81a.2. Mahāmāyā-sādhana of Ratnākaraśānti. • Sgyu ma chen mo’i sgrub thabs. Sde dge bstan ‘gyur, rgyud, ya, ff. 269b-273b. • Sādhanamālā #221. Benoytosh Bhattacarya, ed. Baroda: Oriental Institute, 1968, pp. 458-64.

• Mahāpratisararārakṣā attr. Buddhajñānapāda.

• So sor ‘brang ma chen mo’i bsrung ba. Tōh. 3124. Sde dge bka’ ‘gyur, rgyud, pu, ff. 220a.6-224b.3. • Pe king bstan ‘gyur, rgyud ‘grel, tu, ff. 248b.5-254b.5. Mahāvairocana-tantra-vṛtti of Buddhaguhya. 452 • Rnam par snang mdzad mngon par byung chub pa’i rgyud chen po’i ‘grel bshad. Tōh. 2663. Sde dge bka’ ‘gyur, rgyud, nyu, ff. 76b.7-337a.3. Mahāyānalakṣaṇasamuccaya of Buddhajñānapāda. • Theg pa chen po’i mtshan nyid kun las bsdus pa. Tōh. 3905. Sde dge bka’ ‘gyur, dbu ma, a, ff. 296a.3-305a.7. • Pe king bstan ‘gyur, dbu ma, ha, ff. 402a.1-413b.8. Maṇimālā of Nāgabodhi. Pañcakramaṭīkā-maṇimāḷā-nāma, • Rim pa lnga pa’i bshad pa’i nor bu’i phreng ba zhes bya ba. Tōh. 1840. Sde dge bstan ‘gyur, rgyud, chi, ff. 14a.6-157a.7. • Pe king bstan ‘gyur, rgyud ‘grel, ngi, ff. 9b.2-174b.6. Mañjuvajrodaya of Anonymous. Śrīmañjuvajrodaya-maṇḍalopāyikā-sarvasattvahitāvahā -nāma. • Dpal ‘jam pa’i rdo rje ‘byung ba’i dkyil ‘khor gyi cho ga sems can thams cad kyi bde ba bskyed pa zhes bya ba. Tōh. 2590. Sde dge bstan ‘gyur, rgyud, ngu, ff. 225a.5-274a.7. Mgon po dmar po’i tshe bsgrub kyi zhal gdams of Śāvaripa. • P 4927. Pe king bstan ‘gyur, rgyud ‘grel, zu, ff. 269b.5-280b.8. Mukhāgama of Śaykamitra (often attributed to Buddhajñānapāda) • Zhal gyi lung. Tōh. 1854. Sde dge bstan ‘gyur, rgyud, di, ff. 17b.3-28b.6. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 20a.5-33b.3. Muktitilaka of Buddhajñānapāda. Muktitilaka-nāma. • Grol ba’i thig le. Tōh. 1859. Sde dge bstan ‘gyur, rgyud, di, ff. 47a.1-52a.7. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 56a.4-62b.1. Muktitilaka-vyākhyāna of Vaidyapāda. Muktitilaka-nāma-vyākhyāna. • Grol ba'i thig le zhes bya ba'i rnam par bshad pa. Tōh. 1870. Sde dge bstan ‘gyur, rgyud, ni, ff. 45b.4-59a.7. • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 330a.7-347b.6. Nāmamantrārthāvalokinī of Vilāsavajra. Āryanāmasaṃgīti-ṭikā-nāmamantrārthāvalokinī -nāma. • ‘phags pa mtshan yang dag par brjod pa’i rgya che ‘grel pa mtshan gsang sngags kyi don du rnam par lta ba zhes bya ba. Tōh. 2533. Sde dge bstan ‘gyur, rgyud, khu, ff. 27b.1-115b.3. Niṣpannayogāvalī of Abhayākaragupta. • Yong-Hyun Lee (ed.) 2004. Niṣpannayogāvalī by Abhayākaragupta: A New Critical Edition of the Sanskrit Text (Revised Edition). Baegun Press: Seoul. Pradīpoddyotana of Candrakīrti. 453 • “Guhyasamājatantrapradīpodyotanaṭīkā Ṣaṭkoṭivyākhyā of Ācārya Candrakīrti (Chapters 7-9)” In Dhīḥ, Vol. 50, Kārtika Pūrṇimā, 2010., ed. Ngawang Samten and S. S. Bahulkar. Sarnath: Central University of Tibetan Studies. Pramāṇavārttika of Dharmakīrti. • Pandeya, Prof Ram Chandra, ed. Pramāṇavārttikam of Ācārya Dharmakīrti with the Commentaries of Svopajñavṛtti of the Author and Pramāṇavārttikavṛtti of Manorathanandin. India: Motilal Banarsidass, 1989. Ratnāvalī of Ratnākaraśanti. Piṇḍīkṛtasādhanopāyikāvṛtti-ratnāvalī-nāma • Mdor bsdus pa’i sgrub thabs kyi ‘grel pa rin chen phreng ba zhes bya ba. Tōh. 1826. Sde dge bstan ‘gyur, rgyud, ci, ff. 1b.1-95a.6. • Pe king bstan ‘gyur, rgyud ‘grel, gi, ff. 273a.7-370b.8. Pañcakrama of Nāgārjuna. • Mimaki, Katsumi and Tōru Tomabechi. Pañcakrama: Sanskrit and Tibetan Texts Critically Edited with Verse Index and Facsimile Edition of the Sanskrit Manuscripts. 1994. Tokyo: The Centre for East Asian Cultural Studies for Unesco. Prajñāpāramitāpiṇḍārtha of Dignāga. • http://gretil.sub.uni-goettingen.de/gretil/1_sanskr/6_sastra/3_phil/buddh/bsa011_u.htm Prañjāpradīpamūlamadhyamakavṛtti of Bhāviveka. • Dbu ma’i rtsa ba’i ‘grel pa shes rab sgron ma. Tōh. 3853. Sde dge bstan ‘gyur, dbu ma, tsha, ff. 45b.4-259b.3. Ratnakaraṇḍodghāta of Atīśa. Ratnakaraṇḍodghāta-nāma-madhyamakopadeśa. • Dbu ma’i man ngag rin po che’i za ma tog kha phye ba zhes bya ba. Tōh 3930. Sde dge bstan ‘gyur, dbu ma, ki, 96b.1-116b.7. Ratnāmati of Vaidyapāda. Ratnāmati-nāma-sādhana. • Mi shigs pa’i rin po che zhe bya ba’i sgrub thabs. Sde dge bstan ‘gyur, rgyud, pi, ff. 81a.2-84b.2. Ratnāvalī of Ranākaraśānti. Piṇḍīkṛtasādhanopāyikāvṛtti-ratnāvalī-nāma. 􀀀Mdor bsdus pa’i sgrub thabs kyi ‘grel pa rin chen phreng ba zhes bya ba. Tōh. 1826. Sde dge bstan ‘gyur, rgyud, ci, ff. 1b.1-95a.6. Sahajāloka of Śrīdhara. Śrīyamāritantrapañjikāsahajāloka. • Dpal gshin rje gshed kyi rgyud kyi dka’ ‘grel lhan cig skyes pa’i snang ba zhas bya ba. Tōh. 1918. Sde dge bstan gyur, rgyud, bi, ff. 81b.-123b.7. Samantabhadrārthasaṃgraha of Buddhajñānapāda(?). • Kun tu bzang po’i bsdus don. Peking Tengyur 2744. This title appears in the Peking Tengyur index but the text itself is absent in the appropriate place in the Tengyur. 454 Samantabhadra-sādhana of Buddhajñānapāda = Caturaṅga-sādhana of Buddhajñānapāda4 • Samantabhadra-nāma-sādhana. Kun tu bzang po zhes bya ba'i sgrub pa'i thabs. Tōh1855. Sde dge bstan ‘gyur, rgyud, di, ff. 28b.6-36a.5. • Kun tu bzang po zhes bya ba'i sgrub pa'i thabs. Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 33b.3-42b.5. • Caturaṅgasādhanopāyikā-samantabhadrī-nāma. Yan lag bzhi pa'i sgrub thabs kun tu bzang mo zhes bya ba. Tōh. 1856. Sde dge bstan ‘gyur, rgyud, di, ff. 36a.5-42b.5. • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 42b.5-51a.4. Samantabhadrasādhana-vṛtti of Śrīphalavajra. • Kun tu bzang po’i sgrub pa’i thabs kyi ‘grel pa. Tōh. 1867. Sde dge bstan ‘gyur, rgyud, di, ff. 139b.3-187b.3. Samantabhadrī-ṭīkā of Vaidyapāda. Caturaṅgasādhanopāyikāsamantabhadrī-nāma-ṭīkā. • Yan lag bzhi pa'i sgrub thabs kun tu bzang mo zhes bya ba'i rnam par bshad pa. Tōh. 1872. Sde dge bstan ‘gyur, rgyud, ni, ff. 130b.1-178b.7. - Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 440b.8.498a.6. Saṃkṣiptābhiṣekhaprakriyā of Advayavajra. • dbang gyi bya ba mdor bsdus pa. Tōh 2244. Sde dge bstan ‘gyur. Rgyud, wi, ff. 125b.7- 134b.3. Saṃyagvidyākara of Vaidyapāda. Saṃyagvidyākara-nāma-uttaratantra-vyākhyāna. • Yang dag rig byed ces bya ba phyi ma’i rgyud kyi rnam par bshad pa. Tōh. 1850. Sde dge bstan ‘gyur, rgyud, ti, ff.170a.3-202a.7 • Pe king bstan ‘gyur, rgyud ‘grel, ji, 208a.7-202a.7. Saṃcayagāthā-pañjikā of Buddhajñānapāda. • Sdud pa tshigs su bcad pa’i dka’ ‘grel. Tōh. 3798. Sde dge bka’ ‘gyur, sher phyin, nya, ff. 235b.5-222a.6. • Pe king bstan ‘gyur, mdo ‘grel, nya, ff. 135b.5-223a.8. Sāramañjarī of Samantabhadra. • Szántó, Péter Daniel, ed. (unpublished). The Sāramañjarī of Samantabhadra, a Commentary to the Samantabhadrasādhana of Jñānapāda: Critical Edition of the ‘Pāla Recension.’ • Yan lag bzhi pa’i sgrub thabs kyi rgya cher bshad pa snying po snye ma. Tōh. 1869. Sde dge bka’ ‘gyur, rgyud, ni, ff. 1b.1-45b.4. • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 274b.4-330a.6. • See Tanaka, 2010, pp. 505-550. • See Kano 2014. Saptaṅga of Vāgīśvarakīrti. • Yan lag bdun pa. Tōh. 1888. Sde dge bstan ‘gyur, rgyud, pi, ff. 190a.3-203a.3. 