Academia.eduAcademia.edu
LUMINOSITY Reflexive Awareness in Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way by Alexander Yiannopoulos © 2012 Acknowledgments This thesis would not exist were it not for the kindness and the wisdom of Chökyi Nyima Rinpoche, who has provided so many Western students with the ability to study the Buddhadharma in depth and detail. I am also deeply indebted to Khenpo Sherab Sangpo, who first opened my eyes to the profound Middle Way; Khenpo Sherab Dorje, whose explanations of the Three Natures dispelled my misunderstandings; and Khenpo Jampa Dönden, whose patience knows no bounds. Perhaps more than anyone else, however, I am in the debt of my advisor and editor, Karin Myers, who directed me towards the research materials I needed to bring this project to a successful conclusion, and tirelessly read through revision after revision, always providing extremely detailed and helpful feedback. I also give heartfelt thanks to Klaus-Dieter Mathes, who encouraged me to translate Dol po pa’s commentary to the Abhisamayālaṃkāra. The ensuing search for the intellectual genealogy of Dol po pa’s ideas, that is, for Indian sources supporting Extrinsic Emptiness (gzhan stong), led eventually to Ratnākaraśānti. In a similar vein, I would like to thank my friend Leo Brouwer, who initially pointed me in the direction of the Pith Instructions, and provided me with the sDe dge rescension of the text. Special thanks as well to Cyril Kassoff, Anders Bjonback, and Magda Galitou, for their invaluable moral support. The present work is the culmination of the better part of a decade spent studying at KaNying Shedrup Ling monastery in Boudhanath, Kathmandu, Nepal. In that time I listened to lectures and received instruction from many of the Western scholars whose names are found in the Bibliography, including: Georges Dreyfus, Tom Tillemans, John Makransky, William Waldron, Jay Garfield, Douglas Duckworth, and Sara McClintock. I thank them all. But I would especially like to thank John Dunne, for his help in going through the Tibetan translation of the Pith Instructions, and his elucidation of many of the Sanskrit-isms it exhibits, as well as the finer details of pramāṇa theory; my own Sanskrit is quite weak, and I would never have caught the more technical aspects of Ratnākaraśānti’s arguments without his help. Thanks as well to Dan McNamara, who sent me the recording of John’s read-through, and Rory Lindsay, who provided me with the Peking edition of the text. Finally, I thank my father, without whose generous support I could not have completed my studies. May whatever merit there is in the composition of this translation and commentary bring benefit to limitless beings. SARVA MAṆGALAṂ For my teachers TABLE OF CONTENTS I. Introduction and Methodology………………………………………………………………….1 A. The Life and Works of Ratnākaraśānti………………………………………………….1 B. Methodology…………………………………………………………………………….7 1. Problems of Radical Incommensurability in the Study of Buddhist Philosophy..7 2. Approaching Ratnākaraśānti’s Pith Instructions…………………………….…18 II. The Intellectual Background of the Pith Instructions for the Ornament of the Middle Way....26 A. The Two Truths………………………………………………………………………..26 1. The Abhidharma Context……………………………………………………….26 2. “Emptiness” in the Great Vehicle………………………………………………32 3. “Emptiness” and the Root Verses of the Middle Way..........................................35 4. Three Approaches to the Two Truths...................................................................41 5. The Two Truths in the Pith Instructions………………………………………..49 B. The Three Natures……………………………………………………………………..60 1. The Constructed Nature………………………………………………………..61 a) Etymology of the Constructed Nature…………………………………61 b) Characteristics of the Constructed Nature……………………………..61 c) Phenomenology and the Constructed Nature…………………………..67 2. The Dependent Nature…………………………………………………………71 a) Etymology of the Dependent Nature…………………………………..71 b) Characteristics of the Dependent Nature……………………………....72 c) Idealism and the Dependent Nature………………………………….......81 3. The Perfected Nature…………………………………………………………..86 a) Etymology of the Perfected Nature……………………………………86 b) Characteristics of the Perfected Nature………………………………..86 c) Two Models of the Perfected Nature…………………………………..90 C. The Means of Reliable Knowledge (Pramāṇa)………………………………………..100 1. The Indian Context………………………………………………………….....100 2. The Pramāṇa of Direct Perception…………………………………………...107 a) Perception in the Abhidharma………………………………………...107 b) Buddhist Pramāṇa Theory…………………………………………….109 c) External Realism vs. Epistemic Idealism……………………………...116 3. False vs. True Images…………………………………………………………124 III. Reflexive Awareness………………………...……………………………………………141 A. Reflexive Awareness in Buddhist Perceptual Theory……………………………….144 B. Ratnākaraśānti’s Presentation of Reflexive Awareness in the Pith Instructions…….161 1. Reflexive Awareness and Pramāṇa Theory……………………………….…161 a) Reflexive Awareness as Direct Pramāṇa……………………………..161 b) Reflexive Awareness as Luminosity………………………………….167 c) The Four Yogas as Objects of Knowledge…………………………....172 2. Reflexive Awareness and the Three Natures……………....………………….182 3. Reflexive Awareness and the Two Truths…………………………………….188 IV. Conclusion………………………………………………………………………………….198 Appendix: Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way…………..220 Bibliography………………………………………………………………………………...….248 List of Abbreviations LAS – Laṇkāvatāra Sūtra (Descent into Laṇka Sūtra) MMK – Mūlamadhyamakakārikā (Root Verses of the Middle Way) of Nāgārjuna MSA – Mahāyānasūtrālaṃkāra (Ornament of the Great Vehicle Sūtras) of Asaṇga MV – Madhyāntavibhāga (Distinguishing the Middle from the Extremes) of Maitreya PS – Pramāṇasamuccaya (Compendium of Pramāṇa) of Dignāga PV – Pramāṇavarttika (Pramāṇa Commentary) of Dharmakīrti TK – Triṃśikāvijñaptikārikā (Thirty Verses on Consciousness) of Vasubandhu TSN – Trisvabhāvanirdeśa (Explaining the Three Natures) of Vasubandhu VK – Viṃśatikākārikā (Twenty Verses) of Vasubandhu VKV – Viṃśatikākārikāvṛtti (Commentary on the Twenty Verses) of Vasubandhu Abstract Despite his stature within the Buddhist commentarial lineage, the works of Ratnākaraśānti remain little-known and even less studied. This is unfortunate, since as a late Indian Yogācāra commentator, Ratnākaraśānti’s articulation of the “False Imagist” view (alīkākāravāda) is worthy of study in its own right, and provides a valuable window into the philosophical debates which shaped later Tibetan polemics. In no wise is this more true than with respect to Ratnākaraśānti’s presentation of “reflexive awareness” (svasaṃvedana/svasaṃvitti). The reflexive nature of awareness was the centerpiece of Ratnākaraśānti’s ontology, phenomenology, epistemology, and praxeology; however, despite its importance to Ratnākaraśānti and other similarly influential commentators, it remains only poorly addressed within the Western literature. In order to address this deficiency, my thesis is a complete translation of Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way (Madhyamakālaṃkāropadeśa), appended to a commentary elucidating this complex work’s most important themes, particularly and especially the reflexive nature of awareness. Translator’s Note: The translation of the Pith Instructions is made on the basis of five separate manuscripts: sDe dge (D), Co ne (C), Peking (P), sNar thang (N), and gSer bris (S). 1 I. Introduction and Methodology Before examining the Pith Instructions for the Ornament of the Middle Way in detail, it is appropriate to say a few words by way of introduction: first, a brief biographical sketch of Ratnākaraśānti, and second, an overview of the hermeneutical and methodological considerations informing this thesis. These methodological considerations are further subdivided into two parts. First, problems typical of Western academic engagement with “Buddhist Philosophy” in general are identified. Finally, the specific issues related to the study of the Pith Instructions for the Ornament of the Middle Way in particular are considered. A. The Life and Works of Ratnākaraśānti Biographical information on Ratnākaraśānti, also known as the Mahāsiddha Śānti-pā, is in short supply. The Biography of the Eighty-Four Siddhas records that he was born a brāhmaṇa.1 The Biography also records that Ratnākaraśānti’s tantric teachers were two other Mahāsiddhas, Thagana-pā (ca. 950) and Nāro-pā (ca. 956-1041),2 the latter of whom was abbot at the great monastic university of Vikramaśīla. Interestingly, the Biography of the Eighty-Four Mahāsiddhas reports that Ratnākaraśānti was a knowledgeable and well-respected scholar, but lacked the “accomplishment of the Great Seal” (mahāmudrā-siddhi), a synonym for 1 Abhayadhatta, trans. James B. Robinson, Buddha’s Lions: The Lives of the Eighty-Four Siddhas (Caturaśītisiddha-pravṛtti, Tib. trans. sMon grub Shes rab, grub thob brgyad cu rtsa bzhi’i lo rgyus) (Berkeley: Dharma Publishing, 1979), 60. 2 Keith Dowman, Masters of Mahāmudrā (Albany: State University of New York Press, 1985), 99. 2 enlightenment, until instructed in meditation by a former student, Koṭāli-pā (ca. 1000).3 Tāranātha (1575-1634) does not mention either Thagana-pā or Koṭāli-pā in relation to Ratnākaraśānti, but reports that Vagīśvarakīrti (ca. 950) taught him the Dharma, 4 and expands on the relationship between Nāro-pā and Ratnākaraśānti: Along with his disciples, ācārya [master] Śānti-pā was once engaged in an act of worship. He sent one of his disciples with offerings, who saw a terrible-looking yogi on the bali [ritual cake, gtor ma] altar. He was panic-stricken and just threw the offerings, came back and reported it to the ācārya [Śānti-pā]. The ācārya realised that this must have been Nāro-pā and so he invited him. He sat at his feet and received many abhiṣekas [ritual empowerments, dbang], sermons and explanations of the sermons etc. He bowed down before him again and again. Later on, when Śānti-pā was about to attain siddhi [accomplishment]… there was a body of a dead elephant on the cross-road. [Nāro-pā] performed the rite of entering into it [projecting his consciousness, pho ba] and it walked into the crematorium. When Śānti-pā came across it, Nāro-pā’s voice said, “Such is the mark of my being a yogi. Will not the great scholar also show some such mark?” Ācārya Śānti-pā said, “What can a person like me perform afterall? However, if permitted by a person like you, I may perform something.” Some people were approaching the place with pitchers full of water. By his magic spell, the water was transformed into liquid gold. He then distributed it among the monks and bhrāmaṇas.5 Multiple sources attest that Ratnākaraśānti was one of the “Guardians of the Gate” (Skt. dvārapāla, Tib. sgo srung) or chief debate-masters at Vikramaśīla, where he received instruction and empowerment from Nāro-pā. There is some uncertainty as to exactly which gate he 3 Abhayadhatta, trans. James B. Robinson, Buddha’s Lions: The Lives of the Eighty-Four Siddhas, 62-64. Koṭāli-pā was also one of the eighty-four Mahāsiddhas. 4 Tāranātha, Lama Chimpa and Alaka Chattopadhyaya, trans., Tāranātha’s History of Buddhism in India (rgya gar chos ‘byung, dgos ‘dod kun ‘byung), 299. 5 Ibid., 299-300. 3 guarded,6 however both Tarānātha in his History of Buddhism in India7 and ‘God Lo tsa ba gZhon nu dpal (1392-1481) in the Blue Annals8 agree that Ratnākaraśānti guarded the eastern gate. Tāranātha may have only been following ‘God Lo tsa ba, but the colophon to the Pith Instructions describes Ratnāksaraśānti as having been the primary of the four Guardians (go srung chen po bzhi las kyang gtso bor gyur pa)9—that is, the guardian of the eastern gate. Certainly he was among the most prolific and highly respected authorities of his age. It is, however, difficult to provide exact dates for Ratnākaraśānti. Tāranātha divides the chapters of his History by the reign of the king, and places Ratnākaraśānti during the regency of Canaka (r. 955-983).10 In the Blue Annals, ‘Gos Lo tsa ba records that he was among the teachers of Maitrī-pā (ca. 1007-1085) and Atiśa (986-1065),11 extending his academic career into the middle part of the eleventh century. According to Keith Dowman, “If he gave Atiśa permission to leave the monastery [for Tibet] he was in office in AD 1040.”12 Tāranātha reports that after leaving (or being expelled)13 from Vikramaśīla, Maitrīpā eventually returned, but by that time 6 Jean Nadou, Buddhists of Kashmir (Delhi: Agam Kala Prakashan, 1980), 159-160. 7 Tāranātha, History of Buddhism in India, 295. 8 ‘God Lo tsa ba gZhon nu dpal, The Blue Annals (deb ther sngon po), trans. G. N. Roerich (Delhi: Motilal Banarsidass, 1976), 206. 9 Ratnākaraśānti, trans. Śāntibhadra and Śākya ‘Od, Pith Instructions for the Ornament of the Middle Way (Madhyamakālaṃkāropadeśa, dbu ma rgyan gyi man ngag). D: 4085 (sDe dge hi: vol. 138, ff. 223.b-231.a), ff. 231.a.3. 10 Tāranātha, History of Buddhism in India, 294. 11 ‘God Lo tsa ba, Blue Annals, 842. 12 Keith Dowman, Masters of Mahāmudrā, 99. 13 Mark Tatz, “Maitrī-pa and Atiśa,” Tibetan Studies: Proceedings of the 4th Seminar of the International Association for Tibetan Studies, ed. Helga Uebach and Jampa L. Panglung (Munchen: Kommission für Zentralasiatische Studien Bayerische Akademie der Wissenschaften, 1998): 473. “Prominent in Tibetan accounts of the life of Maitrī-pā is a tale of how he was expelled from the monastic university of Vikramaśīla by Atiśa (Dīpaṃkaraśrījñāna). The story does not appear in any non-Tibetan source.” 4 Ratnākaraśānti’s tenure as Guardian of the Gate was finished.14 Meanwhile, the Biography of the Eighty-Four Siddhas records that Ratnākaraśānti was invited to Śrī Laṇkā by an otherwise unknown “King Kapina” during the reign of Devapāla (r. 810-850).15 Concerning this story, Dowman notes, “As history of Śrī Laṇkā the legend is incomprehensible. There is no King Kapina in the lists of Siṇghala kings… [and] there is no evidence of a Śāntipa contemporary with the Pāla Emperor Devapāla.”16 Most likely, this episode is a corruption of the story related by Tāranātha, that Ratnākaraśānti was invited to the Pāla court by Canaka. The colophon to the Pith Instructions also records that Ratnākaraśānti was invited to Oḍḍiyāna by wisdom ḍākinīs. At Vikramaśīla, Ratnākaraśānti was a member of the “Ārya School” of tantric practitioners and commentators, the lineage stemming from Saraha/Rāhulabhadra through Nāgārjuna(pāda) and Śavaripa. While according to Dowman “the bulk of Ratnākaraśānti’s literary output deals with the Prajñāpāramitā”17 or Perfection of Wisdom literature, the importance of Ratnākaraśānti’s contributions to tantric or esoteric exegesis cannot be understated. Indeed, of Ratnākaraśānti’s exoteric corpus, only two texts, a technical treatise on logic, the Justifying Inner Pervasion (Antarvyāptisamarthana), and a commentary on the eight thousand line Perfection of Wisdom, the Sarātamā18, are extant in Sanskrit; the latter survives only incompletely. His other exoteric works—the Pith Instructions for the Perfection of Wisdom 14 Tāranātha, History of Buddhism in India, 304. 15 Abhayadhatta, trans. James B. Robinson, Buddha’s Lions: The Lives of the Eighty-Four Siddhas, 61. 16 Keith Dowman, Masters of Mahāmudrā, 98. 17 Ibid., 99. 18 Padmanabh S. Jaini, “The ‘Ālokā’ of Haribhadra and the ‘Sāratamā’ of Ratnākaraśānti: A Comparative Study of the Two Commentaries of the ‘Aṣṭasāhasrikā,” Bulletin of the School of Oriental and African Studies, University of London 35, no. 2 (1972): 272. Jaini notes that “The name of the work as given in both MSS, at the end of the parivartas and in the colophon, is Sāratamā... However, The Tibetan translation of this work… calls it Sārottamā. Although Sārottamā would appear to be a better title we have followed the evidence of our MSS and have accepted Sāratamā as the original title of this work.” 5 (*Prajñāpāramitopadeśa), the Pith Instructions for Meditating on the Perfection of Wisdom (*Prajnāpāramitābhāvanopadeśaī), Establishing Mental Representation-Only (*Vijñaptimātratāsiddhi), and two commentaries on the Ornament of the Middle Way19 (*Madhyamakālaṃkāropadeśa and *Madhyamakālaṃkāravṛtti)—are only extant in Tibetan translation. Although these texts are interesting both in their own right and as examples of late scholastic Indian Yogācāra, institutional monastic Indian Buddhism had collapsed within a few centuries of Ratnākaraśānti’s active period, and these exoteric works had no direct impact on Tibetan scholastic philosophy. By contrast, Ratnākaraśānti’s three separate commentaries on the esoteric Gathering of Secrets (Guhyasamāja) Tantra, as well as his commentaries on the Great Illusion (Mahāmāyā) Tantra and the Hevajra Tantra, are all extant in complete Sanskrit manuscripts; a number of shorter tantric commentaries and ritual texts (sādhanās) survive in Tibetan translation. Ratnākaraśānti’s exegeses of the Gathering of Secrets and Hevajra tantras were extremely influential in Tibet, where they were considered authoritative by all sides of otherwise intractable sectarian divisions. Concerning those sectarian divisions, later Tibetan tradition often follows ‘God Lot sa ba, who reported that Ratnākaraśānti was defeated in debate by his student Maitrī-pā.20 Purportedly, Ratnākaraśānti defended the “False Imagist” (alīkākāra) Yogācāra view, while Maitrī-pā argued from a Madhyamaka perspective. However, Mark Tatz points out that this story was in fact 19 Shinya Moriyama, “On Ratnākaraśānti’s Theory of Cognition with False Mental Images (Alīkākāravijñānavāda),” JIABS 33 (forthcoming). Shinya Moriyama asserts that these are commentaries to an “independent treatise” named the Ornament of the Middle Way, as opposed to the Ornament of the Middle Way composed by Śāntarakṣita. While it is true that Ratnākaraśānti does not directly cite Śāntarakṣita’s Ornament in these texts, Moriyama provides no additional evidence to support this rather implausible claim. 20 ‘God Lo tsa ba, Blue Annals, 842. 6 “interpolated in the Atiśa biographies,”21 and translates Tāranātha’s strong disputation of the tale’s accuracy: The purportedly historical accounts (lo rgyus ‘chad pa) of Tibetans—that [Maitrī-pā] debated with Śānti-pā—appear to make no sense, and they have no counterpart whatsoever in the oral tradition of Āryadeśa. Since it is so reported in Tibet, I submit that… “Tibetans snap their fingers at the [samayas]22 by yelping like dogs.” Who are the accomplished ones verses the ordinary practicioners? That is to say, the lies of Tibetan fools should be considered as dramatization.23 Tatz further notes that the Sham Shere manuscript, which “may be the earliest witness to the life of Maitrī-pā,” describes Maitrī-pā’s instruction in “False Imagist” Yogācāra under Ratnākaraśānti, but makes no mention of this episode.24 All sources agree that Ratnākaraśānti lived a very long life. The Biography of the EightyFour Siddhas reports that he lived for seven hundred years.25 The oral tradition maintains that he is still alive, practicing meditation in a temple underneath the great Swayamabhūnāth Stūpa in Kathmandu, Nepal. 21 Mark Tatz, “Maitrī-pa and Atiśa,” Tibetan Studies: Proceedings of the 4th Seminar of the International Association for Tibetan Studies, ed. Helga Uebach and Jampa L. Panglung (Munchen: Kommission für Zentralasiatische Studien Bayerische Akademie der Wissenschaften, 1998): 476. 22 Samaya (Tib. dam tshig) is a general term referring to tantric commitments, above all the commitment to always respect and never denigrate one’s teacher. 23 Ibid., 476-477. 24 Ibid., 473. 25 Tāranātha, History of Buddhism in India, 295 n. 12. 7 B. Methodology 1. Problems of Radical Incommensurability in the Study of Buddhist Philosophy The problems faced by Western students of Asian philosophical and religious traditions are well-attested. In general they fall into two broad categories: on the one hand, the temptation to see any and every philosophical school or stance in terms of some familiar Western source, and on the other, the temptation to view the object of one’s study as being irreducibly alien and therefore incomprehensible without an in-depth knowledge of Asian culture, history, languages, and so on. The latter tendency, in essence a species of Orientalism, is most pronounced in the popular literature. Academics are generally loath to admit that there exists any subject area they may be incapable of comprehending through textual study, even where textual study is inadequate to the task of meaningful comprehension—as, most saliently, in the case of contemplative practices such as the meditation on emptiness. Accordingly, it is the former temptation, to over-determine the meaning of an argument or text with respect to tropes from the Western intellectual tradition, which dominates the academic study of Buddhism. Although this temptation manifests in a wide variety of forms, it has two primary consequences. The first major consequence has been the tendency, especially since the latter half of the 19th century, for Westerners to measure the doctrines, practices, and philosophical claims of Buddhism against the intellectual priorities of the so-called European “Enlightenment,” particularly with respect to the latter’s emphasis on rational and material categories of truth, but 8 also as regards its humanist and secularist bent. Thus Colonel Henry Steel Olcott was perhaps the first, but certainly not the last, to present Buddhism as a consummately scientific and rational religion. In keeping with this tendency, any Buddhist texts or narratives involving magical, supernatural, or otherwise immaterial phenomena were either rejected on their face or else discredited and downplayed as deviations from some putative “pure” or “original” Buddhism. We see this, for example, in Lamotte’s History of Indian Buddhism, where he writes: The increasing success of [Mahāyāna] propaganda had the effect of transforming Buddhism from the philosophico-mystical message which it was to begin with into a true religion including a God (more exactly, a deified Buddha), a pantheon, holy ones, a mythology and a worship. This religion soon infiltrated the monasteries and influenced, though only slightly, the learned scholars.26 Following this line of thought, some authors such as Stephen Batchelor have attempted to articulate a “Buddhism without beliefs,” stripped of “supernatural” elements such as karma and rebirth. Although he purports to aid his readers in understanding emptiness, his conception of emptiness differs radically from that of traditional Buddhist sources. According to Batchelor, emptiness “…is not something we ‘realize’ in a moment of mystical insight,” 27 where “mystical insight” presumably denotes a type of experience that transcends language and conceptual thought. Batchelor has gone so far as to claim that “We can no more step out of language and 26 E. Lamotte, trans. Sara Webb-Boin, The History of Indian Buddhism (Louvain-La-Neuve: Publications de l’Institut Orientaliste de Louvain, 1988), 620. 27 Stephen Batchelor, Buddhism Without Beliefs: A Contemporary Guide to Awakening. (New York: Riverhead Books, 1997), 81. 9 imagination than we can step out of our bodies.”28 As will become clear, Ratnākaraśānti asserts alongside Nāgārjuna (ca. 100-200) and literally the entire Buddhist tradition that the contemplation of emptiness necessarily and by definition requires “step[ping] out of language,” and more broadly out of conceptual thought altogether. The second, related consequence of this temptation has been a reductive approach to Buddhist philosophy in which the views of Buddhist philosophers are seen as inchoate articulations of the views held by Western philosophers. Thus T.R.V. Murti29 explained Middle Way philosophy by way of constant reference to Kant and Hegel, while in a similar vein, but with respect to a tradition of philosophy that practically defined itself in opposition to both Kant and Hegel, Streng30 apparently believed that Nāgārjuna is best understood in relation to Ludwig Wittgenstein and the Vienna Circle. Chris Gudmensen made an even more radical, and problematic, claim: “We cannot say that Wittgensteinianism and the Mādhyamika are ‘just the same’; all modern adherents of the Mādhyamaka ought, in my submission, to be Wittgensteinians, but followers of Wittgenstein need not become Buddhists.”31 Thankfully, more recent comparative studies, approaching Nāgārjuna and the Madhyamaka or “Middle Way” school from the perspective of e.g. Hume and Sextus Empiricus32 or Derrida,33 have at least tended to qualify the comparison by noting the most 28 Ibid., 39. 29 T.R.V. Murti, The Central Philosophy of Buddhism: A Study of the Mādhyamika System (London: C. Tinling & Co., Ltd., 1955). 30 Frederick Streng, Emptiness: A Study in Religious Meaning (New York: Abingdon Press, 1967). 31 Chris Gundmundsen, Wittgenstein and Buddhism (London: Macmillan, 1977), 115. 32 Jay Garfield, The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā (Oxford: Oxford University Press, 1995). 33 Jin Y. Park, ed., Buddhisms and Deconstructions (Lanham: Rowman and Littlefield, 2006). 10 significant points of divergence and incommensurability between Buddhist and Western systems of philosophical discourse. At the same time, if, as Jan Westerhoff writes, “There is no more need to legitimate a study [of Nāgārjuna] by setting out to show him to be a proto-Kant, protoWittgenstein, or proto-Derrida,”34 the same cannot apparently yet be said of other Buddhist philosophical traditions. Thus, for example, the philosophy of the Yogācāra or “Yogic Practice” tradition is still predominantly seen, depending upon the proclivities and academic background of the Western interpreter, as having exactly the same ontological underpinnings and epistemological implications as either Berkeley’s idealism35 or Husserl’s phenomenology36— despite the fact that, like Kant’s transcendental idealism and Wittgenstein’s philosophy of language, these are radically opposite. It is important to note here that various similarities in approach or content may well exist between various elements of Western philosophy and any number of Buddhist philosophical traditions. Certainly Derrida, with no prior knowledge of Nāgārjuna, stumbles into the same “tetralemma” of existence/nonexistence/both/neither;37 certainly some aspects of Yogic Practice philosophy, which will be attended to below, strongly imply an idealistic ontology. But it remains to be demonstrated that comparative philosophy in the mode of articulating supposed “similarities,” especially across enormous cultural and historical chasms, can be performed 34 Jan Westerhoff, Nāgārjuna’s Madhyamaka: A Philosophical Introduction (Oxford: Oxford University Press, 2009), 12. 35 Jadunath Sinha, Indian Realism (Delhi: Motilal Banarsidass, 1999). Cf. also Ashok Kumar Chatterjee, The Yogācāra Idealism (Delhi: Motilal Banarsidass, 1975). According to Chatterjee (p. 204), “Nothing has done more injustice to the Yogācāra than the line of interpretation which makes it an Indian edition of Berkeley.” 36 Dan Lusthaus, Buddhist Phenomenology: A Philosophical Investigation of Yogācāra Buddhism and the Ch’eng Wei-shih lun (New York: RoutledgeCurzon, 2002). 37 Ian W. Mabbett, “Nāgārjuna and Deconstruction,” Philosophy East and West 45 (1995): 203-225. 11 without distorting at least one (and quite possibly both) of the philosophical systems under consideration. For example, in the introduction to his translation of and commentary upon Nāgārjuna’s Root Verses of the Middle Way, Jay Garfield emphasizes the importance of not stretching the text to fit our own intellectual and cultural presuppositions: Now a word about the methodology and intent of this commentary: Since the intended audience is Western philosophers and students of philosophy whose primary study has been in the Western tradition, I have tried throughout, insofar as that is possible without distortion of the meaning of the text, to explain Nāgārjuna’s arguments and positions in language familiar to Western philosophers. I have occasionally used analogies to positions and arguments found in Western texts, but have avoided doing so where I thought that the comparisons might force a Procrustean38 analysis of Nāgārjuna’s own views. And it is, of course, impossible and pointless to completely recast Nāgārjuna’s positions as those with which we in the West are familiar and to replace his technical terminology with ours. For Nāgārjuna is not a Western philosopher. He is an Indian Buddhist philosopher whose work we approach through a vast Asian Buddhist commentarial literature. And while many of his concerns, problems, theses, and arguments are recognizable cousins of ours, many are not, and there are genuine differences in outlook.39 Garfield’s point here is well-taken. It is, indeed, extremely important for the Western student of Buddhist philosophy to remember that Buddhist philosophy radically diverges from Western philosophy in any number of ways. However, while Garfield manages to avoid any gross oversimplification of Nāgārjuna’s view, he repeatedly falls prey to a more subtle form of 38 Procrustis was a figure from Greek mythology, who strapped passing travelers to an iron bed, then stretched them out to fit it. 39 Garfield, The Fundamental Wisdom of the Middle Way, 95. 12 “Procrustean” distortion, as seen in his discussion of the first chapter of the Root Verses, the discussion of causality: To assert the emptiness of causation is to accept the utility of our causal discourse and explanatory practice, but to resist the temptation to see these as grounded in reference to causal powers or as demanding such grounding. Dependent origination simply is the explicability and coherence of the universe. Its emptiness is the fact that there is no more to it than that.40 The problem with Garfield’s exegetical approach to emptiness here (and elsewhere) is that he mischaracterizes some absolutely crucial elements of Nāgārjuna’s presentation. To be sure, Nāgārjuna does not deny the practical utility of ordinary discourse, causal or otherwise. But the claim that emptiness is nothing more than “the explicability and coherence of the universe” in the absence of “grounded” causal powers is strictly false. Whatever other role(s) emptiness plays in his philosophy, Nāgārjuna clearly states that emptiness serves as the means by which karma, affliction, and conceptual structuring are all brought to cessation: Karma and affliction [result] from mental construction; Liberation [results] from the elimination of karma and affliction. That conceptual structuring, [which results] from conceptual structuring, Is brought to cessation by emptiness.41 [MMK 18.5] 40 Ibid., 122 (emphasis added). 41 Nāgārjuna, Root Verses of the Middle Way (Mūlamadhyamakakārikā). Ed. P.L. Vaidya (Darbhanga: Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning, 1960). karmakleśakṣayānmokṣaḥ karmakleśā vikalpataḥ| te prapañcātprapañcastu śūnyatāyāṃ nirudhyate|| It should be noted that, in both the original Sanskrit and the Tibetan translation of this passage, emptiness (Skt. śūnyatā, Tib. stong pa nyid) serves as the grammatical instrument, i.e. that by means of which an action is accomplished (stong pa nyid kyis ‘gag par ‘gyur/). Instrumentality is an extremely important topic in Sanskrit grammar, and there will be more to say about it later. 13 Obviously, the mere “explicability and coherence of the universe” is incapable of serving as the means by which conceptual structuring, karma, and affliction are all eliminated. Indeed, as Garfield rightly notes, one of Nāgārjuna’s main points is that the attempt to articulate a coherent and consistent explanation of the universe is precisely the type of conceptual activity that must be pacified through the contemplation of emptiness. But Garfield is already committed to the view that emptiness is nothing more than the fact that the universe is coherent and explicable; that, in rough agreement with Siderits’ famous phrase, “the ultimate truth is that there is no ultimate truth.”42 This is probably the reason why, in his gloss of MMK 18.5, Garfield supplements Nāgārjuna’s words with a subtle interpolation: “The realization of emptiness eliminates [the] fabrication of essence, which eliminates grasping, contaminated action, and its pernicious consequences.”43 But this is not, of course, what Nāgārjuna actually says in the verse. Nor is it entirely clear how the “realization” of the fact that there is nothing more to dependent origination than the explicability and coherence of the universe, can lead to the cessation of all conceptual activity—to say nothing of ultimate liberation. The underlying problem here is that the emptiness of the phenomenal world, far from being nothing more than the fact that ordinary discourse can function in the absence of metaphysical speculation or grounding, directly and necessarily entails that the phenomenal world itself does not exist in the way that it appears to exist, even with respect to mere linguistic conventions. From a Mahāyāna Buddhist perspective, apprehending the phenomenal world as though it truly exists is already utterly mistaken. Traditionally, this point is expressed in terms of 42 Mark Siderits, Personal Identity and Buddhist Philosophy: Empty Persons (Burlington: Ashgate Publishing Company, 2003), 133. 43 Garfield, The Fundamental Wisdom of the Middle Way, 248 (emphasis added). 14 the distinction, addressed in greater detail below, between “conventional” and “ultimate” truth. But Western scholars all too commonly interpret this distinction as an unqualified affirmation of conventional reality, coupled to a negation of only the “ultimate existence” of that conventional reality, which negation constitutes the ultimate truth of “emptiness.” In the following passage, for instance, Garfield’s a priori equation of conventional reality with ultimate truth reduces the scope of emptiness to a matter of the ontological mode of existence that is asserted (or denied) for a world that is itself taken to be prima facie existent: To perceive conventional phenomena as empty is just to see them as conventional and dependently arisen. The difference—such as it is—between the conventional and the ultimate is a difference in the way phenomena are conceived/perceived… We typically perceive and conceive of external phenomena, ourselves, causal powers, moral truths, and so forth as independently existing, intrinsically identifiable, and substantial. But though this is, in one sense, the conventional character of conventional phenomena—the manner in which they are ordinarily experienced—to see them in this way is precisely not to see them as conventional. To see that they are merely conventional… is thereby to see them as empty, and this is their ultimate mode of existence.44 Again, there will be more to say shortly concerning the distinction between the “conventional” truth and the “ultimate” truth. Here it may simply be noted in passing that Garfield’s entire argument rests on his contention that there is no intelligible difference between the conventional and the ultimate, that the ultimate reality of phenomena is the fact that they are merely conventionally existent. Not only does this view obscure the very real and salient difference(s) between conventional and ultimate truth, it is also entirely out of keeping with Nāgārjuna’s 44 Ibid., 319-320. Emphasis added. 15 intent. But it perfectly fits the perspective of an academic philosopher who wants the “realization of emptiness” to be compatible with ordinary language, with everyday perceptions of the conventional world. This is hardly Garfield’s fault; it is, rather, a thoroughly predictable result of the realist commitments endemic to Western scholastic methodology in general. To reiterate, the fundamental problem is that realist commitments as such are radically incompatible with the Buddhist doctrine of emptiness. Insofar as such commitments to the reality of ordinary phenomena have served as the implicit, unquestioned context for explicating emptiness within the Western academy, Western academicians have failed even—and especially—at the most basic level of communicating that these types of realist commitments are precisely what Nāgārjuna intends to refute. Thus the wider point missed by Garfield is that emptiness, itself, is not merely the “conventional” coherence of the universe. Nor is seeing phenomena as “merely conventional” equivalent to seeing them as empty. On the contrary, realizing emptiness necessitates a radical re-constitution of our entire cognitive and perceptual apparatus, including changes such that we no longer perceive phenomena to be existent (or nonexistent) in any sense at all. To see phenomena as “merely conventional” is still to see them as being existent in some meaningful sense—namely, in the sense of being existent or established “as conventional.” But this is precisely the kind of view that Nāgārjuna critiques in the Root Verses. Jan Westerhoff touches on this point when he distinguishes between two different senses of “inherent existence” (Skt. svabhāva, Tib. rang bzhin), the ontological and the cognitive: If we conceive of the Madhyamaka arguments about svabhāva solely in ontological and semantic terms we are likely to miss one important dimension of the concept that occupies a central place in the Buddhist understanding of emptiness. This is the idea that the 16 purpose of determining the existence or non-existence of substance-svabhāva is not just to arrive at a theoretically satisfying understanding of the fundamental objects which make up the world, or of the relation between words and their referents, but that it is supposed to have far more comprehensive implications for how we interact with the world.45 Westerhoff illustrates here one of the primary differences between Buddhist and Western philosophy. While Western philosophy, particularly in its post-“Enlightenment” formulations, is essentially a descriptive project, Buddhist philosophy is inseparable from prescriptive orthopraxis. In other words, while Western philosophy attempts to describe reality, but generally makes no claims as to the efficacy of any particular method for coming into contact with the nature of reality—and, indeed, frequently proceeds on the assumption that direct contact with the nature of reality is impossible or impracticable as a goal of human endeavor—such an approach is, from a Buddhist perspective, wholly unsatisfactory, and furthermore fails at the most fundamental purpose of philosophy: to facilitate unmediated contact with, or direct and nonconceptual understanding of, the nature of reality, “suchness” or “thusness” (tathatā, Tib. de bzhin nyid). The governing commitment to the notion that Buddhist philosophy, qua “philosophy,” is fundamentally about the articulation and argumentation of ideas, instead of about complete and perfect enlightenment, liberation46 from suffering, as a real and actionable 45 46 J. Westerhoff, Nāgārjuna’s Madhyamaka, 46-47. This dimension of religious or philosophical projects is sometimes called “soteriology.” The term comes from Christian theology, and while it has been used with increasing regularity in texts concerned with the study of Buddhism, this seems to be an unhelpful development for what may be called, pace Makransky, “Buddhist theology.” For the aim of Buddhist theology, or the aim of Buddhist practice, is not “salvation” but liberation (Skt. mokṣa, Tib. thar pa). “Soteriology” requires a sotēr (Greek, “savior”), Christ the King. The Buddha is many things to many people, but from an orthodox Buddhist perspective, in no wise does the Buddha “save” anyone or perform the essential work of salvation. Thus throughout this thesis there appears the neologism eleutheriology (from Greek eleutheria, “freedom”) to refer to the final goal of Buddhist praxis. This is not intended in any way as a slight on Christian practices or beliefs, only as a means to ensure that what is separate is treated separately, and what is the same is treated the same. 17 goal, hopelessly mires the Western interpreter in a mess of contradictions that are absolutely and fundamentally irresoluble so long as that Western interpreter does not take seriously the possibility of unmediated access to the nature of reality, to say nothing of complete and perfect enlightenment. 18 2. Approaching Ratnākaraśānti’s Pith Instructions Translation is an inevitably precarious endeavor. This precariousness is reflected in the untranslateable title of Walter Benjamin’s famous essay, “Die Aufgabe des Übersetzers,”47 usually rendered in English as “The Task of the Translator.” But aufgabe does not only mean task or mission—it also means surrender, abandonment, failure. To translate is to fail to translate, if success is a one-to-one reproduction of the original work in a different system of language. Thus for Benjamin, the task of the translator is not mere fidelity to the original; it is, rather, to surrender to the pure or absolute language (reine Sprache) that is “spellbound” (gebannt) and imprisoned within the source language, to set this pure language free in a new language-act that breaks the barriers and expands the boundaries of the target language. 48 Translation, according to Benjamin, is therefore a productive and creative act, that when properly performed will illuminate at least as much as it necessarily obscures. Although Benjamin’s essay only entered the Anglophone after being translated and reproduced in the compilation Illuminations, it was originally the preface to a volume of poetry, Baudelaire’s Les fleurs du mal, which he had translated from French into German. Benjamin’s views, right or wrong, apply universally; but they were originally articulated with respect to a particular source-language (nineteenth century French) and target-language (early twentieth century German) that in the grand scheme of things are quite close to one another, both linguistically and temporally. The present effort, by contrast, aims to translate into twenty-first 47 Walter Benjamin, “Die Aufgabe des Übersetzers,” in Illuminationen: Ausgewählte Schriften 1 (Frankfurt: Suhrkamp, 1977). 48 Ibid., 61. 19 century English a text, written in Sanskrit some thousand years ago, that only survives in its Tibetan translation. Even granting Benjamin’s overall point, this situation presents both the translator and the reader with unique challenges. To begin with, Ratnākaraśānti was a highly educated scholar writing for other highly educated scholars. Moreover, the system of scholarship in which he participated, the discourse of pramāṇa or the “means of reliable knowledge,” was highly technical and possessed its own peculiar vocabulary, which in many respects was deeply informed (if not determined) by the structure of Sanskrit grammar. The key terms of this style of discourse are considered at some length below. But this is only the most superficial level of difficulty, and furthermore plays to a systematic advantage that the translator of philosophy possesses over the translator of poetry; no matter how ad hoc or arbitrary, the substitution of one technical term for another does not necessarily obscure the point at stake, provided the semantic range and import of the substituted term are considered fully. There is, of course, only so much tolerance for this type of contextual glossing that a translator can expect from his audience, but that is fundamentally an issue of style and flow as opposed to exegetical exigency. Far more problematic, and demanding of attention, is the wider context of academic discourse in which the present effort is located. The primary impediment to a proper interpretation of the Pith Instructions is neither the technical vocabulary of pramāṇa theory, nor the absence of an original source text, but the enormous gulf between the presuppositions of contemporary Western scholarship on the one hand, and the concerns which animate Ratnākaraśānti’s exposition of emptiness, epistemology, and reflexive awareness 49 on the other. 49 Throughout this thesis, the terms “reflexive awareness,” “reflexivity of awareness,” and “reflexive nature of awareness” are used interchangeably. Only the first of these (“reflexive awareness”) is strictly justifiable as a translation of the Sanskrit term svasaṃvedana/svasaṃvitti, however in addition to stylistic concerns these other 20 For, while Ratnākaraśānti is no stranger to logical argumentation, and emphatically acknowledges the importance of correct conceptual understanding, he also explicitly contends that a genuine understanding of emptiness and ultimate truth requires nonconceptual contemplative practice. Similarly, his idealistic ontology (that is, his view that the entire cosmos is nothing other than mind) thoroughly contradicts the physicalist worldview taken more or less for granted these days, just as his anti-realist epistemology runs counter to the realist scientific materialism so frequently and uncritically adopted within the academy today. The challenge for any Western interpreter of Ratnākaraśānti, then, is to present his views cogently and coherently, without eliding those aspects of his views which challenge the basic validity of the post-“Enlightenment” paradigm, or distorting/diluting his views with the Western academic orthodoxies of metaphysical realism and “scientific” materialism. Rather than attempting to relate these aspects of his work to some parallel in the Western intellectual tradition, when in fact no suitable analogue exists, or—worse—to blunt the impact of his arguments by adjusting them to conform with realist and materialist prejudices, the goal of this thesis is to approach Ratnākaraśānti’s work on its own terms. Primarily, this requires an appreciation of the fact that Ratnākaraśānti, like all Buddhist philosophers, must be understood as a Buddhist if he is to be understood as a philosopher: Ratnākaraśānti employs a sophisticated philosophical toolbox, but his purpose is not merely to provide an account of “how things are.” His prime concern is, rather, to guide his readers along the path to enlightenment. Beyond these challenges, however, there is a further problem compounding the difficulty of understanding and interpreting the Pith Instructions for the Ornament of the Middle Way: the forms serve to emphasize the fact that reflexive awareness does not denote an ordinary type of entity; in many cases, when used by Ratnākaraśānti in the Pith Instructions or explained in this thesis, svasaṃvitti refers to a facet of consciousness. 21 pressure of Tibetan doxography. In the title, and throughout the work, Ratnākaraśānti places his exegesis within the broader context of the Middle Way tradition; but his style or articulation of the Middle Way was marginalized within Tibet. Although the exact sequence of events is poorly understood, at some point in Tibetan history the commentaries of Candrakīrti became the primary lens through which Nāgārjuna’s Middle Way philosophy was understood. Indeed, the works of Candrakīrti were championed as representing the summit of Indian exegesis on the Root Verses of the Middle Way, despite the fact that there is no evidence of any Sanskrit commentaries on Candrakīrti’s Clear Words, and the Entry into the Middle Way seems not to have even been mentioned by any Indian scholars for a period of at least three hundred years following its composition. Only one Indian commentary on the Entry, composed by Jayānanda some five or six centuries after its first appearance, is attested in the textual record. Candrakīrti’s philosophy has been treated in excruciating detail elsewhere. Accordingly, it will not be considered in any great depth here. It should suffice to note that while various Tibetan commentators have raised any number of issues with Candrakīrti’s interpretation of Nāgārjuna over the centuries, a doctrinal conformity eventually emerged in direct correlation with the political ascendance of the dGe lugs school, Candrakīrti’s main Tibetan champions. This historical fact has had far-reaching consequences for the history of the Western academic study of Buddhist philosophy, particularly with respect to the Tibetan tradition, as up until very recently Western scholars of Tibetan Buddhism have almost exclusively trained under the dGe lugs. Thus the focus of dGe lugs Middle Way exegesis on Candrakīrti, and the consequent focus of the Western academy on the same, has resulted in a situation where views of emptiness that differ from Candrakīrti’s—most relevantly, the views of Ratnākaraśānti—have struggled to gain a Western audience. 22 This is deeply ironic, and not simply on account of the historical contingency of Candrakīrti's late rise to prominence. While Candrakīrti’s own views were anti-rationalist to a fault, founded as they were on a total rejection of pramāṇa theory and especially of the logical syllogisms employed therein, dGe lugs commentators synthesized Candrakīrti’s Middle Way philosophy with the epistemological theory of perhaps the greatest (and certainly the most prolific) Buddhist contributor to the pramāṇa discourse, Dharmakīrti. On the one hand, Candrakīrti’s view of emptiness—which amounts, as the Tibetan translators of Ratnākaraśānti’s Pith Instructions maintain, to nihilism—informed an uncharitable and hermeneutically suspect critique of the Yogic Practice tradition, which in his time was largely synonymous with the Buddhist pramāṇa tradition; despite arguing primarily from the type of perspective most commonly associated with the Sautrāntika or “Sutra-Follower” tradition of exegesis, Dharmakīrti makes use of such distinctively Yogic Practice concepts as the storehouse consciousness (ālayavijñāna) and reflexive awareness (Skt. svasaṃvedana/svasaṃvitti, Tib. rang rig). On the other hand, the dGe lugs synthesis of Candrakīrti and Dharmakīrti proceeded by maintaining the conventionally real existence of external phenomena. This is doubly ironic, since neither Candrakīrti nor Dharmakīrti would have agreed with this position, albeit for different reasons: Dharmakīrti did not accept external objects, while Candrakīrti did not accept any form of phenomenal realism. Given the above discordance, perhaps the most salient point of continuity between Candrakīrti and his dGe lugs commentators, not to mention dGe lugs-influenced Western scholars such as Jay Garfield, is the categorical rejection of Yogic Practice theory. The strong emphasis on the intellectual rejection of inherent existence or svabhāva, in more or less exclusively ontological terms, has led certain strands of traditional and modern Middle Way 23 exegesis—primarily, those strands indebted to Candrakīrti’s Entry into the Middle Way—to cast Yogic Practice philosophy as a strictly inferior or problematic ontology, and to overlook or even ignore the important contributions made by Yogic Practice scholars such as Ratnākaraśānti. These contributions are not all philosophical; Yogic Practice is first and foremost concerned with yogic practice, that is, with contemplative praxis. The four stages of yoga, of which the idealistic ontology of “Mind Only” (Skt. cittamātra, Tib. sems tsam) is but one level, are the indispensable praxeological framework without which Yogic Practice philosophy is only so much sophistry. It should come as no surprise, then, that the Ratnākaraśānti finds much to critique in the Entry into the Middle Way. In fact, Ratnākaraśānti engages with several arguments from that text, before dismissing them as silly and ill-informed. Ratnākaraśānti declines to refer to its author by name; as mentioned above, however, the Tibetan translators of the text specifically state that Candrakīrti was a “nihilist” (med par smra ba) who “perverted the intention of Nāgārjuna” (klu grub kyi dgongs pa ‘chal bar ‘gyur). The reason is clear: Candrakīrti considered both the Yogic Practice tradition in general, and reflexive awareness in particular, to be hopelessly mired in self-contradictory incoherence. In Buddhist epistemological theory, reflexive awareness refers to the fact that consciousness is immediately or reflexively aware of its own contents. For Candrakīrti, and the so-called *Prāsaṇgikas50 or “Consequentialists” who followed him, the idea that an object could act upon itself violated the most basic insights of Nāgārjuna’s Middle Way philosophy—although, as will become clear, this formulation of the doctrine of reflexive awareness thoroughly misses the point at stake. Ratnākaraśānti, by contrast, considered the reflexive nature of awareness to be an absolutely essential element of any coherent epistemology, especially regarding epistemic access to ultimate truth. Ratnākaraśānti accordingly 50 This term is not attested in the Sanskrit literature. 24 spends most of the Pith Instructions explaining and defending this doctrine, which is treated at length below. Here, it need only be understood that the crux of the disagreement is that Ratnākaraśānti did not consider emptiness to be strictly negative. Critics of the doctrine of reflexive awareness, and the Yogic Practice tradition more broadly, have as a rule employed a peculiar reading of emptiness in the service of their cause. Specifically, their interpretation of emptiness has centered on its anti-foundational character, to the exclusion of any non-negative dimensions. This method of interpretation explains emptiness primarily or even exclusively in terms of an antifoundationalism that must always remain silent about anything other than a particular kind of ontology: an anti-foundationalism, that is, which precludes the existence of reflexive awareness, and the “luminosity” (Skt. prakāśa, Tib. gsal ba) that Ratnākaraśānti identifies with the reflexive nature of awareness. In particular, Candrakīrti’s critique of reflexive awareness centers on the conceit that it exhibits precisely the same characteristics as the essences or self-entities (i.e. the svabhāva) refuted by Nāgārjuna. This is despite the fact that even the earliest proponent of the doctrine of the reflexivity of awareness, Dignāga, was careful to note the grammatical or metaphorical nature of the trope of reflexivity. For his part, Ratnākaraśānti argues throughout the Pith Instructions that reflexive awareness does not contradict emptiness, by maintaining their union within a tradition he calls the “Middle Way of the Three Natures.” Rather than describing emptiness as a mere vacuity, Ratnākaraśānti stresses the positive dimensions of the direct, nonconceptual experience of emptiness through “transcendent wisdom” (Skt. lokottaraprajñā, Tib. ‘jig rten las ‘das pa’i shes rab). Indeed, it is this wisdom, or the nondual luminosity of reflexive awareness, which forms the basis of Ratnākaraśānti’s entire presentation in the Pith Instructions. 25 Thus, from my perspective, the single most striking characteristic of the few contemporary treatments of the topic of the reflexive awareness to date has been the extent to which they are dominated by the concerns of traditions, both Buddhist and Western, which deny the existence of reflexive awareness, and are furthermore so often colored by an a priori rejection of the very possibility of nondual mental states. If, then, the goal of this thesis is to provide a coherent and charitable account of what Ratnākaraśānti wrote, for an audience to whom even Ratnākaraśānti’s basic terminology is opaque, despite—and in some cases directly owing to—half a century of Western scholarship, “success” at the task set before this translator first of all requires an approach that, at the very least, takes seriously the argument that emptiness and luminous reflexive awareness are not, in fact, contradictory; as well as the argument that dualistic intentionality as such is a form of cognitive error or distortion. In this sense, the methodology governing the present translation of and commentary upon Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way proceeds from an acceptance of the basic validity of his project and its goals. It is, in other words, an articulation of the Buddhist philosophy of a Buddhist philosopher, in a manner that is consistent both with Buddhism, and with philosophy. 26 II. The Intellectual Background of the Pith Instructions for the Ornament of the Middle Way The Pith Instructions are structured in terms of three general topics: the two truths, the three natures, and the reflexive nature of awareness. Although Ratnākaraśānti devotes the least space to the two truths, they are the first topic he treats, and the conceptual bedrock upon which he builds his later presentation. They are also the main philosophical touchstone of Śāntarakṣita’s Ornament of the Middle Way, the text to which the Pith Instructions is ostensibly a commentary. Therefore, our first concern is to understand the two truths in broad outline, and to see how they relate to the project of the Pith Instructions. A. The Two Truths 1. The Abhidharma Context According to Newland and Tillemans, the distinction between two different types of truth was “initially a construct for reconciling apparently contradictory statements in scripture.”51 Thus some statements were in need of further interpretation (Skt. neyārtha, Tib. drang don), while other statements were definitive (Skt. nītārtha, Tib. nges don). Initially a hermeneutic framework intended to guide disciples toward a proper interpretation of the Nikāya scriptures, the concept gained additional currency during the formative period (ca. 100-200 CE) during which the Great Commentary on Abhidharma (Skt. Mahāvibhāṣa) traditionally ascribed to five hundred Arhats, as well as Nāgārjuna’s Root Verses of the Middle Way, were composed. While the hermeneutic dimensions of the two truths distinction never entirely disappeared, its first 51 Tom Tillemans and Guy Newland, “An Introduction to Conventional Truth” in Cowherds, Moonshadows: Conventional Truth in Buddhist Philosophy (Oxford: Oxford University Press, 2011), 5. 27 articulation in terms of “ultimate” (Skt. paramārtha, Tib. don dam) and “conventional” (Skt. saṃvṛti, Tib. kun rdzob) truth appears to have been within the Root Verses, where it was posited as part of a general critique of Abhidharma ontology and epistemology—particularly, though not exclusively, the Sarvāstivāda or “Pan-Realist” Abhidharma as found for example in the Great Commentary. In order to explain the two truths distinction, then, it is first necessary to say a few words about Abhidharma ontology and epistemology. The Abhidharma project was first and foremost analytical, in the literal sense of “pulling apart” (from Greek lyein “to loosen”) phenomenal appearances (Skt. dharma, Tib. chos) in order to arrive at an understanding of the true nature of reality. If a phenomenal form or dharma in the most general sense can be analyzed into component elements, or constitutes in any way a composite entity, it is only “designated to exist” (prajñaptisat, Tib. btags su yod pa). On the other hand, if—and only if—a phenomenon proves analytically irreducible, it is “substantially existent” (dravyasat, Tib. rdzas su yod pa). Within the Abhidharma system, the only entities that exist substantially are dharmas in a more restricted sense: the irreducible elements of phenomena. Equivalently, any and all compound entities are only “existent” in a secondary, parasitic or provisional manner. There is a close relationship here between the ontological implications of “existence” and the epistemological implications of “truth.” As Tillemans and Newland write, “Truth” here translates the Sanskrit satya… and it is certainly neither a wrong nor a sloppy translation. Nonetheless, it is problematic. The problem is that rendering satya, or its Tibetan, Pali, and Chinese equivalents, as “truth” naturally suggests that conventional and ultimate truths are all truthbearers, that is, statements, propositions, or, if we take a larger perspective, beliefs, ideas, and theories—in short, the sort of things to which we can 28 properly attribute truth and falsity. …. However, satya pertains not only to such truthbearers; things like pots and atoms can just as well be satya, and it is at least anomalous to ascribe truth to them. In short, because satya means “truth” but can also mean “real” and “what is existent,” translational problems are unavoidable.52 Dharmas themselves are defined in terms of their etymology, from the Sanskrit √dhṛ meaning “to hold.” A dharma is that which “holds” its own characteristics (Skt. svalakṣaṇa, Tib. rang mtshan) or its own existence (Skt. svabhāva, Tib. rang bzhin). In other words, dharmas are defined by their quality of self-identity, their self-nature. There is an inescapable ambiguity here, which Nāgārjuna exploited to great effect. On the one hand, dharmas are defined in terms of their intrinsic properties; for example, it is the nature (i.e. the “property-svabhāva”) of fire to be hot and burning, thus fire-dharmas have the svalakṣaṇa or svabhāva of being hot and burning. On the other hand, the fact that an instance of (“designated”) fire can be reduced to individual (“substantial”) fire-dharmas necessarily entails that, in contradistinction to ordinary phenomenal fire, fire-dharmas are finally real, and therefore possess an “essence-svabhāva” or really-existing self-nature. Thus an epistemological point about property-predication becomes inextricably bound with an ontological point about what is and is not real. Although Nāgārjuna intentionally conflates these two points in his critique of svabhāva, some interpreters53 read his refutation of inherent existence or essence-svabhāva in more or less strictly ontological terms, preserving the 52 53 Ibid., 4. Jeffrey Hopkins, Meditation on Emptiness (Boston: Wisdom Publications, 1996), 392. “Since in Prāsaṇgika [sic] emptiness—the absence of inherent existence (svabhāvasiddhi, rang bzhin gyis grub pa)—is the nature (svabhāva, rang bzhin) of all phenomena, it should not be thought that svabhāva is refuted in all its meanings. Svabhāva meaning svabhāvasiddhi or “inherent existence” is refuted, but svabhāva as ‘final nature’ or just ‘character’ [svalakṣaṇa] (such as heat and burning as the character of fire) is not refuted.” On the distinction between different senses of svabhāva, cf. also Jan Westerhoff, Nāgārjuna’s Madhyamaka, 19-52. What is called “property-svabhāva” above, Westerhoff terms “substance-svabhāva,” but the sense is more or less the same; “property-svabhāva” was chosen because it is more neutral with respect to the ontological issues at stake. 29 epistemologically grounded fact that fire has the property-svabhāva of being hot and burning as a facet of conventional truth. There will be more to say on this point below. Insofar as the distinction between “designated” and “substantial” dharmas concerns two types or modes of existence, it may be considered properly ontological. However, it is important to note that dharmas come in two basic “flavors”: material and mental. Physical objects are considered to be composed of material dharmas, while mental events are considered to be composed of mental dharmas. Aside from the interesting issues of mind-body dualism that this raises, it is also noteworthy for introducing a phenomenological dimension to the project; Abhidharma philosophy explicitly concerns the basic elements and categories of conscious experience, defined in terms which are primarily phenomenological as opposed to ontological. The distinction between designated and substantial existence was the axis around which the central questions of Abhidharma theory and contemplative practice turned. But it is essential to remember the latter, because like all Buddhist philosophy, Abhidharma literature did not exist independently of a well-structured praxeology. The point at stake in the Abhidharma style of practice was the foundational Buddhist teaching of anātman or “no-self.” The purpose of delineating all the various categories of material and mental phenomena was to lead the meditator on a complete analysis of everything in the cosmos, particularly and especially the meditator’s own causal continuum. By analyzing and observing each of the psycho-physical “bundles”54 (Skt. skandha, Tib. phung po), the meditator arrives at a direct, experiential 54 This term is sometimes translated as “aggregates,” however Hume makes essentially the same point in the Treatise on Human Nature (§ I.IV.VI “Of Personal Identity”), but calls the psychophysical constituents “bundles”: “There are some philosophers who imagine we are every moment intimately conscious of what we call our SELF…. But setting aside some metaphysicians of this kind, I may venture to affirm of the rest of mankind, that they are nothing but a bundle or collection of different perceptions, which succeed each other with an inconceivable rapidity, and are in a perpetual flux and movement. …. There is properly no simplicity in it at one time, nor identity in different; whatever natural propension we may have to imagine that simplicity and identity” (emphasis added). In Abhidharma theory, perception (Skt. samjñā, Tib. ‘du shes) is only one of the five “bundles,” while Hume 30 realization of the fact that the self is only designated to exist, i.e. that the word “self” has no real referent. In effect, the Abhidharma simultaneously served as an ontological account of what is real, an epistemological guide for contemplative practices aimed at ascertaining reality, and a phenomenological description of that reality—indeed, advanced practitioners of Abhidharmastyle meditation are said to be able to directly perceive individual atoms or dharmas. The slippage between ontological, epistemological, and phenomenological categories will be a recurring theme throughout the discussion of the Pith Instructions. In any case, the concept of two basic ontological and epistemological categories— categories which are distinguished, furthermore, on the basis of primary vs. secondary existence, and which thus form an explicit hierarchy—is found in Nāgārjuna’s Root Verses of the Middle Way as well as in Abhidharma works such as the Great Commentary. In the Root Verses, however, both the terminology of the distinction and the nature of the categories in question are radically different. As described above, mainstream Abhidharma distinguished these two categories on the basis of a relatively straightforward type of mereological reduction, or analysis into component parts. By contrast, Nāgārjuna distinguished between relative (Skt. saṃvṛti, Tib. kun rdzob)55 or conventional (Skt. vyavahāra, Tib. tha snyad)56 truth and ultimate truth apparently considered the illusion of identity to be sustained on the basis of a bundle of perceptions alone: “The mind is a kind of theatre, where several perceptions successively make their appearance… [but] the comparison of the theatre must not mislead us. They are the successive perceptions only, that constitute the mind.” While this does represent a significant difference with Abhidharma theory, the underlying point is more or less the same; and, as “bundles” thus evokes an appropriate pre-existing interpretive framework, while “aggregates” does not mean much of anything to the ordinary English-speaker, skandha/phung po has been translated as “bundle” throughout. 55 Tillemans and Newland, “An Introduction to Conventional Truth,” 12-13. “But do traditional sources actually confirm that the term saṃvṛti—be it used to designate linguistic usage, ideas, or objects—does indeed mean what we understand by ‘conventional,’ i.e. ‘agreement governed’? It is not immediately obvious that they do. As Edgerton has suggested and Karunadasa seconds, what the Pali renders as ‘consensus’ or ‘agreement’ (sammuti), based on √MAN ‘to think,’ is unexpectedly rendered in Sanskrit by saṃvṛti, which comes from the Sanskrit √VṚ vṛṇoti, ‘to cover, conceal.’ …. The first usage of saṃvṛti clearly relies on an etymological understanding in terms of [√VṚ], resulting in saṃvṛtisatya having the sense of ‘true-for-the-ignorant,’ ‘true-for-the-obscured,’ or ‘true-for-thebenighted.’ This saṃvṛti has only a remote connection, if any, with what we understand as ‘convention’ in the sense 31 (paramārthasat, Tib. don dam par yod pa) on grounds which are still disputed some two thousand years later. Although the precise delineation of the difference between these two would become one of the most hotly contested battlegrounds of intra-doctrinal Buddhist polemics, particularly in Tibet, Nāgārjuna himself mentions the distinction only once in the Root Verses. He does so near the end, in response to a Buddhist opponent who claims that Nāgārjuna’s teachings on “emptiness” (Skt. śūnyatā, Tib. stong pa nyid), particularly his contention that the Three Jewels and the Four Noble Truths are “empty,” obviate the Buddhist path: “If the Dharma57 and the Saṃgha do not exist, how can the Buddha exist? Speaking of emptiness in this way maligns the Three Jewels. The reality of causal results, what is Dharma and not Dharma, And the ordinary conventions of the world—you malign them all.”58 [MMK 24.5-6] Nāgārjuna responds first by arguing that the opponent has failed to appreciate what emptiness really means, and then by distinguishing between “two truths”: of rules and rule-guided activities, ways of thinking, and so on. It does, however, capture the aspect of the English term ‘conventional’ when it is used to say that something is superficial, shallow to the point of being misleading, and thus deceptive.” 56 John Dunne, Foundations of Dharmakīrti’s Philosophy (Boston: Wisdom Publications, 2004), 258. “In the classical Sanskrit of Dharmakīrti’s time, the term [vyavahāra] was used outside of philosophy in the sense of a ‘business agreement,’ ‘transaction,’ ‘litigation,’ and other related senses…. The meaning of vyavahāra as an agreement resonates with what may be the best known sense of vyavahāra in Buddhist philosophy—namely, vyavahāra in the sense of a ‘convention.’ 57 In addition to the prior technical definitions, “Dharma” also refers to the teachings of the Buddha, as the second of the “Three Jewels” of Buddha, Dharma, and Saṃgha or community of disciples. 58 dharme cāsati saṃghe ca kathaṃ buddho bhaviṣyati| evaṃ trīṇyapi ratnāni brūvāṇaḥ pratibādhase|| śūnyatāṃ phalasadbhāvamadharmaṃ dharmameva ca| sarvasaṃvyavahārāṁśca laukikān pratibādhase|| 32 This is a mistake. You have not understood the point of emptiness; In this way, you destroy both emptiness itself and the purpose of emptiness. The Dharma taught by the Buddha completely relies on two truths, The conventional truth of the world and the truth which is ultimate. Those who do not understand the division between the two truths, They do not understand the profundity taught by the Buddhas.59 [MMK 24. 7-9] In order to understand this objection, as well as the distinction between ultimate and relative truth that Nāgārjuna posits in response, it is first necessary to understand emptiness in the context of Mahāyāna or “Great Vehicle” Buddhism, as it is precisely in terms of emptiness that Nāgārjuna articulates the distinction between the two truths. 2. “Emptiness” in the Great Vehicle Originally, the term “emptiness” was essentially a synonym for the lack of self. But along with many other traditional Buddhist terms and categories, emptiness underwent a revolutionary shift during the genesis of the Mahāyāna. In the Perfection of Wisdom literature, emptiness became the centerpiece of a radical new ontology, an ontology which stood in stark contrast to the ontology of pre-Mahāyāna Buddhism. The Buddha taught that the world as it appears to ordinary beings is māya or illusion. But this illusory quality was not at first explained in terms of any kind of metaphysical anti-realism. 59 atra brūmaḥ śūnyatāyāṃ na tvaṃ vetsi prayojanam| śūnyatāṃ śūnyatārthaṃ ca tata evaṃ vihanyase|| dve satye samupāśritya buddhānāṃ dharmadeśanā| lokasaṃvṛtisatyaṃ ca satyaṃ ca paramārthataḥ|| ye'nayorna vijānanti vibhāgaṃ satyayordvayoḥ| te tattvaṃ na vijānanti gambhīraṃ buddhaśāsane|| 33 It was, instead, held to be the product of ignorance, primarily the ignorance of belief in a transcendental or absolute “Self.” This belief was considered ignorant or deluded since (according to the Buddha’s teaching) no such self exists to be found; the apprehension of one’s self, the sensation of being a particular individual, is only a type of cognitive error. In the Abhidharma tradition, as previously mentioned, this ignorance was to be eliminated through the examination of each and every phenomenon in the cosmos, in order to determine if the self could be located among the irreducible, “substantially existent” phenomena. In seeing that the self does not exist substantially, and is therefore only “designated to exist,” the cognitive error of apprehending a self could be corrected. However, within this framework, the ontological status of the self qua not-substantiallyexistent entity did not fundamentally differ from the ontological status of any other composite entity that was merely designated to exist. The non-existence of the self was only one special case of the non-substantial-existence of everything other than partless, irreducible dharmas. The self was only designated to exist in dependence upon the psycho-physical bundles, but had no existence independently of those bundles; just so, designated phenomena are, in general, designated to exist in dependence upon the aggregated causal interactions of innumerable individual dharmas, which form the ontological foundation for all of appearance and existence. In this manner, “designated” phenomena were held to exist, just not in the same way that their component dharmas were held to exist. In fact, dharmas were maintained to be the only things which truly existed, and the entire world of designated phenomena held to be nothing other than their physical agglomeration and causal interaction (which in many cases were the same thing, since the junction of one dharma with another is precisely a causal interaction). But even those phenomena merely “designated to exist” were understood have some type of positive 34 ontological status; it was simply a secondary or derivative type of existence, as opposed to nonexistence. In the Perfection of Wisdom literature, however, these traditional Abhidharma notions of reality and existence were turned upside-down, and the basic ontological paradigm of the Abhidharma was completely and utterly rejected. Where the Abhidharma had identified, classified, and described all the categories of phenomena in the universe, and had done so in positive ontological terms, the Perfection of Wisdom rejected the ascription of any positive ontological status to anything whatsoever: “there is no eye, no ear, no nose…” and so on, “from form until omniscience.” The assertion that the constituents of traditional Abhidharma classification schemes such as the eighteen elements (dhātu, Tib. khams) were in some sense non-existent, and this non-existence itself, was expressed in the Perfection of Wisdom as the “emptiness” of dharmas—such that, for example, “form is empty of form.” But what, precisely, this “emptiness” meant in the context of the Perfection of Wisdom was not entirely clear. For example, in the Perfection of Wisdom, there are passages which say that form is empty of form, consciousness empty of consciousness, as well as lists of sixteen or eighteen or twenty-two different kinds of “emptiness,” alongside enigmatic statements such as that “form is emptiness, emptiness is form.” Emptiness was explained to be like a magical illusion, or a flower in the sky, as well as in terms of other, similarly colorful language.60 60 Thanissaro Bhikku, trans., “Phenna Sutta: Foam” (Access to Insight, 2010). In fact, these metaphors were extensively used even in the early Nikāyas. In the Phena Sutta (SN 22.95), for instance, the Buddha explains the “emptiness” (Pali suññatā) of the bundles in very similar metaphorical terms, comparing the form-bundle to a glob of foam, the perception-bundle to a desert mirage, and so on. Of particular note is the description of the emptiness of the conditioning-bundle (Skt. saṃskāra-skandha, Pali saṇkhāra-khanda, Tib. ‘du byed kyi phung po) in analytical terms: "Now suppose that a man desiring heartwood… were to go into a forest carrying a sharp ax. There he would see a large banana tree: straight, young, of enormous height. He would cut it at the root and, having cut it at the root, would chop off the top. Having chopped off the top, he would peel away the outer skin. Peeling away the outer skin, he wouldn't even find sapwood, to say nothing of heartwood. Then a man with good eyesight would see it, observe 35 3. “Emptiness” and the Root Verses of the Middle Way It was therefore in order to clarify the meaning of “emptiness” as found in the Perfection of Wisdom literature that Nāgārjuna composed his Root Verses of the Middle Way. While it is not necessarily the case that “emptiness” means the same thing every time it appears in the Root Verses, it is nevertheless evident that the Root Verses is primarily concerned with opposing emptiness to what Nāgārjuna refers to as “established” (Skt. siddha, Tib. grub pa) or “established existence.” Whatever other dimensions Nāgārjuna’s arguments may have hadf—and these will be touched upon shortly—it is also undoubtedly true that his primary purpose was the deconstruction of the Abhidharma ontological framework, as well as any and every other system or set of ontological claims, most especially those of a certain kind of naïve metaphysical realism. But Nāgārjuna specifically targets the realism of the Abhidharma model of substantially-existing dharmas: for Nāgārjuna, any and all phenomena, most definitely including irreducible dharmas, are precisely unreal. So just what does Nāgārjuna mean by “emptiness”? Consider, for example, the analysis of causes and conditions in the very first chapter of the Root Verses. In this chapter, Nāgārjuna examines perhaps the most centrally important term in Buddhism, “dependent origination” (Skt. pratīyasamutpāda, Tib. rten cing ‘brel bar ‘byung ba), a concept most commonly explained with some form of the aphorism, “Because this comes to be, that comes to be; because this ceases, it, & appropriately examine it. To him… it would appear empty, void, without substance: for what substance would there be in a banana tree? In the same way, a monk sees, observes, & appropriately examines any fabrications [saṇkhāra] that are past, future, or present; internal or external; blatant or subtle; common or sublime; far or near. To him… they would appear empty, void, without substance: for what substance would there be in fabrications?” 36 that no longer exists.” Nāgārjuna begins by questioning the ostensible basis of the relationship between so-called causes and effects: “Conditions” are so called because things are considered to depend on them. As long as nothing has arisen, why are they not “non-conditions”? Neither existent things nor nonexistent things have conditions. What “condition,” for the nonexistent? Why a “condition,” for the existent?61 [MMK 1.5-6] That is to say, at the moment of the cause, there is no effect, and so no point to speaking of it as a cause for anything, strictly speaking. Likewise, at the time of the effect, the putative cause no longer exists, and so there is no way to distinguish the non-existent cause from a non-existent non-cause. How can something nonexistent have a cause or condition? And why does something which already exists need a cause or a condition? By definition, it already exists! Therefore, according to Nāgārjuna, designations like “cause” and “effect” are only provisional, always dependent upon an uncritical or pre-analytic attitude for whatever conceptual or linguistic force they possess. By demonstrating the logical problems inherent in any putative relationship between the “cause” or “condition” of a phenomenon and the “effect” of that phenomenon itself, Nāgārjuna undermines the ultimate foundation of any realist view, namely causality. This is particularly and especially true with respect to the causal interactions of dharmas, the ontological bedrock of the Abhidharma model. Insofar as a “cause” can only over be designated as such with reference to an “effect,” and vice versa, the metaphysical proposition that one is able to serve as the 61 . utpadyate pratītyemānitīme pratyayāḥ kila| yāvannotpadyata ime tāvannāpratyayāḥ katham|| naivāsato naiva sataḥ pratyayo'rthasya yujyate| asataḥ pratyayaḥ kasya sataśca pratyayena kim|| 37 ontological foundation of the other—the warrant or justification for being able to speak of either as substantially or truly existing—must always fail. It is extremely important, however, to note the phenomenological dimensions of this argument. Nāgārjuna is saying is that it is impossible to establish that a particular phenomenon is the “cause” of a particular “effect,” even on the basis of sensory evidence. At the time of a sensory apprehension of the “cause,” by definition there can be no sensory apprehension of the “effect,” and vice versa. He does not deny that phenomena are presented to consciousness, nor does he deny that one may speak of causes and effects in a conventional or relative sense. What he denies is the possibility of providing a coherent account of the relationship of cause to effect, since there is not, nor could there ever be, any phenomenal evidence on which to ground or “establish” this relationship. Of course, Nāgārjuna’s point was not strictly phenomenological. The issue at stake is the (onto)logical coherence of any type of causal theory that proceeds on the basis of a putative relationship between a “cause” at time t0 and an effect at time t1. But the opponent is forced to appeal to the ostensible reality of a relationship which is never, in fact, observed. Thus it was against the very possibility of any kind of ontological foundations whatsoever, including a foundation or warrant to “establish” anything as non-existent, that Nāgārjuna directed his critique. In the next verse, Nāgārjuna writes When a dharma comes forth as neither existent nor non-existent, nor both, How is a “cause which brings it about” logical?62 [MMK 1.7] 62 na sannāsanna sadasan dharmo nirvartate yadā| kathaṃ nirvartako heturevaṃ sati hi yujyate|| 38 In other words, the point of the previous analysis was not to say that the disjunction between cause and effect implies non-existence; it was to deny that their conjunction provides a firm ontological foundation on which any permutation of the existence or nonexistence of phenomena may be predicated. Indeed, Nāgārjuna is pointedly demonstrating that phenomena cannot be established as nonexistent, any more than they can be established as existent. Any and all types of ontology are susceptible to this type of analysis. In this way, the opposite of emptiness is not existence, or nonexistence, or both existence and nonexistence, or neither existence nor nonexistence, but the “establishment” (Skt. siddhi, Tib. sgrub pa) of phenomena as any of these. Unfortunately, Nāgārjuna never fully explains what he means by “establishment.” But in the most basic terms, the idea appears to be that an “established” phenomenon exists independently of anything else. This is closely related to Nāgārjuna’s treatment of self-nature (svabhāva); while he never explicitly relates these two concepts, it is clear from context that an entity with real self-nature would perforce be established as truly or ultimately existent. Nāgārjuna’s refutation of establishment is therefore intimately connected to his refutation of selfnature. Recall that self-nature or svabhāva is precisely what defines dharmas; in refuting “establishment” and self-nature, Nāgārjuna refutes not only ontological foundations in the abstract, but also the fundamental underpinnings of Abhidharma philosophy as it was developing in the mainstream Buddhist commentarial literature (as opposed to the literature of the Mahāyāna, which in Nāgārjuna’s day was still a minority movement). The primary problem Nāgārjuna identifies with self-natures (essence-svabhāvas) is their violation of the basic principles of causality and dependent origination. As Nāgārjuna writes in the opening verses of his examination of self-natures, 39 It is illogical for a self-nature to be produced from causes and conditions. A self-nature which arises from causes and conditions has been made. How, moreover, can a self-nature come to be made? Self-natures are not made and do not require anything else.63 [MMK 15.1-2] What distinguishes a self-nature or essence-svabhāva is that it is what it is, independently of anything else; this is just as true in the contemporary Western philosophical tradition as it was for Nāgārjuna. Nāgārjuna simply pushes this argument to its logical conclusion: if a self-nature must always be what it already is, self-nature cannot be causally conditioned. It must therefore be its own, self-caused, ontological foundation; in this way, the notion of essence-svabhāva necessarily involves a tautological commitment to “established” existence. Furthermore, by definition, self-natures necessarily remain the same throughout the various permutations and transformations of the phenomena of which they are the self-nature. Therefore, whatever is affected by causes and conditions does not remain the same, and cannot serve as a self-nature. Thus, on the one hand, the only possible cause for a self-nature (i.e. an essence-svabhāva) is that self-nature, itself. But, on the other hand, being a self-nature precludes any type of causal interaction with other entities. And since according to realists all entities posses a real self-nature precisely insofar as they are real, the realist position necessitates an eternal stasis that is, in reality, unobserved: Everything would become not-born, un-destroyed, and immutable, Cut off from different circumstances because of self-nature.64 [MMK 24.38] 63 na saṃbhavaḥ svabhāvasya yuktaḥ pratyayahetubhiḥ| hetupratyayasaṁbhūtaḥ svabhāvaḥ kṛtako bhavet|| svabhāvaḥ kṛtako nāma bhaviṣyati punaḥ katham| akṛtrimaḥ svabhāvo hi nirapekṣaḥ paratra ca|| 64 ajātamaniruddhaṃ ca kūṭasthaṃ ca bhaviṣyati| vicitrābhiravasthābhiḥ svabhāve rahitaṃ jagat|| 40 On the contrary, then, Nāgārjuna argues that phenomena do not possess self-nature, and are in this sense “empty”—empty, in Westerhoff’s terminology, of “essence-svabhāva,”65 if not necessarily (at this stage) empty of property-svabhāva. Since they lack self-nature in this way, phenomena are able to perform functions and causally interact. This is because, on Nāgārjuna’s account, possessing self-nature is the opposite of being conditioned; self-natures are, by definition, unchanging and unconditioned. If you perceive entities as existent on account of their own self-nature, You perceive entities that exist without causes or conditions. The agent, the instrument, the acted-upon, the action, its cause— You deny [these], production, cessation, and results.66 [MMK 24.16-17] Thus: If emptiness is logical, everything is logical. If emptiness is illogical, everything is illogical.67 [MMK 24.14] This is closely related to his statement, a few verses later, that Whatever is dependently-originated, that is said to be empty. It is designated in dependence; this is the path of the Middle Way.68 65 Jan Westerhoff, Nāgārjuna’s Madhyamaka, 19-52. 66 svabhāvādyadi bhāvānāṃ sadbhāvamanupaśyasi| ahetupratyayān bhāvāṃstvamevaṃ sati paśyasi|| kāryaṃ ca kāraṇaṃ caiva kartāraṁ karaṇaṃ kriyām| utpādaṃ ca nirodhaṃ ca phalaṃ ca pratibādhase|| 67 sarvaṃ ca yujyate tasya śūnyatā yasya yujyate| sarvaṃ na yujyate tasya śūnyaṃ yasya na yujyate|| “Logical” here may also be understood as “possible”; the point is that the opponent of emptiness has no way to explain reality. 41 [MMK 24.18] By equating dependent-origination and emptiness, Nāgārjuna insists that all phenomena are only able to appear due to their essential insubstantiality. To whatever extent there is a “phenomenal world” to speak of, this world is always and everywhere unreal. 68 yaḥ pratītyasamutpādaḥ śūnyatāṁ tāṃ pracakṣmahe| sā prajñaptirupādāya pratipatsaiva madhyamā|| 42 4. Three Approaches to the Two Truths Following his investigations and refutations of causes and effects, characteristics and characterized, and similarly basic, mundane subjects of investigation, Nāgārjuna also examines the Buddha and Nirvāṇa along similar lines, and applies the exact same logical analysis to the supermundane: The self-nature of the Tathāgata69 is the self-nature of the world. The Tathāgata has no self-nature; the world has no self-nature.70 [MMK 22.16] And: The pacification of all perceptions, of all conceptual structuring, is good.71 The Buddha never taught any Dharma whatsoever, anywhere, any time.72 [MMK 25.24] Now it is finally time to return to the previous objections lodged by Nāgārjuna’s Buddhist opponent, that such an approach is finally incompatible with mainstream Buddhism: “If the Dharma and the Saṃgha do not exist, how can the Buddha exist? 69 This is an epithet for the Buddha, sometimes translated as “Thus-Gone One,” reading gata as “gone” and tathā as a contraction of tathatā, “thusness” or the nature of reality. However, gata may also mean “know” so Tathāgata may also be glossed as the “One Who Knows Thusness.” 70 tathāgato yatsvabhāvastatsvabhāvamidaṃ jagat| tathāgato niḥsvabhāvo niḥsvabhāvamidaṃ jagat|| 71 The word śiva has a wide semantic range, not remotely captured by “good.” Implications include happiness, goodness, and—especially—peace. However, saying that “pacification… is peace” (prapañcopaśamaḥ śivaḥ) sounds ridiculous in English. 72 sarvopalambhopaśamaḥ prapañcopaśamaḥ śivaḥ| na kvacitkasyacitkaściddharmo buddhena deśitaḥ|| 43 Speaking of emptiness in this way maligns the Three Jewels. The reality of causal results, what is Dharma and not Dharma, And the ordinary conventions of the world—you malign them all. [MMK 24.5-6] As previously mentioned, Nāgārjuna first explains that this view arises from misunderstanding emptiness, and then responds by distinguishing between two different categories of truth: This is a mistake. You have not understood the point of emptiness; In this way, you destroy both emptiness itself and the purpose of emptiness. The Dharma taught by the Buddha completely relies on two truths, The conventional truth of the world and the truth which is ultimate. Those who do not understand the division between the two truths, They do not understand the profundity taught by the Buddhas. [MMK 24. 7-9] Nāgārjuna goes on to explain that emptiness is dangerous to those of little intelligence, that misapprehending emptiness is similar to grabbing a poisonous snake in the wrong way. The key here is that when he applies his analysis to the Buddha, Nāgārjuna is not arguing that the Buddha is nonexistent, inasmuch as he never argues that anything is nonexistent, since this would require the “establishment” of nonexistence with respect to phenomena and therefore contradict his own arguments. Unfortunately, he does not clarify further what is meant by “conventional” and “ultimate” truth. It is left for the reader to determine which parts of Nāgārjuna’s words concern the relative truth, and which parts concern the ultimate. It is in any event surely not the case that Nāgārjuna intends to claim that conventional phenomena are ultimately existent (“as conventional,” in Garfield’s rendition), or that their ultimate existence is nothing other than the mere 44 conventionality of their appearance—else why posit the distinction, or frame it in terms of what is “worldly” as opposed to what is “profound”? It is reasonable to understand this verse, as certain later Tibetan commentators73 did, as an assertion that the manner in which the Buddha is empty is fundamentally different from the manner in which, say, mid-size dry objects such as pots and chairs are empty—even given that both are empty. Nāgārjuna’s entire point in this verse is that although both the Buddha and ordinary phenomena are empty, still it does not follow that the doctrine of emptiness maligns or denigrates the Three Jewels. In this way, there is a direct continuity between the notion of “provisional” as opposed to “definitive” meaning, and the “conventional” as opposed to the “ultimate” truth. The precise nature of the division between these two truths, however, is left unstated. This proved a fertile ground for future commentators. Some later Tibetan commentators would interpret Nāgārjuna’s “emptiness” as referring simply to the non-existence of self-natures, while others would affirm the need to speak of emptiness in terms that were not exclusively negative. Historically, in India and later in Tibet, this has frequently been linked to a debate over the status of conventional truth: what kind of “existence” may be predicated of it, the exact sense in which dependent-arising is identified as emptiness (and vice versa), whether such 73 Dol po pa Shes rab rGyal mtshan, Shes rab kyi pha rol tu phyin pa man ngag gi bstan bcos mngon par rtogs pa’i rgyan gyi rnam bshad mdo’i don bde blag tu rtogs pa. In gsung ‘bum/ shes rab rgyal mtshan (‘dzam thang). TBRC W21208. vol. 5 (ma): ff. 123.a-310.b. This line of Tibetan commentary is usually attributed to Dol po pa Shes rab rGyal mtshan, founder of the Jo nang school. Dol po pa is commonly described as the originator of the tradition of exegesis called the “Great Middle Way of Extrinsic Emptiness” (gzhan stong dbu ma chen po). The distinction between “extrinsic emptiness” (gzhan stong) and “intrinsic emptiness” (rang stong) turns on just this point, i.e. that the Buddha, Nirvāṇa, and so on are empty in a way that ordinary phenomena are not. In his commentary to the Ornament of Manifest Realization (Skt. Abhisamayālaṃkāra, Tib. mngon rtogs rgyan), in the discussion of the sixteen or eighteen different kinds of emptiness taught in the Perfection of Wisdom literature (ff. 203.a-207.a), Dol po pa likens the emptiness of ordinary phenomena, “the emptiness of one thing in another” (gcig gis cig stong pa nyid) to “a barn that does not have a cow in it” (khang pa ba lang gis stong pa) (ff. 206.b.2). He opposes this “intrinsic emptiness” of ordinary phenomena to the profound “extrinsic emptiness” of sublime phenomena such as Nirvāṇa (ff. 207.a.1), which are only empty “extrinsically”—that is, empty of anything not similarly sublime and thus “extrinsic” to themselves. 45 conventional phenomena can be said to exist “from their own side,” and so on. Typically these debates centered around what, exactly, the “object of refutation” (dgag bya) was with respect to the ultimate truth—that is, what exactly was being negated by Middle Way analysis. 74 However, despite or across these disputes, there are several commonly agreed-upon senses in which the ultimate may be distinguished from the relative. One common usage involves the same process of intellectual analysis that distinguished substantial from designated existence in the Abhidharma: the relative truth is defined as that which is supposed to be true prior to any such analysis, but is refuted during analysis, while the ultimate is defined as that which can “bear analysis” or is “unharmed” by analysis. Of course, such a definition turns on precisely what is meant by “analysis,” and the manner in which this “analysis” differs from the analysis of dharmas in the Abhidharma. In the context of Nāgārjuna’s Root Verses, this would be an analysis of the relationship of part to whole, characteristic to characterized, or any of the other dyadic and triadic relationships examined over the course of Nāgārjuna’s verses. The prototypical example of this type of analysis is the “neither one nor many” argument (Skt. ekānekaviyogahetu, Tib. gcig du ‘bral gyi gtan tshigs, literally “evidence75 of not being free from one or many”), which was employed by Śāntarakṣita in his Ornament of the Middle Way to the near exclusion of any other form of 74 Sara McClintock, “The Role of the ‘Given’ in the Classification of Śāntarakṣita and Kamalaśīla as SvātantrikaMādhyamikas,” in The Svātantrika-Prāsaṇgika Distinction: What Difference Does a Difference Make?, ed. Sara McClintock and Georges Dreyfus (Boston: Wisdom Publications, 2003), 132. As McClintock informs us, “Tsong kha pa and many of his followers typically discuss the object of refutation in terms of six nearly equivalent expressions: they say that the object of refutation is that which is ‘truly established’ (bden par grub pa), ‘ultimately established’ (don dam par grub pa), ‘really established’ (yang dag par grub pa), ‘established from its own side’ (rang ngos nas grub pa), ‘established by its own character [i.e. svalakṣaṇa]’ (rang gis mtshan nyid kyis grub pa), and ‘established by its own nature’ (rang bzhin gyis grub pa).” On establishment from “its own side,” cf. also Sonam Thakchoe, The Two Truths Debate: Tsongkhapa and Gorampa on the Middle Way (Boston: Wisdom Publications, 2007), 175 n. 46. 75 For more on the technical meaning of hetu (Tib. gtan tshigs) see note 343. 