4 What might otherwise appear to be two different compositions, preserved with different titles in the Tibetan canon, are, in fact, the same text. 455 • Pe king bstan ‘gyur, rgyud ‘grel, thi, ff. 224b.5-238b.8. Sarvasamayasaṃgraha of Atīśa. • Dam tshig thams cad bsdus pa zhes bya ba. Tōh. 3735. Sde dge bstan ‘gyur, rgyud, tshu, ff. 44a.1-49b.1. • Pe king bstan ‘gyur, rgyud ‘grel, nu, 253b.6-259b.8. Sarvayānālokakaravaibhāṣya of Subhūtighoṣa • Theg pa thams cad snang bar byed pa’i bye brag tu bshad pa zhes bya ba. Tōh. 3907. Sde dge bstan ‘gyur, dbu ma, a, ff. 306a4-313a.7. Siddhisaṃbhavanidhi of Vaidyapāda. Śrīguhyasamājasādhana-siddhisaṃbhavanidhi. • Dpal gsang ba ‘dus pa’i sgrub pa’i thabs dngos grub ‘byung ba’i gter. Tōh. 1874. Sde dge bstan ‘gyur, rgyud, pi, ff. 1a.1-69b.6. • Pe king bstan ‘gyur, rgyud ‘grel, thi, 1a.1-83a.4. Śrīguhyārthaprakāśasamahādbhūta of Gambhīravajra. • Dpal zab mo’i don gsla ba rmad du byung ba chen po zhes bya ba. Tōh. 1200. Sde dge bstan ‘gyur, rgyud, ja, ff. 111a.2-154b.1. • Pe king bstan ‘gyur, rgyud ‘grel, wa, ff. 121a.8-170a.1. Śrīguhyasamājamaṇḍalaviddhi-ṭīkā of Ratnākaraśānti. • Dpal gsang ba ‘dus pa’i dkyil ‘khor gyi cho ga’i ‘grel pa. Tōh.1871. Sde dge bstan ‘gyur, rgyud, ni, ff. 59a.7-130a.7. • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 347b.6-440b.7. Śrīguhyasamājamaṇḍalopāyikā of Kṛṣṇācārya. • Dpal gsang ba ‘dus pa’i dkyi ‘khor gyi cho ga. Tōh. 1819. Sde dge bstan ‘gyur, rgyud, ngi, ff. 247b.1-258b.1. • Pe king bstan ‘gyur, rgyud ‘grel, gi, ff. 228a.3-242a.4. Śrīguhyasamājatantranidāṇagurūpadeśabhāsya of Vilāsavajra. • Gsang ba 'dus pa'i rgyud kyi gleng gzhi bla ma'i man ngag gi bshad pa. Tōh. 1910. Sde dge bstan ‘gyur, rgyu, phi, ff. 89b.1-97b.5. • Pe king bstan ‘gyur, rgyud ‘grel, di, ff. 87b.6-96b.6. Śrīguhyasamājatantravivaraṇa of Thagana. • Dpal gsang ba 'dus pa'i rgyud kyi 'grel pa. Tōh. 1875. Sde dge bstan ‘gyur, rgyud, ji, ff. 161b.1-244a.7. • Pe king bstan gyur, rgyud ‘grel, ci, ff. 185a.7-244a.7. Śrīherukasādhana of Buddhajñānapāda. • In Sādhanamālā. Ed. Bentoysh Bajracharya. Vadodara: Oriental Institute, 1968, No. 243. • Dpal he ru ka’i sgrub pa’i thabs. Tōh. 1857. Sde dge bstan ‘gyur, rgyud, di, ff. 42b.6- 43a.7. 456 • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 51a.4-51b.7. Śrīherukasādhana-vṛtti of Anonymous. • Dpal he ru ka’i sgrub thabs kyi ‘grel pa. Tōh. 1858. Sde dge bstan ‘gyur, rgyud, di, ff. 43a.7-47a.1 • Pe king bstan gyur, rgyud ‘grel, ti, ff. 51b.7-56a.4. Śrīraktayamāri-sādhana attr. Buddhajñānapāda. • Dpal gshin rje gshed dmar po’i sgrub thabs. Tōh. 2084. Sde dge bstan ‘gyur, rgyud, tsi, ff. 160a.6-161a.5. • Pe king bstan ‘gyur, rgyud ‘grel, zhu, ff. 177a.1-178a.8. Śrīraktayamāri-sādhana of Śrīdhara. • Dpal gshin rje gshed dmar po’i sgrub thabs. Tōh. 2023. Sde dge bstan ‘gyur, rgyud, tsi, ff. 88a.5-95a.1. • Pe king bstan ‘gyur, rgyud ‘grel, pi, ff. 387a.2-395b.6. Śrīsamantabhadrasādhana-vṛtti of Thagana • Dpal kun tu bzang po'i sgrub thabs kyi 'grel pa. Tōh. 1868. Sde dge bstan ‘gyur, rgyud, di, ff. 187b.4-231a.7. • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 224a.2-274b.4. Sukusuma of Vaidyapāda. Sukusuma-nāma-dvitīyakramatattvabhāvanāmukhāgama-vṛtti. • Mdzes pa'i me tog ces bya ba rim pa gnyis pa'i de kho na nyid bsgom pa zhal gyi lung gi 'grel pa. Tōh. 1866. Sde dge bstan ‘gyur, rgyud, di, ff. 87a.3-139b.3. • Pe king bstan ‘gyur, rgyud ‘grel, ti, ff. 104b.2-168a.6. Sūtrālaṃkāravṛttibhāṣya of Sthiramati. • Mdo sde rgyan gyi ‘grel bshad. Tōh. 4034. Sde dge bstan ‘gyur, sem tsam, mi, ff. 1b.1- 283a.7. • Pe king bstan ‘gyur, sems tsam, mi and tsi, 1a.1-298. Trikāyavākcittadhiṣṭḥanoddeśa attr. Buddhajñānapāda. • Lus ngag yid gsum byin gyis brlab pa’i man ngag ces bya ba. Tōh. 2085. Sdge dge bstan ‘gyur, tsi, ff. 161a.6-161b.5. • Pe king bstan gyur, rgyud ‘grel, zhu, ff. 178a.8-179a.2. Trisattvasamādhisamāpatti attr. Buddhajñānapāda. • sems dpa’ gsum gyi ting nge ‘dzin la mnyam par bzhag pa. Tōh. 2086. Sde dge bstan gyur, rgyud, tsi, ff. 161b.5-126b.5. • Pe king bstan ‘gyur, rgyud ‘grel, zhu, ff. 179a.2-180a.7. Triśaraṇasaptati of Candrakīrti. • Gsum la skyabs su ‘gro ba bdun cu pa. Tōh. 3971. Sde dge bstan gyur, dbu ma, ki, ff. 251a.1-253b.2. • Pe king bstan ‘gyur, dbu ma, khi, ff. 291a.7-295a.2. 457 Vistarajambhalasādhana of Buddhajñānapāda. • In Sādhanamālā. Ed. Bentoysh Bajracharya. Vadodara: Oriental Institute, 1968, No 285. • Rgya che ba'i dzam bha la'i sgrub pa'i thabs. Tōh. 1863. Sde dge bstan ‘gyur, rgyud, di, ff. 65a.7-66b.2. • Pe king bstan gyur, rgyud ‘grel, ti, ff. 78b.6-80a.4. Yogasapta of Vaidyapāda. Yogasapta-nāma-caturabhiṣekaprakaraṇa • Sbyor ba bdun pa zhes bya ba dbang bzhi’i rab tu byed pa. Tōh. 1985. Sde dge bstan ‘gyur, rgyud, pi, ff. 69b.6-75b.4. • Pe king bstan ‘gyur, rgyud, thi, ff. 83a.4-90a.1. Indigenous Tibetan Sources Chos ‘byung me rtog snying po brang rtsi’i bcud of Nyangral Nyima Oser. Gangs can rig mdzod, vol. 5. Lhasa: Bod lojongs mi dmangs dpe skrun khang, 1988. Deb ther sngon po of Gös Lotsāwa Zhonnu Pel. Vols I and II. Chengdu: Si khron mi rigs dpe skrun khang, 1984. (TBRC W1KG5762). Dpal grol ba’i thig le’i khrid yig mi shigs pa’i rinpoche che sgrub pa’i thabs of Tāranātha. In Jo nang rje btsun tA ran ā tha’i gsung ‘bum. Dpe sdur ma, Vol. 9. pp.236-60. (BDRC W1PD45495). Grub mtha’ kun shes ‘grel ba of Taktsang Lotsāwa. In Grub mtha’ kun shes rtsa ‘grel. Thimpu: Kunzang Tobgyel and Mani Dorji. 1976. Gsang ‘dus chos ‘byung of Amye Zhab. Dpal gsang ba ‘dus pa’i dam pa’i chos ‘byung ba’i tshul legs par bshad pa gsang ‘dzus chos kun gsal ba’i snyin byed of A mye Zhabs Nga dbang kun dga’ bsod rnams. Dehradun, UP: Sakya Centre, 1985. Gsang ‘dus ye shes zhabs kyi rnam thar dang brgyud pa’i rim pa of Chögyal Phagpa in ‘Gro mgon chos rgyal ‘phag pa’i gsung rab, Vol II, pp 609-618. Bejing: Dpe bsdur ma. (TBRC W2DB4571). Gshin rje chos ‘byung of Amye Zhab. Dpal gshin rje gshad skor gyi dam pa’i chos byung tshul legs par bshad pa ‘jam dpal chos kun gsal ba’i nyin byed of A mye Zhabs Nga dbang kun dga’ bsod rnams. in Bod kyi lo rgyus rnam thar phyogs bsgrigs, vol 20. Zi ling: Mtsho sngon mi rigs dpe skrun khang, 201. Man ngag lta ba’i ‘phreng ba of Padmasambhava. Bylakuppe: Ngagyur Nyingma Institute, 2001. Rdo rje theg pa snags kyi gso sbyong bdud rtsi’i rol mtsho zhes bya ba bzhugs so. In Chokgyur 458 Lingpa et al.The treasury of revelations and teachings of Gter chen mchog gyur bde chen gling pa. Vols 1-39. Expanded redaction of the Mchog gyur bde chen gling pa revelations. New Delhi: Chos Spyod Publications, Tsikey Chokling Rinpoche. 2004. Rin chen gter mdzod chen mo of Jamgön Kongtrül Lodro Thaye. New Delhi: Shechen Publications, 2007-08. Dpal grol ba’i thig le’i khrid yig mi shigs pa’i rinpoche che sgrub pa’i thabs of Tāranātha. In Jo nang rje btsun tA ran ā tha’i gsung ‘bum. Dpe sdur ma, Vol. 9. pp.236-60. (BDRC W1PD45495). Secondary Sources5 Abe, Ryūichi. 1999. The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse. New York: Columbia University Press. Adamek, Wendi L. 2007. The Mystique of Transmission: On Early Chan History and its Contexts. New York: Columbia University Press. Ali, Daud. 2000. “Royal Eulogy as World History: Rethinking Copper-plate inscriptions in Cola India.” in in Inden et al 2000. Querying the Medieval: Texts and History of Practices in South Asia. New York: Oxford UP, pp. 165-229.