46 argumentation. The general thrust of this argument is that phenomena cannot be established as either singular or manifold; they can be reduced to parts, but those parts are only “parts” in relation to some predetermined “whole.” In the example above, cause and effect are asserted as the “parts” of the “whole” causal act or relationship. However, because reasoned analysis can dismantle or deconstruct the relationship of cause to effect, causality itself is unable to “bear analysis” and therefore is only relatively or contingently true. Cause and effect only “exist” relative to each other, insofar as they are un-analyzed, to the extent that such an analysis would reveal a structural incoherence in the way that they are conceptually understood prior to analysis. The most obvious sign of this incoherence is their mutual dependence, which precludes either from occupying a place of ontological primacy. The second major sense in which the relative truth is distinguished from the ultimate truth concerns the manner in which phenomena such as pots and Buddhas exist. Instead of being defined in terms of analysis, the distinction is presented as two separate ontological categories or modes. For something to be ultimately existent, it must exist by virtue of its own self-nature, without reference or recourse to any other phenomenon. Otherwise, to whatever extent a phenomenon that does not meet this criterion may be said to exist, it can only be relatively existent. This is the primary sense of the distinction operative in Candrakīrti’s Entry to the Middle Way and Clear Words, although he also employs the former. This “ontological” sense of the two truths distinction differs from the “analytic” sense in subtle, yet critical, ways. The analytic distinction brackets the question of just what characteristics a phenomenon that could withstand analysis (and thus be “ultimately existent”) might have. In this account, causes and effects are within the domain of the relative truth since the idea of something being a “cause” or “effect” cannot withstand reasoned analysis. The latter 47 distinction hinges on a particular set of predicates which are considered sine qua non for ultimate existence, most particularly that an ultimately-existent phenomenon not rely on anything extrinsic in any way, shape, or form. As soon as a phenomenon is understood to depend upon anything at all, it can only exist in terms of the conventional or relative truth. Thus, in this latter account, causes and effects are within the domain of the relative truth since theirs is a relationship of mutual dependence: each depends upon the other, and neither can serve as a firm ontological basis on which to “establish” the other. Or, as Nāgārjuna says, If something is established in dependence upon something else, and establishes the other through being dependent on it, If that which is being depended upon must itself be established, what is depending on what?76 [MMK 10.10] To review, we have seen two main senses of the distinction between relative and ultimate truth so far. The first, “analytic” sense of the distinction is that whatever is ultimately true is able to withstand mereological analysis (i.e. analysis into parts). This is effectively identical to the criterion of distinction between substantial and designated existence in the Abhidharma, although as noted above the ontological stakes are quite different. In Abhidharma systems, individual indivisible property-possessing dharmas are able to withstand analysis and are therefore “substantially” existent. For Nāgārjuna, those same dharmas are dependently-originated, and therefore empty insofar as the fact of their having been dependently-originated entails their inability to withstand the analysis of, for example, the “neither one nor many” argument. The second, “ontological” sense hinges on Nāgārjuna’s treatment of self-nature or essence-svabhāva as a general type of property-svabhāva: the property or predicate of real or 76 yo'pekṣya sidhyate bhāvastamevāpekṣya sidhyati| yadi yo'pekṣitavyaḥ sa sidhyatāṃ kamapekṣya kaḥ|| 48 established existence.77 This application of the two truths distinction radically departs from the Abhidharma, as it necessarily entails (or is simply equivalent to) a decisively anti-realist metaphysics. Here, even the property-svabhāva of a dharma is not actually inherent to that dharma; such inherence would preclude the possibility of change, transformation, or indeed any type of causal interaction, because for Nāgārjuna property-svabhāva is conflated with essencesvabhāva and thereby with established, non-empty existence. Therefore, all dharmas—even and especially those dharmas maintained in the Abhidharma to be substantially existent on account of their ability to withstand analysis, or their self-identification with their own inherent propertysvabhāva—are empty, where “emptiness” denotes some as yet undetermined unreality or lack of genuine (“established”) ontological status. As a means for understanding the philosophical investigations in the Root Verses, the distinction between conventional and ultimate truth is usually interpreted according to either analytic or ontological criteria. However, there is another manner in which this distinction is traditionally laid out, which is not explicitly articulated in the Root Verses but is nevertheless quite important, as it directly informs Ratnākaraśānti’s ontology. This third, “teleological”78 sense of the distinction between the two truths is that what is ultimately existent can perform a function, while what is relatively existent cannot. This was the definition put forward by Dharmakīrti in his Commentary on Pramāṇa (Skt. Pramāṇavarttika, Tib. tshad ma rnam ‘grel): 77 This is, of course, one of the more pronounced ways in which (contra Murti) Nāgārjuna contradicts Kant. For Kant, existence can never be a predicate. For Nāgārjuna, possessing any type of inherent property—and what other type of property is there?—additionally entails possessing the property of truly existing. Indeed, much of Nāgārjuna’s argument turns precisely on the point that possessing an essential property (= svabhāva) is tantamount to possessing the property of inherent existence (= svabhāva). 78 “Teleological” is by way of reference to the technical term arthakriyā. Literally, this means “purposeful activity,” hence its translation by Dunne (2004) and others as “telic efficacy.” The Tibetan translation of this term (don byed nus pa) explicitly captures the causal capacity or capability (nus pa) implied in the Sanskrit. 49 In this context, that which is capable of telic function is said to be ultimately real. The other one is said to be conventionally real. They are, respectively, the particular and the universal.79 [PV 3.3] This is, incidentally, the only articulation of the two truths distinction which is even theoretically capable of justifying the notion that the ultimate truth is mere conventionality. The point is that causal efficacy serves as a guarantee of (some type of) reality. Conventional phenomena function; therefore, they are in some (as yet undetermined) sense real. Like other critics of the Yogic Practice tradition, Candrakīrti focused on the contention of Yogācārins that the mind is supposed to exist in some substantial sense. To Candrakīrti, this was extremely problematic. However, the Yogic Practice tradition was in many ways a continuation of the Abhidharma within the Mahāyāna, and in this last “teleological” sense, it was unproblematic for the adherents of the Yogic Practice tradition to assert that the mind is “ultimately existent,” insofar as otherwise nothing could either appear or function. For a commentator of the Yogic Practice tradition to say that the mind or more precisely the mental dharmas “exist ultimately” was not to say that they exist inherently or independently of anything else, nor that they possess either property-svabhāva or essence-svabhāva, only that they are capable of performing a function. Unfortunately, this point has been lost on the vast majority of Western interpreters. With respect to the Root Verses, while Nāgārjuna did not explicitly advance the notion that performing a function implies the ultimate existence of that which performs the function, his assertion of a fundamental identity between dependent origination and emptiness both necessitates and hinges upon the idea that empty phenomena can perform functions—indeed, 79 trans. Dunne in John Dunne, Foundations of Dharmakīrti’s Philosophy, 392. 50 that only what is empty can perform a function, and that anything which performs a function is necessarily empty. Insofar as the “other-dependent nature” (Skt. paratantra-svabhāva, Tib. gzhan gyi dbang gi ngo bo yid) is held to be both the sum total of dependently-originated causal interactions, and empty of the “constructed nature” (Skt. parikalpita-svabhāva, Tib. kun brtags pa’i ngo bo nyid) of duality, this point had important ramifications for Ratnākaraśānti’s “Middle Way of the Three Natures.” But this will be addressed in greater detail below, in the section concerned with the theory of the three natures. 5. The Two Truths in the Pith Instructions It was primarily by eschewing the third “teleological” sense of the distinction between ultimate and relative truth, and by re-interpreting Yogic Practice notions of ultimate existence, that later commentators were able to synthesize the Middle Way approach to emptiness with that of the Yogic Practice method, as well as with the Buddhist Epistemological (Pramāṇa) tradition, which was closely associated with Yogic Practice. For example, in the Ornament of the Middle Way, Śāntarakṣita argues strongly in favor of the view that causally-functioning phenomena are only relatively existent. Śāntarakṣita acknowledges the value of the Yogic Practice presentation, but emphasizes that this is not a description of ultimate truth. Adopting a slightly different stance, Ratnākaraśānti argues in the Pith Instructions for the Ornament of the Middle Way that the causally-produced phenomena of dependent origination—which he maintains to be nothing other than the mind—are neither ultimately existent nor merely designated to exist, but are instead “substantially” existent: 51 It is also said that there is designated existence, substantial existence, and ultimate existence.80 Ratnākaraśānti ’s three-fold division of ontological categories was not universally employed among adherents of the Yogic Practice tradition; but, as will be seen in the next section, the distinctions he draws with respect to “substantial existence” in the sense of performing a function or possessing property-svabhāva, between designated (non-)existence on the one hand and true ultimate existence on the other, follows naturally from Yogic Practice ontology. Furthermore, Ratnākaraśānti presents both the ultimate and the relative truth in terms of three categories: The “two truths” refers to the ultimate truth, and the relative. The ultimate has three aspects: (1) The ultimate itself is suchness, (2) The ultimate means of accomplishment is authentic wisdom, and (3) The ultimate to be attained is Nirvāṇa… Similarly, the relative has three aspects: (1) The constructed [brtags] relative is the nature of all phenomena which are entirely imagined [kun brtags]; (2) Relative consciousness is all mistaken cognitions, and (3) The expressive relative is that which indicates the ultimate through sounds and thoughts.81 In the Pith Instructions, Ratnākaraśānti’s remarks on the ultimate truth as a topic in and of itself are limited to this passage. He only treats the ultimate indirectly, in terms of the three natures and the reflexive nature of awareness. 80 Ratnākaraśānti, Pith Instructions, 225.b.4. yang brtags (D: btags) par yod pa dang rdzas su yod pa dang/ don dam par yod pa dang zhes kyang gsungs so/ 81 Ratnākaraśānti, Pith Instructions, ff. 223.b.3-5. de la bden pa gnyis ni don dam bden pa dang kun rdzob bo/ don dam pa yang rnam pa gsum ste/ don dam pa ni de bzhin nyid do/ sgrub pa don dam pa ni yang dag pa'i ye shes so/ thob par bya ba'i don dam pa ni mya ngan las 'das pa'o/ … kun rdzob kyang rnam pa gsum ste/ brtags pa'i kun rdzob ni chos thams cad kun brtags pa'i rang bzhin no/ shes pa’i kun rdzob ni log pa'i shes pa thams cad do/ brjod pa'i kun rdzob ni sgra dang (P/N/S: em. rtog pas) don dam pa ston par byed pa'o/ 52 It is, however, worth noting that Ratnākaraśānti’s presentation of the relative truth implies that the relative is not necessarily or completely false. While he states that the relative consists in both imputed or unreal phenomena as well as mistaken cognitions, he also asserts that there is a mode of relative truth which bears a connection with the ultimate, the so called “expressive relative” (brjod pa’i kun rdzob). Although Ratnākaraśānti does not specifically refer to Bhāvaviveka’s notion of the “categorized ultimate” (Skt. paryāya-paramārtha, Tib. rnam grangs pa’i don dam), this is obviously a similar idea, as Bhāvaviveka’s notion also described a conceptual mode that was somehow in accordance with the necessarily nonconceptual ultimate truth. Regarding the conceptual status of the “expressive relative,” Ratnākaraśānti states: The expressive relative is neither conceptual nor nonconceptual. Why is it nonconceptual? Because it is conducive to the production of nonconceptuality. Why is it not nonconceptual? Because it is a conceptualization of the authentic ultimate.82 For Ratnākaraśānti, then, in one sense the relative truth simply encompasses all imputed phenomena and all mistaken cognitions; but in another sense, it also encompasses those concepts which indicate the ultimate, however indirectly or provisionally. This may also be understood as highlighting two different senses of “truth.” Specifically, in this passage, there is an ontological sense and a phenomenological sense of “truth.” To say that relative phenomena are “imputed” is to describe their ontological mode of existence, and distinguishes between relative and ultimate (i.e. non-imputed) truth on the basis of this 82 Ratnākaraśānti, Pith Instructions, ff. 223.b.5-6. brjod pa'i kun rdzob ni rnam par rtog pa ma yin zhing rnam par mi rtog pa yang (P/N/S: em. ma) yin no/ ci'i phyir rnam par rtog pa ma yin zhe na/ rnam par mi rtog pa bskyed (P/N/S: skyes) pa dang rjes su mthun pa nyid kyi phyir ro/ ci'i phyir rnam par mi rtog pa ma yin zhe na/ yang dag pa'i don la rnam par rtog pa'i phyir ro/ 53 ontological mode. But to say that they are “mistaken cognitions” or that the relative truth includes conceptual methods for approaching nonconceptual ultimate truth is to make the phenomenological claim that the ultimate truth is a type of unmistaken cognition. Although it does not appear that he ever cited Ratnākaraśānti, the nineteenth century rNying ma scholar and polymath Ju Mi pham (1846-1912) articulated a very similar view. According to Douglas Duckworth, Mi pham used two different models for explaining the two truths. In the first model, the relative truth is appearance and the ultimate truth is emptiness; in this model, “the ultimate truth of emptiness is not qualitatively different from the relative truth of appearance…. Without being regarded with a qualitative difference, both of these are equally applied [to all phenomena] from form to omniscience.”83 All phenomena, then, are both apparent and empty. As Duckworth writes, The two truths are actually inseparable here; they are not really different, but are only conceptually distinct. An important feature of this model is that the relationship is not hierarchical. There is no appearance-reality distinction such that ultimate truth is understood as hidden behind the concealing veil of relative truth. In this ontological model, appearing phenomena are necessarily the relative truth in contrast to the ultimate truth, which is exclusively emptiness.84 Earlier, an “ontological” sense of the distinction between the two truths was considered. In that sense of the distinction, whatever lacks a self-nature (essence-svabhāva) is only relatively existent. But Nāgārjuna made it clear that the very concept of self-nature was internally 83 Douglas Duckworth, “Two Models of the Two Truths: Ontological and Phenomenological Approaches,” Journal of Indian Philosophy 38 (2010): 520. 84 Ibid., 521. 54 inconsistent, contradictory, and absurd. Thus all appearing phenomena lack self-nature, and are accordingly “empty” just to the extent that they are causally produced. Ratnākaraśānti does not formally argue that appearance and emptiness are an indivisible unity. He does, however, quote a selection from the Root Verses, already mentioned above: Whatever is dependently-originated, that is said to be empty. It is designated in dependence; this is the path of the Middle Way. [MMK 24.18] However, while Ratnākaraśānti does occasionally quote Nāgārjuna, he cites the works of Asaṇga and Vasubandhu with much greater regularity, in particular Distinguishing the Middle from the Extremes (Skt. Madhyāntavibhāga, Tib. dbu mtha’ rnam ‘byed). In that work, “duality” is usually glossed as the phenomenal duality of subject and object. But “duality” is also glossed in terms of existence and nonexistence (bhūta and abhūta), as well as in terms of the phenomenal subject and object. In keeping with Vasubandhu, Ratnākaraśānti maintains, like Mi pham, that causally-produced appearances are empty of existence and nonexistence, and that this emptiness is furthermore an inseparable quality of those appearances. Apart from the “ontological model,” however, Mi pham also used a “phenomenological model” in which the two truths are two different modes of experience, mistaken and unmistaken or authentic and inauthentic. In this model, the phenomena of Nirvāṇa, i.e. experiences which are in accord with reality, are ultimate; by contrast, the phenomena of Saṃsāra, i.e. experiences which are not in accord with reality, are relative. Thus, as Duckworth writes, “the two truths in the phenomenological model are such that the ultimate truth is reality experienced without duality or reification, while the relative truth is the world experienced within a distorted framework.” He continues: 55 In contrast to the ontological model of the two truths as appearance and emptiness, the dichotomy of the two truths as authentic and inauthentic experience sustains a qualitative distinction between two truths (as disparate modes of experience). We can see that the ultimate truth here is something positive and the relative truth is something negative; they are two qualitatively different ways of experiencing reality. In contrast to the ontological model, the ultimate truth in this context is not simply emptiness (the lack of true existence in things), because here the experiential presence of reality, known as it is, is the ultimate. Thus, the two truths in this latter model are distinguished based on the way the world is present in experience (phenomenologically) rather than the way it is absent (ontologically).85 As stated in the Introduction, one of the key errors frequently made by Western commentators on the Buddhist doctrine of emptiness is the notion that emptiness or the twotruths distinction only concerns a certain type of ontological negation. To be sure, the negation of inherent existence or self-nature is an extremely important element of traditional presentations on emptiness; but even in the Root Verses of the Middle Way, Nāgārjuna himself notes that properly understanding emptiness entails knowing the nature of reality unadulterated, exactly as it is, and thereby construes emptiness in terms of some unspecified but positive (“the pacification of all perceptions… is good”) phenomenological dimension. If anything, the phenomenological model is the primary framework adopted by Ratnākaraśānti, who consistently identifies the nondual “luminosity” (Skt. prakāśa, Tib. gsal ba) of reflexive awareness as precisely this type of authentic experience: There is nothing that can refute the luminous nature of reflexive awareness. Reflexive awareness is direct [mngon sum, *pratyakṣa], 85 Douglas Duckworth, “Two Models of the Two Truths,” 521. 56 and it is an authentic experience, because there are no means of reliable knowledge [tshad ma, *pramāṇa] apart from it.86 And: The experience of an imputed self-nature is distorted; therefore, all distorted dualistic marks should gradually be abandoned. Like endless space without a single stain, the experience of the luminous body of all phenomena, devoid of duality, is the authentic realization of the ultimate.87 Ratnākaraśānti’s overall approach heavily favors the phenomenological model, although whether this is a question of emphasis as opposed to substantive argument remains somewhat unclear. It may well be irrelevant, though, since for Ratnākaraśānti, the phenomenological point about authentic vs. inauthentic experience is closely intertwined with an epistemological point about what constitutes a “means of reliable knowledge” (Skt. pramāṇa, Tib. tshad ma), about which there will be much more to say below. In any case, as Duckworth points out, Mi pham considered both the ontological and phenomenological models necessary for arriving at a proper ascertainment of ultimate truth: The ontological model by itself, with solely an absence as the ultimate truth that ignores the phenomenological experience of unity, is simply an abstraction. Emptiness, understood as solely an absence, reflects a description divorced from the event of understanding; on its own it 86 Ratnākaraśānti, Pith Instructions, ff. 226.a.1-2. yang (D/C: em. rang) rig pa gsal ba'i ngo bo la gnod par byed pa ni med de/ de las lhag pa'i tshad ma gzhan med pa'i phyir ro/ de ni rang rig pa'i mngon sum yin pa dang/ yang dag tu myong ba'o/ 87 Ratnākaraśānti, Pith Instructions, ff. 229.a.3. sgro btags pas (D/C: pa) rang gi ngo bo myong bas na 'khrul pa yang yin no/ de bas na 'khrul pa'i mtshan ma de thams cad rim gyis (P/N/S: rims kyis) dor te/ nam mkha' gcig tu dri ma med cing mtha' med pa lta bur chos thams cad kyi (D/C: kyis) gsal ba'i lus gnyis kyis stong pa myong ba gang yin pa de nyid don dam pa yang dag par rtogs pa yin gyi/ 57 represents a sterile view of metaphysical absence, isolated from the dynamic reality of a lived-world. In other words, it is a view of nihilism. Yet the phenomenological model alone is not sufficient either. Its appeal to the ultimate truth of an experiential unity that dispenses with analysis and ontological critique tends toward reification. Without being curtailed by the negative dialectic of emptiness, the ultimate truth of authentic experience easily congeals into a naïve, mystical realism, otherwise known as “the extreme of eternalism.”88 Of course, the “Middle Way” tradition is self-consciously defined as being a “middle way” between eternalism (Skt. śāśvatavāda, Tib. rtag par smra ba) and nihilism89 (Skt. ucchedavāda, Tib. med par smra ba). It is reasonable to expect any and all adherents of the Middle Way tradition to chart an exegetical course between these two extremes. This naturally applies as well to such adherents as Ratnākaraśānti, who aimed to combine Middle Way and Yogic Practice methodology in what he called the “Middle Way of the Three Natures.” Nevertheless, there is significant controversy concerning the precise boundaries of what constitutes “eternalism” and “nihilism.” And what was acceptable for Ratnākaraśānti would not necessarily be acceptable to Mi pham: The extreme of eternalism is at play when the two models of truth get conflated and the boundaries of phenomenology stray into the realm of ontological claims. For instance, in the case of the Yogācāra idealism represented in the philosophical system known in Tibet as “Mind-Only”: 88 89 Douglas Duckworth, “Two Models of the Two Truths,” 525. The Sanskrit term for this view (ucchedavāda) literally means the view of “annihilation” or, even more literally, being completely cut off or ceased. However, I have chosen the translation “nihilism” out of deference to previous works, and also because for Buddhist interlocutors this is more or less the metaphysical position implied by the view in question. For Nāgārjuna, the point at stake concerns what happens to a causally produced phenomenon after the causal production process is complete: does the phenomenon hang around forever even in the absence of its cause (“eternalism”) or does it cease completely (“annihilationism”)? To Nāgārjuna, both of these are unacceptable consequences. The point is that, as expressed in the homage at the beginning of the Root Verses, there is neither cessation, nor arising; neither nonexistence, nor existence. 58 To claim that the mind alone is (ontologically) real in the wake of a phenomenological discovery that the mind is constitutive of any experience of phenomena is to confuse levels of discourse; it is akin to making a category mistake.90 Ratnākaraśānti, by contrast, openly and explicitly maintained an idealistic ontology. Indeed, this is one of the most important points covered in the next section, on the theory of the three natures. It is, of course, tempting to see this embrace of idealistic ontology as an instance of just the type of “naïve, mystical realism” from which Mi pham distanced himself. However, for Ratnākaraśānti, ontological and phenomenological concerns regularly bleed into each other. More broadly, one of the major questions at stake in the Pith Instructions concerns the extent to which ontology, phenomenology, and epistemology can in fact be rigorously separated, if indeed such a rigorous separation is, in the end, possible at all. With respect to Mi pham and others, though, there remains the question of whether Ratnākaraśānti’s idealistic ontology in fact amounted to an eternalist view. As will shortly become clear, at least part of what motivated the argument for idealistic ontology was the principle of ontological parsimony—that is, the concern on the part of Buddhist philosophers to reduce as far as possible the number of candidates for real entities, in the service of a general and over-arching tendency toward anti-realism. As Duckworth alludes, the argument for metaphysical idealism was driven by a universal acknowledgement that the mind is “constitutive of any experience of phenomena,” and as will be seen below, this had deep implications. Furthermore, according to Ratnākaraśānti, the move to “Mind-Only” (Skt. cittamātra, Tib. sems tsam) was only the second in a four-stage process of contemplative practice. By the 90 Douglas Duckworth, “Two Models of the Two Truths,” 525. 59 classical standard, this places Mind-Only firmly within the realm of “provisional” as opposed to “definitive” teaching. The question, then, is not whether idealistic ontology is ultimately correct; the question is only whether or not idealistic ontology is more defensible than the alternatives, and serves as a better description of reality from a point of view of one who has still not ascended to the highest and most rarefied levels of spiritual attainment. In other words, the question of whether or not metaphysical idealism implies a type of eternalism misses the most important point, which is that for Ratnākaraśānti and the adherents of Yogic Practice, ontological idealism plays a crucial, but only temporary, role in contemplative development. On this note, we turn to Ratnākaraśānti’s articulation of the three natures theory, hallmark of the Yogic Practice tradition. 60 B. The Three Natures The three natures—constructed, dependent, and perfected—comprise the most basic exegetical scheme of the Yogic Practice tradition. Although, as carefully documented by Boquist, 91 the precise articulation of the three natures did change and develop over time, their most salient features remained for the most part constant across the various texts and contexts in which they appear. Interestingly, the teaching of the “three natures” also appeared in the Perfection of Wisdom literature, albeit in what is perhaps an even more inchoate form than what is otherwise generally considered their earliest appearance in the Sutra Unraveling the Intent (Saṃdhinirmocana Sūtra). 92 However, these verses only appear at the very end of the Large Sutra on Perfect Wisdom,93 and are only extant in Tibetan.94 This may indicate that the passages in question are, at least in part, a later interpolation into the Perfection of Wisdom literature. In any case, given the wide range of source material, and our overarching concern to elucidate the Pith Instructions for the Ornament of the Middle Way on its own terms, in general this analysis of the three natures will be limited to the material cited by Ratnākaraśānti, and to Ratnākaraśānti’s own presentation, apart from our consideration of the relevant Western secondary literature. 91 Ake Boquist, “Trisvabhāva: A Study of the Development of the Three-nature-theory in Yogācāra Buddhism,” Lund Studies in African and Asian Religions 8 (1993): 1-159. 92 Ming-Wood Liu, “The Three-Nature Doctrine and Its Interpretation in Hua-Yen Buddhism,” T’oung Pao 68, no. 4 (1982): 182 n. 2. “The [Sutra Unraveling the Intent] is commonly regarded as the forerunner of Yogācāra Buddhism because it is the earliest extant piece of writing which attempted to define in some detail such concepts as the "storehouse-consciousness (ālaya-vijñāna),"seeds" (bīja) and "three natures" (trisvabhāva), concepts which later formed the basic furniture of Yogācāra philosophy. However, it should be stressed that its treatment of these concepts is far from systematic, and despite the fact that Asaṇga and Vasubandhu quoted it as authority on many occasions, it is basically a text on religious practice and differs considerably both in style and content from such well- known Yogācāra classics as the Yogācārabhūmi-śāstra, the MS and the Triṃśikāvijñapti-kārika.” 93 Edward Conze, trans., The Large Sutra on Perfect Wisdom: with the divisions of the Abhisamayālaṇkāra (Berkeley: University of California Press, 1975), 644-652. 94 Ibid., ix. 61 1. The Constructed Nature a) Etymology of the Constructed Nature The term “constructed nature” translates the Sanskrit parikalpita (Tib. kun brtags). This word is derived from the root √kḷp, which has the primary sense of “array” or “arrange,” as well as the secondary sense of “construct.” By extension, in certain contexts parikalpita means “imagined,” as for example an idle daydream is constructed by the imagination; Boquist translates parikalpita as “imagined nature” in his survey of the three nature theory in early Yogic Practice literature. However, regarding the Tibetan translation of parikalpita (kun brtags), the Tibetan root btags can mean “imputation,” or the literal superimposition of something upon something else, as well as “imagine.” The Tibetan translation also captures the sense of the prefix pari- as completely or totally; kun literally means “all.” b) Characteristics of the Constructed Nature Ratnākaraśānti defines the constructed nature as the phenomenal feature of duality in consciousness: The defining characteristic of the entirely-imagined nature is that it does not exist in the way that it appears, namely as the duality of apprehended and apprehender. Although duality does not exist, 62 dualistic appearances arise through the force of prior psychological imprints.95 This definition is consistent with the opening verse of the Distinguishing the Middle from the Extremes, the primary of Ratnākaraśānti’s Yogic Practice source texts (by frequency of citation). In this verse, the imputed nature is referred to as “duality”: The imagination of the unreal [abhūta parikalpa] exists; in it, duality is not observed. However, its emptiness is observed, and it is observed in that [emptiness].96 [MV I.1] Somewhat counterintuitively, in this verse the “imagination of the unreal” refers to the “otherdependent nature” (Skt. paratantra, Tib. gzhan dbang), the second of the three natures. As briefly mentioned above, and as further developed below, this refers to the sum total of dependent origination. Here, the constructed nature is simply nonexistent “duality” (Skt. dvayaṃ, Tib. gnyis). By contrast, the dependent nature, though qualified as “unreal” (Skt. abhūta, Tib. yang dag pa ma yin), is also maintained as existent (Skt. asti, Tib. yod). It is worthwhile to note the subtle difference between asti (“exist”) and vidyate (“observed”) in the verse above. The Tibetan translation does not differentiate between these two, rendering them both in terms of yod, the generic word for existence. And indeed, the word vidyate may also simply mean “exist.” However, it is a simple derivative of the Sanskrit root 95 Ratnākaraśānti, Pith Instructions, ff. 224.b.1-2. 'di gnyis kyi mtshan nyid ji lta ba bzhin du med pa'i phyir kun brtags pa'i ngo bo nyid do/ de yang gzung ba dang 'dzin pa gnyis so/ gnyis med kyang bag chags kyi dbang gis gnyis su snang bar skye ba'o/ 96 Maitreya, Distinguishing the Middle from the Extremes (Madhyāntavibhāga, dbu mtha’ rnam ‘byed). Ed. Ramchandra Pandeya, Madhyāntavibhāgaśāstra (Delhi: Motilal Banarsidass, 1971). Ed. Gadjin M. Nagao, Madhyāntavibhāga-bhāṣya: A Buddhist Philosophical Treatise Edited for the First Time from a Sanskrit Manuscript (Tokyo: Suzuki Research Foundation, 1964). abhūtaparikalpo’sti dvayaṃ tatra na vidyate | śūnyatā vidyate tvatra tasyāmapi sa vidyate. 63 √vid, which has the primary sense of seeing or observing—this root √vid comes from the same Proto-Indo-European form as the Latin videre “to see. Thus the “non-observation” of the constructed nature is differentiated, in the original Sanskrit, from the “existence” (asti) of the abhūtaparikalpa. The root √vid can also denote knowledge, and while there are some minor differences in how √vid qua knowledge and √vid qua observation behave grammatically, Whitney notes that these two senses were “doubtless originally the same” and that “in some of their meanings, the two are so close together as hardly to be separable; and there are instances, from the Veda down, of exchanges of form between them.”97 In this way, the phenomenological and epistemological dimensions of ontological claims made by the Yogic Practice tradition in general and Ratnākaraśānti in particular are deeply embedded in the Sanskrit language. In any event, what does not exist (or “is not observed”) is duality, the constructed nature. This is usually glossed as the duality of phenomenal subject (grāhakākāra, Tib. ‘dzin rnam, literally “aspect of the grasper”) and object (grāhyākāra, Tib. gzung rnam, literally “aspect of the grasped”), and Ratnākaraśānti initially glosses “duality” in precisely this way. Gold, however, informs us that, Whereas commentaries almost universally gloss the term dvaya (‘duality’) with some version of the phrase grāhyaṃ grāhakaṃ ca (lit. ‘grasped and grasper’, but usually translated as ‘subject and object’), in fact this gloss is absent from the earliest strata. The term and its gloss are derived from separate streams of Yogācāra reasoning—one from discussions of linguistic conceptualization and the other from discussions of perception. Once we see that these two are distinct, it becomes clear that the commentarial literature asserts their identity in order to philosophically 97 William Dwight Whitney, The Roots, Verb-forms, and Primary Derivatives of the Sanskrit Language: A Supplement to his Sanskrit Grammar (1885; repr., Delhi: Motilal Banarsidass, 2005), p. 160. 64 unify Yogācāra thought.98 And, in fact, it should be noted that the glosses of the three natures in the first verse of the Distinguishing the Middle—“duality” for the imputed, “imagination of the unreal” for the dependent, and “emptiness” for the perfected—are not necessarily representative of how the three natures are presented in the Yogic Practice literature taken as a whole. For example, the other main source Ratnākaraśānti uses for his presentation of Yogic Practice doctrine is the Sutra Unraveling the Intent (Skt. Saṃdhinirmocanasūtra, Tib. dgongs ‘grel). In Chapter VI of this sutra, it is taught that “the [constructed] nature is the attribution of own-being (svabhāva) to phenomena by means of language which enables us to talk about them,” and that the constructed nature “consists of names attached to marks [nimitta, Tib. mtshan ma].”99 Ratnākaraśānti implicitly adopts this “linguistic” sense of the imputed nature in his previously-cited remark that phenomena are “nothing more than the imputation of names and characteristics.” It is beyond the scope of this project to trace the development of Yogic Practice doctrine, or how the diverse senses of “duality” eventually came to be subsumed under the grasper/grasped paradigm within the commentarial literature. But it is worth making a brief note of how duality, or more precisely non-duality, is presented within the Descent into Laṇka Sūtra: Again, Mahāmati, what is meant by non-duality? It means that light and shade, long and short, black and white, are relative terms, Mahāmati, and not independent of each other; as Nirvāṇa and Saṃsāra are, all things are not-two. There is no Nirvāṇa except where there is Saṃsāra; there is no Saṃsāra except where there is Nirvāṇa; for the condition of existence is 98 Jonathan Gold, “No Outside, No Inside: Duality, Reality and Vasubandhu’s Illusory Elephant,” Asian Philosophy 16, no. 1 (2006): 1. 99 Boquist, p. 24. 65 not of mutually-exclusive character.100 Thus in the Descent into Laṇka Sutra, as in the Yogic Practice tradition generally, “duality” did not exclusively refer to the opposition between phenomenal subject and object, but somewhat more broadly encompassed any and all linguistic or conceptual oppositions. Certainly the phenomenal sense was always the most common manner of interpreting “duality” when it appeared in the Yogic Practice literature, but Vasubandhu for example was also careful to stress that “duality” referred as well to the opposition between existence (bhūta) and non-existence (abhūta). In other words, from the very beginnings of the Yogic Practice tradition, the concept of non-duality had ontological as well as phenomenological ramifications. It should also be noted that the above identification—or, more precisely, non-opposition—between Saṃsāra and Nirvāṇa, as well as existence and non-existence, is in keeping with the Perfection of Wisdom literature, particularly as commented upon in the Root Verses of the Middle Way. Ratnākaraśānti does not explicitly relate the constructed nature qua duality to the opposition between existence and nonexistence. However, in his account, the First Yoga—the first in a four-fold process of contemplative development—is the contemplation of the “two extremes of entities” (dngos po mtha’ gnyis). This is ambiguous, perhaps deliberately so; in the context of Middle Way exegesis the “two extremes” usually refers to existence and nonexistence, while in Yogic Practice exegesis the “two extremes” more commonly refers to phenomenal subject and object. Although the Four Yogas are only mentioned at the end of the text (in both the Pith Instructions and his longer Commentary on the Ornament of the Middle Way), they clearly structure his entire presentation, and will be revisited in greater detail below. Here, the 100 D.T. Suzuki, trans., The Laṇkāvatāra Sūtra: A Mahāyāna Text (1932; repr. Delhi: Motilal Banarsidass, 1999), 67-68. 66 key point to understand is that Ratnākaraśānti’s presentation of the First Yoga draws attention to both the ontological and epistemological senses of duality. There is additional evidence for this point in his unified approach to the Middle Way and Yogic Practice traditions, via what he called the “Middle Way of the Three Natures”: Concerning the statement that this is the “Path of the Middle Way which Possesses the Three Natures,” there is no existence in terms of the constructed nature, and there is no non-existence in terms of the dependent and perfected natures. Therefore, it is free from the two extremes.101 His assertion that “there is no non-existence” of the dependent and perfected natures will be addressed below. The salient point is that Ratnākaraśānti here maintains that the imputed nature does not exist at all. In this, he follows the next verse of the Distinguishing the Middle, which he also quotes: Thus, everything is explained as not empty and not non-empty, Because it is existent, non-existent, and existent. This is the Middle Path.102 [MV I.2] There is a certain unaddressed tension here, between the broader sense of duality—particularly regarding the opposition of existence and nonexistence—and the contention that the imputed nature is non-existent. If duality refers inter alia to the opposition between existence and nonexistence, in what sense can it meaningfully be said not to exist? One answer is: duality appears 101 Ratnākaraśānti, Pith Instructions, ff. 225.b.5. rang bzhin gsum dang ldan pa de ni dbu ma'i lam yin no zhes gsungs pa ni/ kun brtags pa'i rang bzhin gyis ni yod pa ma yin zhing/ gzhan gyi dbang dang yongs su grub pa'i rang bzhin gyis ni med pa ma yin te/ de bas na mtha' gnyis dang bral ba'o/ 102 Ratnākaraśānti, Pith Instructions, ff. 225.b.6. na śūnyaṃ nāpi cāśūnyaṃ tasmāt sarvaṃ vidhīyate| sattvādasattvāt satvācca [Nagao: satvād asatvāt satvāc ca] madhyamāpratipacca sā|| 67 to exist conventionally, but is ultimately non-existent. Ratnākaraśānti hints in this direction when he writes that the imputed nature “does not exist in the way that it appears,” i.e. that it is apparent but not (truly or really) existent. However, the operative opposition here between appearance and existence, or phenomenology and ontology, obscures one of the main points that Ratnākaraśānti is making in the Pith Instructions, which is that the phenomenal experience of being a conscious observer aware of an object always already implicates the “subject” in a web of imputed, conceptual/linguistic designations. As Gold puts it, duality “signifies the mistaken character of all ordinary conceptual construction, all ordinary language, and all ordinary perception.”103 Therefore “the denial of duality fundamentally consists in a denial of the entire concept of reality as ordinarily conceived—or, to say the same thing, it is a denial of any and all products of conceptual-linguistic construction.”104 c) Phenomenology and the Constructed Nature The Yogic Practice tradition is sometimes assessed in terms of the Husserlian tradition of phenomenology. Dan Lusthaus, for example, has attempted what he calls a “philosophical translation” of Yogic Practice philosophy into the Husserlian tradition (“and vice versa”). 105 A detailed analysis of his project is outside the scope of the present effort. But, while there is much 103 Jonathan Gold, “No Outside, No Inside,” 28. 104 Ibid., 2. 105 Dan Lusthaus, Buddhist Phenomenology, viii. 68 to recommend in Lusthaus’ work, it should be noted that many of his assertions are problematic. For example, Lusthaus simply states that “Yogācāra is Buddhist phenomenology.”106 But Husserlian phenomenology is predicated on a move “back to the things themselves,” a direct mental encounter with the essences of entities (the “eidetic reduction”), and constitutes thereby a rejection of Kant’s position that the noumenon or thing-in-itself must always remain finally inaccessible to both the mind and the senses. However, as already noted, Mahāyāna Buddhist philosophy in general totally rejects the existence of any such essence or “thing-initself,” whether eidos or noumenon. Similarly, the existence of a Transcendental Subject is absolutely essential to Husserl’s phenomenological project, and yet such a Subject would be precisely the type of Self or ātman that the Buddha taught did not exist. Most problematically, however, the “phenomenological reduction,” the most basic methodological tool of Husserlian phenomenology, is predicated on an absolute and rigid divide between the phenomenological (and, ultimately, Transcendental) subject on the one hand, and the phenomenological object on the other. In the most basic sense of Brentano and Husserl’s “phenomenological revolution,” which sought to restore the understanding that to be aware is necessarily to be aware of something, the first of the three natures does invoke the phenomenological distinction between subject and object as being constitutive of consciousness. Crucially, however, this only applies to the consciousnesses of ordinary, unenlightened beings. Like Husserl, Ratnākaraśānti and other Yogic Practice adherents maintained that consciousness ordinarily possesses the intentional structure of a subject apprehending and object. Contra Husserl, however, Ratnākaraśānti maintained that this dualistic feature of consciousness was only apparent, that phenomenal duality itself is in fact deceptive, illusory, and ultimately 106 Ibid., 11. 69 nonexistent. According to Ratnākaraśānti and the Yogic Practice tradition, since duality is unreal, the true nature of consciousness is nondual. Needless to say, this is a major difference between the Yogic Practice school of thought and the tradition of Husserlian phenomenology. Nor is this the only category error that Lusthaus makes in his presentation of Yogic Practice philosophy. Lusthaus strenuously argues against interpreting Yogic Practice philosophy as any type of idealism, particularly as anti-materialist ontological idealism.107 But a strictly phenomenological approach to Yogācāra is fundamentally misguided, since the foundational texts of Yogic Practice doctrine explicitly embrace an idealistic ontology. This is not to say that Yogācāra is first and foremost a system of ontological speculation; on the contrary, in the hands of pramāṇa theorists like Dignāga and Dharmakīrti, the Yogic Practice tradition developed a highly sophisticated idealistic epistemology, while praxeological/eleutheriological concerns were of necessity always in the background (and occasionally in the foreground). Some Western authors also describe Yogic Practice theory as a kind of epistemic108 or “cognitive closure”109 concerning what is knowable, casting the point at stake in strictly epistemological as opposed to ontological terms. This is an understandable exegetical tack, but misses some key features of Yogācāra. Specifically, for Ratnākaraśānti as for most classical Yogācārins, the division between ontological and epistemological idealism is either unintelligible, or irrelevant, or both. 107 Ibid., 4. “What is Yogācāra? It has generally been mislabeled as ‘idealism.’” 108 Alan Fox, “Empty Logic: Mādhyamika Buddhism from Chinese Sources by Hsueh-li Cheng” (review), Journal of Chinese Philosophy 13 (1986): 361-365. “[Cheng] accepts the popular misconception [sic] of Yogācāra as ‘Idealism’…. We are just starting to understand that Yogācāra was not presenting a system of ontology, but rather a system of closure.” 109 Dan Lusthaus, Buddhist Phenomenology, 26-28. Lusthaus translates vijñapti-mātra as “cognitive closure,” which is perhaps justifiable if somewhat unorthodox. However in the same passage, and elsewhere (pp. 351 ff.), he also translates saṃvṛti (“relative” truth) as “cognitive closure,” which is untenable. 70 Consider what would be required for epistemological idealism without ontological idealism: the existence of extra-mental objects which, in principle and a priori, are incapable of ever leaving a perceptual trace in consciousness. Earlier, the deep relationship between ontology and epistemology in the Sanskrit root √vid was observed. In the pramāṇa theory of Dignāga and Dharmakīrti, as previously noted, what makes an entity real is its ability to perform a causal function. As Georges Dreyfus points out, the type of “function” meant ultimately comes down to the ability to cause a cognition in the mind of an observer;110 to be real is to be perceived.111 In what sense does a phenomenon exist, if it is never observed? There is tremendous difficulty involved with framing this problem, even in English; phenomenon itself derives from the Greek root phainein “to appear.” Every “phenomenon,” by definition, appears for some subject. The notion of a phenomenon which does not appear or is never observed is self-contradictory on its face. And, at least for Dharmakīrti and his Yogācārin successors such as Ratnākaraśānti, performing a (causal) function means being able to be perceived; conversely, being unable to be perceived means being unable to perform a causal function, and thus by definition, not being real. Even if the idea of a rigid distinction between idealistic ontology and idealistic epistemology could theoretically have been parsed, Ratnākaraśānti disputes the relevance of such a distinction when he contrasts phenomena which have the “nature of consciousness” (rnam par shes pa’i bdag nyid) to phenomena which are “external” (phyi rol): 110 Georges B.J. Dreyfus, Recognizing Reality: Dharmakīrti’s Philosophy and its Tibetan Interpretations (Albany: State University of New York Press, 1997), 290. “One implication of Dharmakīrti’s account is that the objects with respect to which cognitions are deceptive are nondeceptive must be real, because only real things, that is, causally effective phenomena, have the capacity to perform functions.” 111 John Dunne, Foundations of Dharmakīrti’s Philosophy, 85 n. 52. “Perhaps Dharmakīrti’s most concise statement of this point is: ‘Existence is just perception’ [PVSV ad PV 1.3…: sattvam upalabdhir eva…]. This claim, which might even be rendered ‘To exist is to be perceived,’ indicates the close relation between perception and the ultimately real in Dharmakīrti’s philosophy.” 71 The flaws which follow from the claim that blue-patches and so on are external are the same for a blue-patch which has the nature of consciousness, because there is no difference in the unacceptable conclusions that follow. The distinction between [external] objects and consciousness simply does not amount to anything at all.112 In other words, whether observed phenomena are asserted to be internal (i.e. mental) or external (i.e. extramental) is irrelevant, since they are both susceptible to analysis, and neither is in fact ultimately real. Thus Ratnākaraśānti clearly embraces both an anti-realist idealistic ontology and an anti-realist idealistic epistemology. In this, he was not alone, but following the general trends of Indian Buddhist scholastic philosophy, as defined by the encounter between the Yogic Practice and pramāṇa traditions. 2. The Dependent Nature a) Etymology of the Dependent Nature The second of the three natures is the paratantra (Tib. gzhan dbang) or dependent nature. Para means “other.” Tantra has a wide semantic range: continuity, connection, dependence, and so on. At one point in the Pith Instructions, Ratnākaraśānti plays on the 112 Ratnākaraśānti, Pith Instructions, ff. 227.a.7-227.b.1. ji ltar sngon po la sogs pa phyi rol gyi las (P/N/S: la) thal bar 'gyur pa'i skyon (P: rkyen) brjod pa 'di dag ni/ rnam par shes pa'i bdag nyid kyi sngon po la sogs pa la yang mtshungs te/ thal bar 'gyur ba la bye brag med pa'i phyir ro/ don dang rnam par shes pa'i bye brag tsam gyis ni cir yang mi 'gyur ro/ 72 multiple meanings of tantra, drawing out the sense of “connection.”113 The Tibetan translation of the term literally means “other-powered,” which emphasizes the dependent nature’s relationship to causes and conditions; it is, quite literally, dependent upon or “powered by” them. Ratnākaraśānti states that “the defining characteristic of the otherdependent nature is that it depends on causes and conditions.”114 This is a standard gloss of the dependent nature, which is often explained (e.g. by Mi pham) as having a meaning “similar to dependent origination.”115 It is, in essence, nothing more than causality writ large. To the extent that, as described previously, causal interaction implies some type of perceptual cognition for some being somewhere, the dependent nature also refers to the sum total of phenomenal appearances. Again, given the general equivalence in Yogic Practice theory between “what exists” and “what is observed,” the global process(es) of causality and the sum total of phenomenal appearances should not be considered in any way distinct. b) Characteristics of the Dependent Nature There has been some terminological imprecision concerning the dependent nature, stemming from the opening verse of the Distinguishing the Middle from the Extremes, where it is 113 See Appendix, note 321. 114 Ratnākaraśānti, Pith Instructions, ff. 224.b.2. rgyu rkyen la rag las pa’i phyir gzhan gyi dbang gi ngo bo nyid do/ In this gloss, Ratnākaraśānti draws attention to the categorical difference between the dependent nature, which is (by definition) conditioned, and transcendent wisdom, which is (by definition) unconditioned. 115 Ju Mi pham, trans. Thomas Doctor, Speech of Delight: Mipham’s Commentary on Śāntarakṣita’s Ornament of the Middle Way (Ithaca: Snow Lion Publications, 2004), 42-43. gzhan dbang gi rnam shes zhes bya ste rten ‘brel dang don ‘dra’o/ 73 referred to as the construction or “imagination of the unreal” (Skt. abhūta parikalpa, Tib. yang dag ma yin kun rtog). This is not to be confused with the “constructed” or “imagined” nature (Skt. parikalpita, Tib. kun brtags). The constructed nature is ontologically qualified as strictly nonexistent. The dependent nature, by contrast, is existent, but “unreal” (abhūta); the Tibetan translation, yang dag ma yin, also carries the semantic force of being “inauthentic” or “not genuine.” Thus there is something of a double motion with respect to the dependent nature. On the one hand, it is asserted as real or existent, at least to the extent that the causal processes which undergird (or are simply identical with) it are real. On the other hand, it is also in some sense inauthentic or unreal. Stefan Anacker explains the significance of this point: Abhūta parikalpa denotes “a construction of that which is not.” This term could also be rendered “a construction of the unreal”, since the Sanskrit word “abhūta” comes to mean “unreal”, and it is so rendered by the Chinese and Tibetan translators. But this translation ignores the peculiar manner in which this term is used in [Distinguishing the Middle], and makes impossible puns which Vasubandhu himself makes on “bhūta” and “abhūta”. In fact, for Vasubandhu any duality “real/unreal” cannot exist. It would be “an extreme relating to being and non-being”. What the construction of that which was not constructs is something which “was not” before it was mentally constructed, which “is” as long as it is believed, and which “isn’t” as soon as it is dropped from consciousness.116 So the dependent nature is only provisionally or relatively real; it is real in the sense that it appears, but (“ultimately”) unreal insofar as it is eventually brought to cessation or “dropped from consciousness” as a result of contemplative practice. 116 Stefan Anacker, Seven Works of Vasaubandhu: The Buddhist Psychological Doctor (Delhi: Motilal Banarsidass, 1984), 195. 74 Why is the imagination or “construction of the unreal,” i.e. the dependent nature, said to exist, but the imagined or “constructed nature” said not to exist? The dependent nature refers, as Vasubandhu says in Defining the Three Natures (Skt. Trisvabhāvanirdeśa, Tib. rang bzhin gsum nges par bstan pa), to that which appears, while the imputed or imagined nature refers only to the manner in which it appears: What appears is the dependent (nature); as it appears is the imaginary (nature), (the first one being so called) because it exists subordinated to causes, (the second one being so called) because its existence is only a mental creation.117 [TSN 2] In other words, although duality itself lacks any existence, phenomenal appearances— which are ordinarily experienced dualistically—do, in some sense, exist. The point is to distinguish between, on the one hand, the undeniable fact that there is some kind of phenomenal appearance, and, on the other hand, the illusory or unreal manner in which phenomena ordinarily appear. This, of course, raises the question of whether or not there exists some nonillusory manner in which phenomena may appear. In such a case, the word “phenomenon” is not necessarily appropriate, since such an appearance would not be an appearance to or for some phenomenal subject. But Ratnākaraśānti certainly pushes the metaphor of “luminosity” quite far, possibly even as far as asserting some type of ultimately luminous appearance; at one point, he distinguishes between three different senses of luminosity, the luminosity of “supreme appearances” (rab tu snang ba), “individual appearances” (so sor snang ba) and “luminous-like” (gsal bzhin pa) 117 Vasubandhu, trans. Fernando Tola and Carmen Dragonetti, Defining the Three Natures (Trisvabhāvanirdeśa). In Fernando Tola, “The Trisvabhāvakārikā of Vasubandhu,” Journal of Indian Philosophy 11 (1983): 252. yat khyāti paratantro’sau yathā khyāti sa kalpitaḥ| pratyayādhīnavṛttitvāt kalpanāmātrabhāvataḥ|| 75 appearances.118 Clearly, this distinction hinges on the differentiation of everyday phenomenal appearances from some other unspecified type of “supreme” appearance. Still, according to Ratnākaraśānti, the fourth and final stage of Yogic Practice is “nonappearance” (snang med), so even if there is some type of non-dual appearance, presumably even this appearance ultimately ceases. In any event, as is well-attested, the Yogic Practice school represented a continuation of mainstream Abhidharma philosophy within the Mahāyāna. Within this continuity, the paradigm of “designated” vs. “substantial” existence was eventually sublated into the Middle Way ontology of “relative” vs. “ultimate” existence, such that what was “substantially” existent from the point of view of the Abhidharma was only the “relative” truth of dependent origination. In other words, from the point of view of Yogic Practice theorists, causally-efficient dharmas were still understood to be constitutive of gross phenomena like pots and mountains, but were no longer understood to exist “ultimately” or inherently insofar as ultimate or inherent existence necessitates a commitment to metaphysical realism. This required an explicit distinction between “substantial” and “ultimate” existence, which is precisely the move that Ratnākaraśānti makes: It is also said that there is designated existence (btags pa’i yod pa, *prajñaptisat), substantial existence (rdzas su yod pa, *dravyasat), and ultimate existence (don dam par yod pa, *paramārthasat). 118 Ratnākaraśānti, Pith Instructions, ff. 226.b.7-227.a.1. chos rnams rang nyid gsal bzhin par bdag nyid kyis gsal ba'i ngo bo nyams su myong bas grub pa yin la/ gsal ba de yang rab tu snang ba dang/ so sor snang ba dang/ gsal bzhin ba zhes bya’o/ 76 Although Ratnākaraśānti does not explicitly equate “substantial existence” with relative truth, nor indicate whether “designated” phenomena exist only relatively or not at all, it is clear that his presentation of the three natures is based on three separate ontological modes, only one of which is the ultimate truth according to the Middle Way. Naturally, this schema hinges on the idea that phenomena can exist “substantially” without existing “ultimately” or being ultimately real. Nāgārjuna, to recall, far from attempting to negate dependent arising as such, only demonstrated that the causal process and its products are always and everywhere empty. In the Root Verses, the “object of emptiness,” that of which the subjects under consideration were understood to be empty, was primarily understood as self-nature or established/inherent existence, essence-svabhāva. The emptiness of dependent-arising was understood, then, not to deny the fact of phenomenal appearance, but to pre-emptively preclude the possibility of phenomena ever existing by virtue of their own identity. As phenomenal appearances depend upon causes and conditions in order to manifest—which fact was used by Nāgārjuna in order to demonstrate their essential emptiness—the full range of phenomenal appearances is coextensive with the sum total of the relationships of those causes and conditions. And, since the dependent nature refers to the sum total of dependent origination, it maps onto the concept of relative truth, to the extent that relatively-existent phenomena are dependently-arisen. The dependent nature’s dependency upon causes and conditions therefore entails its unreality or emptiness from the perspective of the Middle Way; the Yogic Practice tradition maintains that the dependent nature both exists and is empty, just as in the Perfection of Wisdom and the Root Verses of the Middle Way it is asserted that dependent-arising is not merely or exclusively non-existent, but rather is that which is empty. The crucial point here is that, to Ratnākaraśānti and Yogic Practice 77 theorists, “being empty of subject-object duality” must in some sense mean “being empty of inherent or established existence.” Given the “substantiality” of the dependent nature, in relation to the constructed nature, the obvious question is what nature the “substance” (Skt. dravya, Tib. rdzas) being causally conditioned has. It is clear that “substance” here does not mean physical matter. Ratnākaraśānti explains that the “substance” of the dependent nature is entirely mental, as it is exhaustively described in terms of three different categories of consciousness: The dependent nature also has three aspects: the storehouse consciousness, afflicted mind, and active cognition. The first is called “storehouse” because it is the basis for the “seeds” of all the phenomena of emotional affliction, and “consciousness” because it takes itself as its support. Afflicted mind is “afflicted” since it is always concomitant with the four emotional afflictions, and “mind” because it constantly thinks of itself: “I am.” Finally, active cognition continually arises, over and over again, through engagement with various objects. It takes these objects as its support, thus it is a consciousness.119 In this scheme, what Ratnākaraśānti terms “active consciousness” refers to the traditional six consciousnesses according to the Abhidharma: the five sense-consciousnesses, plus the sixth “mental” consciousness. In keeping with the Yogic Practice tradition of maintaining eight collections of consciousness, Ratnākaraśānti explains that the other two aspects of the dependent nature are the seventh (“afflicted”) and eighth (“storehouse”) consciousness. 119 Ratnākaraśānti, Pith Instructions, 224.b.3-5. gzhan gyi dbang de yang rnam pa gsum pa ste/ kun gzhi rnam par shes pa dang/ nyon mongs pa can gyi yid dang/ 'jug pa'i rnam par shes pa'o/ kun nas nyon mongs pa'i chos thams cad kyi sa bon gyi rten du gyur bas kun gzhi yin la rang gi ngo bo la nye bar dmigs pas rnam par shes pa'o/ nyon mongs pa bzhi dang rtag tu mtshungs par ldan pas nyon mongs pa yin la/ bdag go zhes rtag tu rlom sems pa'i bdag nyid yin pas yid do/ yul so so la yang dang yang du 'byung bas 'jug pa yin la/ yul la nye bar dmigs pas rnam par shes pa'o/ 78 The latter two—particularly the “all-ground” or “storehouse” consciousness—were the subject of polemical disputes in both India and Tibet. These disputes were highly abstract and technical, and do not directly concern the Ornament of the Middle Way or Ratnākaraśānti’s Pith Instructions.120 Briefly, however, it may be noted that the “storehouse” consciousness originated as an exegetical device for explaining the metaphor of karmic “seeds” (Skt. bīja, Tib. sa bon). According to the traditional Abhidharma framework, the six consciousnesses were strictly momentary. As William Waldron explains, the problem was that this framework had no power to explain how actions from the distant past, as for example from a previous life, could ripen and be experienced at a later point in time. Abhidharma philosophy also could not explain precisely what happened to meditators during certain high-level contemplative practices, wherein all psychophysical activity was brought to a halt; according to standard Abhidharma theory, it should not have been possible to re-emerge from such a state. The “storehouse” consciousness was therefore posited as the repository for the “seeds” created by various actions, and a way to explain the “attainment of extinction” (nirodha-samāpatti).121 Ratnākaraśānti only asserts that the seeds are exhausted in those who have attained realization, without treating this issue in any further detail. One of the main facets of the debate surrounding the storehouse consciousness concerned whether or not it could truly be counted as a consciousness, since, according to the Abhidharma, a consciousness must take an object in order to be a consciousness. Ratnākaraśānti alludes to this 120 Ratnākaraśānti acknowledges that the storehouse consciousness exists, but does not describe it in any detail nor relate it to the theory of Buddha-Nature (Skt. tathāgatagarbha, Tib. de bzhin gshegs pa’i snying po) as did some Yogācāra commentators; in fact the term “Buddha-Nature” is not found anywhere in the Pith Instructions. Ratnākaraśānti mentions the storehouse consciousness only twice, first as one of the aspects of the dependent nature, and later on in the context of explaining that enlightenment is defined as the exhaustion of karmic “seeds” in the storehouse. 121 William Waldron, The Buddhist Unconscious: The ālaya-vijñāna in the context of Indian Buddhist thought (London: RoutledgeCurzon, 2003), 78-87. 79 debate when he justifies the inclusion of the storehouse consciousness among the eight collections of consciousness by “taking itself as an object.” In other words, he maintains that the storehouse consciousness is indeed a consciousness, since it has an object: itself, or perhaps more precisely, the karmic “seeds” stored there. Ratnākaraśānti makes a similar point with respect to the seventh “afflicted” consciousness and the cognitive habit of self-centeredness: “Afflicted mind is ‘afflicted’ since it is always concomitant with the four emotional afflictions, and ‘mind’ because it constantly thinks of itself: ‘I am.’”122 This brings us to a crucial point regarding consciousnesses and their objects. Previously, while examining the imputed nature, it was noted that there is a dominant tendency among Buddhist scholars and commentators to gloss “duality” (in the context of Yogic Practice literature) as the phenomenological apprehended and apprehender. It must be understood here, however, that “apprehender” does not refer to an ontologically singular subject, since this would be precisely the kind of metaphysical self that the Buddha taught to be non-existent, and is refuted through e.g. the analysis of dharmas. The “apprehender” is, instead, one of these six cognitive/perceptual faculties. As Gold writes, Traditionally, both grasper and grasped consist primarily in elements of physical form. These terms most centrally refer, respectively, to the six sensory organs and the six sensory ‘fields’ (viṣaya) or objects—all but two of which are counted as dharmas of physical form. To deny grasper and grasped is thus to deny the full apparatus of sensory contact. Yes, this includes and begins with a denial of what we take to be external reality (bāhyārtha); but more importantly, this first denial leads to and culminates in a denial of what we take to be internal reality: The sense organs that are the basis for our false imagination of the self (ādhyātmikāyatana). What makes this good Buddhism, then, instead of nihilistic solipsism, is that 122 Ratnākaraśānti, Pith Instructions, ff. 224.b.4. nyon mongs pa bzhi dang rtag tu mtshungs par ldan pas nyon mongs pa yin la/ bdag go zhes rtag tu rlom sems pa'i bdag nyid yin pas yid do/ 80 subjectivity must be denied along with objectivity.123 In the Abhidharma, each moment of sensory consciousness is structured by the relationship between the (momentary) sensory faculty and the (momentary) object of sensation. The stream of perception, one of the five psychophysical bundles, is—qua bundle—ontologically manifold. What appears, to ordinary observers, as a single phenomenon is in fact the complex interplay of each sensory faculty contacting each object of sensation and producing each senseconsciousness, moment by moment. Ratnākaraśānti’s point is that even this basic level of subjectivity, the fact that sensory consciousness has been structured in this way, always already constitutes cognitive distortion (Skt. bhrānti, Tib ‘khrul pa), an important term that will be discussed in greater detail below. The bottom line is that the denial of grasper and grasped is as much a denial of the ontological reality of the internal psychophysical bases of perception as it is a denial of the ontological reality of the external objects perceived. In other words, to deny the existence of grasper and grasped is to deny the very bases of conscious experience. To be sure, for Ratnākaraśānti the primary point at stake is phenomenological: “grasper” and “grasped” refer to the subjective and objective dimensions of ordinary experience. But critiques to the effect that this amounts to solipsism are fundamentally misguided. To say that the dependent nature is composed of different types of consciousness is, pointedly, not to say that it is some type of arch-subjectivity. It is to point out that the process of causal interaction which defines dependent origination does not exist independently of the mind—that dependent origination itself is nothing more than the causal interactions of different types of consciousness 123 Jonathan Gold, “No Outside, No Inside,” 3. Emphasis original. 81 producing different types of appearances, which are only ever mistakenly apprehended in terms of a phenomenal subject and object. “Subjectivity” is therefore every bit as problematic as “objectivity.” And, as Gold points out, the denial of the “grasper” entails the denial of the ontological reality of any strictly “physical” foundation for conscious experience—up to and including the physical sense-organs and the brain. Needless to say, this has important consequences, which are the next topic. c) Idealism and the Dependent Nature The philosophical position that “material” phenomena are nothing other than mind or consciousness is usually called idealism. This is, for example, the stance articulated in the third verse of Distinguishing the Middle from the Extremes [MV I.3]: “Consciousness arises, appearing as objects, sentient beings, the self,” and so on. 124 In explicitly equating the dependent nature or dependent origination with different types of consciousness, Ratnākaraśānti thereby adopts a type of idealistic ontology or metaphysics. He makes this even more clear when he quotes Vasubandhu’s argument for idealism: If they are joined by their six sides, then “partless” particles would have six parts! If the six are together in one single place, masses would be merely a single particle!125 [VK 12] 124 artha sattvātma (Nagao: satvātma) vijñapti pratibhāsaṃ prajāyate| vijñānaṃ nāsti cāsyārthas tadabhāvāt tad apyasat|| 125 Vasubandhu, Twenty Verses (Viṃśatikākārikā), ed. Ram Shankar Tripathi (Varanasi: Sampurnananda Sanskrit University, 1992). ṣaṭkena yugapadyogāt paramāṇoḥ ṣaḍaṃśatā| ṣaṇṇāṃ samānadeśatvāt piṇḍaḥ syādaṇumātrakaḥ|| 82 In other words, materiality as such requires the local existence of partless particles, the buildingblocks of all matter. However, in order for the product of their agglomeration to possess spatial extension, these particles must all have a part that is joined to at least one other particle, and are therefore not “partless.” On the other hand, if they are all joined and still do not possess any parts, then they must all be localized in the same space. Of course, it is possible to argue against this position in any number of ways, but Ratnākaraśānti clearly found the above argument compelling. The most important thing to understand is that this is an argument against the reality of material phenomena as such, that is, an argument against the position that phenomena are composed of matter, as opposed to having the nature of consciousness. Some Western scholars have argued that the Yogic Practice philosophical system does not advance an idealistic ontology, but instead describes an idealistic epistemology: what is negated by the three natures model is not materiality or externality as such, but the possibility of having epistemic access to an “outside” world. This question will be addressed in the next section, on pramāṇa or the means of reliable knowledge, the dominant (and predominantly epistemological) scholarly discourse of India. Certainly, there is something to the argument that certain Yogic Practice theoreticians—for example, Dharmakīrti—argued explicitly in favor of this kind of epistemic closure, and only implicitly, or not at all, for an idealistic ontology. At the same time, it is extremely important to understand that the Yogic Practice tradition is not merely concerned with philosophy as a speculative or descriptive endeavor,126 as even its 126 John Makransky, Buddhahood Embodied: Sources of Controversy in India and Tibet (Albany: State University of New York Press, 1997), 79. “Yogācāra Buddhism has sometimes been described in modern scholarly works as an ontological idealism that speculatively reduces all phenomena to the nature of consciousness alone. But we can see from the descriptions of the meditation practice stages above that the yogic process itself does not comprise a speculative philosophy of any kind, let alone a speculative ontology. In Western philosophical terms, it correlates 83 signature doctrine of “mind-only” (Skt. cittamātra, Tib. sems tsam) or “mental representationonly” (Skt. vijñaptimātra, Tib. rnam rig tsam) is only the second step in the four-stage process of internal development known as the “Four Yogas.” In general, the Four Yogas are taught to proceed from an absorption in no external world, to the absorption of mind-only, to the absorption in thusness (Skt. tathatā, Tib. de bzhin nyid) or the nature of reality, to no absorption whatsoever. Different commentators parse the four stages slightly differently, but the basic structure is found in the Descent into Laṇka Sūtra,127 which Ratnākaraśānti cites as the canonical source for his presentation. To whatever extent the Yogic Practice tradition adopted the philosophical position of ontological idealism, then, this position was, first of all, only intended to be understood from within a framework of contemplative practice and, furthermore, was only relevant to the second stage of that practice. However, it should also be clear from the preceding analysis that Ratnākaraśānti does in fact adopt an idealistic ontology, as he considered dependent origination and the eight collections of consciousness to be literally equivalent. In this way, the move from ordinary perceptual cognition to “mind-only,” and from “mind-only” beyond, were distinct and incommensurate steps along a “sliding scale of analysis.”128 better with the project of phenomenology…. But it also goes far beyond Western phenomenology insofar as the four yogic states lead the practitioner into deeper and deeper strata of consciousness, through an analytical and meditative process that permanently alters, at its most fundamental level, the affective and cognitive structures of consciousness itself.” 127 D.T. Suzuki, trans., The Laṇkāvatāra Sūtra, 176. “When the Samskrita [i.e. the causally conditioned} is seen as devoid of qualified and qualifying, all predicates are discarded and thus the world is seen as of Mind itself. When the [Yogin] enters upon Mind-only, he will cease discriminating an external world; establishing himself where suchness has its asylum he will pass on to Mind-only. By passing on to Mind-only, he passes on to the state of imagelessness, he sees not [even] the Mahāyāna.” In a note (n. 1) on this last verse, Suzuki asks, “‘Not’ (na) is replaced by ‘he’ (sa) in one MS. May this be a better reading?” Ratnākaraśānti apparently would have answered in the affirmative, since he maintains that the Fourth Yoga is “seeing the Great Vehicle” (theg pa chen po mthong). 128 Sara McClintock, “The Role of the ‘Given’,” 139 ff. Dreyfus (1997) and Dunne (2004) use the concept of a “scale of analysis” to describe Dharmakīrti’s shift from External Realism to Epistemic Idealism, which will be 84 On this note, one of the most common rebuttals of ontological idealism is that it necessitates a commitment to solipsism, the idea that only oneself or one’s own mind exists. It is certainly a logical objection: if someone claims that the entire cosmos is nothing other than consciousness, and the person making the claim possesses a uniquely individual conscious experience, then is the claim not tantamount to an assertion that the entire cosmos is nothing other than the experience of the person making the claim? While there are a variety of ways to address this rebuttal, it is absolutely crucial to understand that this argument is invalid as applied against the Yogic Practice model, for the reasons outlined by Gold in the above passage: the denial of “grasper” and “grasped” entails the denial of the internal sense-faculties—particularly insofar as these internal sense-faculties are naïvely assumed to be the physical basis of our “subjective” experience—as well as the denial of external, “objective” reality.129 For Ratnākaraśānti to argue for ontological idealism, then, is manifestly not for him to argue for any form of solipsism, nihilistic or otherwise. It is, rather, to argue that both internal and external phenomena, as “internal” phenomena (such as sense-faculties or sense-data) and “external” phenomena (such as pots and mountains), are equally unreal. The dependent nature exists, but the dualistic mode of its appearance does not. Thus the dependent nature does not strictly refer either to “subjective” mental states or to an “objective” world. Sometimes the dependent nature is depicted as a “nondual flow”130 of mental causation, for just this reason. examined in greater detail below. Ratnākaraśānti makes a very similar shift; the primary difference is that, unlike Dharmakīrti, he articulates (if only elliptically) a stage of “non-appearance” beyond even “mind-only.” 129 Vasubandhu, Twenty Verses, v. 9. yataḥ svabījād vijñaptiryadābhāsā pravartate| dvividhāyatanatvena te tasyā munirabravīt|| Cf. also VKV ad VK 9. 130 Paul Williams, Mahāyāna Buddhism: The Doctrinal Foundations (New York: Routledge, 1989), 96. 85 Ratnākaraśānti does not use the exact phrase, but the “nondual flow” reading of the dependent nature is in agreement with his statement that, “Although duality does not exist, dualistic appearances arise through the force of prior psychological imprints.”131 In other words, neither phenomena nor the subjects perceiving them have any independent existence; phenomena are only perceived dualistically due to the cognitive habit of perceiving them dualistically. Phenomena, whatever it is that appears, is neither internal nor external, neither existent nor nonexistent. Nor, ultimately, are they “mind” as opposed to “matter”; recall Ratnākaraśānti’s earlier point: The flaws which follow from the claim that blue-patches and so on are external are the same for a blue-patch which has the nature of consciousness, because there is no difference in the unacceptable conclusions that follow. The distinction between [external] objects and consciousness simply does not amount to anything at all. Now, someone might say, “External particles are surrounded by six directions, but the individual moments of consciousness have four points of connection.”132 However, because they are the same in terms of having parts, these [differences] do not amount to anything at all, either. As there is no third kind of bundle apart from something which is either singular or manifold, blue-patches and so on are not established either as internal or external objects, and are therefore just false.133 131 Ratnākaraśānti, Pith Instructions, ff. 224.b.2. gnyis med kyang bag chags kyi dbang gis gnyis su snang bar skye ba’o/ 132 Material particles are described in terms of three dimensions, for a total of six points of connection (top, bottom, left, right, forward, and backward). Mental particles are described in two dimensions, for a total of only four points of connection (top, bottom, left, and right). 133 Ratnākaraśānti, Pith Instructions, ff. 227.a.7-227.b.3. ji ltar sngon po la sogs pa phyi rol gyi las (P/N/S: la) thal bar 'gyur pa'i skyon (P: rkyen) brjod pa 'di dag ni/ rnam par shes pa'i bdag nyid kyi sngon po la sogs pa la yang mtshungs te/ thal bar 'gyur ba la bye brag med pa'i phyir ro/ don dang rnam par shes pa'i bye brag tsam gyis ni cir yang mi 'gyur ro/ ci ste don gyi rdul phra rab ni phyogs drug nas ni bskor ba yin la/ rnam par shes pa'i rdul phra mo ni phyogs bzhi pa'o zhe na/ cha shas dang bcas pa mtshungs pa'i phyir des kyang cir yang mi 'gyur te/ gcig dang du ma las ma gtogs pa'i dngos po phung po gsum pa gzhan dag kyang med pas sngon po la sogs pa dag ni phyi rol gyi dngos po dang nang gi dngos por ma grub pas brdzun pa nyid do/ 86 In fact, Ratnākaraśānti argues that in fact all phenomenal appearances only arise due to psychological imprinting; this is precisely the “epistemic closure” referenced above. But this will be addressed with greater detail in the next chapter, concerning the pramāṇa discourse in relation to the Yogic Practice theory of perception. 87 3. The Perfected Nature a) Etymology of the Perfected Nature The third and last of the three natures is the “perfected nature” (pariniṣpanna, Tib. yongs grub). It is composed of the adverbial prefix pari, together with the past participle niṣpanna, meaning “completed” or “perfected” (the same word as the tantric practices of niṣpannakrama or “perfection stage”). b) Characteristics of the Perfected Nature The perfected nature is usually explained by way of reference to the other two natures; recall the first verse of the Distinguishing the Middle from the Extremes, which states that emptiness—that is, the perfected nature—is observed in the imagination of the unreal. In other words, the perfected nature is the dependent nature’s lack of duality. It is the fact that, although phenomenal appearances do arise, the dualistic manner in which they arise lacks any positive ontological status whatsoever; or, equivalently, that those appearances themselves lack any selfnature or ontological establishment as “existent” entities. As Ratnākaraśānti states, At all times, the dependent nature is empty of, isolated from, and without the constructed nature. This is the defining characteristic of the perfected nature. All consciousnesses are always empty of even the slightest trace of duality. And this emptiness never 88 changes: it is of one taste everywhere, like space. This is how the perfected nature is explained.134 Sometimes, the relationship among the three natures, and therefore the ontology of Yogic Practice philosophy in toto, is explained to hinge on the dependent nature. Boquist, for example, writes: The dependent nature is invariously [sic] held to be the sole existent entity within the Yogacara system, be it defined as fundamental causation or as ideas or both. The imagined and the [perfected] natures are not alternative categories on the same level, as the dependent nature, but modes which the dependent may assume. Hence, the latter is often described as the constant flow of awareness which is maintained as really existing. This stream of perceptual images may bifurcate into a subjective and an objective part, which accounts for the every-day experience which is illusive. When this way of beholding is eliminated, pure awareness is left. Therefore the [dependent nature] is construed from an ontological point of view while the other two are epistemological modes of the former.135 In this schema, the dependent nature is held to really exist qua causally-efficacious particulars (i.e. svalakṣaṇa) and/or nondual flow of mental events; the imputed nature is one, mistaken “epistemological mode” of accessing the dependent nature, while the perfected nature is simply a different—but, crucially, ultimately accurate—“epistemological mode.”136 Rather 134 Ratnākaraśānti, Pith Instructions, ff. 224.b.5-6. gzhan dbang de kun brtags pa des dus thams cad du stong pa nyid dang/ dben pa nyid dang/ bral ba nyid du gyur pa de ni yongs su grub pa'i ngo bo nyid do/ shes pa gang nam (P/N/S: rnams) yang gnyis kyis mi stong pa de ni cung zad tsam yang med de/ stong pa de yang nam yang 'gyur ba med par thams cad nas thams cad du nam mkha' lta bur ro gcig pa nyid ste (D/C: de)/ de bas yongs su grub pa'i mtshan nyid du brjod do/ 135 136 Boquist, “Trisvabhāva,” 15. Of course, this argument rests on the possibility of being able to rigorously distinguish between an “epistemological mode” and the ontological reality of that which is accessed via the “epistemological mode” in 89 than constituting a separate ontological or metaphysical category entirely, in this exegetical framework the perfected nature is simply the opposite of the imputed nature: the imputed nature is the dualistic apprehension of the dependent nature, while the perfected nature is the non-dual apprehension of that same dependent nature, what Boquist calls “pure awareness” above. A contemplative practicing emptiness meditation can thus metaphorically “pivot” from the imputed nature, to the perfected nature, while the underlying substratum (i.e. the dependent nature) remains unchanged; what shifts is the epistemic mode of access to the dependent nature, not the dependent nature itself. It is easy to understand the origin of this view, which D’Amato calls the “standard interpretation”137 of the three natures theory. If the perfected nature is nothing other than the lack of the imputed nature within the dependent nature, which is not ordinarily cognized as being empty in this way, it would appear reasonable to maintain that the difference between the imputed and the perfected natures is constituted in terms of an epistemic or even a phenomenological (if it is in fact possible to speak of non-dual phenomenology) approach to the dependent nature. D’Amato claims that “there is indeed strong support for the standard interpretation of the three-nature theory as presented in the classical texts of Asaṇga and Vasubandhu,” citing as evidence v. 21c-d of Vasubandhu’s Thirty Verses on Consciousness, which he translates as “The perfected is the perpetual absence of the former [i.e. the imputed] in that [i.e. the dependent].”138 question; this will be an important theme in Ratnākaraśānti’s account of perception and, later, the reflexive nature of awareness. 137 Mario D’Amato, “Three Natures, Three Stages: An Interpretation of the Yogācāra Trisvabhāva-Theory,” Journal of Indian Philosophy 33 (2005): 199. 138 Ibid. 90 Appearances, however, can be deceiving. It is doubtless the case that this perspective captures some important facets of the relationship between the perfected and the dependent natures. But Boquist’s assertion that the dependent nature is “invariously held to be the sole existent entity” within the Yogic Practice system is, as has already been observed, false: Ratnākaraśānti quotes a crucially important verse from Distinguishing the Middle, where it is explained that the three natures are “existent, non-existent, and existent,” in other words that both the imagination of the unreal (i.e. the dependent nature) and emptiness (i.e. the perfected nature) exist. This point is further reinforced in another verse cited by Ratnākaraśānti: The non-existence of duality, and the existence of that nonexistence, is the characteristic of emptiness.139 [MV I.13a] Of course, asserting that emptiness “exists” is problematic from a strict Middle Way perspective, and this issue will be addressed shortly. The point, however, is that Yogic Practice theory, as articulated in such seminal (and early) works as the Distinguishing the Middle from the Extremes, is not necessarily committed to the position that the perfected nature is only an “epistemological mode” on an equal ontological footing with the imputed nature, far less to the position that the dependent nature is the “sole existent entity.” 139 Maitreya, Madhyāntavibhāga. dvayā'bhāvo hyabhāvasya bhāvaḥ śūnyasya lakṣaṇam| 91 c) Two Models of the Perfected Nature There is, however, another approach, for example the position articulated by D’Amato in his sweeping exegesis of the three natures in the context of the Treasury of Great Vehicle Sūtras (Mahāyānasūtrālaṃkāra, “MSA”). D’Amato argues that, in this text at least, the three natures form a hierarchy or progressive series, with the imputed nature at the bottom and the dependent nature at the top of what he calls a “soteriologico-ontological model” of “three progressive stages of ontological realization, culminating in the perfected, non-conceptually-constructing awareness of thusness.”140 D’Amato clearly considers the “standard interpretation” to be normative, and describes this view as an “alternative interpretation.” Accordingly, he strictly limits his analysis to the Ornament to the Great Vehicle Sūtras and its Commentary (Mahāyānasūtrālaṃkārabhāṣya, “MSABh”), contrasting its “progressive” model of the three natures to the “pivot” model outlined above: Briefly stated, the difference between the MSA/Bh’s interpretation of the three-nature theory and the standard interpretation is that the MSA/Bh does not emphasize the ultimate existence of the dependent nature. In the standard account offered above, the dependent nature is understood to be ultimately real since it is the basis or substratum of reality itself: although the dependent nature is empty of inherent nature, it does ultimately exist. In the MSA/Bh, however, the existence of the dependent nature is not ultimately affirmed (although it is, of course, conventionally affirmed). In discussing the non-duality of the ultimate, the commentary to 6.1 states that the ultimate is not existent because of the imagined and dependent [natures], but it is not non-existent 140 Ibid., 185. 92 because of the perfected [nature]. This implies that the dependent [nature] does not ultimately exist.141 By drawing a clear line between the ultimate existence of the perfected nature and the lack of ultimate existence in the constructed and perfected natures, the author of the MSA/Bh was setting up an explicit contrast between what Ratnākaraśānti called “substantial existence” (i.e. the conventional reality of causes and effects at the level of dependent origination) and ultimate truth. This implies that, as the focus of contemplative practice progresses from the relative to the ultimate, an initial abandonment of duality leads eventually to the abandonment of dependently-originated appearances: At 11.13, the commentary states that the dependent nature is to be abandoned (praheyam). Such a statement would be unreasonable if the dependent nature were understood to be the basis or substratum of reality, given that the ultimate goal posited by the text is assimilation to reality itself. Furthermore, the commentary to 19.51 states that thorough knowledge of the three natures brings about the termination of the dependent nature. Again, this would not be reasonable if the dependent nature were understood to be the ultimate substratum of reality.142 In this regard, Ratnākaraśānti is clearly following the intent of the Ornament of the Great Vehicle Sūtras. The larger question concerns just how normative this model of contemplative progression was within three natures theory. 141 Ibid., 200. 142 Ibid., 200-201. 93 D’Amato’s reluctance to step outside the bounds of the Ornament of the Great Vehicle Sūtras is entirely justifiable on the grounds that his focus is the presentation of the three natures in that text, and that text alone. But it is also clear, from his above remarks, that he considers this “alternative interpretation” to be something of an anomaly within the Yogic Practice tradition. At the same time, he quotes the Sūtra Unraveling the Intent in support of his argument: Further support for these claims may be found in the Saṃdhinirmocana-sūtra’s trope of the three natures in terms of having diseased vision, where the imagined is “like the defects of clouded vision,” the dependent is “like the appearance of the manifestations of clouded vision… which appear as a net of hairs,” and the perfected is “like the unerring objective reference, the natural objective reference of the eyes when that person’s eyes have become pure and free from the defects of clouded vision.”143 This support for the “alternative interpretation” of the three natures from the Sūtra Unraveling the Intent and its commentary is intriguing enough on its own, and is particularly compelling when considered in conjunction with the previously-quoted evidence from the Ornament to the Great Vehicle Sūtras and Distinguishing the Middle from the Extremes. But it is still far from obvious that Vasubandhu, even in the Thirty Verses, subscribed to this “standard” interpretation. Consider verse 21: Conceptual construction (vikalpa)144 has the nature of being otherdependent, as it arises from conditions. The perfected is the permanent lack of the former [i.e. vikalpa] in the latter.145 [TK 21] 143 144 Ibid., 201 n. 15. It should be noted that the kalpa of vikalpa is the same kalpa as parikalpita. Vi- suggests dualistic bifurcation or conceptual activity; compare with vijñāna, “[dualistic] consciousness,” from vi- plus jñāna, which by itself simply means “awareness” or sometimes, in certain contexts, “[nondual] wisdom.” Both senses are present in the Pith Instructions; Ratnākaraśānti refers to “pure worldly awareness” (‘jig rten pa’i ye shes, *laukika-jñāna), the notquite-entirely undistorted perceptual awareness of the Buddhas, but he also frequently speaks of “authentic wisdom” 94 Here, Vasubandhu explains that dualistic “conceptual imputations” arise due to unspecified causes and conditions, presumably the prior psychological imprint or cognitive habit of thinking dualistically. Therefore it “has the nature of being other-dependent.” There is some slippage here between the imputed and the dependent natures: Vasubandhu is asserting that imputation has the nature of arising in dependence upon causes and conditions; therefore, imputation itself is dependently-originated. And yet, as he states, the dependent nature is permanently lacking dualistic imputation, and this very lack constitutes the perfected nature. Leaving aside the logical difficulties posed by this apparent contradiction, clearly Vasubandhu did not intend to assert that the imputed nature and the perfected nature have the same ontological status with respect to the dependent nature. It is true that the dependent nature is held to be the locus of emptiness, that which is (ultimately) empty of duality, but it is far from obvious that this constitutes an argument to the effect that the dependent nature is the “substratum of reality” and the perfected nature is nothing but one of two possible “epistemological modes” for accessing that substratum. If anything, in the Thirty Verses, Vasubandhu presents the three natures in terms of three separate, though inter-related, ontological modes—that is, three distinct ways in which things exist (or do not exist). In TK 20b, Vasubandhu posits that what is imagined or imputed has no self-nature at all.146 In TK 24, he explains how the three natures lack self-nature in three different kinds of ways, what he calls in (yang dag pa’i ye shes, *yathābhūta-jñāna). Thus vikalpa, which is frequently rendered “thought,” “discrimination,” or simply “concept” in English translation, can also mean “conceptual imputation” or “conceptual construction”; the vi- prefix only occasionally makes a difference. 145 Vasubandhu, Thirty Verses on Consciousness (Triṃśikāvijñaptikārikā), ed. Stefan Anacker, Seven Works of Vasubandhu. paratantrasvabhāvastu vikalpaḥ pratyayodbhavaḥ| niṣpannastasya pūrveṇa sadā rahitatā tu yā|| 146 parikalpita evāsau svabhāvo na sa vidyate|| 95 TK 23 the “three-fold lack of self-nature” (trividhāṃniḥsvabhāvatām, Tib. ngo bo nyid med rnam gsum): The defining characteristic of the first is that it lacks self-nature. The next [lacks self-nature] because it does not arise from itself. The next is that very lack of self-nature.147 The constructed nature, then, is defined by the fact that it does not possess any self-nature. The dependent nature has no self-nature because it is dependent upon causes and conditions; this may be understood as a nod in the general direction of the Root Verses of the Middle Way, as it is more or less the same argument made by Nāgārjuna. Finally, the perfected nature is not something that “lacks” self-nature—rather, it “is” the lack of self-nature. There are, in other words, three separate ontological senses of “lacking self-nature,” corresponding to the three natures. And while the exigencies of Sanskrit grammar do not necessitate the sequential structure of this verse to be interpreted as indicating a progression from the first to the second to the third, it should be clear from context that the three different senses of lacking self-nature are not co-equal. According to Vasubandhu, the imputed nature lacks any self-nature by definition, and the dependent nature lacks self-nature on account of the mode of its production; the perfected nature, by contrast, occupies a separate ontological category entirely, as it does not “lack” self-nature but is that lack of self-nature, itself. That is to say, the way in which the constructed nature lacks self-nature is not the same as the way in which the dependent nature lacks self-nature, which is, in turn, not the same as the way in which the perfected nature lacks self-nature. The imputed nature is strictly nonexistent: its lack of self-nature entails a 147 prathamo lakṣaṇenaiva niḥsvabhāvo’paraḥ punaḥ| na svayaṃbhāva etasyetyaparā niḥsvabhāvatā|| 96 complete denial of its reality. The dependent nature is “real” in some sense, but its lack of selfnature is the necessary correlate of its being causally-produced, and therefore impermanent or momentary, just as dharmas are impermanent or momentary. On the other hand, the perfected nature was described in v. 21 as being “permanent,” or more precisely, being the permanent fact that dualistic construction or imputation itself is utterly unreal. In other words, the perfected nature is not momentary/impermanent (Skt. kṣaṇika, Tib. skad cig ma); this is a major ontological difference, which is even further developed in the Commentary to the Ornament of Great Vehicle Sutras (MSABh): If the [perfected] nature (pariniṣpannasvabhāva) is the complete absence in the dependent nature (ātyantikābhāvalakṣaṇa) of this imaginary nature, how is it [perfect] and why is it called [perfect]? - Because it is immutable (avikāra), it is [perfect]. Because it is the object of the purified mind (viśuddhālambana) and the quintessence of all good dharmas (kuśaladharmaśreṣṭha), it rightly called "perfect."148 Clearly, the characterization of the perfected nature as the “quintessence of all good dharmas” strongly suggests that the perfected nature is being construed in a manner that is fundamentally different from the imputed or dependent natures. And indeed, D’Amato informs us that the Ornament consistently equates the perfected nature with ultimate truth or the ultimate nature of reality: In the MSA/Bh, the perfected nature is the only one of the three natures that is posited as ultimately existent. In the commentary to 11.41 of this text, the perfected [nature] is equated with thusness 148 Étienne Lamotte, trans., Mahāyānasaṃgraha: La Somme du Grand Vehicule d’Asaṇga, vol. 2 (trans. Gelongma Karma Migme Chodron), 150-51. 