---

. 2004. Courtly Culture and Political Life in Early Medieval India. Cambridge:

Cambridge UP.

---

. 2011. “Padmaśrī’s “Nāgarasarvasva” and the World of Medieval Kāmaśāstra.” Journal of

Indian Philosophy, vol 39, No. 1, pp. 41-62. Almogi, Orna. 2014. “The Eighteen Mahāyoga Tantric Cycles: A Real Canon or the Mere Notion of One?” Revue d’Etudes Tibetaines, no. 30, Ocober 2014, pp. 47-110. Arnold, Edward A. 2009. As Long as Space Endures: Essays on the Kālacakra Tantra in Honor of H.H. the Dalai Lama. Boston: Snow Lion Publications. Asher, Catherine B. and Cynthia Talbot. 2006. India Before Europe. Cambridge: Cambridge UP. Bahulkar, S.S. (ed.) 2010. Śrīguhyasamājamaṇḍalavidhiḥ of Ācārya Dīpaṅkarabhadra. Rare Buddhist Texts Series 31. Sarnath: Central University of Tibetan Studies. Bajracarya, Bentoysh, ed. 1968. Sadhanamāla Vols I and II. Baroda: Oriental Institute. 5 Including primary sources in translation, listed by translator. 459 Bentor, Yael. 2010. “The Convergence of Theoretical and Practical Concerns in a Single Verse of the Guhyasamāja Tantra.” In Tibetan Ritual. edited by José Ignacio Cabezón. New York: Oxford University Press.

---

. 2015. “Interpreting the Body Mandala: Tsongkhapa versus Later Geluk Scholars.”

Revue d’Etudes Tibetaines. No. 31, Feb. 2051, pp. 63-74. Beyer, Stephan. 1973. The Cult of Tara: Magic and Ritual in Tibet. Berkeley: University of California Press. Campbell, John R.B. 2009. Vajra Hermeneutics: A Study of Vajrayāna scholasticism in the Pradīpoddyotana. PhD Dissertation, Columbia University. Chattopadhyaya, Brajadulal. 1994. (Reprint 1997.) The Making of Early Medieval India. Delhi: Oxford UP.

---

. 2005: Studying Early India: Archaeology, Texts and Historical Issues. Delhi:

Permanent Black. Chimpa, Lama and Alaka Chattopadhyaya, trans. 1970. Tāranātha’s History of Buddhism in India. Simla: Indian Institute of Advanced Study. Chaudhuri, K.N. 1985. Trade and Civilization in the Indian Ocean: An Economic History from the Rise of Islam to 1750. Cambridge: Cambridge UP. Cohn, Bernard S. 1987. An Anthropologist among the Historians and Other Essays. New York: Oxford University Press. Cozort, Daniel. 1986. Highest Yoga Tantra: An Introduction to the Esoteric Buddhism of Tibet. Boston: Snow Lion. Dachille, Ray Erin. 2015. The Body Mandala Debate: Knowing the Body through a Network of Fifteenth-Century Tibetan Buddhist Texts. PhD Dissertation, UC Berkeley. Dalton, Catherine. 2014. “Some Observations on the Locus Classicus of the Two Stages of Tantric Practice and its Implications regarding the Relationship between the Samājottara and the Early Jñānapāda School.” Paper delivered at Buddhist Studies Graduate Student Conference, UC Santa Barbara, Feburary, 2014.

---

. forthcoming. The Samājottara. Translated by the Dharmacakra Translation Committee

for 84000: Translating the Words of the Buddha. Dalton, Catherine and Szántó, Péter-Dániel. forthcoming. “Jñānapāda.” Brill Encyclopedia of Buddhism, Volume II. Dalton, Jacob P. 2002. The Uses of the Dgongs pa ‘dus pa’i mdo in the Development of the Rnying-ma School of Tibetan Buddhism. Ph.D. dissertation. University of Michigan 460

---

. 2004. “The Development of Perfection: The Interiorization of Buddhist Ritual in the

Eighth and Ninth Centuries.” In Journal of Indian Philosophy 32: 1-30.

---

. 2005. “A Crisis of Doxography: How Tibetans Organized Tantra during the 8th-12th

Centuries” In Journal of the International Association of Buddhist Studies 28.1: 115-181.

---

. 2011a. The Taming of the Demons: Violence and Liberation in Tibetan Buddhism. New

Haven: Yale UP.

---

. 2011b. “Mahāyoga Ritual Interests at Dunhuang: A Translation and Study of the Codex

IOL TIB J 437/Pelliot Tibétain 324.” in New Studies of the Old Tibetan Documents: Philology, History and Religion. edited by Yishiro Imaeda, Matthew T. Kapstein and Tsuguhito Takeuchi. Tokyo: Research Institute for languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies. pp 293-313.

---

. 2012. “The Questions and Answers of Vajrasattva.” in Yoga in Practice edited by David

Gordon White. Princeton: Princeton UP, pp. 185-203.

---

. 2016. “How Dhāraṇīs WERE Proto-Tantric: Liturgies, Ritual Manuals, and the Origins

of the Tantras.” In David B. Gray and Ryan Richard Overbey, ed., Tantric Traditions in Transmission and Translation. Oxford: Oxford University Press, pp. 199-229. Dalton, Jacob and Sam van Schaik. 2006. Tibetan Tantric Manuscripts from Dunhuang: A Descriptive Catalogue of the Stein Collection at the British Library. Leiden and Boston: Brill. Damron, Ryan. 2014. The Great Illusion: A Preliminary Study of the Mahāmāyā Tantra and its Corpus. M.A. Thesis at U.C. Berkeley. Datta, Amaresh. 1988. Encyclopedia of Indian Literature, Volume II. 2005 reprint. Delhi: Wellwish Printers. Davidson, Ronald M. 1982. “The Litany of Names of Mañjusri: Text and Translation of the Mañjusri-namasaṃgiti,” in Michel Strickmann, ed., Tantric and Taoist Studies in Honour of R.A. Stein, Mèlanges chinois et bouddhiques, vol. XX, pp. 1-69. Brussels: Institut Belge des Hautes Études Chinoises.

---

. 1995. “The Litany of Names of Mañjusri.” in Religions of India in Practice. ed. Donald

Lopez. Princeton: Princeton UP, pp. 104-25.