97 (tathatā) or the ultimate nature of reality. And in the commentary to 11.13, the perfected nature is said to be naturally purified (prakṛtyā viśuddha) from defilements, which is itself a characteristic of ultimate reality, as described by the MSA/Bh. And at 9.78, the perfected nature is described as ultimately existent (paramavidyamānatā), while the imagined nature is described as non-existent. Also, the commentary to 19.79 equates the comprehension of the perfected with buddhahood; in that buddhahood may be understood as the awareness of thusness, here the perfected is again related to thusness. Furthermore, the commentary to 20-21.60-61 states that purified thusness (viśuddhā tathatā) is the perfected (niṣpanna), which is the ultimate (paramārtha). So here it can be seen that the perfected nature is brought together with terms for ultimate reality. Therefore, according to the MSA/Bh’s interpretation of the three natures, it is not the dependent nature that is ultimately real or that functions as the substratum of reality; it is the perfected nature that is ultimately real.149 Insofar as “thusness” is universally held to be truly and ultimately real, this would seem to imply a hierarchy or progression, from the lowest of the three natures (the constructed nature), which does not exist at all; through the dependent nature, which exists provisionally or relatively or in the manner of dependent origination; to the perfected nature—the supreme of the three natures— explained to be emptiness or thusness, the sublime object of noble beings, which exists ultimately and “immutably,” unlike the dependent nature. As D’Amato concludes, In the context of the [Treasury of Great Vehicle Sutras], the threenature theory is properly understood not as an ontological model simpliciter, but rather as a soteriologico-ontological model, a model presenting stages of gnosis that must be traversed in order to obtain buddhahood—an attainment which may be interpreted in 149 Mario D’Amato, “Three Natures, Three Stages,” 201-202. 98 terms of a thusness.150 non-conceptually-constructing awareness… of But there is no need to restrict this conclusion to the Treasury; there is good evidence for a “progressive” model of the three truths throughout the earliest strata of Yogic Practice scripture and commentary, even within Vasubandhu’s Thirty Verses. This should not, however, be understood as an argument against the “pivot” model in and of itself. D’Amato rightly points out that this view has much to recommend it as a method for approaching Yogic Practice philosophy. Indeed, in TK 22, Vasubandhu notes that the perfected nature is “neither different nor non-different” from the dependent nature, which is “not observed as long as [the perfected nature] is not observed.”151 Clearly, then, the perfected nature is being understood in some sense as the true form of the dependent nature, as opposed to its false representation as the imputed nature. The problem with the “standard interpretation” is not that it is wrong or inaccurate, but that it is incomplete and, furthermore, that it is not the sole or even necessarily the normative interpretation of the three natures, even in the works ascribed to Asaṇga and Vasubandhu. It should also be noted that the “pivot” model does not necessarily contradict the “progressive” model. Ratnākaraśānti, for one, clearly embraces both hermeneutical frameworks. At one point, he writes that “[Nirvāṇa] results from ascertaining the perfected nature of the otherdependent; [Saṃsāra] results from ascertaining the entirely-imputed nature.” Nirvāṇa and Saṃsāra are held to have the same locus: the dependent nature. Seeing the essential emptiness of 150 Ibid., 206. 151 ata eva sa naivānyo nānanyaḥ paratantrataḥ/ anityatādivad vācyo nādṛṣṭe’smin sa dṛśyate// 99 that locus leads to enlightenment, while being tricked by the dualistic mode of its appearance leads to further rebirths in the wheel of suffering. On the other hand, Ratnākaraśānti’s articulation of three different ontological categories for the three natures—“ultimate existence, substantial existence, and imputed existence”— clearly implies a progression from what is merely “imputed” to exist, through what exists “substantially” but not “ultimately,” to what does in fact “ultimately” exist. Ratnākaraśānti also posits that there are two aspects to the perfected nature, the “unmistaken” and the “unchanging,” and defines them in terms of the end result of the process of abandoning duality: Since they appear dualistically, [phenomena] are said to be the “mark of duality.” However, they come to cessation in transcendent wisdom. This wisdom is authentic and undistorted, thus it is the unmistaken perfected nature. The suchness of that wisdom is the unchanging perfected nature.152 That is to say, dualistic phenomena “come to cessation” in transcendent wisdom, which is explicitly identified with the “unmistaken” perfected nature. Suchness itself is furthermore identified as the other aspect, the “unchanging” perfected nature. This is clear evidence that Ratnākaraśānti subscribed to the “progressive” model of the three natures; and, taken in concert with his earlier characterization of the dependent nature as the locus of both Saṃsāra and Nirvāṇa, demonstrates that, at least as far as he was concerned, the “pivot” model and the “progressive” model were not contradictory. 152 Ratnākaraśānti, Pith Instructions, ff. 226.a.2-3. gnyis ltar snang ba'i phyir gnyis kyi mtshan ma zhes kyang bya ste/ de 'jig rten las 'das pa'i ye shes la 'gag par 'gyur ba dang/ de ni ma 'khrul pa dang yang dag pa'i ye shes yin pas phyin ci ma log par yongs su grub pa'o/ de'i de bzhin nyid ni mi 'gyur ba yongs su grub pa'i ngo bo nyid do/ 100 Perhaps the best way to harmonize these two different accounts is to see the pivot model as emphasizing the intersubjective nature of our common, seemingly-“objective” reality, but the progressive model as emphasizing the increasingly rarefied manner in which a practitioner sees (or, more precisely, no longer sees) that reality. On the one hand, the dependent nature is the dependently-originated flow of the eight collections of consciousness; seeds and imprints in the storehouse consciousness produce the phenomenal experiences of sensory consciousness, which are similar enough for beings with similar karmic conditioning (i.e. similar imprints) that those sensory experiences appear to be experiences of “the same” objects. On the other hand, as a being ascends through the Bodhisattva stages, those seeds are exhausted, so the appearances and experiences of the dependent nature are also gradually exhausted. Early in the process of contemplative development, “emptiness” means the suspension of phenomenological bifurcation; dependently-originated appearances still arise, but the manner in which they are experienced is nondual. Later, as the Bodhisattva attains higher and higher levels of realization, the perfected nature—thusness—is experienced ever more purely and directly, until eventually appearances dissolve into non-appearance. While Ratnākaraśānti does not explicitly equate thusness or the perfected nature with the luminosity of reflexive awareness, it is worth noting that the semantic range of these two are quite close, particularly since both are ultimately explained in terms of transcendent wisdom. But in order to understand the reflexive nature of awareness, particularly in relation to thusness and the ultimate truth, it is first necessary to make a few remarks concerning the means of reliable knowledge (Skt. pramāṇa, Tib. tshad ma). 101 C. The Means of Reliable Knowledge (Pramāṇa) The topic of pramāṇa (Tib. tshad ma) or the “means153 of reliable knowledge” is the last major element of Śāntarakṣita’s Middle Way-Yogic Practice-Pramāṇa synthesis from the Ornament of the Middle Way and Ratnākaraśānti’s commentary in the Pith Instructions. This subject is extremely broad, as the history of the pramāṇa discourse begins in the second century and spans approximately a thousand years of continual exegesis among almost all the different Indian religious and scholastic traditions, both Buddhist and non-Buddhist. In addition to its wide range, pramāṇa literature is also known for its dense and highly technical style. Fortunately for our purposes, Ratnākaraśānti’s arguments in the Pith Instructions usually draw upon the more basic features of the pramāṇa discourse, and are oriented toward his presentation of the reflexive nature of awareness. In order to understand Ratnākaraśānti’s arguments, however, as well as his presentation of the reflexive nature of awareness, it is first necessary to say a few words about the pramāṇa discourse in general. 1. The Indian Context The discourse of the means of reliable knowledge began with the appearance of Akṣapāda Gautama’s Logical Scriptures (Nyāyasūtras) circa 150 CE. In this seminal volume, Gautama laid out the basic governing principles of formal scholastic debate, primarily the structure of logical syllogisms (avayava). Gautama distinguished between the “means of reliable knowledge” (pramāṇa) and the “objects of reliable knowledge” (prameya); in this, he was following the “kāraka system,” a “formulaic way of analyzing the ‘functional elements’ or kārakas that contribute to an action (kriyā).”154 This was a common convention stemming from the tradition of Sanskrit grammatical analysis, as found for example in the 153 Throughout this chapter, and the thesis as a whole, “means” and “instrument” are interchangeable in the context of pramāṇa theory (i.e. “means of reliable knowledge” and “instrument of reliable knowledge” are both translations of pramāṇa). “Instrument” is preferable in certain contexts, as for example when addressing the intricacies and standard translations of Sanskrit grammatical categories; “means” is used elsewhere, in order to avoid stuffy formalism. But this is purely a stylistic choice and does not reflect the meaning of the term in question. 154 John Dunne, Foundations of Dharmakīrti’s Philosophy, 17. 102 Great Commentary (Mahābhāṣya) of Patañjali. Generally speaking, the Sanskrit grammarians identified three karakas: the “agent” (kratṛ), the “instrument” (karaṇa), and the “object” (karman). This analytical framework was adapted by Vātsyāyana, the first to write a commentary on the Nyāyasūtras, such that the act of knowing (pramiti) was understood to have three components: the “knower” (pramatṛ), the “known” (prameya), and the “means” or “instrument” of knowledge (pramāṇa). Future pramāṇa theorists, both Buddhist and nonBuddhist, universally adopted this basic paradigm. There are two crucially important things to understand about the Indian context of pramāṇa literature. The first is that the meaning of “knowledge” in this literature is not identical with the contemporary, inherited Western notion of knowledge. Generally speaking, “knowledge” in contemporary Western epistemology means something akin to “justified true belief”: we have “knowledge” when we believe proposition P(x) to be true, while P(x) is a “true” statement about x—that is to say, a statement which is justifiable on the basis of some demonstrable and “objective” property of x. The central conceit of this theory is the assumption that knowledge concerns propositional content. Participants in the pramāṇa discourse did not share this assumption, or this definition of knowledge. As John Dunne writes, On most Euroamerican accounts, “knowledge” is a belief or attitude that is true (under some set of conditions or truth theory). As a belief or attitude, “knowledge” is dispositional, and it therefore cannot be an act in itself. But on the account of Pramāṇa Theory… “knowledge” (pramiti or pramā) is the act (kriyā) of “knowing indubitably” that is constituted by a process involving the interaction of an agent, instrument, and object of knowledge. This model requires that the “action of knowledge” (pramā or pramiti) be a cognitive event occurring in a particular person’s mind within a particular set of circumstances. A theory of knowledge must therefore take 103 into account any relevant aspect of those circumstances that, for example, might distort a cognitive event in such a manner that we should not consider it knowledge… Physical infirmities such as jaundice or cataracts might distort cognitive events: a person with jaundice will see conch shells as yellow; a person with cataracts thinks that his water-jug is filled with small pieces of hair. They also generally maintain that intense emotions such as anger or lust so strongly affect the mind that all cognitions occurring with those emotions are necessarily distorted.155 This is an absolutely critical point. Rather than defining knowledge strictly in relation to the truth-value of some proposition, pramāṇa theory incorporates the affective disposition of the “knower” into its definition of knowledge. Even a “justified true belief” does not count as knowledge, if there is some type of cognitive distortion (Skt. bhrānti, Tib. ‘khrul pa) affecting the epistemic subject. If a conch shell appears yellow in the perceptual cognition of someone with jaundice, that person’s propositional statement “This conch shell is white” does not have the same truth-value, for a pramāṇa theorist, that it would have for a Western epistemologist. Nor is such “cognitive distortion” to be understood exclusively in relation to the sensory content of perceptual judgments. For Ratnākaraśānti, as for Dharmakīrti and Buddhist pramāṇa theorists generally, the phenomenological bifurcation into subjective and objective aspects of cognition constitutes the most basic form of “distortion,” and is intimately related to the false belief in a self. In fact, for most (if not all) Buddhist pramāṇa theorists, the beginningless cognitive habit of phenomenological bifurcation is the most significant manifestation of the beginningless cognitive habit of false belief in the self, since the false belief in the self manifests experientially just as phenomenological duality. 155 Ibid., 18-19. 104 This point is worth considering in some greater detail. As already discussed, perhaps the central teaching of Buddhism is “no-self” or anātman: that the apprehension of oneself, as a “self,” is a type of cognitive error. In fact, the classical Buddhist definition of the “ignorance” (Skt. avidyā, Tib. ma rig pa) that is the root of suffering is precisely this mistaken belief in the self. Although this mistaken belief is certainly evinced in propositional statements such as “I believe there is such a thing as the self,” it was always understood that correcting this cognitive error required more than simply making the opposite propositional statement—saying “I believe there is no such thing as the self” is very different from actually, experientially realizing the absence of self. Indeed, the purpose of Abhidharma meditation was precisely to arrive at such an experiential realization; self-grasping (Skt. ātmagrāha, Tib. bdag ‘dzin) was always cast as a nonconceptual cognitive error, which could only be remedied through correct practice. In keeping with this understanding, one of the key features of Buddhist pramāṇa theory was an explicit relationship between the nonconceptual cognitive error of self-grasping, and the type of nonconceptual cognitive distortion (bhrānti) that prevents a knowledge-act from being epistemically reliable. But unlike classical Abhidharmikas, Buddhist pramāṇa theorists expressed the ignorance of self-grasping in phenomenological terms, as the duality of subject and object. In effect, the most important type of cognitive distortion—the error that conditions each and every moment of consciousness for ordinary, unenlightened beings, preventing those consciousnesses from serving as sources of reliable information—is phenomenological bifurcation. Needless to say, this runs wildly contrary to contemporary Western traditions of epistemology, to say nothing of Husserlian phenomenology. Classically, in post- “Enlightenment” Western epistemology, the mode of epistemic access to phenomena has little or 105 nothing to do with whether or not that epistemic access constitutes “knowledge,” such “knowledge” being essentially a matter of correspondence between some proposition and an objective state of affairs. Furthermore, in typical Western epistemological paradigms, there is nothing “distorted” about dualistic cognition, and in Husserlian phenomenology subject-object intentionality is held to be the only possible type of epistemic access to anything at all—indeed, the only possible type of consciousness at all. In addition to the radically different conception of what constitutes knowledge—first and foremost, that knowledge is an “act of knowledge” (pramiti) which requires an “agent of knowledge” (pramatṛ) free from cognitive distortion—there is, additionally, a second major difference between mainstream Western epistemological traditions and Buddhist pramāṇa theory. This is the relationship between philosophy and grammar. Although the kāraka system was primarily grammatical in origin, the line between grammar and philosophy in Sanskrit was never particularly clear, especially during the period in which the Nyāyasūtras first appeared. This was roughly the same period as the Root Wisdom of the Middle Way, which directly engages the tradition of Sanskrit grammatical philosophy at several points; its second chapter is notoriously inaccessible without at least a working knowledge of the Sanskrit grammaticalphilosophical tradition. The essential point is that, for those who participated in the Sanskrit grammatical discourse, especially the non-Buddhists, grammatical categories such as the kārakas were real entities. The focus on the “instruments” of knowledge accordingly derived from philosophical considerations that were deeply informed by the role of the instrumental case in Sanskrit grammatical inflection. Therefore, in order to understand the epistemology of pramāṇa theory, it is necessary to understand the role and function of the instrumental case. Most basically, the 106 instrumental case denotes that by means of which an action is performed. In standard Sanskrit grammatical philosophy, the instrument determines the type of action, as both the relationship between the agent and the object, and the action performed, changes in direct dependence upon the instrument employed. Just so, in pramāṇa theory, the “knowledge-instrument” (pramāṇa) determines the quality of the knowledge-act (pramiti). As Dunne writes, The comparatively early works of Vātsyāyana and Uddyotakara… claim that the emphasis on analysis of the instrument of knowledge derives from its primacy in the process of knowing. To use the analogy of a person cutting a tree with an axe: the person and the tree can be identified as the “cutter” and the “cut object” only when the action of cutting occurs; and that action can only occur when a cutting instrument—the axe—is employed. It is only by changing the type of instrument used that the action then becomes a different action. That is, if we replace the agent with some other person, or if we can direct the axe against some other object, the action is still the action of cutting. In short, neither the agent nor object can change the character of the action. If, however, some other kind of instrument, such as a yardstick, is used, then the agent (“the cutter”), object (“that which is cut”) and action (“cutting”) all take on a different character: they become the “measurer,” the “measured” and the action of “measuring.” Hence, inasmuch as the character of the instrument determines the character of the other three factors, the instrument is primary. This way of understanding the instrument as primary appears to have been widely accepted among Pramāṇa Theorists, including Dharmakīrti.156 Similarly, for all pramāṇa theorists, the quality of a knowledge-act is finally determined by the quality of the knowledge-instrument: the pramāṇa. The question of whether or not a given cognition yields epistemically reliable information becomes a question of whether or not the “instrument” of the knowledge-act is epistemically reliable, that is, whether or not the 156 Ibid., 20. 107 knowledge-instrument is in fact a valid (as opposed to an invalid) knowledge-instrument. Hitting a tree with an axe is cutting; hitting a tree with a dead fish is not. Just so, employing a valid pramāṇa is knowing, while employing an invalid pramāṇa—which is to say, employing something that is not, in fact, a pramāṇa—is not. The central questions of the pramāṇa discourse therefore turned on what, precisely, constituted an instrument of reliable knowledge. Some participants in this discourse argued that the Vedas were the supreme means of reliable knowledge; others, including some Buddhist pramāṇa theorists (although, notably, not Dharmakīrti) considered their own, non-Vedic scriptures to be means of reliable knowledge, or considered the testimony of the Buddha or other authorities to be a means of reliable knowledge. Everyone, however, agreed upon two basic means of reliable knowledge: perception (pratyakṣa, Tib. mngon sum) and inference (anumāna, Tib. rjes dpag). Many Buddhists accepted only these two. For his part, Ratnākaraśānti clearly embraces this two-instrument model. Furthermore, apart from a single mention, in the Pith Instructions he does not comment upon inference at all, far less delve into any of the inscrutably technical arguments concerning this contentious topic (these he reserved for a separate text, the Justifying Inner Pervasion). Instead, he focuses more or less exclusively on direct perception, the next topic for consideration. 108 2. The Pramāṇa of Direct Perception a) Perception in the Abhidharma In order to explain the perspective of Buddhist pramāṇa theorists such as Dignāga and Dharmakīrti, it will be helpful to first consider how perception was treated differently in the system of pramāṇa as compared with classical Abhidharma notions. As discussed above, Abhidharma literature describes the typology of phenomena, particularly in terms of the most basic phenomenal constituents of reality. It did not usually include anything like a systematic, consistent, or even necessarily coherent treatment of epistemology. More relevantly to the present discussion, most formulations of the Abhidharma (and commentaries thereupon) did not include anything like a detailed account of the perceptual process, even where it did articulate the “most basic constituents” of perception. One of the main Abhidharma schematics was that of the eighteen “elements” (Skt. dhātu, Tib. khams). These are the six sense-faculties (Skt. indriya, Tib. dbang po) which allow contact (Skt. sparśa, Tib. reg pa) with the six types of sense-objects—visible form, tactile form, and so on—giving rise to the six types of consciousness (Skt. vijñāna, Tib. rnam shes). Another common scheme was that of the twelve āyatanas, or the six sense-faculties together with the six types of sense-object. However, it is not at all clear how this model accounts for perception. Abhidharma literature typically treats each of these eighteen “elements” in great detail, listing every possible facet of all eighteen: for example, every different possible kind of visual, auditory, or tactile sense-object. While some Sautrāntika-Sarvāstivāda commentators such as 109 Śrīlāta157 did make attempts to sketch out the mechanism or mechanisms by which sense-faculty contacts sense-object, giving rise to sensory cognition, in general the Abhidharma literature was vastly more concerned the ontology of the elements of perception, rather than the epistemology of how those elements interact to produce knowledge. In non-Buddhist traditions, the tendency was to see the sense-faculties in and of themselves as the means by which reliable knowledge about the external world is attained, i.e. as pramāṇas. Here, it is important to recall the kāraka system. Implicit in the claim that the sensefaculties are the karaṇa (the instrument) is the claim that the one who possesses those sensefaculties is the kratṛ (the agent). Apart from its simplicity and intuitiveness, from a non-Buddhist perspective this formulation had the additional benefit of supporting the existence of the ātman or self. The eleutheriology of Vedantic religion, for example, centers around the relationship between one’s own ātman with the transcendent ātman of Brahman, the Absolute; liberation (Skt. mokṣa, Tib. thar pa) results from unifying the one’s own ātman with the ultimate ātman. In this case, one’s own ātman is identified as the proper agent, the kratṛ, of one’s actions. Thus, if one’s own sense-faculties are the karaṇa/pramāṇa (instrument) of the pramiti (knowledge-act), it logically follows that the ātman (self) is the kratṛ/pramatṛ (agent)—exactly the Vaiśeṣika (i.e. Vedic/Vedantic) eleutheriological point Akṣapāda Gautama was making in the Nyāyasūtras. This was, obviously, a view that the Buddhists could not accept, insofar as it is maintained by the Buddhist tradition that the self does not exist, and hence that the means of truly reliable knowledge could not be anything which acknowledged or entailed the existence of such a self. For Buddhist pramāṇa theorists, then, admitting that the sense-faculties are in and of 157 Venerable K.L. Dhammajoti, “Ākāra and Direct Perception (Pratyakṣa),” Pacific World: Journal of the Institute of Buddhist Studies 3, no. 9 (2007): 245-49. 110 themselves means of reliable knowledge is tantamount to admitting that there is an individual agent or self—the one who possesses those faculties—for whom the sense-faculties function in this “instrumental” capacity. It was therefore incumbent upon Buddhist interlocutors to demonstrate just how one could attain reliable knowledge, without positing the existence of a truly-existing self: without, in other words, arguing that the sense-faculties themselves are the means of reliable knowledge. This necessitated the articulation of what amounted to a new theory of perception. b) Buddhist Pramāṇa Theory Buddhist contributions to the formal pramāṇa discourse began with the Pramāṇasamuccaya of Dignāga (ca. 480-540), but reached their full flowering with the Pramāṇavārttikakārika of Dharmakīrti. Space prohibits an extensive consideration of Buddhist pramāṇa theory, particularly in its historical dimensions, e.g. the extent (if any) to which Dharmakīrti’s theory differed with Dignāga’s, the various lineages of commentary on Dharmakīrti, and so on. For our purposes, it is sufficient to provide a general sketch of Dharmakīrti’s perspective, since Ratnākaraśānti clearly embraces this perspective in its broad outlines. Furthermore, with the sole exception of the reflexive nature of awareness—which will be addressed shortly—he does not concern himself with describing the intricate mechanics of perception. In effect, Ratnākaraśānti takes Dharmakīrti’s account of perception as background knowledge, and so that is the manner in which it will be treated here. In brief, then, Dharmakīrti’s solution to the problem outlined above was to shift the discussion away from the sense-faculties. Rather than maintaining a direct connection between 111 sense-faculties and the objects of perception, Dharmakīrti instead followed the SautrāntikaSarvāstivāda in positing that sensory contact produces a phenomenal “image” (Skt. ākāra, Tib. rnam pa) in consciousness, the features of which depend upon the characteristics (Skt. lakṣaṇa, Tib. mtshan nyid) or causal powers of the object observed. As Matthew Mackenzie writes, Dharmakīrti defended a basically representationalist theory of perception. According to Dharmakīrti, the object of perception leaves its ‘mark’ on awareness, and awareness thereby captures or reflects the form of the object. Hence, awareness takes on a certain structure or aspect (ākāra) due to a causal link between an object and a sense organ.158 For example, phenomena—specifically, dharmas—only have the characteristic of being blue inasmuch as they possess the power to cause the perception “blue.” On this model, when there is a perception of a blue patch, what is being perceived is not the particles of blue, nor the blueness of the particles, but the sensory-cognitive image of “blue” that arises in consciousness as a result of the interaction of sense-faculty and the relevant causal properties (viz., being able to give rise to a perception of “blue”) of the phenomena in question.159 This image, not the sense-faculty, serves as the means by which reliable knowledge is attained; since all the sense-data of a particular sensory cognition are contained in the cognitive “image” produced by the interaction between sense-faculty and object of sensation, anything beyond or besides that image is unnecessary in order to obtain reliable sense-data. 158 Matthew Mackenzie, “The Illumination of Consciousness: Approaches to Self-Awareness in the Indian and Western Traditions,” Philosophy East and West 57, no. 1 (2007), 47. 159 “Phenomena” (for dharmas, plural) is the correct term here, as opposed to “phenomenon” (for dharma, singular). Ordinary beings can never have sensory access to individual dharmas; the image that is produced in the mind of an ordinary being results from the causal properties of multiple dharmas acting in concert, giving rise to the illusion that there is a single distributed entity—as opposed to a multiplicity of infinitesimal particles—being perceived. Cf. Dunne (2004, 100-13) for greater detail on this point. 112 Mackenzie’s point about “representationalism” is generally accurate. Ratnākaraśānti describes this perceptual theory, shared by several schools of Buddhist philosophy, in terms of the relationship between the object of perception and its “reflection” (gzugs rnyan, *pratibimba) in consciousness. However, it is also worth specifying that the cognitive image only “reflects” the causal properties of what is perceived. Even more precisely, the image is not a “reflection” at all, so much as it is an effect that is produced as a result of the causal powers or properties of what is perceived, in combination with the causal properties (particularly and especially the causal conditioning; recall the yellow conch) of the sensory faculty. It is certainly not the case that the cognitive image of e.g. a blue circle is a “representation” of a really-existing blue circle, since what the image captures is not the form of what is perceived, only the causal structure of its components. The cognitive image of a blue circle does not “represent” a blue circle. It only re-presents those features of a particular causal complex which have the capacity to produce the perceptual judgment “blue circle,” in a manner that is directly accessible by sensory consciousness (which the causal complex itself is not). But the causal complex itself is neither blue nor circular. Furthermore, according to Dharmakīrti, the production of the phenomenal image in the mind of the observer is only the first step in a three-stage process. The second step ordinarily occurs directly following the first; at this stage, the phenomenal image is automatically and subliminally processed into a useable form, specifically by being represented in consciousness as the “objective aspect” (Skt. grāhya-ākāra, Tib. gzung rnam) of a dualistic cognition. The reason this image must be processed is that it is, in the first moment, nondual or nonconceptual, and so remains inaccessible to ordinary (i.e. dualistic) cognition. As Dreyfus writes, 113 Like in other forms of representationalism, Dharmakīrti’s perception has access to external objects only via the mediation of directly apprehended objects, the aspects [i.e. images, ākāra]. These aspects are perceptual objects with which we have direct acquaintance, not unlike Locke’s ideas. But whereas Locke holds these ideas to stand for external objects and their apprehension to be by themselves cognitive, Dharmakīrti posits a more complex relation between direct and external objects. For him, the apprehension of these directly given objects is not by itself cognitive, because perception is contentless. Perception passively holds its object without determining it and merely induces appropriate conceptualizations. For example, we sense a blue object that we categorize as blue. The perceptual aspect (the blue aspect) is not yet a representation, since its apprehension, the perception of blue, is not yet cognitive. Only when it is interpreted by a conception does the aspect become a full fledged intentional object standing for an external object.160 In other words, it is only by being represented in consciousness as the “objective aspect” of a cognition that an ordinary observer has epistemic access to the information contained within the phenomenal image. Such cognitive processing takes place subconsciously, and is not ordinarily noticed as an element of the perceptual process. In the above passage, Dreyfus refers to the pre-cognitive contents of sense perception as “directly given.” There is some question, however, as to the precise sense in which these contents are “given,” since they are not epistemically accessible as initially present. Sensory information can only be accessed subsequent to cognitive processing, which is by definition conceptual. Therefore, conceptuality is deeply ingrained in the ordinary perceptual process, since 160 Georges Dreyfus, Recognizing Reality, 337. Emphasis added. Cf. also 312: “Since it accurately reflects reality and since reality is reducible to bare particulars, perception cannot provide any cognitive articulation and boils down to a passive encounter with things in their momentariness.” And 352-53: “It is important to keep in mind that Dharmakīrti does not claim that bare contact with reality, which happens in perception, can explain human knowledge on its own.” 114 without such conceptual processing, ordinary beings would be unable to form what Dreyfus calls “determinations,” also called “definitive ascertainments” (Skt. niścaya, Tib. nges pa), of the contents of sense-perception. These “definitive ascertainments” are correct judgments about the objects of perception—for example, the correct perception of a red-patch as a fire. If that same red-patch were perceived as a bouquet of poppy flowers, the sensory content would be the same red-patch, but there would not be any definitive ascertainment. Thus, as Sara McClintock writes, Although perception itself is nonconceptual, it still serves as the primary, though not exclusive, cause for the conceptual judgments that arise from it. At the same time, perception qua means of trustworthy awareness, or pramāṇa, is said to occur only when a corresponding correct judgment arises directly or indirectly as its result. Now, only those perceptual judgments will arise for which the conditions exist in the mindstream wherein the perceptual awareness occurs. In this sense, we only see (or know) what we are already conditioned to see (or know).161 Thus the ability of the perceptual faculties to produce reliable knowledge, for ordinary beings, is determined by the karmic and cognitive conditioning of the particular being in question. Two observers can perceive the same red-patch, but form radically different judgments about the nature of the red-patch: one sees poppies, while another sees fire. To repeat, the conditioning of the being in question is an essential element of this process;162 perhaps the one loves flowers, or the other is a pyromaniac. Conditioning also determines whether or not, or which parts of, the contents of sensory consciousness are epistemically accessible. Depending upon the respective conditioning of two beings, one can form the definitive ascertainment of a 161 Sara McClintock, “The Role of the Given,” 129. Emphasis original. 162 See Dunne (2004, 184-185) for an illuminating, if somewhat salacious, example. 115 red-patch, while the other does not consciously perceive a red-patch at all. In the latter case, the red-patch is perceived, insofar as it produces sensory content in accord with its causal capacities; but it is not observed, insofar as the contents of the produced sense-perception never reach the threshold of liminal—which is to say, conceptual—consciousness. Crucially, the “objective aspect” never arises alone, but is always conjoined with a “subjective aspect” (grāhaka-ākāra, Tib. ‘dzin rnam). This is the sense of being a “self” or the “one who is perceiving.” Deeply embedded within every instance of an ordinary being’s stream of perception is a subtle, yet constant, sense of being at the <0, 0, 0, 0>-point of a fourdimensional Cartesian coordinate system, the center of an entire perceptual universe. Although the language used is more explicitly physicalist (i.e. in terms of “brains” as opposed to “minds”), this is something like the position articulated by Antonio Damasio in his landmark work on cognitive science and perception, The Feeling of What Happens: As the brain forms images of an object—such as a face, a melody, a toothache, the memory of an event—and as the images of the object affect the state of the organism, yet another level of brain structure creates a swift nonverbal account of the events that are taking place in the varied brain regions activated as a consequence of the object-organism interaction. The mapping of the object-related consciousness occurs in first-order neural maps representing the proto-self and object; the account of the causal relationship between object and organism can only be captured in second-order neural maps.163 What Damasio calls the “proto-self” is, more or less, the “aspect of the grasper.” Ordinarily, in the cognitive mode called the “awareness of objects” (Skt. arthasaṃvit, Tib. don rig),164 this 163 164 Antonio Damasio, The Feeling of What Happens (Boston: Harcourt, 1999), 170. The other cognitive mode is “awareness of awareness” (Skt. jñānasaṃvit, Tib. shes rig), in which the cognitive image is no longer construed as an external object, but is recognized as being just cognitive. See below. 116 “subjective aspect” goes unnoticed, as the “objective aspect” is simply taken to be the object itself, and all other considerations—specifically, the fact that what is being accessed is not the object in and of itself, but a cognitive “representation” or phenomenal “image” of the object, as well as the fact that this representation qua “objective aspect” arises simultaneously with the “proto-self,” the “subjective aspect” or sense of being an observer—are pushed to the cognitive background. Technically, this picture is still incomplete, as the formation of the subjective- and objective-aspects is only the second stage of the perceptual process according to Dharmakīrti. The third stage, however, is extremely complicated, and involves the much-debated process of “other-exclusion” (anyāpoha, Tib. gzhan sel), which Ratnākaraśānti does not comment upon in any extensive way, and is thus irrelevant for the present purposes. With respect to apoha, it should suffice to note briefly that the “objective-aspect” still contains all the sense-data or causal information contained in the initial phenomenal image. While there is a measure of epistemic access to the “objective-aspect,” ordinarily this aspect will undergo a further round of subliminal cognitive processing, in which only those causal features which are relevant for the purpose of some particular identification or practical goal are picked up by supraliminal consciousness, by “excluding” (Skt. apoha, Tib. sel) all the “other” (Skt. anya, Tib. gzhan), irrelevant features. In other words, phenomena are identified to possess certain properties,165 only by excluding all of the other causal properties they possess from perceptual consciousness. Thus, in rough terms, the “blueness” of a large blue sphere can only be perceived by excluding its size and shape from supraliminal perceptual judgment, while its “sphericity” can only be perceived by excluding its color and size from supraliminal perceptual judgment, and so on. But this example is only 165 Technically, property-universals (Skt. ārthasāmānya, Tib. spyi don). 117 approximate since, as noted, apoha is a highly technical subject.166 As Ratnākaraśānti by and large only concerns himself with a discussion of phenomenal images and the twin aspects of the second stage of perception, the analysis of perception in Buddhist pramāṇa theory will accordingly be limited to these topics, and nothing more will be said concerning apoha. c) External Realism vs. Epistemic Idealism To review, for Dharmakīrti and all subsequent Buddhist pramāṇa theorists, very much including Ratnākaraśānti, it is the phenomenal image produced in consciousness that serves as the means by which knowledge is gained. And since the sense-faculties are only one causal condition for the pramāṇa, rather than being self-identical with it, there is no need to posit the existence of a truly-existent self as the agent of perception. But Dharmakīrti’s assertion that cognitive “images” are the means of reliable knowledge had another important consequence. At a certain level of discourse, or from a certain perspective, it is assumed that the phenomena—i.e. the dharmas—which produce these images exist materially and outside of the mind. At this level of discourse, which in keeping with common doxography167 Ratnākaraśānti 166 Those interested are advised to consult Mark Siderits, Tom Tillemans, and Arindam Chakrabarti, Apoha: Buddhist Nominalism and Human Cognition (New York: Columbia University Press, 2011). Cf. also Lawrence McCrea and Parimal Patil, Buddhist Philosophy of Language in India: Jñānaśrīmitra on Exclusion (New York: Columbia University Press, 2010). 167 It is important to note at the outset of any discussion of the “Sautrāntika” school that this designation appears to be largely a product of the classification schemes of later Indian doxographers. Lamotte (1988, 526) explains that “The Sautrāntikas… represented a philosophical movement rather than a homogeneous sect: up until now the existence of Sautrāntika monasteries has not been attested by any inscriptions.” Nevertheless, key features of reflexive awareness can be traced back to “Sautrāntika”-style epistemological analysis. As is well known, the pramāṇa theory of both Dignāga and Dharmakīrti is heavily indebted to Sautrāntika epistemology, so much so that Tsong kha pa, the founder of the dGe lugs school of Tibetan Buddhism, considered Dharmakīrti to be a Sautrāntika as opposed to an adherent of the Yogic Practice tradition (despite the latter’s use of the concept of the storehouse consciousness). 118 associates with the tradition of the Sūtra Followers (Skt. Sautrāntikas, Tib. mdo sde pa), the internal image is like a “reflection” of the external phenomena, meaning that it possesses features in accordance with the causal capacities or characteristics of those phenomena. Dunne terms this perspective “External Realism,” the view “that there are entities—most notably the objects of perception—that exist external to any mind.”168 However, given that perceptual awareness or sensory cognition only ever has access to these images, it is quite natural to apply Occam’s Razor and ask if external phenomena are necessarily required in order to account for perception. That is to say, since there is never any direct epistemic access to external phenomena, and they are only involved in the process of perception insofar as they are the putative cause which gives rise to sensory-cognitive images, anything else which is capable of producing a sensory-cognitive image might fulfill this role just as well as external entities. For example, in the Yogic Practice theory of Vasubandhu, “cognitive representations” (Skt. vijñapti, Tib. rnam rig) arise from the interaction of psychological imprints (Skt. vāsanā, Tib. bag chags) with consciousness. These imprints, always tainted by ignorance, are stored as “seeds” in the storehouse consciousness. When the seed “ripens,” a sensory-cognitive image corresponding with the causal properties of the seed is produced, just as in the External Realist model the properties of the sensory-cognitive image correspond with the properties of the external causal complexes. The key point here is that in both models, only the sensory-cognitive image is ever perceived; the sole difference lies in whether the causal complex that produces sensory perceptions is asserted to be external dharmas, or nothing other than consciousness in the form of psychological imprints or karmic seeds. This is, at bottom, a difference in ontology; but, 168 John Dunne, Foundations of Dharmakīrti’s Philosophy, 58. 119 by accepting that there is never any direct access to the objects of perception, the External Realist has already conceded the epistemological point at stake: it is a priori impossible for an observer to distinguish between sensory cognitions caused by external entities and sensory cognitions caused by the interaction of psychological imprints with sense-consciousness. In the latter Epistemic Idealist model, phenomena appear to have objective properties independently of the subject observing them, but this is only an illusion owing to the strongly intersubjective nature of the storehouse consciousness. Similar types of beings possess similar karmic “seeds,” producing similar sensory cognitions under similar circumstances. Moreover, as mentioned above, these psychological imprints are always conditioned by ignorance. In Buddhist pramāṇa theory, ignorance manifests as phenomenological duality. Thus for Dignāga and Dharmakīrti, who follow Vasubandhu’s idealistic epistemology, the reason phenomena appear to be external is because the sensory cognition of those phenomena is deeply structured by ignorance; properly speaking, the object of perception is only the seed or imprint, but the ignorance which always and everywhere conditions those imprints results in the cognitive distortion of duality, concomitant (or simply identical) with a mistaken sense that “what is perceived” lies outside the mind. As Dunne writes, The erroneous belief in the existence of extra-mental matter is eliminated through the realization that the subject/object duality apparent in awareness is actually due to the influence of ignorance (avidyā). As such, that duality is erroneous, and any determinations based upon it, such as the notion that the cause of the objective appearance [i.e. the “aspect of the grasped” or grāhya-ākāra] in sensory awareness is due to extra-mental particles, is also false.169 169 Ibid., 59. 120 Dunne makes a further point of noting that “Dharmakīrti’s critique of extra-mental entities arises in the context of determining what it is that we know in perception.” The result of this critique is, according to Dunne, an “Epistemic Idealism” (i.e. a claim about what is knowable), instead of an ontological or metaphysical idealism (i.e. a claim about what is real). But this point should be examined in closer detail. Generally speaking, contemporary scholarship has resolved the dichotomy between External Realism and Epistemic Idealism in Dharmakīrti’s philosophy by framing his presentation in terms of a “sliding scale of analysis,” whereby Dharmakīrti spends most of his time arguing from a Sautrāntika External Realist position, only to ultimately abandon this position in favor of Yogācāra-style epistemic (and possibly ontological) idealism. This shift is more explicitly present in the works of Dignāga, when Dignāga argues that the cognitive image, rather than the external object (Skt. artha, Tib. don), is properly speaking the object of knowledge (prameya, Tib. gzhal bya).170 This contrasts with the Sautrāntika position, held by Dharmakīrti at a lower level on the “sliding scale,” that the external object itself is the object of knowledge, while the image is only the “instrument” or “means” (pramāṇa, Tib. tshad ma) by which reliable knowledge about the external object is attained. In theory, even given that there is only ever epistemic access to cognitive images, it may still be possible to posit the existence of extra-mental entities. On such a reading, what is being refuted or denied is not the existence of external entities as such, but rather the possibility of epistemic access to those hypothetical extra-mental entities. In practice, however, truly extra170 Dignāga, Compendium of Pramāṇa (Pramāṇasamuccaya), ed. Ernst Steinkellner, “Dignāga’s Pramāṇasamuccaya, Chapter 1: A hypothetical reconstruction of the Sanskrit text with the help of the two Tibetan translations on the basis of the hitherto unknown Sanskrit fragments and the linguistic materials gained from Jinendrabuddhi’s Ṭīkā,” PS 1.1.10: yadābhāsaṃ prameyaṃ tat pramāṇaphalate punaḥ| grāhakākārasaṃvittyos trayaṃ nātaḥ pṛthak kṛtam. Cf. also Tola and Dragonetti, “Dignāga’s Ālambanaparīkṣāvṛtti.” 121 mental entities—that is, entities which lie wholly outside the bounds of cognition, and cannot ever give rise to a mental representation or cognitive image—are a priori epistemologically inaccessible, and therefore serve absolutely no function. While hardly a strong metaphysical argument against their existence, the principle of ontological parity demands that phenomena which are and must always remain outside the bounds of knowledge, even inferential knowledge, are best considered nonexistent. Moreover, as Gold elucidated, the specific object of the Yogācāra critique of external objects is not just externality but the entire sensory-cognitive framework of dualism, very much including the “physical” bases of perception, the sense-faculties.171 Properly speaking, phenomenal images do not arise as the result of either external or internal causes, since the internal/external opposition itself is nothing more than a deluded construction or imputation. “What appears” is only the dependent nature, the causal interactions of dependent origination qua the eight collections of consciousness, which is ultimately lacking the imaginary duality of the constructed nature. Appearances are causally produced, and are thus “substantially existent” (rdzas su yod pa, *dravyasat), but the internal/external dichotomy is not substantial in the same way—it is only “imputedly existent” (brtags su yod pa, *prajñaptisat). Clearly it is selfcontradictory to assert that the notion of externality as such is incoherent, but that particles or other material phenomena nevertheless exist in a manner that is entirely independent of the mind. 171 It is also worth noting that classical Abhidharma theory considers the sixth “mental consciousness” (Skt. manovijñāna, Tib. yid kyi rnam par shes pa) in the same terms as the other five sensory consciousnesses, viz. as arising due to the contact between a sense-faculty and an object of sensation. Thus the refutation of the mental faculty (Skt. mano-indriya, Tib. yid kyi dbang po) was a refutation of the physical basis of mental consciousness— that is to say, roughly speaking, the brain. Physicalist reductionist arguments against idealism (or the existence of the mind more generally) usually turn on the success of neuroscience in explaining mental processes. But even supposing that a 1:1 correspondence between neurological and mental events (which has yet to be demonstrated) exists, the point here is that even the brain does not exist as extra-mental matter. 122 In this way, Ratnākaraśānti embraces an idealistic ontology that implicitly follows from his idealistic epistemology, irrespective of his metaphysical commitments. And for Ratnākaraśānti, as for any and every other Yogācārin, Epistemic Idealism is not a matter of asserting the primacy of some phenomenal subject; subjectivity is every bit as much on the chopping-block as objectivity. The key point is that the true nature of the mind is nondual; the causes of perceptual events are neither internal nor external, because the entire framework of internal vs. external causes for perceptions is fundamentally mistaken. There is, accordingly, no meaningful way to separate the epistemological claim that the only knowable entities are mental representations (Skt. vijñaptimātra, Tib. rnam rig tsam) from the ontological claim that the only real entities are mental (Skt. cittamātra, Tib. sems tsam), since any such separation presupposes the very opposition that is being refuted. This is perhaps the most striking instance of fundamental irreconcilability between the Western and Buddhist traditions of philosophy. In the contemporary Western tradition, ontological questions are rigidly distinguished from epistemological and phenomenological questions; any slippage between or among categories is generally considered sloppy, and leaves the philosopher open to critique. In the Buddhist tradition, however, such distinctions often serve essentially no purpose. Thus even the attempt to rigidly distinguish ontology, phenomenology, and epistemology in works of Buddhist philosophy frequently leads into hermeneutical aporias. Of course, later opinions differ greatly as to whether Dharmakīrti actually promulgated ontological idealism, or a closed system of epistemic idealism in which it is only ever possible to have access to mental representations. However, as Dunne points out, Dharmakīrti, like Vasubandhu, certainly posits that sensory perceptions only occur due the presence of prior psychological imprints in the storehouse consciousness, in other words that the entire sensory 123 world can—and, ultimately, can only—be explained solely in terms of mental events. Therefore, in the Epistemic Idealist model, the psychological imprint of beginningless ignorance is responsible for the phenomenological bifurcation of ordinary experience. But it is also responsible for the content of perception, since the phenomenal characteristics of sensorycognitive images are nothing other than similar imprints, stored in the “storehouse” consciousness. As Ratnākaraśānti explains, Even though, for example, a conceptualized blue-patch172 may exist, the characteristic “blue” does not exist, because it cannot withstand logical analysis. There is cognitive distortion due to the contaminating force of the “blue” psychological imprint. Since it arises in this way, the experience is distorted, and it is experienced as though one were experiencing something else.173 Thus the “psychological imprint of ‘blue’” is inextricably intertwined with dualistic “distortion,” and is furthermore responsible for the perception of blue patches. In other words, the “psychological imprint of ‘blue’” (or any other phenomenal characteristic) is already dualistic. The phenomenological structure of subject/object dualism is deeply embedded in any and all “psychological imprints,” which are themselves necessary correlates of not having yet been enlightened or liberated from cyclic existence; in the standard Yogācāra Buddhist account, liberation means liberation from psychological imprints or habits. 172 The word lus here in the Tibetan translation can have many different meanings in Sanskrit (and English). In this context, Ratnākaraśānti is referring to an entity—probably some type of extended entity—characterized by being blue. Since this is the function performed by the term “patch” in contemporary philosophy, the existing terminology has been adopted and applied to the Pith Instructions. 173 Ratnākaraśānti, Pith Instructions, 225.b.7-226.a.1. sngon po la sogs pa'i rnam par rtog pa'i lus ni yod pa'o/ sngon po la sogs pa'i mtshan nyid ni med pa ste/ ji skad du 'chad par 'gyur ba'i gnod pa yod pa'i phyir ro/ de bas na sngon po la sogs pa'i bag chags kyis bslad pa'i dbang gis 'khrul pa'o/ de ltar gyur pas de myong ba yang 'khrul pa dang gzhan myong ba lta bur myong ba'o/ 124 This is an absolutely crucial point, since—like Nāgārjuna, Asaṇga, Vasubandhu, Dignāga, Dharmakīrti, and every other contributor to the Mahayana philosophical tradition— Ratnākaraśānti’s main purpose in writing the Pith Instructions is helping others attain liberation. To that end, as previously mentioned, he structures his presentation in terms of the Four Yogas. The First Yoga is the recognition of the “two extremes of entities,” existence and nonexistence or phenomenal subject and object. The recognition that all phenomena are only mind (Skt. cittamātra, Tib. sems tsam) or only mental representations (vijñaptimātra, Tib. rnam rig tsam) is the second of the Four Yogas. In other words, the idea that e.g. a blue patch is not an extra-mental entity only holds at a provisional level of analysis. At the next level, the Third Yoga, the practitioner moves beyond even “mind-only.” This is the recognition of suchness; at this level, according to Ratnākaraśānti, there is no longer any differentiated phenomenal content, only the nondual, undifferentiated luminosity of reflexive awareness. As will become clear, for Ratnākaraśānti, the reflexivity of awareness is the pramāṇa by which epistemic access to the ultimate object of knowledge—suchness, the nature of reality—is granted. This is a major shift, as at the level of the Second Yoga, the pramāṇa is asserted to be the sensory-cognitive image (although, in a trivial sense, even at this level the pramāṇa is ultimately reflexive awareness; see below). In other words, while the cognitive image of “blue” is a means of reliable knowledge at the second stage of Yogic Practice, it is not a means of reliable knowledge at the third stage of development. At this later stage, only the reflexive nature of awareness is a means of reliable knowledge. Finally, according to Ratnākaraśānti, the Fourth Yoga is non-appearance. Ratnākaraśānti does not elaborate on this formulation, but it is clear that he means the absence of any cognitive images, insofar as these cognitive images are necessarily “contaminated” (bslad) by ignorant 125 distortion, specifically the distortion of phenomenal duality. This seems to entail the lack of any and all phenomenal content, although following the Descent into Laṇka Sūtra he maintains that at this level the yogi “sees the Great Vehicle.” Presumably, Ratnākaraśānti did not intend to assert that the Fourth Yoga is merely a blank void, particularly since he critiques Candrakīrti’s idea of Buddhahood as precisely such a blank void. However, these issues will be addressed further on. 3. False vs. True Images As mentioned above, Ratnākaraśānti denies that the cognitive “images” are ultimately reliable means of knowledge or pramāṇas. At a relative level, in terms of gaining more or less reliable information about a world that is provisionally accepted to exist outside of the mind of the observer, the image does serve as a pramāṇa. However, insofar as the production of the image is always concomitant with phenomenological duality, the image itself bifurcated into a subjective and an objective aspect, each and every image is contaminated with ignorance. If the image were ultimately epistemically reliable, its dualistic structure would have to capture some real feature of the world. But, according to Ratnākaraśānti, this dualistic structure is just an imaginary construction that has no basis whatsoever in reality. Given the play between his ontological and epistemological stances, this is equivalent to the view that images are not actually real. Indeed, in the literature this position is variously termed “False Imagism”174 (alīkākāravāda) and “Non-existent Imagism” (Skt. nirākāravāda, 174 Yuichi Kajiyama, “Controversy between the Sākāra- and Nirākāra-vādins of the Yogācāra School—Some Materials,” Journal of Indian Buddhist Studies 14, no. 1 (1965): 429. This is sometimes translated as “False 126 Tib. rnam pa med par smra ba), as opposed to “True Imagism” (satyākāravāda) or “Real Imagism” (Skt. sākāravāda, Tib. shes pa rnam pa dang bcad par smra ba) or the view that sensory-cognitive images are real and undeceptive.175 Crucially, however, the unreality of cognitive images was only held at the level of the ultimate truth, the perfected nature. The cognitive images themselves are causally produced, substantially existent, facets of the dependently-originated dependent nature. The question is what happens when the dependent nature is emptied of the constructed nature: do the images completely disappear, or does the consciousness of enlightened beings also have images? Thus, according to Kajiyama, the dispute between these two points of view only concerned the status of Buddhas and high-level Bodhisattvas, since both sides acknowledged that the images exist by definition for ordinary beings who have not yet transcended dualistic fixation: The Sautrāntika thought that what we perceive is not an external reality itself, the existence of which can be known only by inference, but the impress or image which is left by the external reality upon our consciousness. The Yogācārin advanced a step farther and said that the external reality is not existent at all, the world being nothing but our ideas which are the sole reality. Therefore, to the Yogācārin, the image of cognition is the appearance of our mind; and this necessarily implies that a cognition is always endowed with an image which is Aspectarianism,” following the translation of ākāra as “aspect.” However, this translation is misleading or inaccurate, since the ākāra that is asserted to not ultimately exist in consciousness, or equivalently to be false, is the cognitive “image” (ākāra, Tib. rnam pa), not the dual “aspects” (ākāra, Tib. rnam pa) of subject and object. In other words, nirākāravāda or alīkākāravāda is tantamount to a type of anti-realism, where the image that appears in consciousness is strictly unreal, while sākāravāda or satyākāravāda is tantamount to a type of realism, where the image is somehow real. Neither stance is directly related to any particular philosophy of mind, thus Ratnākaraśānti speaks of Yogācārins and Madhyamakas on both sides of the ākāravāda debate. Moreover, Kajiyama attests that non-Buddhist schools also fell on both sides of the debate, Sāṃkhya and Vedānta with sākāravāda and Nyāyavaiśeṣika, Mīmāṃsaka, and Jaina with the nirākāravāda. 175 Shinya Moriyama, “On Ratnākaraśānti’s Theory of Cognition with False Mental Images,” 2 n. 4. “The Sanskrit terms for sub-divisions of the Yogācāra school, sākārajñānavādin and nirākāravādiyogācāra, are found in Advayavajra’s Tattvaratnāvalī 4.20 and 5.5. However, the terms rnam bden pa (*satyākāravādin) and rnam rdzun pa (*alīkākāravādin) are found in Tibetan doxographies.” In the Tibetan translation of Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way, the term for nirākāravādiyogācāra is rendered rnam brdzun pa. 127 represented by our mind. Thus, all Yogācārins must be sākāravādins so far as the cognition of common people is concerned. A problem, however, appears in regard to the emancipated person, who is supposed to have acquired nirvikalpajñāna or non-conceptual, super-mundane knowledge. Some Yogācārins thought that knowledge of an emancipated person is freed from the fetter of cognitum and cognizer and accordingly is clear like a pure crystal without specks. And they held that this clear, imageless knowledge is the essence of cognition, regarding images as false, unreal stains born due to our vāsanā. This is the essential [point] of the nirākārajñānavāda. But others from the same school criticized this theory, saying that what is not real can never be manifested, since otherwise it would entail the unfavourable doctrine of asatkhyāti.176 Every cognition, inasmuch as it is knowledge, must have an image, and yet there is no harm in that an emancipated person’s knowledge is with an image, if he is freed from conceptual thinking, the fundamental of which is the bifurcation of cognitum and cognizer. This is the essential point of the sākārajñānavāda of the Yogācārins.177 Unfortunately, there is no space here to consider the satyākāravāda position in any detail. Interestingly, however, one of the chief defenders of this position appears to have been Jñānaśrīmitra (975-1025), a contemporary of Ratnākaraśānti’s at Vikramaśīla, who wrote a lengthy rebuttal of the latter’s position, the Treatise Establishing Real Images (Sākārasiddhiśāstra). Ratnākaraśānti identifies two main objections to his view, the first of which may be a response to Jñānaśrīmitra, whose position Ratnākaraśānti equates with a type of realism: 176 John Grimes, A Concise Dictionary of Indian Philosophy: Sanskrit Terms Defined in English (Albany: State University of New York Press, 1996), 166. Bimal Krishna Matilal, Perception: An Essay on Classical Indian Theories of Knowledge (Oxford: Clarendon Press, 1986), 188. Grimes defines khyāti-vāda as the theory of error in pramāṇa discourse. According to Grimes, “There are three types of theories of error: theories where the object of error is real (sat-khyāti-vāda), theories where the object of error is unreal (asat-khyāti-vāda), and the theory where the object of error is neither real nor unreal (anirvacanīya-khyāti-vāda). Under the first group we find… Yogācāra’s ātma-khyāti…. Under the second group we find the Mādhyamika’s asat-khyāti.” On this point, Matilal (1986, 188) writes, “The proponent of [asatkhyāti] shows that the nature of an awareness cannot be such that its object-form is always, or is always caused by, an existent entity.” 177 Yuichi Kajiyama, “Controversy between the Sākāra- and Nirākāra-vādins of the Yogācāra School,” 29-30. 128 Some adherents of the Yogic Practice and Middle Way schools, who maintain that consciousness possesses [real] cognitive images, claim that it is unacceptable for luminous blue-patches to have a nature that is both deceptive and non-deceptive, as these qualities are contradictory. If [luminosity] were not their nature, then bluepatches and so on would not be luminous; however, blue-patches are indeed luminous. Since they are not something other than the luminosity which is real or non-imputed, blue-patches and so on are also real or non-imputed.178 In other words, the objection (as Ratnākaraśānti understands it) is that phenomena are either luminous or not luminous. If they are not luminous, they cannot be perceived. This is owing to the identification of “luminosity” with the reflexive nature of awareness, or the fact that the contents of cognition are immediately present to the mind; to be “illuminated” is to be accessible to consciousness. If phenomena are luminous, however, they cannot be totally separate from this luminosity, which both Ratnākaraśānti and Jñānaśrīmitra(?) accept is real. Therefore, according to the unnamed “True Imagist,” phenomena must be real. Thus sākāravāda is understood by Ratnākaraśānti as being tantamount to realism:179 The evidence180 for the prior objection was that anything which is not different from luminosity—even a false image—is a real entity. But if [luminosity] is not different from the nature of being a real 178 Ratnākaraśānti, Pith Instructions, ff. 228.a.3-5. rnal 'byor spyod pa pa dang/ dbu ma pa shes pa rnam pa dang bcas par smra ba kha cig na re/ sngon po gsal bdag (P/N/S: ba dag) brdzun pa dang brdzun pa ma yin par 'gyur na chos 'gal ba nyid kyis na de'i bdag nyid du 'thad par mi 'gyur la/ de'i bdag nyid ma yin na/ sngon po la sogs pa gsal bar mi 'gyur na/ sngon po la sogs pa ni gsal ba yin no zhe na/ sgro ma btags pa'am dngos por 'gyur ba'i gsal ba las gzhan ma yin pa'i phyir/ sngon po la sogs pa ni sgro ma btags pa 'am/ dgnos por 'gyur ba yin no zer te/ 179 This is not, of course, to say that Jñānaśrīmitra would have necessarily accepted this characterization of the sākāravāda position. But there is even less research concerning Jñānaśrīmitra than there is concerning Ratnākaraśānti. The details of the debate between these two luminaries of Vikramaśīla, while beyond the scope of this thesis, should prove a fascinating area for further study. 180 See note 343. 129 entity, this evidence does not apply. And if the evidence is mere non-difference then it is inconclusive, since this evidence also holds for apprehended and apprehender and so on, even though they do not exist.181 In other words, if “being luminous” is the same as “being real,” then even totally illusory phenomena which are in no way real—such as the distorted, dualistic subjective and objective aspects of ordinary cognition—are real. Ratnākaraśānti further rebuts this position by pointing out the unacceptable consequences which follow: Well, according to them, because all luminous [phenomena] are experienced in their non-erroneous nature, there would not be any cognitive distortion whatsoever. Therefore, all sentient beings would be eternally liberated, and would always already be completely perfect, genuine Buddhas!182 That is to say, if there is no difference between phenomenal appearances and the luminosity of those appearances, there is no basis upon which to distinguish between erroneous and nonerroneous cognitions, since all phenomena are identical in terms of their luminous nature, and luminosity is non-erroneous. All cognitions must therefore be non-erroneous, precisely to the extent that they are luminous. And since by definition the only beings whose cognitions are all non-erroneous are fully-enlightened Buddhas, all beings must be fully-enlightened Buddhas. 181 Ratnākaraśānti, Pith Instructions, ff. 228.b.1-2. gang yang gsal ba las tha mi dad pa'i phyir rnam pa brdzun pa yang dngos por 'gyur ro zhes pa'i gtan tshigs 'di la gal te dngos po'i rang bzhin du tha mi dad do/ zhe na gtan tshigs ma grub par 'gyur ro/ 'on te tha mi dad pa tsam gtan tshigs yin na ni ma nges par 'gyur te/ gzung ba dang 'dzin pa nyid la sogs pa la de med par gtan tshigs yod pas so/ 182 Ratnākaraśānti, Pith Instructions, ff. 228.a.5-6. de dag gi ltar na gsal ba thams cad phyin ci ma log pa'i rang gi ngo bo myong ba'i phyir 'khrul pa gtan med par 'gyur ro/ des na sems can thams cad gtan du grol zhing yang dag par rdzogs pa'i sangs rgyas nyid du 'gyur ro/ 130 Ratnākaraśānti uses this unacceptable consequence to rebut the “True Imagist” perspective. Whether or not his account is a fair representation of the “True Imagist” perspective is as yet unclear, although one strongly suspects that there is much more to be said from their side. However the important point here is what this objection reveals about Ratnākaraśānti’s own perspective. Clearly, Ratnākaraśānti considered the most salient critique of his position by the “True Imagists” to be their characterization of his stance as the position that the objects of perception are both (in some sense) deceptive and (in some other sense) non-deceptive. Elsewhere, Ratnākaraśānti distinguishes three different types of luminosity. The underlying point, which both Ratnākaraśānti and this interlocutor agree upon, is that all appearances are illuminated and thus reflexively available to consciousness. There must, however, be some way to account for the difference between deceptive and non-deceptive, or erroneous and nonerroneous, cognitions. For Ratnākaraśānti, the criterion of distinction cannot be whether or not phenomena are illuminated, since all phenomena, including erroneous cognitions, are luminous. Ratnākaraśānti is therefore forced to argue that phenomena can be simultaneously deceptive, insofar as their distorted (i.e. dualistic) appearance is erroneous, and non-deceptive, insofar as they are luminous. In effect, Ratnākaraśānti is arguing for “False Imagism” as a type of anti-realism, which is borne out by his characterization of the “True Imagist” perspective to be tantamount to a type of realism (“blue-patches and so on are also real or non-imputed”). The essence of the dispute may in fact be over the dualistic status of cognitive images. Jñānaśrīmitra or the unnamed “True Imagist” probably maintained that nondual cognitive images exist. Recall Kajiyama’s description of the sākāravāda position: “Every cognition, inasmuch as it is knowledge, must have an image, and yet there is no harm in that an emancipated person’s 131 knowledge is with an image, if he is freed from conceptual thinking, the fundamental of which is the bifurcation of cognitum and cognizer.” In other words, the “True Imagists” are arguing that there can exist nondual cognitive images, that Buddhas possess phenomenal content but that this content is not presented dualistically. For Ratnākaraśānti, by contrast, all cognitive images are necessarily dualistic, and therefore necessarily distorted. Therefore, the mind of Buddhas cannot have any cognitive images. For Ratnākaraśānti, the ultimate realization of the Buddhas is nondual, and precisely for that reason contentless. As he writes in verse, In the worldly awareness, Once cognitive images disappear, The Lord of Non-Appearance, Nondual suchness without conceptual structures, arises. Although worldly consciousness has images, Those images are thoroughly resolved To be both false and unreal. Thus it is said to be “without images.”183 In other words, Ratnākaraśānti does not maintain that ordinary beings ordinarily experience the world without cognitive images. On the contrary, ordinary “worldly” awareness has dualistic images. Only when these images disappear (nub) can there be a genuine encounter with “nondual suchness.” The other objection Ratnākaraśānti answers is from someone arguing that the external object is false, while the internal object is not:184 183 Ratnākaraśānti, Pith Instructions, ff. 226.b.2-3. 'jig rten pa’i ye shes la/ rnam pa dag ni nub gyur nas/ de nyid gnyis med spros pa med/ khyab bdag snang ba med pa skye/ 'jig rten shes pa rnam bcas kyang/ rnam pa dag ni brdzun pa dang/ bden min rnam par gcod pas na/ de ni rnam pa med par brjod/ 184 See also John Dunne, Foundations of Dharmakīrti’s Philosophy, 406 n. 15. Devendrabuddhi answers a very similar objection. 132 [Objection:] “While cognitive images are not joined to blue particulars, they are not false. But the external “blue” is false, because its appearance (rang gi snang ba), despite not being an object, is nevertheless positioned185 as an object through an imaginative186 determination.”187 This is almost certainly a reference to Dharmottara (ca. 750). According to Dreyfus, Dharmottara’s epistemology centered around a close relationship between direct perception and conceptual inference.188 One of the central problems in Dharmakīrti’s pramāṇa theory concerned the difficulty in accounting for human knowledge on the basis of sense perception, since the contents of sense perception are pre-cognitive and as such not ordinarily accessible to consciousness. A cognitive image is by definition conceptual and inferential, since it is a processed representation of the contents of sense-perception. The appearances of a cognitive image are gross phenomena, not individual particulars; this is exactly why they are epistemically accessible for ordinary, unenlightened beings, who do not have epistemic access to particulars in and of themselves. Dharmottara’s solution to the radical dichotomy between the nonconceptual contents of perception (i.e. particulars) and the conceptual contents of inferential cognitive images (i.e. universals) was to propose that “the ascertainment {i.e. niścaya} generated by the power of perception takes as its object the continuum of moments held [by perception]. Since this 185 See Appendix, note 338. 186 See Appendix, note 343. 187 Ratnākaraśānti, Pith Instructions, ff. 228.a. 6-7. ji ste blo'i rnam pa ni sngon po la sogs pa'i rang gi mtshan nyid la mi rigs pas brdzun pa ma yin la/ phyi rol gyi sngon po la sogs pa ni brdzun pa yin te/ rang gi snang ba don ma yin pa la don du mgnon par zhen pas bzhag pa'i phyir ro zhe na/ 188 Georges Dreyfus, Recognizing Reality, 356-57. 133 [continuum] is the object ascertained, there is no difference in the object of application conceived by both perception and inference.”189 In other words, while their respective “apprehended objects” are particulars and universals, perception and inference both take the same extended entity, the continuum of the individual moments of e.g. a blue-patch, as their “object of application,” that with respect to which they facilitate useful action in the world. In this way, the resulting “ascertained” or “determined” cognition—the cognitive image—is not “joined” to the particular instances of blue, but is instead only a representation of their continuum. For Dharmottara, then, only the seeming externality of the blue-patch is mistaken. The image itself is unmistaken, insofar as it represents real objects—the real particulars, conceptualized or represented in consciousness as a continuum.190 Thus Dharmottara posits a distinction between the “apprehended object” (the blue particulars), and the cognized “object of application” (the blue patch or continuum of blue particulars), both of which are accessible to perception.191 Or, as Ratnākaraśānti characterizes this position, “cognitive images are not joined to blue particulars,” since only perception is “joined” to particulars; but cognitive images are “not false,” either, because—insofar as those particulars are real—the conceptualized object is real, although it is not external. Again, this implies that “True Imagism” is tantamount to a type of realism. Ratnākaraśānti responds by disputing the basis of the distinction: 189 Ibid., 360. 190 Georges Dreyfus, Recognizing Reality, 349. According to Dreyfus, Dharmottara accepts a radical discontinuity between the “Sautrāntika” and “Yogācāra” positions adopted by Dharmakīrti. The reason is that, from a Yogācāra perspective, a perception which appears to be of an external object cannot be unmistaken; therefore, all ordinary perception is mistaken. Dharmottara maintains that the question of a cognition’s unmistakenness is only applicable at the Sautrāntika level of analysis. Dreyfus writes, “For Dharmottara, Dharmakīrti’s unmistakenness must be understood in relation to the way things appear in consciousness. A perception is unmistaken inasmuch as its object appears as it is.” 191 Ibid., 322. 134 If an object is apprehended, then because it is in every respect the object of that [perceptual apprehension], there is no apprehension when the object does not exist. In such an event, the object is not apprehended, and it is also not cognized. In this manner, mental contamination is the positioning [of objects] as they are generally renowned in the world, which is not the genuine way things actually are.192 In other words, perception necessarily and by definition is the perception of real objects. “To be real is to be perceived.” Therefore, if something is apprehended by perception, it is real. Since even Dharmottara admits that the continuum of particulars is not real, but only imputed,193 there can be no perception of that continuum. Therefore, on Dharmottara’s own account, there can be no cognition of the continuum, given (as he admits) that the continuum is not real, only conceptualized. In other words, since the continuum is not real, it cannot be perceptually apprehended; and since it cannot be perceptually apprehended, it cannot be cognized. This brings Ratnākaraśānti to his overall point, which is that Dharmottara has attempted to rescue a commonsense account of perception, while just such a commonsense account is what is being rebutted. More broadly, Ratnākaraśānti’s point here is that, as he sees it, epistemic/ontological idealism—the position that “external blue is false”—necessarily implies the ultimate 192 Ratnākaraśānti, Pith Instructions, 227.a.7-228.b.1. ci ste don la zhen na ni de'i don rnam pa thams cad du yod par 'gyur bas don med bzhin du zhen par mi 'gyur ro/ de bas na don la yang zhen pa ma yin la/ shes pa yang ma yin no/ de bas na blo'i bslad pa ji ltar grags pa bzhin du rnam par bzhag pa yin gyi yang dag pa ji lta ba bzhin du ni ma yin no 193 Georges Dreyfus, Recognizing Reality, 360. Dreyfus translates from Dharmottara: “Sense perceptions apprehend moments, and the ascertainment generated by its (perception) power ascertains [this object], thinking that it sees a continuum [of] momentary parts. …. Therefore, that object which is held by perception is not obtained, for it is not possible to obtain a [single] moment. Nevertheless, the ascertainment generated by the power of perception takes as its object the continuum of moments held [by perception]. Since this [continuum] is the object ascertained, there is no difference in the object of application conceived {i.e. cognized} by both perception and inference.” 135 nonexistence of cognitive images. To Ratnākaraśānti, it is incoherent to maintain that external objects are false, but the cognitive images of those objects are real. Again, this is perhaps best understood as a consequence of his (unstated) view that there is no such thing as an unbifurcated or nondual cognitive image. As long as cognitive images are appearing, they are by definition appearing as an internal, subjective experience of external objects. To deny externality is, a fortiori, to deny the reality of cognitive images. This relates to the final, and perhaps most important, point. It was earlier noted that Ratnākaraśānti denies that cognitive images are real or reliable, since phenomenal forms cannot ultimately be established as either mental or material/extra-mental. But the same passage continues, explicitly contrasting the epistemic unreliability of cognitive images with the epistemic reliability of luminosity: The flaws which follow from the claim that blue-patches and so on are external are the same for a blue-patch which has the nature of consciousness, because there is no difference in the unacceptable conclusions that follow. The distinction between [external] objects and consciousness simply does not amount to anything at all. Now, someone might say, “External particles are surrounded by six directions, but the individual moments of consciousness have four points of connection.”194 However, because they are the same in terms of having parts, these [differences] do not amount to anything at all, either. As there is no third kind of bundle apart from something which is either singular or manifold, blue-patches and so on are not established either as internal or external objects, and are therefore false. This [false image] also has the nature of mere luminosity, which is not disturbed by the force of prior psychological imprints.195 194 Material particles are usually described in terms of three dimensions, for a total of six points of connection (top, bottom, left, right, forward, and backward). Mental particles are more commonly described in two dimensions, for a total of only four points of connection (top, bottom, left, and right). 195 Ratnākaraśānti, Pith Instructions, ff. 227.a.7-227.b.3. ji ltar sngon po la sogs pa phyi rol gyi las (P/N/S: la) thal bar 'gyur pa'I skyon (P: rkyen) brjod pa 'di dag ni/ rnam par shes pa'i bdag nyid kyi sngon po la sogs pa la yang 136 In other words, the most salient feature of sensory-cognitive images or phenomenal forms such as “blue” is not the manner of their production, but the fact that they are the product of delusion, and therefore false or nonexistent despite their “nature of mere luminosity” (gsal ba tsam gyi bdag nyid)—a statement which will be explained more fully in the next chapter, on the reflexive nature of awareness. Although the Pith Instructions is ostensibly a commentary on Śāntarakṣita’s Ornament of the Middle Way, Ratnākaraśānti does not quote the root text even once; and, while Ratnākaraśānti follows the general structure of a Yogic Practice-Middle Way-Pramāṇa synthesis, on occasion the details of Ratnākaraśānti’s synthesis depart significantly from Śāntarakṣita’s presentation in the Ornament of the Middle Way, for instance regarding the relationship between Yogic Practice ontology and the two truths. Perhaps the most significant of Ratnākaraśānti’s deviations, however, is his open embrace of “False Imagism,” as evidenced by his clear assertion that images are deceptive and/or unreal: “[phenomena such as] ‘blue’ are not established either as internal or as external objects, and are therefore false.” Śāntarakṣita, by contrast, critiques both “True Imagism” and “False Imagism” in the Ornament of the Middle Way. Śāntarakṣita considers various systems of philosophy (both Buddhist and non-Buddhist), dismantling each in turn through the application of Nāgārjuna-style mtshungs te/ thal bar 'gyur ba la bye brag med pa'i phyir ro/ don dang rnam par shes pa'i bye brag tsam gyis ni cir yang mi 'gyur ro/ ci ste don gyi rdul phra rab ni phyogs drug nas (P/N/S: ni) bskor ba yin la/ rnam par shes pa'i rdul phra mo ni phyogs bzhi pa'o zhe na/ cha shas dang bcas pa mtshungs pa'i phyir des kyang cir yang mi 'gyur te/ gcig dang du ma las ma gtogs pa'i dngos po phung po gsum pa gzhan dag kyang med pas sngon po la sogs pa dag ni phyi rol gyi dngos po dang nang gi dngos por ma grub pas brdzun pa nyid do/ 'di yang gsal ba tsam gyi bdag nyid yin yang bag chags kyis dkrugs pa'i dbang gis med du zin kyang/ 137 “neither one nor many” arguments. The last system he considers is the Yogic Practice system, which Śāntarakṣita eventually refutes in the same manner. Although their view (i.e. the Yogacara view) is virtuous, we should think about whether such things [as the images known by consciousness accepted by Yogacaras] actually exist or if they are something contentedly accepted only when left unanalyzed.196 [MA 45] Śāntarakṣita first “analyzes” the Yogic Practice view by questioning whether it is tenable to assert that cognitive images truly exist as a multiplicity: If you accept an equal number of consciousnesses and images, then it would be difficult to overcome the same type of analysis as is made regarding particles.197 [MA 49] Next, Śāntarakṣita explains that the images cannot truly exist as a singular entity, either: If the variety [of images] exists in a single nature, how could they appear in the nature of many, and how could parts such as those being obstructed and those which are unobstructed, etc. be distinguished? 198 [MA 51] Śāntarakṣita thus closes off the possibility that the “images” or phenomenal forms cognized by perceptual awareness can truly exist, either as a singular or as a multiple entity. 196 Śāntarakṣita, Ornament of the Middle Way (Madhyamakālaṃkāra), trans. James Blumenthal in The Ornament of the Middle Way: A Study of the Madhyamaka Thought of Śāntarakṣita (Ithaca: Snow Lion Publications, 2004), 119. 197 Ibid., 125. 198 Ibid., 126. 138 So far, Ratnākaraśānti agrees, as can be seen from his discussion of whether cognitive images are properly singular or manifold: The appearance of cloth and woven grass mats and so on must either be singular or manifold. If it is singular, then the beginning, middle, and end parts would not appear separately, and there would arise the contradiction of what is different being the same. And why is that? Because it is impossible that whatever possesses parts in this way a single, individual entity.199 And elsewhere: One might ask, “Is it not the case that blue-patches and so on are singular in terms of how they are experienced? After all, they are experienced simultaneously.” But this is not so; in such a case, what is different would be undifferentiated…. Because the nature of each thing is experienced individually, and there are no “simultaneous” experiences, a single [experience] does not result in the experience of everything. Therefore blue-patches and so on are neither singular nor manifold, and are similar to hairs in the sky.200 The objection here is that ordinarily, when apprehending e.g. a multicolored object, its various colors seem to be apprehended simultaneously. However, in keeping with the Abhidharma framework that each moment of consciousness only takes one proper object, Ratnākaraśānti 199 Ratnākaraśānti, Pith Instructions, ff. 227.a.4-5. gzhan yang snam bu dang re lde la sogs par snang ba 'di dag gcig gam du ma zhig/ gal te gcig na thog ma dang tha ma dang dbus kyi cha tha dad par snang bar mi 'gyur te/ tha dad pa dang tha dad pa ma yin pa 'gal ba'i phyir ro/ 200 Ratnākaraśānti, Pith Instructions, ff. 227.b.6-7. ci ste sngon po la sogs pa 'di ni ji ltar nyams su myong ba de ltar gcig yin te/ lhan cig nyams su myong ba'i phyir ro zhe na/ ma yin te tha dad pa nyid tha dad pa ma yin no/ zhes smra bar 'gyur ro/ ci ste de'i bdag nyid yin pa'i phyir/ …de dag bdag nyid so sor myong ba'i phyir/ lhan cig tu myong ba med pas gcig gi thams cad myong ba med par thal bar 'gyur ro/ de bas na sngon po la sogs pas ni gcig dang du ma yin pas na nam mkha'i skra'i tshogs bzhin no/ 139 denies that this experience is truly “simultaneous,” as it only appears to be so from the perspective of ordinary, deluded beings. However, while Ratnākaraśānti proceeds from this point to the assertion that cognitive images are ultimately false or nonexistent like “hairs in the sky” (nam mkha’i skra’i tshogs), Śāntarakṣita critiques the notion that cognitive images are nonexistent. As he puts it, Some say that [consciousness] does not naturally possess images of these [objects]. In reality, images do not exist but appear to the consciousness by virtue of a mistake. If [images] do not exist, how can consciousness clearly experience those [objects]? That [clear, non-dual consciousness] is not like a consciousness which is distinct from the entities.201 [MA 52-53] Śāntarakṣita’s point here is that a non-dual consciousness is, by definition, not separate from its intentional “object” (to the extent that the notion of an “object” is intelligible for a nondual consciousness). Thus to deny that the image is real is to deny that consciousness can have any kind of content; this is the asatkhyāti mentioned by Kajiyama, the theory of perceptual error that erroneous perceptions take non-existent entities as their object. If the image is not real, then how can anything ever be experienced? Apart from the fact that images are only nonexistent with respect to the minds of enlightened beings, the explanation Ratnākaraśānti provides for how consciousness can “clearly” experience phenomenal objects is that they are experienced through being illuminated by reflexive awareness—the topic of the next chapter. But this is not the only argument Śāntarakṣita directs against the “False Imagist” position: 201 Śāntarakṣita, Ornament of the Middle Way, trans. Blumenthal, 127-28. 140 If there were no cause [for images], how is it suitable that they arise only on occasion? If they have a cause, how could they not have an otherdependent nature (paratantra-svabhāva, Tib. gzhan gi dbang gin go bo)?202 [MA 58] This point was mentioned above, in the discussion of the three natures. Unfortunately, Ratnākaraśānti does not directly address this objection. But it may be surmised that Ratnākaraśānti saw images as only arising or existing due to the contamination of ignorance and delusion. According to Ratnākaraśānti, only the luminosity of the reflexive nature of awareness is a truly reliable means of knowledge, and everything else exists as delusion, similar to an assortment of hairs floating in the sky. In other words, just as the appearance of dark floating marks is a distortion of the visual consciousness, and there are in fact no floating hairs “out there,” so too is the cognitive image nothing but a type of distortion. As long as cognitive images appear, the one to whom they appear is deluded. Cognitive images disappear completely upon the attainment of liberation; only the nondual luminosity of reflexive awareness remains. The phenomenal forms of cognitive images are thus “occasional” or adventitious in the sense that they dissolve upon the attainment of the Fourth Yoga, the stage of non-appearance.203 This discussion highlights one of the main differences between Śāntarakṣita and Ratnākaraśānti. While Ratnākaraśānti purported to comment upon Śāntarakṣita’s Ornament, and took great pains to reconcile his approach with both the Middle Way and Yogic Practice traditions, at the end of the day he tended toward a “False Imagist” Yogic Practice view, over and above a Middle Way view. Śāntarakṣita, by contrast, saw fit to refute both “True Imagist” and “False Imagist” Yogic Practice views. In other words, Śāntarakṣita’s primary concern was 202 Ibid., 132. 203 See below, § III.B.1.c. 141 ontological; his eleutheriological approach centered around the refutation of any and all possible existents (such as true images) and non-existents (such as false images). Ratnākaraśānti, on the other hand, clearly considered both a positive ontology with respect to the luminosity of reflexive awareness, and a negative phenomenology with respect to cognitive images, essential to his eleutheriological project. Be that as it may, unlike the other systems considered—specifically, unlike the other Buddhist systems considered—Śāntarakṣita finds a place for Yogic Practice philosophy, by stating that the latter is eminently suitable for describing the relative truth of causal interaction or dependent origination, thus sublating the view of “mind only” within his Middle Way-Yogic Practice-Pramāṇa synthesis: That which is cause and result is mere consciousness only. Whatever is established by itself abides in consciousness. By relying on the Mind Only [system], know that external entities do not exist. And by relying on this [Madhyamaka] system, know that no self at all exists, even in that [mind]. Therefore, due to holding the reins of logic as one rides the chariots of the two systems (i.e., Yogacara and Madhyamaka), one attains [the path of] the actual Mahayanist.204 [MA 91-93] Of course, Śāntarakṣita could only maintain the consistency of his viewpoint by denying that the reflexive nature of awareness is ultimately real; as he writes in his autocommentary to the Ornament, reflexive awareness is “classified as conventional truth because it cannot bear an analysis which looks for a singular or a manifold nature.”205 As will momentarily become clear, this is the argument with which Ratnākaraśānti finds the most to disagree. 204 Ibid., 169-72. 205 Ibid., 169. 142 143 III. Reflexive Awareness The reflexivity of awareness (svasaṃvitti or svasaṃvedana, Tib. rang rig) is a complex term that relates Buddhist praxis and eleutheriology with epistemology and, on some accounts, with ontology as well. In its most basic formulation, reflexive awareness is the property that consciousness has of being lucid and aware. In phenomenological language, svasaṃvedana accounts for—or just is—the present quality of what is presented to consciousness. One of the most important and hotly-debated topics in Indian Buddhist scholastic philosophy, it has only recently begun to receive significant attention in Western scholarship. Williams (1998) treated the internal polemics of the so-called “Consequentialist” or *Prāsaṇgika Middle Way tradition in far greater detail than the reflexive nature of awareness itself. 206 Yao (2005), an exhaustive survey of the early sources for this doctrine as well as its presentation in the works of Dignāga (ca. 480-540), is most valuable in elucidating the profusion of early views regarding the reflexive nature of awareness; while he presents all of the elements that influenced its place in the philosophy of Dignāga, Yao deliberately avoids extended commentary on any Indian sources later than Dharmakīrti (ca. 600-660).207 Apart from these two monographs, no sustained treatment of reflexive awareness exists in the Western literature. A recent (2010) volume of the Journal of Indian Philosophy208 was devoted to this topic; however, not all of the essays concerned reflexive awareness in Buddhist 206 Paul Williams, The Reflexive Nature of Awareness: A Tibetan Madhyamaka Defense (London: Routledge, 1997). 207 Zhihua Yao, The Buddhist Theory of Self-Cognition (New York: Routledge, 2005). 208 Journal of Indian Philosophy 38 (2010). 144 philosophical traditions, and Dan Arnold’s concluding article,209 in particular, is problematic for reasons which will be explained below. Dreyfus (1997) and Dunne (2004), while important contributions to the understanding of Dharmakīrti’s philosophy, do not treat the reflexive nature of awareness in any great detail. Blumenthal (2003), like Williams (1997), addresses the reflexive nature of awareness in the context of Tibetan scholastic commentaries to the Ornament of the Middle Way of Śāntarakṣita (725-88), although Blumenthal also engages with Śāntarakṣita’s autocommentary. Academic scholarship on late Indian Buddhist accounts of reflexive awareness, such as the presentations of Ratnākaraśānti (ca. 10th century) and Ratnakīrti (ca. 11th century), remains practically nonexistent. The present effort is directed toward addressing this deficit, with one important caveat. Given the wide variety within and disagreement among the various sources for the doctrine of the reflexive nature of awareness, a comprehensive treatment of svasaṃvitti is beyond the scope of this essay. As the primary concern is the philosophy of Ratnākaraśānti, the analysis of Dharmakīrti’s philosophy will, in general, be limited only to those of its facets which demonstrably contribute to Ratnākaraśānti’s presentation of the reflexive nature of awareness in the Pith Instructions. By no means should this be taken as an implicit argument that Ratnākaraśānti is in agreement with Dharmakīrti on every point of the latter’s presentation of reflexive awareness. Certainly, if Bodhibhadra and Mokṣākaragupta are correct that Dignāga and Dharmakīrti are in some sense “True Imagists,”210 Ratnākaraśānti’s self-identification as a “False Imagist” indicates at least one significant point of dissension. 209 Dan Arnold, “Self-Awareness (svasaṃvitti) and Related Doctrines of Buddhists Following Dignāga: Philosophical Characterizations of Some of the Main Issues,” Journal of Indian Philosophy 38 (2010): 323-378. 210 See Yuichi Kajiyama, “Controversy between the Sākāra- and Nirākāra-vādins of the Yogācāra School,” 28. 