---

. 2002. Indian Esoteric Buddhism: A Social History of the Tantric Movement. New

York: Columbia University Press.

---

. 2002b. “Reframing Sahaja: Genre, representation, ritual and lineage. Journal of Indian

Philosophy 30 (1), 45-83.

---

. 2005. Tibetan Renaissance: Tantric Buddhism in the Rebirth of Tibetan Culture. New

York: Columbia UP. 461

---

. 2011. “Sources and Inspirations: Esoteric Buddhism in South Asia.” in Esoteric

Buddhism and the Tantras in East Asia edited by Orzech, Sorenson and Payne, pp 19-26. Boston: Brill.

---

. 2012.“Canon and Identity in Indian Esoteric Buddhism.” in Dynamics in the History of

Religions between Asia and Europa: Encounters, Notions, and Comparative Perspectives. ed Volkhard Krech and Marion Steinicke. Boston: Brill, pp. 321-342. Decleer, Hubert. 1998. “Two Topics from the Swayaṃbhū Purāṇa: Who was Dharma-shrī-mitra? Who was Shāntikāra Achārya?” Paper presented at the Conference on the Buddhist Heritage of Nepal Mandala. http://www.aioiyama.net/lrc/papers/cbhnm-ppr.2.htm. De Jong, J.W. 1984. “A New History of Tantric Literature in India.” Acta Indologica VI. Narita: Naritasan Shinshoji. De La Vallée Poussin, Louis. 1936-7. “Musīla et Nārada.” Mélanges chinois et bouddiques 6, pp. 189-222. Desmond, Laura. 2011. “The Pleasure is Mine: The Changing Subject of Erotic Science.” Journal of Indian Philosophy, Vo. 39, No. 1, pp. 15-39. Dharmachakra Translation Committee. 2007. Deity, Mantra, and Wisdom: Development Stage Meditation in Tibetan Buddhist Tantra, with texts by Jigme Lingpa, Paltrul Chökyi Wangpo, and Getse Mahāpaṇḍita. Ithaca: Snow Lion. Dharmachakra Translation Committee for 84000. Tantra of the Blue-Clad Blessed Vajrapāṇi. http://read.84000.co/translation/toh498.html. Doctor, Andreas. 2005. Tibetan Treasure Literature: Revelation, Tradition, and Accomplishment in Visionary Buddhism. Ithaca: Snow Lion. Dorje, Gyurme. 1987. The Guhyagarbhatantra and its XIVth Century Commentary Phyogs bcu mun sel. PhD Dissertation School of Oriental and African Studies. Dorje, Rangjung and Jamgön Kongtrul Lodrö Taye. 2014. The Profound Inner Principles With Jamgön Kongtrul Lodrö Taye’s Commentary Illuminating “The Profound Principles.” Translated, annotated, and introduced by Elizabeth M. Callahan. Boston: Snow Lion. Dudjom Rinpoche. 1991. The Nyingma School of Tibetan Buddhism: its Fundamentals and History. Boston: Wisdom. Eastman, Kenneth W. 1979. “The Dun-huang Tibetan Manuscript of the Guhyasamājatantra” Paper presented at the Japanese Association for Tibetan Studies. Kyōto, Japan

---

. 1981. “The Eighteen Tantras of the Vajraśekhara/Māyājāla” Paper presented at the 26th

International Conference of Orientalists in Japan. Tōkyo, Japan. 462

---

. 1983. “Mahāyoga Texts at Tun-huang.” Bulletin of Institute of Buddhist Cultural

Studies Ryukoku University, No. 22, pp. 42-60. English, Elizabeth. 2002. Vajrayoginī: Her Visualizations, Rituals, and Forms. Boston: Wisdom Publications. Farrow, G.W. and I. Menon. 2001. (First edition 1992). The Concealed Essence of the Hevajra Tantra. Delhi: Motilal Banarsidass. Faure, Bernard. 1991. The Rhetoric of Immediacy: A Cultural Critique of Chan/Zen Buddhism. Princeton: Princeton UP. Flood, Finbarr. 2009. Objects of Translation: Material Culture and Medieval “Hindu-Muslim” Encounter. Princeton: Princeton UP. Flood, Gavin. 2002. “The Purification of the Body in Tantric Ritual Representation.” Indo- Iranian Journal. Vol. 45, pp. 25-43. Foulk, T. Griffith. 1992. “The Ch’an Tsung in Medieval China; School, Lineage, or What?” Pacific World 8, Fall: 18-31. Fremantle, Francesca. 1971. “A Critical Study of the Guhyasamāja Tantra.” PhD dissertation, University of London. Gardener, Alexander. 2011. Short biography of Rinchen Zangpo. http://www.treasuryoflives.org/biographies/view/Rinchen-Zangpo/10199 Garson, Nathaniel. 2004. Penetrating the Secret Essence Tantra: Context and Philosophy in the Mahāyoga System. PhD Dissertation University of Virginia. Geibel, Rolf W. 1995. “The Chin-kang-ting ching yü-ch’ieh shih-pa-hui chih-kuei: An Annotated Translation.” Journal of the Naritasan Institute for Buddhist Studies No.18, 107-202.

---

. 2001. Two Esoteric Sutras: The Adamantine Pinnacle Sutra [and] The Susiddhikara

Sutra. BDK English Tripiṭaka 29-II, 30-II. Numata Center for Buddhist Translation and Research. George, Christopher. 1974. The Caṇḍamahāroṣaṇa Tantra Chapters I-VIII: A Critical Edition and English Translation. New Haven: American Oriental Society. Gellner, David. 1992. Monk, Householder, and Tantric Priest: Newar Buddhism and its Hierarchy of Ritual. Cambridge: Cambridge UP. Germano, David. 1994. “Architecture and Absence in the Secret Tantric History of the Great Perfection (rdzogs chen).” JIABS 17.2, pp. 203-335. 463

---

. 2002. “The Seven Descents and the Early History of rNying ma transmissions.” in The

Many Canons of Tibetan Buddhism (eds. Helmut Eimer and David Germano). Leiden: Brill. Geslani, Marko. 2011. The Ritual Culture of Appeasement: Śānti Rites in Post-Vedic Sources. Ph.D. dissertation. Yale University. Goodall, Dominic. 2016. “How the Tattvas of Tantric Śaivism Came to Be 36: The Evidence of the Niśvāsatattvasaṃhitā.” Tantric Studies: Fruits of a Franco-German Collaboration on Early Tantra, Vol II, ed. Dominic Goodall and Harunaga Isaacson. pp. 77-111.

---

. 2012. “Śānti Rites in the Development of the Purāṇic Rājyābhiṣeka.” Indo-Iranian

Journal 55, pp. 321-377. Granoff, Phyllis. 2000. “Other People’s Rituals: Ritual Eclecticism in Early Medieval Indian Religions.” Journal of Indian Philosophy 28, pp. 399-424.

---

. 2001. “My Rituals and My Gods: Ritual Exclusiveness in Medieval India.” Journal of

Indian Philosophy 29, pp. 109-134.

---

. 2004. “Mahākāla’s Journey: From Gaṇa to God.” Rivista Degli Studi Oriental, Vol.

LXXVII. Rome: Instituti Editoriali e Poligrafici Internazionali. Gray, David. 2007. The Cakrasaṃvara Tantra (The Discourse of Śrī Heruka): A Study and Annotated Translation. Treasury of the Buddhist Sciences Series. New York: American Institute of Buddhist Studies.

---

. 2009. “On the Very Idea of a Tantric Canon.” JIATS, no 5: 1-37.

Gray, David and Ryan Richard Overbey, ed. 2016. Tantric Traditions in Transmission and Translation. New York: Oxford UP. Guenther, Herbert. 1963 (Reprint 1987). The Life and Teaching of Naropa. Boston: Shambhala Gyatso, Geshe Jampa. 2006. FPMT Basic Program Grounds and Paths of Secret Mantra: The Ilumination of the Texts of Tantra by Ngawang Palden with Oral Commentary by Geshe Jampa Gyatso. Boston: Wisdom Publications. Gyatso, Janet. Gyatso, Janet. 1986. “Signs, Memory and History: A Tantric Buddhist Theory of Scriptural Transmission.” Journal of the International Association of Buddhist Studies. 9 (2): 7-35.

---

. 1992. “Genre, Authorship and Transmission in Visionary Buddhism: The

Literary Traditions of Thang-stong rGyal-po.” in Ronald Davidson and Steven Goodman, eds. Tibetan Buddhism: Reason and Revelation. Albany: SUNY Press. pp 96-106.

---

. 1993. “The Logic of Legitimation in the Tibetan Treasure Tradition.” History of

Religions. 33 (2): 97-134. 464

---

. 1994. “Guru Chos-dbang’s gTer ‘byung chen mo: an early survey of the treasure

tradition and its strategies in discussing Bön treasure.” Occasional Papers—Institute for Comparative Research in Human Culture I (I): 275-287.

---

. 1996. “Drawn form the Tibetan Treasury: The gTer ma Literature.” in José

Cabezón and Roger Jackson, eds. Tibetan Literature: Studies in Genre. Ithaca: Snow Lion. 147-169.

---

. 1998. Apparitions of the Self: The Secret Autobiographies of a Tibetan Visionary.