145 This chapter begins with an examination of the place of reflexive awareness in Buddhist perceptual theory, particularly as articulated by Dignāga and Dharmakīrti. Special attention is paid to the relationship between reflexive awareness and the other constituents of the knowledgeact, particularly the “aspect of the grasper” and the “aspect of the grasped.” Following this overview, Ratnākaraśānti’s presentation of reflexive awareness is considered in depth, with respect to the major traditions of Buddhist philosophy synthesized in the Pith Instructions: pramāṇa theory, Yogic Practice, and the Middle Way. 146 A. Reflexive Awareness in Buddhist Perceptual Theory As previously discussed, up until the shift to Epistemic Idealism on the “sliding scale,” Dharmakīrti followed the External Realist epistemology typically associated with the Sautrāntika sub-school of the Sarvāstivāda. Sautrāntika epistemology maintained that external objects, although really existing, can only ever be perceived in the form of a sensory-cognitive image. This image results from the interaction of the particles (i.e. dharmas) comprising those objects with the sensory faculties of the observer. In the Sautrāntika system, like all systems derived from the Sarvāstivāda Abhidharma, each particle is momentary, fading out of existence just as soon as it arises. Thus, in the time it takes for the interaction between sense-faculty and object of sensation to produce the sensorycognitive image, the object of sensation—the external particles—have always already faded away. The particles themselves are never observed. Therefore the process of perceiving external objects is entirely mental; perception is the mind taking itself, in the form of the sensorycognitive image or the sense-consciousness(es), as an object. There is, accordingly, always some basic cognitive level at which the objects of sense-perception are known, since those objects are not in fact separate from the mind which knows. This is one type of theory of mind exemplifying what contemporary philosopher Matthew Mackenzie terms the “Self-Awareness Thesis”: “If a subject is aware of an object, then the subject is also aware of being aware of that object.”211 Specifically, it is an example of the “Reflection Thesis,” that “Self-awareness is the product of a second-order awareness taking a 211 Mackenzie, “The Illumination of Consciousness,” 40. 147 distinct, first-order awareness as its intentional object.”212 In Sautrāntika epistemology, perceptual awareness is the second-order mental awareness of one’s own first-order sensorycognitive mental contents. The awareness of one’s own mental contents is reflective insofar as it necessarily takes another, prior mental event as its object. The opposite of the “Reflection Thesis,” according to Mackenzie, is the “Reflexivity Thesis”: that “Conscious states simultaneously disclose both the object of consciousness and (aspects of) the conscious state itself,”213 or, in other words, that consciousness is reflexively, immediately aware of itself and/or its contents. This is the position adopted by Ratnākaraśānti, who follows both Dignāga and Dharmakīrti in this regard. Whereas the Sautrāntikas maintained that the image could only be “reflectively” cognized by the mental consciousness subsequent to its production, Dignāga and Dharmakīrti asserted that the mind is “reflexively” aware of its own contents, including sensory contents. That is to say, in the same moment that the sensorycognitive image is produced, there is some type of awareness of that image. This serves as a backstop against the possibility of an infinite regress: if a second-order cognition is required to establish awareness of a first-order perception, then it stands to reason that awareness of this second-order cognition necessitates a third-order cognition of the second-order cognition, and so on. The contents of consciousness are thus “self-illuminating” (Skt. svaprakāśa, Tib. rang gis gsal ba) as in the simile of the lamp; consciousness illuminates itself at the same time that it illuminates the objects of perception, just as a lamp illuminates itself as it illuminates a darkened room. 212 Ibid. 213 Ibid. 148 Reflexive awareness also addresses the same concern that led some Buddhist schools, such as the Mahāsāṃghikas, to posit the existence of multiple simultaneous mental events,214 as it is not prima facie evident that a sense-perception of e.g. blue is sufficient to establish an awareness of “blue” in the consciousness of an observer. This is so particularly on the Abhidharma model, which posits that the consciousness of mental objects is qualitatively different from the consciousness of visual objects, at least to the extent that the mental sensefaculty is ontologically independent of the visual sense-faculty. To use the common example of a blue-patch, “seeing blue” is at least conceptually distinguishable from “knowing that one sees blue,” as for example when one is preoccupied and does not notice a blue-patch in one’s visual field. In the epistemological theory of Dignāga and Dharmakīrti, the subject in this case fails to form a valid perceptual judgment or “definitive ascertainment” (niścaya, Tib. nges pa) of the blue-patch. But the blue-patch is still present in the visual-sensory image, of which the subject is reflexively aware. In other words, the reflexive nature of awareness does not guarantee that all the contents of the sensory image will be ascertained, but all valid perceptual judgments require reflexive awareness of the contents of sensory or perceptual consciousness, i.e. the sensorycognitive image. It is important to reiterate at this point that, according to Dignāga and Dharmakīrti, every cognitive image possesses two “aspects”, the subjective “aspect of the grasper” (Skt. grāhakākāra, Tib. ‘dzin rnam) and the objective “aspect of the grasped” (Skt. grāhyākāra, Tib. gzung rnam). Dignāga famously identified these two, together with the reflexivity of awareness, 214 Zhihua Yao, The Buddhist Theory of Self-Cognition, 6-33. 149 as the three components of a knowledge-act (pramiti).215 In Dignāga’s system, the object of knowledge (prameya) is first identified as the “aspect of the grasped,”216 while the instrument (pramāṇa) of the knowledge-act is first identified as the “aspect of the grasper.” In other words, the aspect of the grasped is known by means of the aspect of the grasper—a relatively standard and straightforward account of intentional consciousness. Reflexive awareness, on the other hand, is the “result” (pramāṇaphala)217 of the knowledge-act, since every cognition is concomitant with the reflexive awareness of that cognition. In this sense, reflexive awareness is always the result of any cognition. However, Dignāga also defines the instrument of the knowledge-act as the result of the knowledge-act.218 In other words, a pramāṇa is only a pramāṇa if it results in an epistemically reliable cognition; the instrument of an epistemically reliable cognition is the result of an epistemically reliable cognition, or just such an epistemically reliable cognition itself. 219 In effect, as Dignāga explains, these three—object of knowledge, means of knowledge, and result of the knowledge-act—are ultimately inseparable, and are merely conceptualized or 215 PS 1.10 yadābhāsaṃ prameyaṃ tat pramāṇaphalate punaḥ| grāhakākārasaṃvittyos trayaṃ nātaḥ pṛthak kṛtam|| Kellner (2010, 224) translates: “That whose appearance [cognition possesses] is the object that is validly cognized. The form as apprehending and {reflexive} awareness, again, are the means of valid cognition and the result. Therefore these three [aspects of cogntion] are not separate [from one another].” 216 Technically, Dignāga only refers to “that appearance” (yadābhāsaṃ) as the prameya. But the meaning is the same; the point is that the “aspect of the grasped” is “that appearance.” 217 PS 1.9a svasaṃvittiḥ phalaṃ vātra. 218 PS 1.8cd: savyāpārapratītatvāt pramāṇaṃ phalam eva sat|| Kellner (2010, 219 and n. 46) translates: “[Cognition], though it is actually the result, is [metaphorically referred to as] the means of valid cognition because it is held to perform an activity.” Arnold (2010, 87) renders from the Tibetan: “A pramāṇa is only real as a result.” 219 This is why, in many English translations of later Tibetan pramāṇa theory, the Tibetan term for pramāṇa (tshad ma) is often rendered “Valid Cognition.” Tibetan sources tend to blur the distinction between the instrument of a knowledge-act, the result of a knowledge-act, and the knowledge-act (or epistemically reliable = “valid” cognition) itself. 150 linguistically characterized as different entities.220 For Dignāga, then, the reflexivity of awareness is not an independent cognitive act, nor is it simply identified with the subjective “aspect of the grasper.” Reflexive awareness is, rather, the means by which there is awareness of the “aspect of the grasped,” precisely because it is also the result of the subjective aspect’s apprehension of the objective aspect. In terms of the previous discussion of externality, for Dignāga, at the External Realist level, where external objects are accepted for the sake of argument, the aspect of the grasper serves as the instrument and the aspect of the grasped serves as the object. It should be noted, however, that even at the External Realist level, the ostensible object of knowledge—the aspect of the grasped—is reflexively known. Reflexive awareness serves as the means by which the subjective aspect apprehends the objective aspect, the metaphorical “glue” holding the intentional act together. Therefore, if only in a trivial sense, reflexive awareness is the pramāṇa. However, as Birgit Kellner explains, “While the nature of cognition is then also known through self-awareness, this is set aside when the object is external.”221 For this reason, Dharmakīrti did not consider the “aspect of the grasper” to be the instrument, instead arguing that the cognitive image (ākāra) itself is the instrument. But the underlying point is the same in both systems, since even in cases where the cognition is of objects that are provisionally accepted to exist externally to the mind, the image must be 220 Dignāga, trans. Kellner, PSV ad PS 1.10. “In this way, based on the [self-]awareness of cognition with [its] several forms, one metaphorically speaks in this or that way of [one form] being the means of valid cognition [another] being the object. But [in reality], all constituent factors are without activity.” Kellner (2010, 219) writes, “Cognition arises (from an external object) as bearing or containing the form of the object (viṣayākāratayā) that caused it. In fact, however, cognition performs no activity whatsoever. This situation is comparable to that of cause and effect in general.” 221 Birgit Kellner, “Self-Awareness (svasaṃvedana) in Dignāga’s Pramāṇasamuccaya and -vṛtti: A Close Reading,” Journal of Indian Philosophy 38 (2010): 223. 151 reflexively known in order to be epistemically accessible at all. In such cases, reflexive awareness is still the ultimate backstop of the perceptual process; it is only “set aside,” in Kellner’s words, because mere reflexive awareness is not sufficient to account for the relationship between the contents of perception and the objects perceived, at least in terms of ordinary discourse. That is to say, so long as the objects of knowledge are considered to exist externally, the instrument or means by which they are known must be somehow directly connected to those objects. Reflexive awareness is not causally structured or produced.222 Thus, insofar as a perception is ordinarily considered to be produced (in some sense) by the object of perception, reflexive awareness cannot meaningfully be considered as the instrument of knowledge in the mundane cognition of external objects. The situation is rather different, however, when the proper objects of cognition are no longer considered to exist externally. Like Dignāga, Dharmakīrti, as a Yogācārin, did not accept that the phenomenological bifurcation between “internality” and “externality” or “subject” and “object” was in any sense real. Thus, at the Epistemic Idealist level, it makes no sense whatsoever to speak of the aspect of the grasper and the aspect of the grasped as two different entities. They are, rather, two aspects of the same cognitive image. Therefore, at this level, the pramāṇa cannot be anything other than the reflexive nature of awareness, since the only candidate for an object of knowledge is that cognitive image—there is, quite literally, nothing other than the cognitive image to be aware of. The aspect of the grasper cannot be aware of the whole cognitive image; the aspect of the grasper takes the aspect of the grasped, not itself, as an object. Thus the only way to know the cognitive image as a cognitive image—rather than mistakenly apprehending the cognitive image as though it were an external object—is to be 222 See below, note 255. See also Dunne, Foundations of Dharmakīrti’s Philosophy, 407 n. 15. 152 reflexively, nondually aware. Any awareness of the “aspect of the grasper” is necessarily nondual, since a dualistic awareness takes the “aspect of the grasped” as its object. In other words, taking one’s own mental contents as the object of an epistemological action is not the same thing as being reflexively aware of the subjective aspect of a cognitive image. Reflexive awareness is not object-oriented.223 Because it is nondual, reflexive awareness, by definition, does not take either itself or anything else as an intentional object: intentional objects (i.e. aspects of the grasped) are the exclusive purview of the aspect of the grasper, and dualistic consciousness generally. For this reason, reflexive awareness should not in any way be considered a type of “self”-awareness in the sense of a subject’s awareness of that subject’s own self. More to the point, reflexive awareness is not, and should not be understood as, any type of intentional consciousness. This crucial and difficult point has been missed in most contemporary accounts of the reflexive nature of awareness. Reflexive awareness has been systematically misrepresented as a particular kind of consciousness that takes itself as an object, or even as being strictly identical with the “subjective aspect” of ordinary cognition. For example, Paul Williams distinguishes between two different types of reflexive awareness, which he calls self-awareness (i) and (ii). By his definition, self-awareness (i) “in some sense takes an object, and in some sense that object is itself.”224 But as should already be increasingly clear, this is simply false with respect to the 223 John Dunne, Foundations of Dharmakīrti’s Philosophy, 391-92. Dunne provides a translation of the earliest Sanskrit commentary to Dharmakīrti’s Pramāṇavarttika, written by Devendrabuddhi. On PV 3.1: “[Objection:] ‘The subjective aspect is the object of reflexive awareness; therefore, the example is not established.’ This is not so, because the subjective aspect is cognized (rig pa = vedita) in that it itself arises as reflexive awareness.” … Śākyabuddhi comments: “This means the following. [Reflexive] awareness is not something other than the subjective aspect such that its object would be that subjective aspect. Rather, the subjective aspect is cognized in that it arises with the nature of reflexive awareness. If they are not distinct, how can there be the relation of subject and object such that the example would not be established?” Cf. also Ibid., 407 n. 15. 224 Paul Williams, The Reflexive Nature of Awareness, 21. 153 views of Dignāga and Dharmakīrti, and is similarly incorrect with respect to Śāntarakṣita, Kamalaśīla, and Ratnākaraśānti: for all of the above, reflexive awareness is nondual, undifferentiated, and strictly speaking does not take an intentional object at all. Williams’ “selfawareness (i)” was never posited by any of the theorists of self- or reflexive awareness, including Dignāga. Even “self-awareness (ii),” as Williams translates from an explanation by the dGe lugs teacher Thub bstan chos kyi grags pa, is not quite right in terms of classical pramāṇa theory: (ii) Accompanying all the consciousnesses that are aware of others there is also a mere luminosity, a mere awareness, of its own nature, turned solely inwards, without dependence on the external object, and [here] all the dual-appearances of object and subject are posited as a mistake.225 “Mere awareness” and “mere luminosity” are more or less accurate glosses, but “turned solely inwards” (nang kho nar phyogs) is a fatal mischaracterization. More problematic still are various interpretations of reflexive awareness that have proceeded on the basis of faulty assumptions. According to Arnold, who uncritically accepts the distinction between these two different types of self-awareness, self-awareness (i) in Williams’ sense is a “basically perceptual” account, while self-awareness (ii) is “basically constitutive” of consciousness as such, in the manner outlined above.226 It is extremely important to understand, however, that in this presentation of the Buddhist theory of reflexive awareness, Arnold makes two critical errors. First, he defines self-awareness (ii) as “the subjectively known character of 225 Ibid., 5. 226 Dan Arnold, “Self-Awareness and Related Doctrines,” 324. 154 our own experience.” This is false, even on Williams’ account. The subjective aspect’s apprehension of itself is self-awareness (i), while in self-awareness (ii) the “dual-appearances of object and subject” are mistaken, and therefore by definition nothing is “subjectively known”; for there to be such subjective knowledge, there would have to be a knowing subject and a known object, which is precisely the type of dualistic intentionality that is no longer operative within self-awareness (ii). Second, and more importantly, Arnold confuses the “subjectively known character” of experience with the immediately present quality of experience. From the preceding, it should be clear that Dignāga did not consider the reflexivity of awareness to be in any way the same thing as the “subjectively known character of our own experience,” which is nothing more than the subjective aspect of intentional consciousness, Damasio’s “proto-self,” the “aspect of the grasper.” The immediately present quality of experience does not entail that experience is always and everywhere “subjectively known,” i.e. known by some subject. On the contrary, for Yogācārins such as Dignāga and Dharmakīrti, the phenomenological bifurcation of subject and object, or the “subjective character” of any experience which is thus bifurcated, is imaginary and nonexistent. Reflexive awareness, by contrast, is very real. Following Bilgrami, who claims that there is a “governing disjunction”227 between these two accounts, Arnold traces the argument that the reflexivity of awareness cannot be both perceptual and constitutive, since a perception is necessarily of the external world, and “thus defined by its being independent of the things it is about,” while “the subjectively known character of our own experience is not thus independent.”228 To his credit, Arnold identifies the 227 Akeel Bilgrami, Self-Knowledge and Resentment (Cambridge: Harvard University Press, 2006), 29. 228 Arnold, “Self-Awareness and Related Doctrines,” 356. 155 basic problem with Bilgrami’s position, namely that it begs the question by asserting a priori that perception is the sensory cognition of an objectively-existing external world: Note, however, that in characterizing the “lack of independence” that distinguishes self-awareness on a constitutive view thereof, Bilgrami could reasonably thought by Dignāga and Dharmakīrti to beg precisely the question at issue; thus, Bilgrami (for whom it is perhaps axiomatic that idealism is a non-starter) says particularly that cognitions are not independent of our knowing them in the way that “facts or objects in the external world” are. While this characterization surely captures something significant about the phenomenologically distinctive character of perception— specifically, the fact that perceptual objects seem, phenomenologically, to “impinge” upon us—it is just the point of Dignāga (and, following him, Dharmakīrti) to argue that we can account for this strictly phenomenological fact about perception without any reference to external objects… it is just whether we need to say there are such things that is in question for these Buddhists.229 However, Arnold remains committed to the view that reflexive awareness is the “constitutively subjective”230 or “constitutively first-personal character”231 of cognition, and thus misses the exegetical boat. The most basic point that Arnold apparently fails to appreciate is that, according to all Buddhist pramāṇa theorists, intentionality as such—that is to say, the phenomenological 229 Ibid. 230 Dan Arnold, “Is Svasaṃvitti Transcendental? A Tentative Reconstruction Following Śāntarakṣita,” Asian Philosophy 15, no. 1 (2005): 95. “For Śāntarakṣita, [reflexive awareness] simply refers to the constitutively subjective aspect that defines any cognitive as a cognition.” Emphasis original. As will momentarily become clear, this is not true; insofar as he accepted the basic paradigm of Dignāga and Dharmakīrti, Śāntarakṣita did not consider either reflexive awareness or cognition in general to be “constitutively subjective.” 231 Dan Arnold, Brains, Buddhists, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind (New York: Columbia University Press, 2012), 257. 156 experience of being a subject aware of an object—is nothing more than a type of cognitive distortion or error. Dualistic or intentional cognitions are inherently distorted and therefore unreliable; this is literally the entire reason for distinguishing the dependent nature of phenomenal appearances, which does exist, from the constructed nature of phenomenological duality, which does not exist and of which the dependent nature is completely and utterly empty. By contrast, reflexive awareness is not intentional, both by definition and by virtue of the fact that it is not distorted in any way, shape, or form. That is precisely why reflexive awareness is epistemically reliable in a way that phenomenologically bifurcated cognitive images, and especially aspects of the grasper (on account of their particularly close relationship with selfgrasping), are not. Reflexive awareness is always and everywhere a pramāṇa, while the subjective aspect or “subjectively known character” of consciousness is only provisionally a pramāṇa for Dignāga, and never a pramāṇa for Dharmakīrti. And even in contexts where Dharmakīrti maintains that the cognitive image is the pramāṇa, it is still the case that phenomenological bifurcation as such, the fact that for ordinary unenlightened beings cognition appears to have separate objective and subjective aspects, is nothing more than a type of cognitive error, similar to perceiving hairs in the sky. This is precisely why Arnold’s claim, that “Śāntarakṣita can be taken to have understood svasaṃvitti not as exemplifying intentionality—not, that is, as itself simply another kind of intentional cognition—but as intentionality itself—that is, as what it is in virtue of which any cognition could count as a token of that type,”232 is nonsensical. It is certainly true that Śāntarakṣita considered the reflexivity of awareness to be that by virtue of which any cognition is properly cognitive. But Śāntarakṣita, as has already been discussed, accepted the basic validity 232 Ibid., p. 97 157 of the Yogācāra project, and did not accept that phenomenological duality was in any sense real. Moreover, as Arnold himself notes, Śāntarakṣita cast pramāṇa theory in strictly classical terms, accepting on its face Dignāga’s argument that reflexive awareness does not possess a causal structure in terms of the agent-action-object relation. Therefore the notion that, according to Śāntarakṣita, the reflexivity of awareness (which Arnold calls “apperception” in the passage below) is “intentionality itself,” is strictly false: Śāntarakṣita had argued that “cognition does not perceptually cognize (vijānāti) any external object whatsoever.” In this context, the foregoing passage from Śāntarakṣita is introduced by his commentator Kamalaśīla as answering the following question: “But why do these various conceptions not apply as well in the case of apperception? … Kamalaśīla elaborates, “For apperception is not admitted as being intentional (grāhaka); rather, [it is admitted] as intrinsically (svayaṃ)—that is, naturally—being itself luminous, like the light of the atmosphere.” This claim then raises the question (made explicit by Kamalaśīla): “Then why is it not accepted as being intentional?” Śāntarakṣita answers: Its [cognition’s] apperception [does not exist] as being in an actionagent relation, since the threefoldness of [cognition], whose form is partless, does not make sense… How could there be cognition of something distinct, having the nature of an object?233 The fundamental problem here is that Arnold repeatedly makes the mistake of assuming that intentionality, rather than reflexivity, is what these Buddhist pramāṇa theorists assert to be the most basic and constitutive facet of consciousness. To Arnold, the constitutive reflexivity of consciousness simply is the constitutive intentionality of consciousness. This is, of course, a defensible position, articulated by Husserl and a long line of Western philosophers. It is, 233 Arnold, “Is Svasaṃvitti Transcendental?”, 95. Emphasis added. 158 however, not at all what Dignāga or Dharmakīrti were actually arguing. There may be good reason to read Śāntarakṣita and Kamalaśīla as departing from Dignāga and Dharmakīrti in any number of ways, but even (or especially) considering the fact that Śāntarakṣita explicitly described reflexive awareness as that by virtue of which a being has a mind—which, it should be noted, is perfectly compatible with the philosophy of Dignāga and Dharmakīrti—there is simply no reason to assume that Śāntarakṣita was using the concept of reflexive awareness in some essentially different way than Dignāga and Dharmakīrti. Specifically, there is simply no reason to assume that Śāntarakṣita intended reflexive awareness to be understood as any type of intentional (i.e. dualistic) consciousness, far less as “intentionality itself.” And there many good arguments to the contrary, not the least of which is the plain meaning of Śāntarakṣita’s own words. Essentially, by equating reflexivity with intentionality, Arnold has begged the question, since it is just this point—whether or not subject/object intentionality is in fact inherent to or constitutive of consciousness—that is disputed by Yogācārins and Buddhist pramāṇa theorists. By failing to distinguish between the intentional structure and the reflexive nature of consciousness, Arnold effectively imputes his own ideas, his own philosophy of mind, onto the Buddhist pramāṇa tradition. It should therefore be thoroughly clear by now that Arnold’s claim, that “for Śāntarakṣita, [reflexive awareness] simply refers to the constitutively subjective aspect that defines any cognition as a cognition,”234 is both false and incoherent. Arnold’s other major claim, that svasaṃvitti “seems to denote a special kind of (intentional) cognition—that kind, specifically, whose object is other cognitions,” need hardly be mentioned. As discussed above, the type of 234 Ibid., 96. Emphasis added. 159 cognition that takes another cognition as its object is “reflective,” as expressly opposed to “reflexive.” Dreyfus even used this exact terminology, to express this exact point, in 1997: prior to the publication of both of Arnold’s articles on this topic, “Is Svasaṃvitti Transcendental?” (2005) and “Self-Awareness and Related Doctrines” (2010). In the process of revealing external things, cognition reveals itself. Dharmakīrti expresses this idea by saying that cognition is selfluminous (svayam prakāśa, rang gsal ba); that is, self-presencing and, hence, inherently reflexive…. Self-consciousness can be compared to what Western philosophers call apperception; namely, the knowledge that we have of our own mental states. It is important to keep in mind, however, that here apperception does not necessarily imply a separate cognition. For Dharmakīrti, apperception is not introspective or reflective, for it does not take inner mental states as its objects. The self-cognizing or selfpresencing factor of every mental episode brings us a nonthematic awareness of our mental states.235 Arnold does not even acknowledge the existence of Dreyfus (1997) or Dunne (2004), both of whom clearly explain this crucial point. Not coincidentally, Arnold mischaracterizes svasaṃvitti as subjective and reflective, either failing to understand or failing to appreciate the point that svasaṃvitti is nondual and reflexive. However, like Garfield’s mischaracterization of the relationship between emptiness and the two truths, this mischaracterization is perfectly understandable; it is, in a very similar way, the result of pre-existing commitments to the prevailing orthodoxies of the Western academic establishment, in this case the assumption that consciousness is necessarily (or “constitutively”) intentional and dualistic. 235 Dreyfus, Recognizing Reality, 339. Emphasis added. 160 In any event, while Dharmakīrti’s Commentary on Pramāṇa diverges from Dignāga in a few key respects, such as the previously-mentioned difference that Dharmakīrti does not consider the aspect of the grasper to be a pramāṇa, the reflexive nature of awareness serves the same function in both epistemological systems. It also serves the same function for Devendrabuddhi, Śākyabuddhi, and Jinendrabuddhi. More generally, it serves this function for all Indian Buddhist pramāṇa theorists, at least insofar as they accept the basic framework laid out by Dharmakīrti. The nondual reflexivity of awareness is, accordingly, an essential feature of consciousness as such according to Śāntarakṣita, Kamalaśīla, and Ratnākaraśānti. Reflexive awareness undergirds each and every moment of perceptual awareness, allowing sensory contents to be perceived. There is, however, one last point that needs attention. The existence of reflexive awareness does not imply that “the present” is always and everywhere experienced as pure or unmediated presence. Quite the opposite: in ordinary beings, the quality that consciousness has of being im-mediately present to itself goes unnoticed. In other words, reflexive awareness is a necessary element or condition of consciousness, but does not necessarily comprise the sum total of a particular cognitive event. In fact, in the overwhelming majority of cognitive events, it is not noticed at all, and thus is not consciously ascertained. In the wider eleutheriological context, without being trained in recognizing reflexive awareness, one will never be able to even notice it in and of itself as a feature of consciousness. Therefore, in the overwhelming majority of cognitive events it is causally inefficacious, insofar as no valid perceptual judgment or definitive ascertainment (Skt. niścaya, Tib. nges pa) of reflexive awareness is formed, despite the pivotal role reflexive awareness plays in the perceptual process. 161 This is a subtle point, a detailed examination of which is unfortunately beyond the scope of this thesis. The issue concerns the nature of determinate judgment or definitive ascertainment. In the pramāṇa theory of Dignāga and Dharmakīrti, as has already been discussed, such valid perceptual judgments are at least partly conceptual, since nonconceptual perception alone does not produce intelligible cognitive content. This is why, for Dharmakīrti, even “yogic direct perception” (Skt. yogapratyakṣa, Tib. rnal ‘byor mngon sum) is at some level conceptual.236 In order to form a reliable ascertainment, the mind must engage in conceptual “other-exclusion” (Skt. anyāpoha, Tib. gzhan bsal) with respect to the contents of sense-perception. As a Buddhist pramāṇa theorist, Ratnākaraśānti accepts this basic framework—indeed, his contention that the cognitive image is necessarily dualistic, and therefore necessarily distorted, and therefore epistemically unreliable, depends on it. But Ratnākaraśānti explicitly states that emptiness must be the object of a determinate judgment in order to be eleutheriologically efficacious: One might ask: “Since this emptiness of duality is beginningless, doesn’t that mean that we should have been enlightened from that time, at the very beginning?” This does not follow, because even though one has had some experience, still [emptiness] is not definitively ascertained.237 236 John Dunne, “Realizing the Unreal: Dharmakīrti’s Theory of Yogic Perception,” Journal of Indian Philosophy 34 (2006): 503-4. “For [Dharmakīrti] yogic perception paradigmatically amounts to the perception of universals such as “impermanence” (anityatā) and “emptiness” (śūnyatā)…. the obvious question here [is]: if universals are causally inert and ultimately unreal, how can adepts perceive them? How, in other words, can a conceptual cognition become nonconceptual?” See 510 n. 36 and Dunne (2004, 353-60) for an elaboration of Dharmakīrti’s position that the encounter with “emptiness” is initially conceptual; based on the apprehension of a proper concept of “emptiness,” a nonconceptual encounter with emptiness eventually occurs. 237 Ratnākaraśānti, Pith Instructions, ff. 224.b.7-225.a.1. 'o na gnyis kyi stong pa thog ma med pa yin pas de'i tshe na/ thog ma nyid nas byang chub par 'gyur ro zhe na/ mi 'gyur te nyams su myong ba yod du zin kyang nges pa med pa'i phyir ro/ 162 In other words, duality never really existed, and the emptiness of duality has always existed. Yet, up until the present moment, sentient beings have not formed a definitive ascertainment of this emptiness. Therefore, even though everything is always and everywhere empty of duality, beings who have failed to realize emptiness are not yet enlightened. Thus, as Rantākaraśānti writes, “Enlightenment results from ascertaining the completelyperfected nature of the other-dependent; cyclic existence results from ascertaining the constructed nature.”238 The key implication here is that it is possible to form a definitive ascertainment (nges pa, *niścaya) of the perfected nature—and it is not at all clear that Dharmakīrti, or Ratnākaraśānti’s other contemporaries in the scholastic pramāṇa discourse, would have agreed with this stance. Be that as it may, Ratnākaraśānti’s point is fairly straightforward, namely that there are two different and essentially opposed types of determinate judgments: dualistic and nondual. The latter are always available, inasmuch as duality is merely illusory and in no meaningful sense real. But until or unless there is a nondual ascertainment, there is only dualistic ascertainment. Again, the implication here is that it is possible to form a nondual, nonconceptual definitive ascertainment. In the Pith Instructions for the Ornament of the Middle Way, Ratnākaraśānti only describes such ascertainments in terms of emptiness, and does not explicitly relate reflexive awareness to definitive ascertainment. But reflexive awareness occupies a very similar place to emptiness in his ontology. In exactly the same way that all cognitions are always ultimately empty of duality, all cognitions are always ultimately luminous. In exactly the same way that the emptiness of duality is not ordinarily ascertained, nondual luminosity is not 238 Ratnākaraśānti, Pith Instructions, ff. 225.a.1-2. de bas na gzhan dbang la de yongs su grub pa'i rang bzhin du nges nas byang chub par 'gyur ro/ kun brtags pa'i rang bzhin du nges na (C: nas) 'khor bar 'gyur ro/ 163 ordinarily ascertained. Therefore, even though the luminosity of nondual reflexive awareness is the ultimate inherent nature of the mind, the mind is not ordinarily ascertained as luminous, any more than it is ordinarily ascertained as empty. This is an important point, which will be returned to in the conclusion. 164 B. Ratnākaraśānti’s Presentation of Reflexive Awareness in the Pith Instructions Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way follows the Ornament of the Middle Way in synthesizing Buddhist pramāṇa theory with Yogic Practice and Middle Way philosophy. However, while Śāntarakṣita’s root text limits its consideration of the reflexive nature of awareness almost exclusively to Sautrāntika-style epistemological analysis, for Ratnākaraśānti the reflexive nature of awareness is of absolutely central importance to the synthesis. Ratnākaraśānti’s presentation of the reflexive nature of awareness will be examined in terms of pramāṇa theory, the Yogic Practice philosophy of the three natures, and the Middle Way hermeneutic of the two truths. 1. Reflexive Awareness and Pramāṇa Theory a) Reflexive Awareness as Direct Pramāṇa Ratnākaraśānti begins his discussion of reflexive awareness by distinguishing the luminous appearance of e.g. a blue-patch from the luminosity of that appearance. First, Ratnākaraśānti notes that the external appearance of a blue-patch is, in fact, deceptive: Even though, for example, a conceptualized blue-patch may exist, the characteristic “blue” does not exist, because it cannot withstand logical analysis. There is cognitive distortion due to the contaminating force of the “blue” psychological imprint. Since it arises in this way, the experience is distorted, and it is experienced as though one were experiencing something else.239 Ratnākaraśānti’s premise here is that “characteristics” (Skt. lakṣaṇa, Tib. mtshan nyid) such as the blueness of a blue particle are not truly existent, since they are unable to withstand logical 239 Ratnākaraśānti, Pith Instructions, 225.b.7-226.a.1. sngon po la sogs pa'i bag chags kyis bslad pa'i dbang gis 'khrul pa'o/ de ltar gyur pas de myong ba yang 'khrul pa dang/ gzhan myong ba lta bur myong ba'o/ 165 analysis. Recall the “analytic” sense of the two truths distinction from §II.A.4. In MMK V.1-5, Nāgārjuna argues that it is logically impossible to separate the characteristics of a dharma (or any phenomenon) from that dharma, since the characteristic and the characterized are mutuallyimplicating and therefore empty in precisely the same manner as cause and effect. For Ratnākaraśānti, the fact that these characteristics are incapable of withstanding analysis means that they are, furthermore, incapable of serving as a means reliable knowledge about phenomenal appearances. The dharmas being conceptualized as an extended blue-patch, and thus (from a certain perspective) the blue-patch itself, are dependently originated; therefore, the blue patch is “substantially” existent. But the blueness of the blue patch is illusory, since it only arises due to the presence in the storehouse consciousness of the pre-existing imprint to perceive “blue.” As all imprints are contaminated by ignorance, the experience of blue is distorted, just as the experience of hairs floating in the sky is distorted. This distortion manifests as the seeming externality of the blue-patch, the fact that it is “experienced as though one were experiencing something else,” that is, something other than one’s own mind. The perception of blue is therefore not epistemically reliable.240 Ratnākaraśānti next contrasts the epistemically unreliable quality of phenomenal characteristics with the epistemically reliable quality of the reflexive nature of awareness, which he asserts is direct (mngon sum, *pratyakṣa) and unassailable: 240 Although Ratnākaraśānti does not explicitly mention the theory of apoha or “other-exclusion,” this passage is also likely a reference to apoha theory. The phenomena under consideration have many different properties, all the rest of which are excluded by the conceptual process that yields the determination (niścaya) “blue.” So the determination that the blue patch is blue comes at the epistemic cost of being unable to determine other properties of the blue patch, such as its momentariness. In this way, the phenomena being perceived as blue exist, but the visual perception of their blueness is not a means of gathering ultimately or unassailably reliable information about those phenomena. As Ratnākaraśānti will argue, in the final analysis, the only ultimately reliable pramāṇa—as well as the only unassailably reliable information to be gained—with respect to the blue patch is its luminous nature. 166 By contrast, there is nothing that can refute the luminous nature of reflexive awareness. Reflexive awareness is direct [mngon sum, *pratyakṣa], and it is an authentic experience, because there are no means of reliable knowledge [tshad ma, pramāṇa] apart from it. Thus it is established as a means of reliable knowledge. Indeed, this means of reliable knowledge cannot be refuted, even by a hundred means of reliable knowledge—to say nothing of mere extreme assertions, which are not means of reliable knowledge.241 What does it mean to say that reflexive awareness is “direct”? As previously mentioned, there are only two means of reliable knowledge in the pramāṇa theory of Dignāga and Dharmakīrti: the “direct” (Skt. pratyakṣa, Tib. mngon sum) means, and the indirect “inferential” (Skt. anumāna, Tib. rjes dpag) means. Many Western commentators on Buddhist pramāṇa theory have noted that, while inference is considered a pramāṇa, it is strictly dependent upon a prior direct pramāṇa. This is reflected in the etymology of the term for inference: in this context, the Sanskrit prefix anu means “following” or “subordinate to,” while māna comes from the root √mā, “to measure.” Likewise, the Tibetan translation of pramāṇa, tshad ma, comes from the root tshod which means “to measure,” as does the root dpag in the Tibetan translation of anumāna (rjes dpag). In Tibetan, rjes literally means “following” or “after.” The term anumāna or rjes dpag is usually translated into English as “inference,” but in both Sanskrit and Tibetan it literally means “subsequent measurement.” The point is that an epistemically reliable knowledge-act is a cognition that employs an epistemically reliable instrument of measurement. A “direct” pramāṇa is the means for taking a direct measurement, or just such a direct measurement, while an “inferential” pramāṇa is a subsequent measurement following from such a direct measurement; 241 Ratnākaraśānti, Pith Instructions, ff. 226.a.1-2. yang (D/C: em. rang) rig pa gsal ba'i ngo bo la gnod par byed pa ni med de/ de las lhag pa'i tshad ma gzhan med pa'i phyir ro/ de ni rang rig pa'i mngon sum yin pa dang/ yang dag tu myong ba'o/ de bas na 'di ni tshad mas grub pa ste/ tshad ma ni tshad ma brgyas kyang mi gnod de/ tshad ma ma yin pa pha rol gyi ‘dod (P/N: gnod) pa tsam gyis smos kyang ci dgos/ 167 the direct pramāṇa is thus both ontologically and epistemologically prior to the inferential pramāṇa. Ordinarily, the “direct” means is identified with sensory perception, a convention Ratnākaraśānti adopts in his repeated use of the example of the visual perception of a blue-patch to illustrate the direct means of reliable knowledge. Strictly speaking, however, the “direct” pramāṇa is only defined as being “free from conceptual construction” (kalpanā-apodha).242 Moreover, what makes a pramāṇa a pramāṇa according to the definition used by Dignāga and Dharmakīrti is that i) a pramāṇa gives new knowledge about its object and ii) a pramāṇa gives reliable knowledge about its object,243 where this kind of “epistemic reliability” (Skt. prāmāṇya)244 is a function of being able to accomplish some specific goal (arthakriyā, Tib. don byed nus pa).245 Non-Buddhist contributors to the pramāṇa discourse often simply equate the “direct” means of reliable knowledge with sensory perception, and indeed this is the primary force of the term for both Dignāga and Dharmakīrti. But sensory perception, according to the Buddhist pramāṇa theorists, was only one type of “direct” pramāṇa. Dignāga, for example, considered reflexive awareness to be a direct pramāṇa. As explained above, Dignāga equates the instrument (pramāṇa) of the knowledge-act (pramiti), with the result of the knowledge-act (pramāṇaphala). In other words, a pramāṇa is only really a pramāṇa insofar as it is the result of a knowledge-act employing a valid pramāṇa, or as Arnold 242 PS 1.3a 243 Dunne, Foundations of Dharmakīrti’s Philosophy, 254-55. 244 Ibid., 223. “[Prāmāṇya] concerns not the workings of an alleged instrument of knowledge [pramāṇa]—be it a perception, an inference, the sense faculties, a sacred text, or some other candidate—but rather what it is that justifies the claim that one or another of these is indeed an instrument of knowledge.” Dunne translates prāmāṇya as “instrumentality.” “Epistemic reliability” is perhaps somewhat looser, but attempts to relate the explicitly epistemological nature of the point at stake. “Epistemic instrumentality” might be the most literal formulation. 245 Ibid., 287-290. 168 translates, “a pramāṇa is only real as a result.”246 Whether or not something is a pramāṇa is determined by whether or not it serves as an instrument capable of giving new, reliable information about that which it “measures.” Dignāga explains that the result of a knowledge-act is the reflexive awareness of the epistemically contentful cognition (i.e. the knowledge-act): “Cognition arises as appearing twofold: [having] the appearance of itself [as subject], and the appearance of an object. In terms of these two appearances, the one that is apperception (svasaṃvitti) is the one that is the result.”247 Thus reflexive awareness, the direct means (pratyakṣa-pramāṇa) by which the cognizer has access to the contents of cognition, is also the result of that cognition (pramāṇaphala)—that is, the pramāṇa or instrument itself. By virtue of the fact that it gives direct and unmediated access to new sensory or cognitive information, and is furthermore free any conceptual construction, reflexive awareness is a direct pramāṇa. Therefore, as Arnold writes, To the extent, then, that “a pramāṇa is real only as a result,” and to the extent that that “result” is (as Dignāga here says) svasaṃvitti, it turns out that the latter is the only really occurrent pramāṇa in any case—that, in other words, the only indubitably immediate cognition concerns the occurrence of our own mental states. Of course, this is not typically regarded as an example of “perception,” as that word is generally understood in English; but it is important to recall that Dignāga has defined pratyakṣa only as being definitively “free of conceptual elaboration” (kalpanā apodha). To say this much is not, ipso facto, to say that “perception” designates only sensory cognition, but simply that it denotes whatever cognition immediately (that is, without the mediation of any concepts) apprehends a uniquely particular object—which is as much as to say, that kind of cognition whose 246 Arnold, “Is Svasaṃvitti Transcendental?”, 87. 247 Ibid. 169 phenomenological content is at the same time its direct object. And we have now seen Dignāga argue that in the final analysis, svasaṃvitti is the only really occurrent type of such unmediated cognition.248 Being a “direct” pramāṇa, reflexive awareness is already categorically distinguished from inference. The question, then, is in which sense any other pramāṇa is “direct,” since, has already been seen, Ratnākaraśānti does not consider the classical example of a direct pramāṇa—visual perception, as e.g. in the cognitive image of a blue patch—to be ultimately epistemically reliable. In effect, Ratnākaraśānti has claimed that the reflexive nature of awareness is a direct means of reliable knowledge, while denying that the measuring instrument most commonly defined as the direct pramāṇa is, in fact, a pramāṇa at all! Like Dignāga, Ratnākaraśānti’s point is that, in the final analysis, any other type of measuring instrument considered to be “direct” must rely on the reflexive nature of awareness for its cogency qua pramāṇa. Therefore, even sense perception is “subsequent” or “subordinate” to reflexive awareness. For Ratnākaraśānti, as for Dignāga and Dharmakīrti, the reflexive nature of awareness is the ultimate backstop of the perceptual process, a necessary facet of each and every perceptual cognition. Some of Dharmakīrti’s commentators, such as Śākyabuddhi, describe the reflexive nature of awareness as the “ultimate” pramāṇa.249 While Ratnākaraśānti does not use this particular phrase in the Pith Instructions, he makes the same point when he 248 249 Ibid. Dunne, Foundations of Dharmakīrti’s Philosophy, 317. “If we trust Śākyabuddhi’s opinion, the ultimate pramāṇa would be the pure, non-dual, reflexive awareness of the mind itself. But while this ultimate instrumental cognition is the means to Dharmakīrti’s final soteriological [sic] goal, it is not useful for practical action in the world (i.e., saṃsāra). If the ultimate instrument of knowledge is indeed some pure form of reflexive awareness, then there are no longer external objects—or even mental content—on which to act. Hence, it would seem that conventional perceptions and inferences are eventually left behind.” This is precisely what Ratnākaraśānti argues. 