Princeton: Princeton UP. Harrison, Paul. 1978. “Buddhānusmṛti in the Pratyutpanna-buddha-saṃmukhāvasthita-samādhisūtra.” Journal of Indian Philosophy, 6: 35-57.

---

. 2003. “Mediums and messages: Reflections on the Production of Mahāyana Sūtras.”

The Eastern Buddhist vol 35, no2: 115-152. Hadano, Hakuyu. 1986 (1950). “Chibetto Bukkyō Keisei no Ichikadai,” Indo-Chibetto Gaku Shūse: Dai 1 Kan Chibetto-hen 1. Kyoto: Hōzōkan, pp. 3-28.

---

. 1987 (1951). “Himitsushū Tantra ni okeru Jyunyānapāda Ryū ni tsuite.” Indo-Chibetto

Gaku Shūse: Dai 3 Kan, Indo-hen 1. Kyoto: Hōzōkan, pp. 36-49. Hatchell, Christopher. 2009. Naked Seeing: The Great Perfection, The Wheel of Time, and Visionary Philosophy in Renaissance Tibet. PhD Dissertation. University of Virginia. Hatley, Shaman. 2016. “Converting the Ḍākinī: Goddess Cults and Tantras of the Yoginīs Between Buddhism and Śaivism.” in In David B. Gray and Ryan Richard Overbey, ed., Tantric Traditions in Transmission and Translation. Oxford: Oxford University Press, pp. 37-86. Hermann-Pfandt, Adelheid. 2002. “The Lhan kar ma as a Source for the History of Tantric Buddhism.” in The Many Canons of Tibetan Buddhism. Helmut Eimer and David Germano, eds. Boston: Brill, p.129-49. Hodge, Stephen. 1995. “Considerations on the Dating and Geographical Origins of the Mahāvairocanābhisambodhi-sūtra.” In The Buddhist Forum, vol. 3, eds. Tadeusz Skorupski and Ulrich Pagel. London: School of Oriental and African Studies; rpt. New Delhi: Heritage Publishers, pp. 57-83.

---

. 2003. The Mahā-Vairocana-Abhisaṃbodhi Tantra with Buddhaguhya’s Commentary.

New York: Routledge Curzon. Hong, Luo and Toru Tomabechi, ed. 2009. Candrakīrti’s Vajrasattvaniṣpādanasūtra (Vajrasattvasādhana): Sanskrit and Tibetan Texts. Sanskrit Texts from the Tibetan Autonomous Region No.6. Beijing-Vienna: China Tibetology Publishing House and Austrian Academy of Sciences Press. 465 Hopkins, Jeffrey. 1992. Tibetan Arts of Love: Sex, Orgasm and Spiritual Healing. Ithaca: Snow Lion.

---

. 2003. Maps of the Profound: Jam-yang-shay ba’s Outline of the Great Exposition

Buddhist and Non-Buddhist Views on the Nature of Reality. Ithaca: Snow Lion.

---

. 2005. Yoga Tantra: Paths to Magical Feats. Ithaca: Snow Lion.

---

. 2008. Tantric Techniques. Edited by Kevin Vose. Ithaca: Snow Lion.

Inden, Ronald. 2000a. “Introduction: From Philological to Dialogical Texts.” in Inden et al 2000. Querying the Medieval: Texts and History of Practices in South Asia. New York: Oxford UP, pp. 3-28.

---

. 2000b. “Imperial Purāṇas: Kashmir as Vaiṣṇava Center of the Words.” in Inden et al

2000. Querying the Medieval: Texts and History of Practices in South Asia. New York: Oxford UP, pp. 29-98. Isaacson, Harunaga. 1998. “Tantric Buddhism in India (From c AD800 to c AD 1200).” In: Buddhismus in Geschichte und Gegenwart. Band II. Hamburg. pp. 23–49. (Internal publication of Hamburg University.) PDF of a revised edition available on the author’s website: http://www.tantric-studies.org/people/harunaga-isaacson/

---

. 2002. “Ratnākaraśānti’s Hevajrasahajasadyoga. (Studies in Ratnākaraśānti’s Tantric

Works 1).” In Le Parole E I Marmi: Studi in Onore di Raniero Gnoli Nel Suo 70˚ Compleanno, edited by Raffaele Torella, 457-97. Roma: Instituto Italiano per L’Africa e L’Oriente.

---

. 2007. “First Yoga: A commentary on the ādiyoga section of Ratnākaraśānti’s

Bhramahara.” In Pramāṇakīrtiḥ: Papers Dedicated to Ernst Steinkellner on the Occasion of his 70th Birthday. Edited by Birgit Kellner, Helmut Krasser, Horst Lasic, Michael Torsten Much and Helmut Tauscher. Wein: Arbeitskreis fur Tibetische und Buddhistische Studien Universitat Wein.

---

. 2010b. "Observations on the Development of the Ritual of Initiation.” In Hindu and

Buddhist Initiations in India and Nepal. ed. Astrid Zotter and Christof Zotter., 261-279. Wiesbaden: Harrassowitz Verlag.

---

. 2013. “Yogācāra and Vajrayāna according to Ratnākaraśānti.” in The Foundation for

Yoga Practitioners: The Buddhist Yogācārabhūmi Treatise and Its Adaptation in India, East Asia, and Tibet. Harvard Oriental Series, Vol. 75. edited by Ulriche Timme Kragh. Cambridge: Harvard University Press, pp. 1036-1051. Isaacson, Harunaga and Francesco Sferra, ed. 2014. The Sekanirdeśa of Maitreyanātha (Advayavajra) with the Sekanirdeśapañjikā of Rāmapāla: Critical Edition of the Sanskrit and Tibetan Texts with English Translation and Reproductions of the MSS. Manuscripta 466 Buddhica: Jointly Published with the Asien-Afrika-Institut, 2. ed. Harunaga Isaason and Francesco Sferra. Serie Orientale Roma: Fondata da Giuseppe Tucci. Diretta da Gherardo Gnoli, Vo. CVII. Napoli: Universita Degli Studi di Napoli “L’Orientale.” Jaini, Padmanabh S. 2001. Collected Papers on Buddhist Studies. Delhi: Motilal Banarsidass. Kalff, Martin Michael. 1979. Selected Chapters from the Abhidhānottara-tantra: The Union of Female and Male Deities. Ph.D. Dissertation. Columbia University. Kano, Kazuo. 2014. “Fugen jojuho no shinshutsu bonbun shiryo. [Newly Available Materials of Jñānapāda’s Samantabhadrasādhana].” Mikkogaku kenkyū 46, pp 61-73. Kapstein, Matthew T. 2000. The Tibetan Assimilation of Buddhism: Conversion, Contestation, and Memory. Oxford: Oxford UP. Karmay, Samten. 2007. The Great Perfection. Leiden: Brill. Kawasaki, Kazuhiro. 2004. “On a Birch-bark Sanskrit Manuscript Preserved in the Tibet Museum.” Journal of Indian and Buddhist Studies, Vol. 52, No. 2, March 2004, 905-03. Keira, Ryusei. 2004. Mādhyamika and Epistemology: A Study of Kamalaśīla’s Method for Proving the Voidness of All Dharmas. Wein: Arbeitskreis für Tibetische und Buddhistiche Studien Universitat Wien. Kent, Stephen A. 1982. “Early Sāṃkhya in the ‘Buddhicarita.’” in Philosophy East and West, vol 32, No 3, pp. 259-78. Kikuya, Ryūta. 2000. “The Ṣaḍaṅgayoga System of the Jñānapāda School.” (In Japanese) Bulletin of the Graduate School, Koyasan University Graduate Program of Liberal Arts. Vol 4, pp. 45-57.

---

. 2000b. “Notes on Jñānapāda Sādhanas.” (In Japanese) Bulletin of the Graduate School,

Koyasan University Graduate Program of Liberal Arts. Vol 4, p 1.

---

. 2008. “Ju’nyānapādaryū ni okeru sanshu no tekiyuga ni tsuite” [Buddhaśrījñānapāda’s

Binduyoga Theory in the Guhyasamāja Cycle]. Bukkyo-Gaku. Vol. 50, 93-115.

---

. 2010. “Two steps (dvikrama-) in the Jñānapāda School of Indian Tantric Buddhism.” (In

Japanese) in The Annual Reports of Graduate School for Arts and Letters Tohoku University Vol. 60, pp 1-23. (Unpublished English translation prepared by the author for conference at UC Berkeley, March 2014)

---

. 2012. “Reconstruction of Jñānapāda’s *Caturaṅgasādhana-Samantabhadrī.” Journal of

Indian and Buddhist Studies Vol 60, No.3.

---

. 2012b. “The Transmission of the Guhyasamājatantra in Indian Tantric Buddhism.” (In

Japanese) The Journal of the Nippon Buddhist Research Association, Vol. 77, pp. 213- 236. 467 Kilty, Gavin, trans. 2013. Tsongkhapa’s A Lamp to Illuminate the Five Stages: Teachings on the Guhyasamāja Tantra. Boston: Wisdom Publications. Kittay, David R. 2011. Interpreting the Vajra Rosary: Truth and Method Meets Wisdom and Method. PhD Dissertation, Columbia University. Klein-Schwind, Sabine Gudrun. 2012. The Compendium of the Ten Fundamentals: Daśatattvasaṃgraha of paṇḍita Kṣitigarbha. PhD Dissertation, University of Hamburg. Komarovski, Yaroslav. 2007. Echos of Empty Luminosity: Reevaluation and Uniquie Interpretation of Yogācāra and Niḥsvabhāvavāda Madhyamaka by the Fifteenth Century Tibetan Thinker Śākya mchod ldan. PhD Dissertation. University of Virginia. Kongtrül, Jamgön. 1994. The Torch of Certainty. Translated by Judith Hanson. Boston: Shambhala.

---

. 1998. The Treasury of Knowledge: Buddhist Ethics. Translated and edited by the

International Translation Committee founded by the V.V. Kalu Rinpoche. Ithaca: Snow Lion.

---

. 2005. The Treasury of Knowledge: Systems of Buddhist Tantra. Translated by Elio

Guarisco and Ingrid McLeod. Ithaca: Snow Lion.

---

. 2007. Treasury of Knowledge: Frameworks of Buddhist Philosophy. Translated by

Elizabeth Callahan. Ithaca: Snow Lion.

---

. 2009. The Treasury of Knowledge: The Elements of Tantric Practice. Translated by Elio

Guarisco and Ingrid McLeod. Ithaca: Snow Lion.

---

. 2010. The Treasury of Knowledge: Buddhism’s Journey to Tibet. Translated by

Ngawang Zangpo. Ithaca: Snow Lion.

---

. 2011. The Treasury of Knowledge: Journey and Goal. Translated by Richard Barron.

Ithaca: Snow Lion. Kosambi, D.D. 1956. Reprint 1975. An Introduction to the Study of Indian History. Bombay: Popular Prakarshan. Kramer, Ralf. 2007. The Great Tibetan Translator: Life and Works of rNgog Blo ldan shes rab (1059-1109). München: Indus Verlag. Krug, Adam. 2018. The Seven Siddhi Texts: The Oḍiyāna Mahāmudrā Lineage in its Indic and Tibetan Contexts. PhD dissertation, UC Santa Barbara. Kwon, Do-kyun. 2002. Sarva Tathāgata Tattva Saṃgraha, Compendium of all the Tathāgatas: A Study of its Origin, Structure, and Teachings. PhD dissertation, SOAS. 468 Larson, Gerald James. 1969. (Reprint 1998.) Classical Sāṃkhya: An Interpretation of its History and Meaning. Delhi: Motilal Banarsidass.

---

. 1983. “An Eccentric Ghost in the Machine: Formal and Quantative Aspects of the

Sāṃkhya-Yoga Dualism. in Philosophy East and West. Vol. 33, No. 3, pp 219-233. Lee, Yong-Hyun. 2004. The Niṣpannayogāvalī by Abhayākaragupta: A New Critical Edition of the Sanskrit Text (Revised Edition). Baegun Press: Seoul. Lessing, Ferdinand D. and Alex Wayman. 1968. Introduction to the Buddhist Tantric Systems. Mkhas-grub-re’s Rgyud dse spyiḥi rnam par gzag pa rgyas par brjod with original text and annotation. The Hague. Lopez, Donald. 1996. Elaborations on Emptiness. Princeton: Princeton UP. Ludvik, Catherine. 2007. Sarasvatī: Riverine Goddess of Knowledge. Boston: Brill. Maclean, Derryl. 1989. Religion and Society in Arab Sind. New York: Brill. Makransky, John. 1997. Buddhahood Embodied: Sources of Controversy in India and Tibet. Albany: SUNY Press. Matsunaga, Yukei. 1964. “A Doubt to authority of the Guhyasamāja-Ākhyāna Tantras." Journal of Indian and Buddhist Studies [IBK]13, no. 2: 16-25.

---

. 1977a. “A History of Tantric Buddhism with Reference to Chinese Translations.” In

Buddhist Thought and Asian Civilization, edited by Leslie Kawamura and Keith Scott, 167-91. Dharma Publishing.

---

. 1977b. “Some Problems of the Guhyasamāja-tantra.” In Studies in Indo-Asian Art and

Culture, edited by Lokesh Chandra and Perala Ratnam. New Delhi: International Academy of Indian Culture.

---

. 1978. The Guhyasamāja Tantra: A New Critical Edition. Osaka: Toho Shuppan, Inc.

McNamara, Daniel. 2017. “When Madhyamaka is Not the Middle Path: Ratnākaraśānti on Yogācāra, Nāgārjuna, and the Madhyamapratipadā.”JIABS, vol. 40., pp. 189-207. Mimaki, Katsumi and Tōru Tomabechi. 1994. Pañcakrama: Sanskrit and Tibetan Texts Critically Edited with Verse ndex and Facsimilie Edition of the Sanskrit Manuscripts. Tokyo: The Centre for East Asian Culturlal Studies for Unesco. Mori, Masahide. (n.d.) The Vajrāvalī of Abhyākaragupta: A Critical Study, Sanskrit Edition of Selected Chapters and Complete Tibetan Version. Phd dissertation, SOAS.

---

. 2009. Vajrāvalī of Abhayākaragupta. Edition of Sanskrit and Tibetan Versions. 2 Vols.

(Buddhica Britannica Series Continua 11). Tring: The Institute of Buddhist Studies. 469 Negi, Jita Sain. 1993-2005. Tibetan-Sanskrit Dictionary, vols. 1-16. Sarnath: Central Institute of Higher Tibetan Studies. Nicholson, Andrew. 2010. Unifying Hinduism: Philosophy and Identity in Indian Intellectual History. New York: Columbia UP. Nihom, Max. 1995. “On Attracting Women and Tantric Initiation: Tilottamā and “Hevajratantra.” Bulletin of the School of Oriental and African Studies, University of London, Vol 58, No3, pp 521-31. Olivelle, Patrick. 2012. “Kings, Ascetics, and Brahmins: The Socio-Political Context of Ancient Indian Religions.” in Dynamics in the History of Religions between Asia and Europa: Encounters, Notions, and Comparative Perspectives. ed Volkhard Krech and Marion Steinicke. Boston: Brill, pp. 117-136. Onians, Isabelle. 2003. Tantric Buddhist Apologetics or Antinomianism as a Norm. PhD Dissertation, Oxford. Orzech, Charles D. Henrik Sorensen, and Richard Payne. 2011. “Introduction: Esoteric Buddhism and the Tantras in East Asia: Some Methodological Considerations.” in Esoteric Buddhism and the Tantras in East Asia edited by Orzech, Sorenson and Payne, pp 3-18. Boston: Brill. Padoux, André. 2002. “What do we mean by Tantrism?” in The Roots of Tantra. Katherine Anne Harper and Robert L Brown, ed. Albany: SUNY Press, 17-25. Petech, Luciano. 1997. “Western Tibet: Historical Introduction.” In Tabo: A Lamp for the Kingdom. edited by Deborah E. Klimburg-Salter, pp. 229-55. New York: Thames & Hudson. Pollock, Sheldon. 2009. Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India. Berkeley: University of California Press. Roerich, George. 1976. The Blue Annals. Delhi: Motilal Banarsidass. (Originally published in Calcutta, 1949.) Roberts, Peter Alan, trans. 2010. Mahāmudrā and Related Instructions: Core Teachings of the Kagyü Schools. Thuben Jinpa (ed.). Boston: Wisdom Publications in association with the Institute of Tibetan Classics. von Rospatt, Alexander. 2015. “Local Literatures: Nepal” in Brill's Encyclopedia of Buddhism. ed. Jonathan Silk, Oskar von Hinüber, and Vincent Eltschinger. Vol. 1, pp. 819-830. Leiden: Brill. Ruegg, David Seyfort. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Otto Harrassowitz. 470 Sakurai, Munenobu. 1996. Indo Mikkyō Girei Kenkyū: Kōki indo Mikkyō no Kanchōshidai. (A Study on the ritual of Abhiṣeka.) (in Japanese) Kyoto: Kōzōkan.

---

. 2007. “A note on the sbyor ba bdun handed down in Jñānapāda School.” (in Japanese)

Journal of Chizan Studies No 72 pp 359-370.

---

. 2009. “On the pañcākārābhisaṃbodhi transmitted in the Jñānapāda School.” (in

Japanese) Journal of Chizan Studies No 72, pp 37-53. ‘ Samdhong Rinpoche and Vrajvallabh Dwivedi, ed. 1987. Guhyādi-aṣṭasiddhisaṃgraha. Sarnath: Central Institute for Higher Tibetan Studies Rare Buddhist Texts Projext. Samuel, Geoffrey. 2008. The Origins of Yoga and Tantra: Indic Religions to the Thirteenth Century. New York: Cambridge UP. Sanderson, Alexis. 2001. “History Through Textual Criticism.” In: Les Sources et le temps. Sources and Time: A Colloquium, Pondicherry, 11-13 January 1997, edited by François Grimal. Publications du département d'Indologie 91. Pondicherry: Institut Français de Pondichéry/École Française d'Extrême-Orient, pp. 1-47.

---

. 2009. “The Śaiva Age.” In Genesis and Development of Tantrism. ed. Shingo Einoo.

Tokyo: Institute of Oriental Culture, University of Tokyo. Institute of Oriental Culture Special Series, 23, pp. 41-350. Schaeffer, Kurtis. 2009. The Culture of the Book in Tibet. New York: Columbia University Press. van Schaik, Sam. 2004a. “The Early Days of the Great Perfection.” JIABS Vol. 27, No. 1, pp. 165-206.

---

. 2004b. Approaching the Great Perfection: Simultaneous and Gradual Approaches to

Dzogchen Practice in the Longchen Nyingtig. Boston: Wisdom Publications.

---

. 2008. “The Sweet Sage and the Four Yogas.” in JIATS, Issue 4, pp. 1-67.

---

. 2012. “Dzogchen, Chan and the Question of Influence.” Revue d’Etudes Tibetaines, No.

24, October, pp. 5-20. Schmithausen, Lambert. 1981. “On Some Aspects of Descriptions or Theories of ‘Liberating Insight’ and ‘Enlightenment’ in Early Buddhism.” In Studien zum Jainismus und Buddhismus: Gedenkschrift für Ludwig Alsdorf, ed. Klaus Bruhn and Albrecht Wezler. Wiesbaden: Franz Steiner Verlag, pp. 199-250. Seyfort-Ruegg, David. 1981. The Literature of the Madhyamaka School of Philosophy in India. Jan Gonda (ed.). A History of Indian Literature, vol VII, Fasc. 1. Wiesbaden: Otto Harrassowitz. 471 Sferra, Francesco. 1990. The Ṣaḍaṅgayoga by Anupamarakṣita with Raviṣrījñāna’s Guṇabharaṇīnāmaśaḍaṅgayogaṭippaṇī. Rome: Istituto Italiano Per L’Africa e L’Oriente. Sharf, Robert H. 2002. “On Esoteric Buddhism in China.” In Coming to Terms with Chinese Buddhism: A Reading of the Treasure Store Treatise. Kuroda Institute Studies in East Asian Buddhism. Honolulu: University of Hawaii Press.

---

. forthcoming. “Mindfulness and Mindlessness in Early Chan.”

Shinohara, Koichi. 2014. Spells, Images, and Maṇḍalas: Tracing the Evolution of Esoteric Buddhist Rituals. New York: Columbia UP. Shechen Gyaltsab IV and Rinchen Dargye. 2011. A Practice of Padmasambhava: Essential Instructions on the Path to Awakening. Translated by the Dharmachakra Translation Committee. Ithaca: Snow Lion. Skorupski, Tadeusz. 1983. The Sarvadurgatipariśodhana Tantra. Delhi: Motilal Banarsidass.

---

. 2002. Kriyāsaṃgraha: Compendium of Buddhist Rituals, An Abridged Version. Tring,

UK: The Institute of Buddhist Studies. Smith, Frederick. 2006. The Self Possessed: Deity and Spirit Possession in South Asian Literature and Civilization. New York: Columbia UP. Snellgrove, David. 1959. The Hevajra Tantra: A Critical Edition. Oxford: Oxford Up.

---

. 1987. (reprint 2002.) Indo-Tibetan Buddhism: Indian Buddhists and Their Tibetan

Successors. Boston: Shambhala Publications. Sparham, Gareth. 1989. A Study of Haribhadra’s Abhisamayālaṃkārālokā Prajñāpāramitāvyākhyā. Ph.D. Dissertation. University of British Columbia.

---

. 2006; 2008; 2009; 2012. Abhisamayālaṃkāra with Vṛtti and Ālokā Vols 1-4. Jain

Publishing Company. Stearns, Cyrus, tr. and ed. 2006. Taking the Result as the Path: Core Teachings of the Sakya Lamdre Tradition. Boston: Wisdom. von Stietencron, Heinrich. 2007. “Religious Configurations in Medieval India and the Modern Concept of Hinduism.” in The Oxford India Hinduism Reader. ed. Vasudha Dalmia and Heinrich von Stietencron. Oxford: Oxford UP, pp. 50-89. Studholme, Alexander. 1999. On the Early History of the Oṃ Maṇīpadme Hūṃ Mantra: A Study of the Kāraṇḍavyūha Sūtra. PhD dissertation, University of Bristol. Szántó, Péter-Dániel. 2008a. “Antiquarian Enquiries into the Initiation Manuals of the Catuṣpīṭha.” Newsletter of the NGMPP, Number, pp. 2-12. 472

---

. 2008b. “A Critical Edition of the Catuṣpīṭha-tantra I.3 with Three Sanskrit

Commentaries.” In Tantric Studies Vol. 1, pp. 2-21.

---

. 2010. ‘The Case of the Vajra-Wielding Monk.” Acta Orientalia Academiae

Scientiarum Hung. Volume 63 (3), pp. 289-299.

---

. 2012a. Selected Chapters From the Catuṣpīṭhatantra: Introductory study with the

annotated translation of selected chapters. PhD Dissertation, Oxford University.

---

. 2012b. “The sources for utkrānti in Tantric Buddhism.” Paper abstract, Vikramaśīla

Workshop at the Institute for Advanced Studies on Asia, University of Tokyo, September 14-17, 2012. http://www.ioc.u-tokyo.ac.jp/eng/news/news.php?id=20120829142138

---

. 2014. “How to Organize a Gaṇacakra?” Paper delivered at The Evolution of Tantric

Ritual Conference, UC Berkeley, March 2014. ______. 2015a. “Early Works and Persons Related to the So-called Jñānapāda School.” JIABS 36/37, pp 537-561.

---

. 2015b. “Tantric Prakaranas.” In Brill’s Encyclopedia of Buddhism: Literature and

Languages, Vol. I. Handbook of Oriential Studies: Seciton Two: India. ed. Oskar Von Hinuber, Vincent Eltschinger, and Jonathan Silk. Leiden: Brill.

---

. 2015c. “Catuṣpīṭha.” Brill’s Encyclopedia of Buddhism. In Brill’s Encyclopedia of

Buddhism: Literature and Languages, Vol. I. Handbook of Oriential Studies: Seciton Two: India. ed. Oskar Von Hinuber, Vincent Eltschinger, and Jonathan Silk. Leiden: Brill. Tagare, G. V., trans. 1958, reprint 2002. The Brahmānda Purāna, Part I. Delhi: Motilal Banarsidass. Takahashi, Kammie. 2009. Lamps for the Mind: Illumination and Innovation in dPal dbyang’s Mahāyoga. PhD thesis, University of Virginia. Tanemura, Ryugen. 2015. “Guhyasamāja.” in Brill’s Encyclopedia of Buddhism: Literature and Languages, Vol. I. Handbook of Oriential Studies: Seciton Two: India. ed. Oskar Von Hinuber, Vincent Eltschinger, and Jonathan Silk. Leiden: Brill. Tanaka, Kimiaki. 1991. “’Himitsushūe’ Jñānapāda-ryū no shōkishidai caturaṅga no shin-shiryō; National Archives pra.1697 (Kha 2) no kenkyū.” Ga no shisō: Maeda Sengaku hakase kanreki kinen ronshū. Tokyo: Shunjūsha, 1991 (pp 667-683).

---

. 1995. “Some Buddhist Tantric Manuscripts Identified during a Stay at Oxford

University.” Journal of Indian and Buddhist Studies Vol 43, No. 2, March 1995 (45), 1008-04.

---

. 1996. Indo chibetto Mandara no kenkyū [A Study of Indo-Tibetan Maṇḍalas (with

English Chapter Summaries)]. Tōkyo: Hozokan. 473

---

. 1999. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi (10/11) avāhavidhi.

A romanized text of newly discovered Sanskrit manuscript.” Mikkyo Bunka 203, pp. 23- 36.

---

. 2000a. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi 9, raṅgajāpavidhi

and 10 raṅgapātanavidhividhi.” Indogaku Bukkyogaku Kenkyu 203: 23-36.

---

. 2000b. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi, śiṣyapraveśavidhi.”

Mikkyo Bunka 205, pp 209-214.

---

. 2001. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi 8,

akāśasūtrapātanavidhi.” Indogaku Bukkyogaku Kenkyu 50(1), pp. 209-214.

---

. 2002. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi,

śiṣyādhivāsanavidhi.”Mikkyo Bunka 209, pp. 1-15.

---

. 2003a. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi. The Tibetan

translation and Sanskrit text of chapters 13, 14, and 15. The Memoirs of the Institute of Oriental Culture (Toyo Bunka Kenkyusho Kiyo) 142, pp. 193-217.

---

. 2003b. “On the ritual procedures following consecration according to Nāgabodhi’s

Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi.” The Memoirs of the Institute of Oriental Culture (Toyo Bunka Kenkyusho Kiyo) 146, pp. 233-248.

---

. 2004a. “On the five postures (sthānaka) and fifty-three postures (ākṣepa) according to

Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi.” The Memoirs of the Institute of Oriental Culture (Toyo Bunka Kenkyusho Kiyo) 146, pp. 109-130.

---

. 2004b. “Nāgabodhi’s Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi: The Tibetan

translation and Sanskrit text of chapters 5 and 6.” in Three Mountains and Seven Rivers. Prof Musashi Tachikawa’s Felicitaiton Volume. Delhi: Motilal Manarsidass, pp. 857- 870.

---

. 2004c. “The measurement of maṇḍala according to Nāgabodhi’s

Śrīguhyasamājamaṇḍalopāyika-viṃśati-vidhi.” Journal of Buddhist Iconography (Mikkyo Zuzo) 23, pp . 26-39.

---

. 2007a. “Nāgabodhi’s Samājasādhanavyavasthāna: The Tibetan translation and Sanskrit

text of 3-1-3 to 3-3-3.” Memoirs of the Institute of Oriental Culture (Toyo Bunka Kenkyusho Kiyo) 148, pp. 304-282.

---

. 2007b. “The Procedures for Maṇḍala Offerings in Buddhajñanapāda’s

Samantabhadranāma sādhana.” (in Japanese). in The Memoirs of the Institute of Oriental Culture, No. 150, March 2007, pp.168-203. Tokyo: The Institute of Oriental Culture.

---

. 2010. Indo ni okeru Mandala no Kenkyū. (Doctoral Dissertation, in Japanese) Tokyo:

Shunjyūsya. 474

---

. 2017. The Sanskrit commentary on the Samantabhadra nāma sādhana of

Buddhajñānapāda: Introduction, Romanized Sanskrit Text and Translation. Tokyo: Watanabe Publishing.

---

. 2018. Mañjuvajramukhyākhyāna, A ritual manual belonging to the Jñānapāda school of

the Guhyasamāja-tantra: Introduction, Romanized Sanskrit Text and Related Articles. Tokyo: Watanabe Publishing.

---

. unpublished. “Chapter Five: The Emergence of the Guhyasamāja-maṇḍala.” English

Translation (by the author) of Chapter Five of Doctoral Dissertation Indo ni okeru Mandala no Kenkyū. Templeman, David, tr .and ed. 1983; Reprint 2007. The Seven Instruction Lineages. New Delhi: LTWA. Thapar, Romila. 2002. Early India: From the Origins to AD 1300. Berkeley: University of California Press.

---

. 2004. Somnatha: The Many Voices of a History. Delhi: Penguin Viking.

Thurman, Robert. 2010. Brilliant Illumination of the Lamp of the Five Stages. New York: Columbia UP. Todaro, Dale Allen. 1985. An Annotated Translation of the ‘Tattvasamgraha” (Part 1) with an Explanation of the Role of the “Tattvasamgraha” Lineage in the Teachings of Kukai. Ph.D. Dissertation. Columbia University. Tomabechi, Tōru. 1994. “La Formation Textuelle du Pañcakrama.” in ASIEA XLVII.4, pp. 1383-1387.

---

. 2000. “Notes on Robert Thurman’s Translation of the Pañcakrama.” Journal of Indian

Philosophy. 28: 531-548.

---

. 2006. Étude du Pañcakrama: Introduction et Traduction Anotée. Phd Dissertation

University of Lausanne.

---

. 2008. “Vitapāda, Śākyamitra, and Āryadeva: On a Transitional Stage in the History of

Guhyasamāja Exegesis.” in Esoteric Buddhist Studies: Identity in Diversity. Proceedings of the International Conference of Esoteric Buddhist Studies, Koyasan University, 5 Sept.-8 Sept. 2006. Editorial board of the ICEBS (ed.), Koyasan: Koyasan University, pp. 171-77. Torricelli, Fabrizio and Āchārya Sangye T. Naga. 1995. The Life of the Mahāsiddha Tilopa. Dharmasala: Library of Tibetan Works and Archives. Tribe, A.H.F. 1994. The Names of Wisdom: A Critical Edition and Annotated Translation of Chapters 1-5 of Vilāsavajra's Commentary on the Nāmasaṃgīti, with Introduction and 475 Textual Notes. PhD thesis, University of Oxford.

---

. 1997. “Mañjuśrī and ‘The Chanting of Names’ (Nāmasaṃgīti): Wisdom and its

Embodiment in an Indian Mahāyāna Buddhist text.” in Indian Insights: Buddhism, Brahmanism and Bhakti. edited by Peter Connolly and Sue Hamilton. Luzac Oriental, pp. 109-36.

---

. 2000. “Mantranaya/Vajrayana—tantric Buddhism in India.” in Paul Williams with

Anthony Tribe. Buddhist Thought: A complete introduction to the Indian Tradition. New York: Routledge.

---

. 2016. Tantric Buddhist Practice in India: Vilāsavajra’s commentary on the Mañjuśrīnāmasamgīti.

New York: Routledge. Tsuda, Shin’ichi. 1974. The Saṃvarodaya Tantra. Tokyo: The Hokuseido Press.

---

. 1995. “The Original Formation and Performance of the “Secret Assembly”

(guhyasamāja), an Integration of the Guhyasamāja-tantra into the History of Tantric Buddhism in India.” International College for Postgraduate Buddhist Studies. Tucci, Giuseppi. 1998. (1932). Rin-chen-bzan-po and the renaissance of Buddhism in Tibet Around the Millennium. Translated by Nancy Kipp Smith. New Delhi: Aditya Prakashan. Ui, Hakuju, Munetada Suzuki, Yenshō Kanakura, and Tōkan Tada. 1934. A Complete Catalogue of the Tibetan Buddhist Canons (Bkhaḥ-ḥgyur and Bstang-ḥgyur). Sendai: Tōhoku Imperial University. van Schaik, Sam. 2008a. “The Sweet Sage and the Four Yogas: A Lost Mahāyoga Treatise from Dunhuang.” Journal of the International Association of Tibetan Studies 4, 1-67.

---

. 2008b. “A Definition of Mahāyoga: Sources from the Dunhuang Manuscripts.” Tantric

Studies 1, 53-105. Vetter, Tilmann. 1988. The Ideas and Meditative Practices of Early Buddhism. New York: Brill. Vitali. 2004. “The Role of Clan Power in the Establishment of Religion.” in Cüppers, ed., The Relationship Between Religion and the State In Traditional Tibet, pp 105-158. Lumbini: Lumbini International Research Institute. Vogel, Claus. 1965. Surūpa’s Kāmaśāstra. An Erotic Treatis in the Tibetan Tanjur. Helsinki: Studia Orientalia Edidit Societas Orientalis Fennica XXX: 3. Vose, Kevin. 2009. Resurrecting Candrakīrti: Disputes in the Tibetan Creation of Prasangika. Boston: Wisdom Publications. Wallace, 2001. The Inner Kālacakratantra: A Buddhist Tantric View of the Individual. Oxford: Oxford UP. 476

---

. 2010. The Kālacakra Tantra: The Chapter on Sādhanā Together with the Vimalaprabha

Commentary. New York: The American Institute for Buddhist Studies. Wallis, Glenn. 2002. Mediating the Power of Buddhas: Ritual in the Mañjuśrīmūlakalpa. Albany: SUNY Press. Wayman, Alex. 1977. Yoga of the Guhyasamājatantra. Delhi: Motilal Banarsidass, 1991. Wedemeyer, Christian. 1999. Vajrayāna and its Doubles: A critical historiography, exposition, and translation of the Tantric works of Āryadeva. PhD dissertation, Columbia University.

---

. 2006. Review of Indian Esoteric Buddhism. in History of Religions, vol. 45, no. 4, pp.

373–376.

---

. 2007. Āryadeva’s Lamp that Integrates the Practices (Cāryamelapakapradīpa): The

Gradual Path of Vajrayāna Buddhism. Treasure of the Buddhist Sciences Series. New York: American Institute of Buddhist Studies.

---

. 2009. “Pseudepigrapha in the Tibetan Buddhist ‘Canonical Collections’: The case of the

Caryāmelāpakapradīpa Commentary Attributed to Śakyamitra.” Journal of the International Association of Tibetan Studies, no 5, 1-31.

---

. 2013. Making Sense of Tantric Buddhism: History, Semiology and Transgression in the

Indian Traditions. New York: Columbia UP.

---

. 2014. “The Vidyāvrata and its dānavidhi in the Guhyasamāja Tradition.” Conference

paper presented at “The Evolution of Tantric Ritual” Conference held at UC Berkeley in March 2014. Weinberger, Steven N. 2003. The Significance of Yoga Tantra and the Compendium of Principles (Tattvasaṃgraha Tantra) within Tantric Buddhism in India and Tibet. PhD dissertation, University of Virginia. White, David Gordon. 2000. Introduction. Tantra in Practice: Mapping a Tradition.” in David Gordon White, ed. Tantra in Practice. Princeton: Princeton UP, 3-38.

---

. 2003. Kiss of the Yogiṇī. Chicago: Chicago UP.

Wright, Roger. 2010. The Guhyasamāja Piṇḍikṛta-sādhana and its context. MA Thesis, SOAS. Yamada, Isshi. 1981. Sarva-tathāgata-tattva-saṅgraha-nāma-mahāyāna-sūtra: A Critical edition based on a Sanskrit manuscript and Chinese and Tibetan translations. Delhi: Jayyed Press. Yamamoto, Carl. 2009. Vision and Violence: Lama Zhang and the Dialectics of Political Authority and Religious Charisma in Twelfth-Century Central Tibet. PhD Dissertation. University of Virginia. 477 Yampolsky, Philip B. 1967. The Platform Sutra of the Sixth Patriarch: The Text of the Tunhuang Manuscript with Translation, Introduction, and Notes. New York: Columbia University Press. Yeshi, Khenpo and Jacob P. Dalton. 2018. “Signification and Hisotry in Zhang Nyi ma ‘bum’s rDzogs pa chen po tshig don bcu gcig pa.” Revue d’Etudes Tibetaines, no. 43, pp. 256- 273. Yiannopoulos, Alexander. 2017. “Tantric Yogācāra: Reflexive Awareness and the Four Stages in Ratnākaraśānti’s Epistemological Works.” JIABS, vol 40., pp. 239-267. Yonezawa, Yoshiyasu. 1998. “Sanskrit Fragment of the Mahāyānalakṣaṇasammucaya.” Journal of Research Society of Buddhism and Cultural Heritage No. 7, 1998: 36-65.

Source

Wikipedia:Enacting Perfection: Buddhajñānapāda’s Vision of a Tantric Buddhist World