170 argues that there are, in the end, no means of reliable knowledge apart from the reflexivity of awareness. Ultimately, nothing can be known if it is not immediately, which is to say directly, present to consciousness. And, ultimately, only this immediacy is a direct means of reliable knowledge, i.e. a direct pramāṇa, since even visual perception is always and everywhere distorted by phenomenological duality. Unlike every other form of supposedly “direct” perception, reflexive awareness is nondual and undistorted. b) Reflexive Awareness as Luminosity In the Pith Instructions, as in the passage above, Ratnākaraśānti frequently describes reflexive awareness as “luminous” (Skt. prakāśa, Tib. gsal ba) and often refers to it simply as “illumination” or “luminosity.” By no means was he alone in this characterization of reflexive awareness. As Chakravarthi Ram-Prasad writes, Classical Indian theories of consciousness focus on a concept captured by the metaphor of luminosity. Invoking the way light falls, they seek to capture the intuition that consciousness somehow makes something manifest. The metaphor became accepted in Indian thought as the way to talk about the defining feature of consciousness… Luminosity is, intuitively, the manifestedness of consciousness to the conscious subject.250 250 Chakravarthi Ram-Prasad, Indian Philosophy and the Consequences of Knowledge: Themes in Ethics, Metaphysics, and Soteriology (Burlington: Ashgate Publishing Company, 2007), 53. 171 Ratnākaraśānti was thus following in a long line of thinkers, both Buddhist and non-Buddhist, who described the present quality of what is presented to consciousness as luminosity or illumination. For precisely this reason, Ratnākaraśānti considered the luminosity of reflexive awareness to be “irrefutable” (shin tu bzlog par bka’ ba).251 Any attempt to refute the present or “illuminated” quality of what is presented to or “illuminated” in consciousness necessarily relies, for its cogency, on being present or “illuminated” to the mind. There is, quite literally, no such thing as an intelligible counter-argument against luminosity. This is an important aspect of Ratnākaraśānti’s presentation, which will be revisited in the conclusion. This is also the reason why, as Ratnākaraśānti writes, “Phenomena themselves are established as luminous by the experience of their inherently luminous nature.” 252 Phenomena, i.e. appearances, are necessarily “illuminated” by virtue of the fact that they are appearing at all. A really-existing phenomenon that is not “illuminated” is indistinguishable from a phenomenon that does not exist and thus does not appear. Or, as Ratnākaraśānti puts it, “If that [luminosity] were not established, then nothing would appear, and in consequence nothing at all could be established.”253 Ratnākaraśānti develops this point elsewhere: Because luminosity is not positioned with respect to anything else, it is incorrect to say that it is deceptive. Because it is undifferentiated, and because there would arise the unacceptable 251 Ratnākaraśānti, Pith Instructions, ff. 226.a.5. 252 Ibid., ff. 226.b.7. chos rnams rang nyid gsal bzhin par bdag nyid kyis gsal ba’i ngo bo nyams su myong bas grub pa yin la/ 253 Ibid., ff. 227.a.1. de grub pa ma yin na ni gang yang snang bar mi 'gyur bas thams cad ma grub par thal bar 'gyur ro/ 172 consequence of an infinite regress, there can be no act of positioning it. Therefore, this luminosity is a direct means of reliable knowledge, because there is no distortion in its nature. As for positioned entities, these are distorted, similar to the appearance of an assortment of hairs floating in the sky.254 In contrast to phenomena like patches of blue, which are only designated as “blue” due to the force of prior psychological imprints, Ratnākaraśānti argues that the reflexive nature of awareness—“luminosity”—is not positioned with respect to any other entity. In other words, it is a structural feature of consciousness, rather than (per the spurious interpretations of some contemporary exegetes) an independent cognitive act; the reflexivity of awareness is, to reiterate, not causally-produced.255 This recalls Dignāga’s point, that the object, means, and result of a knowledge-act are ultimately indistinguishable. To say that the reflexive awareness of a cognition is the result of that cognition, as well as the means by which that cognition is considered an epistemically reliable knowledge-act, is to deny that reflexive awareness possesses a causal structure at all. Ratnākaraśānti dismisses facile objections to his position as “mere extreme assertions,” which of course he does not consider to be means of reliable knowledge (i.e. pramāṇas). He 254 Ibid., ff. 227.b.4-5. gsal ba ni gzhan gyis bzhag (D/C: gzhag) pa ma yin pa'i phyir/ brdzun par mi rung ngo/ de'i 'jog par byed pa ni med do/ khyad par med pa'i phyir dang/ thug pa med par thal bar 'gyur ba'i phyir ro/ de bas na gsal ba 'di ni rang gi bdag nyid la ma 'khrul pa'i phyir mngon sum gyi tshad ma yin no/ bzhag (D/C: gzhan) pa ni nam mkha' la skra'i tshogs snang ba bzhin 'khrul pa yin no/ 255 John Dunne, Foundations of Dharmakīrti’s Philosophy, 276 n. 93. “That reflexive awareness is noncausal follows from its simultaneity with its object, namely, the awareness that is reflexively perceived itself. Indeed, what can be most confusing about reflexive awareness is the notion that it is a cognition distinct from its object. This distinction is clearly the case for all forms of perception, including mental perception (mānasa-pratyakṣa), for in all cases the object (grāhya) of perception is its cause (see, for example, PV3.224)… In contrast, what Dignāga first identifies as the three aspects of an awareness—namely, reflexive awareness, the objective aspect (grāhyākāra), and the subjective aspect (grāhakākāra)—are all ultimately identical and hence simultaneous. The notion that reflexive awareness is cognizing the subjective- and objective-aspects is merely a way of conceptualizing the process of knowing (see the locus classicus in PS 1.1.10)… Dharmakīrti accepts and elaborates upon Dignāga’s opinion.” 173 does, however, take one question very seriously: if all phenomena have the nature of luminosity, are they not then in some sense all the same? How is it possible to account for the variegation of phenomenal appearances, such as a color-patch containing both blue and yellow? This being the case, do blue-patches and have an identity of one single luminosity? Characteristics such as “blue” and “yellow” are mutually exclusive; when appearing, they are not singular. However, while it is mistaken to say that they have the identity of a single luminosity, they do not have a manifold nature, either.256 By virtue of appearing in a complex cognitive image possessing both subjective and objective aspects, as well as reflexive awareness, the appearance of “blue” in consciousness is always already manifold. Therefore even phenomenal “blue” does not have a truly singular nature, and so there is no question of “blue” possessing the same singular nature as “yellow.” But the “blue” and the “yellow” are both qualified by the same illumination; they are both immediately present to consciousness. Ratnākaraśānti, however, denies that phenomenal “blue” and “yellow” are either the same as or different than the luminous quality of their appearance, using a “neither one nor many”-style argument in a manner very similar to Śāntarakṣita: Are blue-patches and so on the same as luminosity, or different from it? What is different from luminosity cannot appear. On the other hand, if they are not different from luminosity, then because they are this luminous nature, they would be one single thing. And 256 Ratnākaraśānti, Pith Instructions, ff. 227.b.5-6. ‘di (P/N/S: de) ltar sngon po la sogs pa gsal ba gcig gi bdag nyid yin no zer na/ sngon po dang ser po la sogs pa mtshan nyid phan tshun spangs te snang bas gcig ma yin la/ de'i bdag nyid kyis gsal ba gcig dang 'gal ba'i phyir du ma nyid kyang ma yin no/ 174 it is impossible for the variety [of phenomena] to be singular, as was discussed above.257 Ratnākaraśānti also addresses the underlying axiomatic assumption here, that both the “blue” and the “yellow” are being experienced simultaneously: One might ask, “Is it not the case that ‘blue’ and so on are singular in terms of how they are experienced? After all, they are experienced simultaneously.” This is not the case. One has claimed that what is different is undifferentiated. …. Because each thing is experienced individually, and there are no “simultaneous” experiences, a single [experience] does not result in the experience of everything. Therefore “blue” and so on are not one and not many, and are similar to hairs in the sky.258 Here, Ratnākaraśānti clearly denies that a single mental continuum is capable of supporting multiple simultaneous mental events. This is necessary, he argues, in order to avoid the absurd consequence that a single “luminous” experience should be able to give rise to the experience of all “luminous” phenomena. Thus, according to Ratnākaraśānti, the reflexive nature of awareness, like sensory consciousness, is momentary (skad gcig, *kṣaṇika). It is a feature of each individual phenomenal appearance, but does not endure long enough to serve as a means by which any phenomenon other than its “object,” i.e. the “illuminated” cognitive image, may be apprehended. 257 Ratnākaraśānti, Pith Instructions, ff. 228.a.1-3. 'di ltar sngon po la sogs pa gsal ba las tha dad dam/ 'on te tha dad pa ma yin pa zhig yin/ ci ste tha dad pa yin na ni de mi snang bar 'gyur ro/ gal te tha dad pa ma yin na ni de rnams gsal ba'i bdag nyid yin pas gcig tu 'gyur ba 'am/ de yang de rnams dang tha mi dad pas du ma gcig tu mi btub pa yang sngar bshad zin to/ 258 Ratnākaraśānti, Pith Instructions, ff. 227.b.6-7. ci ste sngon po la sogs pa 'di ni ji ltar nyams su myong ba de ltar gcig yin te/ lhan cig nyams su myong ba'i phyir ro zhe na/ ma yin te tha dad pa nyid tha dad pa ma yin no/ zhes smra bar 'gyur ro/ ci ste de'i bdag nyid yin pa'i phyir/ de dang 'dra bar gsal ba yang tha dad do zhe na/ de dag bdag nyid so sor myong ba'i phyir/ lhan cig tu myong ba med pas gcig (P/N/S: em. gi) thams cad myong ba med par thal bar 'gyur ro/ de bas na sngon po la sogs pas ni gcig dang du ma yin pas na nam mkha'i skra'i tshogs bzhin no/ 175 In this respect, it is similar—qua momentary pramāṇa—to sensory consciousness. Unlike sensory consciousness and every other type of entity, however, reflexive awareness is neither causally-produced nor differentiable: Something is asserted to be lacking in “self” if it continuously changes by becoming something else. This refutes phenomena which are permanent and possess parts. However, phenomena possessing the characteristic of being momentary, yet abiding uninterruptedly, cannot be refuted because they are established by a means of reliable knowledge.259 Unlike the phenomenal characteristic of being “blue,” then, the reflexive nature of awareness “cannot be refuted.” It is ontologically simple insofar as it does not possess differentiable parts, and it serves as a means of reliable knowledge. Therefore Ratnākaraśānti holds that, unlike the cognitive images of unreal phenomenal appearances such as a blue-patch, “Luminosity is a direct means of reliable knowledge, because there is no distortion in its nature. [Other] entities… are distorted, similar to the appearance of an assortment of hairs floating in the sky.”260 c) The Four Yogas as Objects of Knowledge 259 Ratnākaraśānti, Pith Instructions, ff. 230.a.1-2. rgyun gyis gzhan du gyur pas ni bdag med pa'i phyir 'dod pa ste de'i cha dang bcas pas ni rtag pa'i dngos po khegs mod kyi chos gang mi rtag pa'i mtshan nyid can rgyun mi 'chad par gnas pa ni dgag par mi nus te/ tshad mas grub pa'i phyir ro/ 260 Ratnākaraśānti, Pith Instructions, ff. 227.b.4-5. de bas na gsal ba 'di ni rang gi bdag nyid la ma 'khrul pa'i phyir mngon sum gyi tshad ma yin no/ bzhag (D/C: gzhan) pa ni nam mkha' la skra'i tshogs snang ba bzhin 'khrul pa yin no/ The two manuscript traditions differ as to whether luminosity is being contrasted to “other” entities or to “positioned” entities. Since luminosity is “unpositioned” (avyavasthā), unlike every other entity, the conclusion is the same. For a detailed description of what it means for something to be “positioned” (Skt. vyavasthā, Tib. rnam par bzhag pa), see Appendix note 338. 176 Before turning to the ontological consequences of this stance, a final note regarding the “instrumental” nature of the reflexivity of awareness is appropriate. Ratnākaraśānti concludes his discussion in the Pith Instructions by distinguishing between two different methods for approaching contemplative praxis (rnal ‘byor = yoga). The first is the “ordinary division” (thun mong gi dbye ba) of calm-abiding (zhi gnas, *śamatha) and clear-seeing (lhag mthong, *vipaśyanā) meditation, the classic division of Buddhist contemplative practice. The second is the “special division” (thun min gyi dbye ba) of the Four Yogas, the eponymous praxeological system of the Yogic Practice tradition. Ratnākaraśānti identifies the Four Yogas in the following manner: As for the special divisions of yoga, these are: the observation of the two limits of entities, the observation of “mind only,” the observation of suchness, and no observation. The first stage of yoga takes as its support the full range of phenomena. The second stage of yoga takes as its support the way in which phenomena arise—namely, as mere mind (sems tsam, *cittamātra). The third stage of yoga takes as its support the exact way in which phenomena are, i.e. suchness. The fourth stage of yoga, in which one sees the Great Vehicle, is non-appearance.261 Ratnākaraśānti does not explicitly apply the paradigm of pramāṇa theory to the Four Yogas. However, the reference to “observation” (dmigs pa, Skt. *ālambana)262 implicates pramāṇa, 261 Ratnākaraśānti, Pith Instructions, ff. 230.b.4-6. rnal 'byor gyi khyad par gyi sa gzhi ni/ dngos po'i mtha' gnyis yod pa la dmigs pa dang/ sems tsam la dmigs pa dang/ de bzhin nyid la dmigs pa dang/ dmigs pa med pa'o/ de yang ji snyed yod pa la dmigs pa ni rnal 'byor gyi sa dang po'o/ ji lta bar gyur ba'i sems tsam la dmigs par gyur pa ni rnal 'byor gyi sa gnyis pa'o/ ji lta ba bzhin du gyur pa'i chos thams cad kyi de bzhin nyid la dmigs pa ni rnal 'byor gyi sa gsum pa'o/ theg pa chen po mthong ba ste/ snang ba med pa ni rnal 'byor gyi sa bzhi pa'o/ 262 In the absence of a Sanskrit manuscript, it is impossible to be certain, but dmigs pa is most likely translating ālambana. In Yogācāra literature, the word ālambana means both “object” and “support.” See Tola and Dragonetti 177 insofar as what is “observed” here is, properly speaking, an object of knowledge (prameya). The point is that the practitioner comes to have knowledge or realization of each of the Four Yogas. The question is the manner in which this knowledge comes about—what it is that serves as the means (pramāṇa) of knowing the Four Yogas. The First Yoga is the “two limits of entities.” Earlier, it was noted that this may either refer to existence and nonexistence, or to the aspect of the grasper and the aspect of the grasped. Ratnākaraśānti cites a verse from the Descent into Laṇka Sūtra that seems to imply that the “two limits” refer to internality and externality: In dependence on “mind only,” External objects become unintelligible. Having focused on suchness, One should also pass beyond “mind only.” Having passed beyond “mere mind,” One should pass into non-appearance. The yogi who abides in non-appearance Sees the Great Vehicle. [LAS IX.256-257] Yael Bentor has translated the apposite passage from Ratnākaraśānti’s Pith Instructions for the Perfection of Wisdom: In the first stage the yogis apply their minds (yid la byed pa [*manasikāra]) to the diversity of phenomena in the world that are the objects of the six senses. Then they apply their minds to the six senses and the six consciousnesses, in order to comprehend the mental activities that engage with the world. By combining calm abiding (zhi gnas) and penetrating insight (lhag mthong) they reach (1982). Ratnākaraśānti is most likely playing on the dual meaning here. This is not the only instance where Ratnākaraśānti plays on language; see Appendix, notes 322, 352, and 356. 178 an understanding of conceptual reflected images [rnam pa, *ākāra] to the extent they exist, and discern the modes of apprehending them through the eighteen spheres [khams, *dhātu] of perception.263 This gloss on the First Yoga contains both ontological and phenomenological elements, as the reference to the “eighteen spheres”—that is, the six sense-faculties, the six types of sensory object, and the six consciousnesses—refers to both cognition and its ontological support. Irrespective of whether the “two limits” refers to ontological or phenomenological duality, however, the bottom line is that the First Yoga concerns duality. In other words, the prameya of the First Yoga is the constructed nature, which is seen to be merely imaginary, or simply nonexistent. The attainment of the First Yoga is by definition a nondual experience. Therefore, the ālambana of the First Yoga, the “full range of [dualistic] phenomena,” can only be cognized through nondual reflexive awareness; reflexive awareness is the pramāṇa by which the prameya of imaginary duality is observed and transcended. The Second Yoga is the observation of “Mind-Only.” From the Pith Instructions for the Perfection of Wisdom: In the second stage the yogis reflect on the perception of all phenomena as products of mental-processes-only (sems tsam), which appear due to habitual tendencies of clinging to objects. Since objects grasped as external to the mind do not exist as they are conceptualized, their grasper cannot exist in that way either. By combining calm abiding and penetrating insight, the yogis understand that the diversity of appearances of the eighteen 263 Yael Bentor, “Fourfold Meditation: Outer, Inner, Secret, and Suchness,” Religion and Secular Culture in Tibet: Tibetan Studies II: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies (Leiden: IATS, 2000): 42. 179 spheres of perception are mental-processes-only, empty of object and subject, and devoid of inherent existence.264 At this level, all appearances are seen as being nothing other than the mind. Crucially, there are still phenomenal appearances. These appearances, however, lack duality. In other words, the prameya of the Second Yoga is the dependent nature, the nondual flow of dependently-originated appearances. Ratnākaraśānti identifies the ālambana or “support” of the Second Yoga as the way in which phenomena arise; at this level, the yogi understands that the causal processes which produce phenomenal appearances are in fact nothing more than the interactions of the eight collections of consciousness. Therefore, all appearances are “mind only.” Elsewhere Ratnākaraśānti describes the “essence of ‘mere representation’” as a “direct means of reliable knowledge,” because it is “undistorted.”265 Assuming that he means this literally, the “essence of mere cognitive representation” presumably functions as the pramāṇa for the Second Yoga, the knowledge that all entities and phenomenal appearances are merely mental representations. However, at the stage of the Second Yoga, the mental representations in question are still reflexively known, as the yogi has realized that nothing exists independently of the mind. The act of observation is constitutive of the reality that is being observed. Therefore, reflexive awareness also serves as the pramāṇa by which the dependent nature, the prameya of the Second Yoga, is known. The “essence of mental representation-only” is, in fact, the reflexive nature of awareness, since the mind which knows that all phenomena are nothing other than mind is the same as the mind which is known as the nature of all appearances. The yogi’s 264 265 Ibid., 42-43. Ratnākaraśānti, Pith Instructions, ff. 229.a.2. rnam par rig pa tsam gyi (P/N/S: gyis) gnyug ma rang gi ngo bo myong ba ni mgnon sum gyi tshad ma yin te/ ma 'khrul ba'i phyir ro/ 180 nondual mind reflexively and immediately knows that all appearances are only mental representations. The difference between the way in which reflexive awareness serves as a pramāṇa at the level of the First Yoga and at the level of the Second Yoga concerns the relative subtlety of the two stages’ respective supports. As already noted, the support for the First Yoga is the “full range of phenomena.” The point here is that the yogi, entering into nondual meditation, directly observes the dualistic construction of all phenomenal experiences, the fact that every ordinary phenomenal experience is known through an intentional or dualistic consciousness. The support for the Second Yoga, by contrast, is the “way in which phenomena arise,” namely as the nondual flow of mental causation. At the first level, the yogi is reflexively aware of dualistic bifurcation; the yogi observes the deep-seated and beginningless psychological imprint of duality, manifest in the yogi’s own mind as the subjective aspect of the grasper and the objective aspect of the grasped. At the second level, the yogi is reflexively aware of consciousness, or awareness, itself; the yogi observes the nature of mind, by means of the nature of mind, which is reflexive awareness. The Third Yoga is the observation of suchness (de bzhin nyid, *tathatā). From the Pith Instructions for the Perfection of Wisdom, In the third stage the yogis apply non-appearance to the false marks of manifest appearances, as meditators on the formless realms pass beyond the perception of form, by perceiving infinite space. Thereby they relinquish all false conceptual marks of the object and subject and view them as space, utterly immaculate and limitless, empty of duality, sheer clarity [i.e. luminosity]. They realise that all phenomena are formless, undemonstrable, and unobstructed, their one essential characteristic being the absence of characteristics. By combining calm abiding and penetrating 181 insight, they realise that all appearances are reflected images of emptiness and apprehend the suchness of all phenomena as they are.266 At this level, the yogi makes direct contact with the ultimate nature of reality, seeing it “just as it is” (ji lta ba bzhin du, *yathābhūta). By definition, the prameya is suchness, the nature of reality, itself. But yet again, this prameya is known by means of reflexive awareness, which is not separate from suchness. As Ratnākaraśānti writes, Luminosity is established as being real, because as an awareness that is free from distortion, it is a direct means of reliable knowledge. Luminosity is the inherent nature which gives illumination, because it is not positioned by the contamination through which some experience would be designated as distorted.267 Above, Ratnākaraśānti argued that there is no pramāṇa apart from the luminosity of reflexive awareness. Here, Ratnākaraśānti maintains that, in addition to being “the direct means of reliable knowledge,” luminosity itself is the “inherent self-nature” (gnyug ma’i rang bzhin) and established as a real thing (dngos po nyid du grub pa). All entities and appearances are only mind, and the ultimate nature of the mind is the luminosity of reflexive awareness. Thus, at the level of the Third Yoga, the pramāṇa and the prameya are the same. The ultimate nature of reality is known by means of the ultimate nature of reality, which is the reflexive nature of 266 267 Yael Bentor, “Fourfold Meditation,” 43. Ratnākaraśānti, Pith Instructions, 229.b.7-230.a.1. yang gsal (D/C: bsal) ba de rig pa'i 'khrul pa dang bral ba nyid du mngon sum yin pa'i phyir dgnos po nyid du grub pa yin no/ gsal ba ni gsal pa sbyin pa nyid kyis gnyug ma'i rang bzhin te/ gang gis myong ba 'di 'khrul par 'jog pa'i bslad pas ma bzhag pa med pa'i phyir ro/ 182 awareness. In this way, because it is the pramāṇa for both relative and ultimate objects of knowledge, reflexive awareness serves as the bridge between relative and ultimate truth. Recall that, according to Ratnākaraśānti, as long as any phenomena appear, there is still cognitive distortion. At the level of the Second Yoga, appearances still arise. Therefore, the Second Yoga is still in some sense contaminated by concepts, at least to the extent that any phenomenal appearance as such is necessarily conceptual, and by extension, distorted. Interestingly, Ratnākaraśānti maintains that this is also true of Buddhas and high-level Bodhisattvas, who must retain some conceptual distortion at the provisional level of dependently-originated appearances in order to communicate with ordinary beings: For one with transcendent awareness, all phenomena are seen as illusory, like in the eight examples. By contrast, pure worldly awareness is “pure” insofar as this awareness thoroughly resolves suchness, but it is “worldly” insofar as it is still distorted; by the force of necessity, even perfect enlightenment at the level of Buddhahood retains a tiny amount of cognitive distortion, because the purity is worldly in nature.268 The Second Yoga is therefore a kind of in-between state. All appearances are reflexively known to have the nature of mental representations only, but they still appear. To the extent that phenomena appear, they are differentiated: some phenomena are blue, other phenomena are yellow, and so on. By contrast, at the Third Yoga, there are no longer any differentiated appearances whatsoever. And because there is no longer any differentiation, only nondual and undifferentiated luminosity remains. The reflexive nature of awareness is the same throughout this progression, but unlike the prior two stages the Third Yoga is fully and truly nonconceptual; 268 Ratnākaraśānti, Pith Instructions, ff. 226.b.1. de bzhin du sangs rgyas kyi sa la dgos pa’i (P/N/S: dgongs pa’i) dbang gis rdzogs pa'i byang chub cung zad 'khrul pa yin te/ dag pa 'jig rten pa'i bdag nyid yin pa'i phyir ro/ 183 it is a genuine encounter with the ultimate nature of reality, or thusness. In this way, reflexive awareness serves as the means by which the mental continuum, though initially distorted, is able to come into direct contact with the nature of reality. Ratnākaraśānti furthermore explains that, at the level of the Third Yoga, all phenomena are seen to be of “one taste” (ro gcig)269 like the “center of the sky” (nam mkha’i dkyil).270 Knowledge based on the characteristics of phenomena is not ultimately reliable, since the perception of characteristics (svalakṣaṇa) or dualistic marks (nimitta) is necessarily distorted, conditioned by ignorance. Having cultivated reflexive awareness, however, the yogi has access to truly reliable knowledge. How is it that all phenomena are within the transcendent wisdom of the Buddhas and the Bodhisattvas? They appear as one taste— suchness—like limitless space utterly without stains. And why is that? Because the marks of duality have disappeared.271 The point here is that, as Ratnākaraśānti has already argued, all phenomena have the nature of luminosity. The luminosity that is the nature of all phenomena is the same luminosity that is the nature of the mind, reflexive awareness. The yogi has direct access to ultimate truth, because the yogi’s cognition (pramiti) is not in any way conditioned by the psychological imprint of ignorance; all differentiated phenomenal content such as blue and yellow has been replaced by nondual luminosity. 269 Ratnākaraśānti, Pith Instructions, 224.b.6. 270 Ratnākaraśānti, Pith Instructions, 226.a.6. 271 Ratnākaraśānti, Pith Instructions, ff. 229.a.5-6. ci ste sangs rgyas dang byang chub sems dpa' rnams kyis (D/C: kyi) 'jig rten las 'das pa'i ye shes la chos thams cad ji ltar zhe na/ nam mkha' dri ma med pa 'ba' zhig mtha' med pa dang 'dra bar de bzhin nyid ro gcig tu snang ngo/ ci'i phyir zhe na/ mtshan ma de rnams nub pa'i phyir ro/ 184 Despite the fact that there is no longer the appearance of any differentiated phenomena such as blue-patches and yellow-patches, the yogi has ultimately reliable epistemic access to information about blue-patches and yellow-patches, because their blueness or yellowness are not in fact epistemically reliable. Only their luminous nature is, in the final analysis, epistemically reliable. Therefore, because the yogi is reflexively aware of the luminous nature of phenomena such as blue-patches and yellow-patches, there is no need for differentiated conceptual knowledge of phenomenal characteristics such as blue and yellow. However, the third Yoga still has a type of phenomenal content: luminosity. There is no longer differentiation, but there is still the “appearance” of luminosity. The Fourth Yoga, on the other hand, is strictly non-appearance. From the Pith Instructions for the Perfection of Wisdom: In the fourth stage, the yogis pass beyond the subtlest conceptualisation of phenomena. Without exertion and without conditioning, they realize experientially, through a direct perception, the suchness of all phenomena. They realise the complete vanishing of the marks of phenomena and the nature of phenomena, the enlightened wisdom, which is non-dual, free of appearances and apprehension, the supra-mundane non-conceptual calm abiding and penetrating insight.272 At the level of the Fourth Yoga, the dependent nature of phenomenal appearances has been completely emptied of any and all distortion, any and all duality, any and all psychological imprints—and any and all appearances. The prameya of the Fourth Yoga is, in other words, the perfected nature. The pramāṇa is the “inherent self-nature” of luminous reflexive awareness, which is not separate from suchness, the nature of reality: by means of reflexively knowing the 272 Yael Bentor, “Fourfold Meditations,” 43. 185 luminous nature of reality, the yogi realizes the non-appearance of even that luminosity. Ratnākaraśānti does not adopt the language of the “union of luminosity and emptiness” (gsal stong zung ‘jug), a common formulation in later Tibetan literature. But this seems to be the general thrust of his point. While it is irrefutable, and there are no means of reliable knowledge that exist independently of it, even the luminous nature of reflexive awareness is empty. 186 2. Reflexive Awareness and the Three Natures As elaborated upon above, Ratnākaraśānti frames his ontology primarily in terms of the three natures: constructed, dependent, and perfected, which he respectively describes as existing “through imputation, substantially, and ultimately.” Ratnākaraśānti relates three natures theory with the pramāṇa discourse by maintaining the epistemic unreliability of any type of phenomenal appearance: “As there is no third kind of bundle apart from something which is either singular or manifold, blue-patches and so on are not established as either internal or external objects, and are therefore false.”273 This tracks very closely to Śāntarakṣita’s position in the Ornament of the Middle Way. As a Yogācārin, Ratnākaraśānti is committed to the unreality of the constructed nature. But Ratnākaraśānti also adopts “False Imagism,” the view that the cognitive “images” produced by sensory consciousness are false and deceptive, or not ultimately real. As should be clear from the preceding discussion, considered in the light of his refutation of even the bare phenomenal blue-patch, Ratnākaraśānti did not ultimately consider the dependent nature any more real than the constructed nature. It is not only duality that is imaginary, illusory, unreal; while the dependent nature is causally-produced, and therefore “substantially existent,” according to Ratnākaraśānti it is also illusory and unreal. Dependently-originated cognitive images—even when nondually or reflexively known—are not actually epistemically reliable, and are therefore “false.” 273 See above, p. 132. 187 Above, it was observed that Ratnākaraśānti considered phenomenal characteristics such as “blue” to be epistemically unreliable, since they only arise due to the force of deluded “psychological imprints,” (Skt. vāsanā, Tib. bags chags), identified with the “seeds” of delusion present in the storehouse consciousness. The luminosity of reflexive awareness, by contrast, is epistemically reliable. As Ratnākaraśānti writes, Although their nature is luminosity, blue-patches and so on are false since they are harmed by analysis. But that luminosity is established as being real, because as an awareness that is free from distortion, it is a direct means of reliable knowledge. Luminosity is the inherent self-nature which gives illumination, because it is not positioned by the contamination through which some experience would be designated as distorted.274 That is to say, phenomenal perception, the apprehension of sensory content, is always unreliable, because it is not actually real. Luminosity, by contrast, is real; therefore, luminosity is reliable. For Ratnākaraśānti, even sensory perception is not really a direct means of reliable knowledge, i.e. a pratyakṣa-pramāṇa. This is because sensory perception is conditioned by the psychological imprint of duality; the causally-produced phenomenal appearances of the dependent nature are nothing more than the product of beginningless ignorance. Definite perceptual judgments (niścaya) of blue-patches and so on are possible, but they do not actually yield ultimately reliable information about the phenomena they are supposedly oriented toward. Only the nondual luminosity of reflexive awareness yields ultimately reliable information; the only ultimately 274 Ratnākaraśānti, Pith Instructions, ff. 229.b.7-230.a.1. yang sngon po la sogs pa de gsal (D/C: bsal) ba'i rang bzhin yang gnod pa yod pa'i phyir brdzun pa'o/ yang gsal (D/C: bsal) ba de rig pa'i 'khrul pa dang bral ba nyid du mngon sum yin pa'i phyir dgnos po nyid du grub pa yin no/ gsal ba ni gsal pa sbyin pa nyid kyis gnyug ma'i rang bzhin te/ gang gis myong ba 'di 'khrul par 'jog pa'i bslad pas ma bzhag pa med pa'i phyir ro/ 188 valid perceptual judgment (niścaya) is the ascertainment (niścaya) of the nondual luminosity of reflexive awareness. In other words, ultimately, all experience is reducible to mind’s reflexive awareness of mind. And ultimately, this reflexive awareness is the only truly direct pramāṇa, as the only truly reliable information about phenomena is their luminous nature; phenomenal characteristics cannot withstand analysis, and are therefore not reliable sources of information. This is because information such as the blueness of a blue-patch is only able to be perceived on the basis of the prior psychological imprint to perceive “blue.” Since this imprint, like all imprints, is contaminated by ignorance qua dualistic fixation, the experience itself is contaminated. And insofar as this distorted experience is “subsequent” to the initial, undistorted, nonconceptual moment of luminous reflexive awareness, any experience of blue is an inferential, “subsequent measurement” on the basis of that first moment of reflexive awareness. Consequently, Ratnākaraśānti draws one of his key conclusions: Luminosity is not insentient, and it is irrefutable, as it has the nature of being utterly and completely luminous. If this luminosity were not established, then nothing would appear, and in consequence nothing at all could be established. And if it is established, then because it itself is consciousness, all phenomena are established as having the nature of consciousness. Needless to say, in asserting that “all phenomena are established as having the nature of consciousness,” Ratnākaraśānti explicitly adopts an idealistic ontology. But his argument rests on epistemological concerns, namely the unreliability of ordinary, dualistic perceptions. 189 Contemporary discussions of Yogic Practice philosophy, such as Lusthaus’ take on the issue,275 have often treated epistemological and ontological issues separately, and many have even gone so far as to claim that Yogic Practice idealism is essentially limited to a kind of epistemic closure. This perfectly describes the Sautrāntika position, but fails to distinguish what separates the Sautrāntika from the Yogācāra ontology. In Sautrāntika epistemology external objects do really exist, in spite of the fact that, first of all, they no longer exist by the time they are perceived, and second of all, there is never any direct access to those external objects independently of the cognitive process. In other words, Sautrāntika philosophy describes precisely this type of epistemic closure, where external objects exist even though they are entirely cut off from the internal subject. The Yogācāra critique of Sautrāntika epistemology, meanwhile, hinged on the rejection of just those external objects, or more precisely the rejection of the opposition between internality and externality (or any other type of duality). More importantly, the rigid exegetical distinction between ontology and epistemology, motivated by the exigencies of the Western philosophical tradition, obscures the point: for Ratnākaraśānti, as for all Yogācārins, epistemology and ontology are not two separate categories. For commentators on both sides of the “Imagism” (ākāravāda) debate, the position that images are false is tantamount to the position that they do not really exist, and vice versa. Therefore, in denying the epistemic reliability and ontological reality of cognitive images, Ratnākaraśānti also argues that phenomenal appearance as such necessarily bears traces of ignorance or delusion. For this reason, Ratnākaraśānti posits a crucial distinction between “worldly awareness” (‘jig rten pa’i ye shes = laukikajñāna) and “transcendent awareness” (jig rten las ‘das pa’i ye shes = 275 Dan Lusthaus, Buddhist Phenomenology, 437-442. 190 lokottarajñāna). As has already been discussed, Ratnākaraśānti explains “False Imagism” with respect to the dis- or non-appearance of cognitive images within the worldly consciousness: This is the tradition which maintains the non-existence of aspects: In the worldly awareness, Once cognitive images disappear, The Lord of non-appearance, Nondual suchness without conceptual structures, arises. Although worldly consciousness possesses images, Those images are thoroughly determined To be both false and unreal. Thus [consciousness] is said to be “without images.” Interestingly, Ratnākaraśānti felt that this position entailed a rather counterintuitive consequence, already mentioned above: Pure worldly awareness is “pure” insofar as this awareness thoroughly resolves suchness, but it is “worldly” insofar as it is still distorted; by the force of necessity, even perfect enlightenment at the level of Buddhahood retains a tiny amount of cognitive distortion, because the purity is worldly in nature.276 As this passage makes clear, Ratnākaraśānti was thoroughly committed to the position that any kind of phenomenal appearance as such is necessarily mistaken, and that even the Buddhas necessarily embrace some small measure of delusion, in order to be able to communicate with ordinary beings for whom phenomenal appearances still seem to be real. If Buddhas always 276 Ratnākaraśānti, Pith Instructions, ff. 226.a.7-226.b.1. de la dag pa 'jig rten pa'i ye shes zhes bya ste/ ye shes des de kho na nyid yongs su gcod pa nyid kyis dag pa yin la/ 'khrul pa nyid kyis 'jig [452] rten pa yin no/ de bzhin du sangs rgyas kyi sa la dgos (P/N/S: dgongs) pa'i dbang gis rdzogs pa'i byang chub cung zad 'khrul pa yin te/ dag pa 'jig rten pa'i bdag nyid yin pa'i phyir ro/ 191 exercised only their perfectly pure, transcendent awareness, they would be literally unable to perceive the world. In this manner, Ratnākaraśānti posits a strict distinction between what is “worldly” and what is “transcendent.” Appearances, by definition, are worldly; only non-appearance, the Fourth Yoga, is transcendent. 3. Reflexive Awareness and the Two Truths Earlier, it was observed that Ratnākaraśānti’s three-fold ontological scheme of “imputed,” “substantial” and “ultimate” existence draws an implicit opposition between “imputed existence” and “substantial existence.” Here, however, Ratnākaraśānti denies that even the “substantial” existence of the dependent nature is truly real. If, then, neither cognitive “images” nor the phenomena they represent to consciousness are real—and, furthermore, at the highest level of contemplative praxis, the ultimate truth is non-appearance—how is the relative or conventional truth to be understood, particularly with respect to the ordinary sensory perceptions of unenlightened beings such as ourselves? Ratnākaraśānti does not directly answer this question, but provides a refutation of two related views that he considers unacceptable. These are the views of a very particular commentarial tradition: Some followers of the Middle Way tradition claim that whatever is generally renowned in the world is the relative truth. External objects are generally known in the world, as are the mind and mental functions; so these, too, exist in terms of the relative truth. [They say] mind and mental functions are distorted, and reflexive 192 awareness is like a sword that cannot cut itself, therefore it is contradictory.277 As is well known, “whatever is generally renowned in the world” is the definition of the relative truth advanced by Candrakīrti in his Entering the Middle Way and autocommentary. Although Ratnākaraśānti does not mention his opponent by name, there is no question that he is here arguing against the views promulgated by the author of the Entry into the Middle Way, as Ratnākaraśānti goes on to describe the latter’s unique and original conception of Buddhahood: Once cognitive distortion has been extinguished, the ultimate is realized. Therefore Buddhas are those in whom the mind and mental functions have been extinguished. At that point, the “arising” of a Buddha is posited as an experience in conformity with the ultimate, which is the non-arising of all phenomena. In order to accomplish the various aims of sentient beings within unlocated Nirvāṇa, imbues form with blessings. Having done that, genuine enlightenment is made manifest.278 According to Candrakīrti, the Buddha has “exhausted all mind and mental functions.” From John Dunne’s translation of Candrakīrti’s autocommentary to the Entry into the Middle Way: Because they continue to live with the ignorance which is defined as an obstruction to knowing the knowable, the Āryas’ sphere of activity includes appearances; the merely spurious appears to them, 277 Ratnākaraśānti, Pith Instructions, ff. 228.b.3-4. gang yang dbu ma pa kha cig na re 'jig rten la (D/C: las) grags pa ni kun rdzob yin te/ phyi rol gyi don yang 'jig rten la grags pa yin pas/ sems dang sems las byung ba bzhin du de yang kun rdzob tu yod pa yin no/ sems dang sems las byung ba ni 'khrul pa yin te/ rang gi rig pa 'gal ba'i phyir ral gri'i sos rang mi gcod pa bzhin no/ 278 Ratnākaraśānti, Pith Instructions, ff. 228.b.5-6. de bas na 'khrul pa zad pas don dam pa rtogs (P/N/S: rtog) pa'i phyir ro/ sangs rgyas pa rnams ni sems dang sems las byung ba zad pa nyid do/ gang gi tshe de rnams kyi skye ba de'i tshe thams cad skye ba med pa'i don dam par rjes su byed pa thob pas myong bar gzhag go/ mi gnas pa'i mya ngan las 'das pa ni sems can sna tshogs pa'i don bya ba'i phyir gzugs byin gyis brlabs te/ gzhag nas/ yang dag pa'i byang chub mngon du byed do zer ba dag ni smad pa las kyang smad par 'gyur ba yin te/ 193 but it does not appear to those whose sphere of activity has no appearances. The Buddhas have fully awakened to (a complete understanding of) all elemental things. Therefore, the fluctuation of mind and mental functions has completely ceased (for them).279 Insofar as the realization of suchness or the nature of reality is in some sense mental, this stance entails that the Buddhas have no realization. Candrakīrti attempts to get around this objection by positing that suchness is strictly an absence, and that consequently “knowledge” of suchness is a contradiction in terms: If suchness (or ultimate reality) is the pacification (of all cognitive structures), then one does not have cognitions of it; and if one does not have cognitions of it, it does not make sense for one to have a determinate cognition [niścaya] whose object is (that) knowable (thing, suchness). How can non-knowledge be knowledge? They contradict each other.280 Having denied that the Buddhas possess any knowledge or realization of suchness, Candrakīrti goes on to describe Buddhas as a type of perpetual motion machine, continually dispensing blessings through the force of habit, without any mental experience whatsoever: The strong potter’s wheel turns very quickly because he has long striven at it. Even though the potter no longer exerts himself, the wheel turns… Likewise, while (a Buddha) makes no conceptual effort, s/he abides in the body whose essence is Dharma, and that [Dharma Body’s] activity is impelled by beings’ distinctive virtue 279 Candrakīrti, Commentary to the Entry into the Middle Way (MAB55a) trans. John Dunne in “Thoughtless Buddha, Passionate Buddha,” Journal of the American Academy of Religion 64, no. 3 (1996), 544. 280 Ibid., 546. 194 and the special prayers (that the Buddha made when s/he was a Bodhisattva)—how inconceivable!281 Although Ratnākaraśānti argues elsewhere that phenomenal appearances are necessarily deluded, he finds the notion that Buddhas are without any mental experience or realization at all—i.e. that Buddhas are mindless, like a potter’s wheel that has been “set up” to dispense blessings—utterly ridiculous, and pokes fun at Candrakīrti’s interjection (“how inconceivable!”): This is even more inconceivable than the inconceivable! Since there are no mind and mental functions, there is not any realization, either—so how can there be any ultimately authentic, perfect enlightenment at all?282 Ratnākaraśānti goes on to argue that the ultimate realization of the Buddhas is necessarily mental: “Realization is just that [absence],” he says. But realization itself is mental, therefore it is not the case that mind and mental functions no longer exist. Furthermore, if the ultimate is experienced as something else,283 then that experience is distorted. And if it is not different, then it is reflexively known.284 281 Ibid., 549. 282 Ratnākaraśānti, Pith Instructions, ff. 228.b..6-7. smad pa las kyang smad par 'gyur ba yin te/ gang gi phyir 'di ltar sems dang sems las byung ba 'di dag med pa'i phyir rtogs (D/C: rtog) pa nyid kyang med na/ don dam par yang dag par rdzogs pa'i byang chub lta ga la yod/ 283 284 That is, something external to the mind. Ratnākaraśānti, Pith Instructions, ff. 228.b.7-229.a.1. ci ste yang 'di ni rtogs pa kho na yin no zhe na/ de nyid sems yin pas sems dang sems las byung ba med pa ma yin no/ gzhan yang don dam pa de gzhan (D/C: bzhin) du gyur pa myong na ni 'khrul par 'gyur la/ tha mi dad na (P/N/S: pa) ni rang rig par 'gyur ro/ 195 Thus, for Ratnākaraśānti, any undifferentiated ultimate realization is necessarily and by definition the realization of reflexive awareness, stripped of differentiated phenomenal appearances and the dualistic distortion this entails. The luminous nature of reflexive awareness, the ultimate nature of the mind, just is suchness. As noted in the previous discussion of the Four Yogas, reflexive awareness remains the same as the yogi transitions from the realization of nonduality, to the realization of mind-only, to the realization of suchness, to the realization of non-appearance. In all of these cases, reflexive awareness is the pramāṇa; only the prameya is different. Thus reflexive awareness is the bridge between relative and ultimate, as in addition to being identical with the ultimate nature of reality it is also the backstop which guarantees the epistemic reliability of cognitions in general—even, and especially, relative cognitions of dependently-originated phenomena. This is why Ratnākaraśānti so flippantly dismisses the claim that the reflexive nature of awareness is contradictory, ostensibly because awareness cannot act upon itself in the same way that a knife cannot cut itself: a denial of the existence or the efficacy of reflexive awareness is tantamount to a denial of any and all phenomenal appearances, i.e. a species of nihilism. Such a denial also contradicts basic Buddhist doctrine regarding the realization of the Buddhas, since it is not taught that the Buddha no longer has any mind at all. Classically speaking, Buddhas have completely purified and perfected their minds, such that they are absolutely selfless and are thus able to spontaneously benefit limitless sentient beings. The mind of the Buddhas is taught to be transcendent, and unintelligible by ordinary sentient beings, but that is a far cry from saying that the Buddhas do not have a mind or that the mind no longer exists upon the attainment of Buddhahood. 196 As for the epistemological dimensions of Candrakīrti’s arguments, Ratnākaraśānti does not directly rebut the claim that the relative truth is “whatever is renowned in the world.” He does, however, engage a related claim: that the Middle Way philosophy of the Two Truths is contained within the “illusoriness of phenomena.” This is a re-statement of Candrakīrti’s view that in Middle Way philosophy phenomena are understood to be illusory and there is no special need to engage in the pramāṇa discourse in order to understand this illusoriness: Some adherents of the Middle Way claim that the Path of the Middle Way is just the illusoriness of phenomena. But if these [phenomena] are proven to be nonexistent by a means of reliable knowledge, then it is impossible to establish their existence by means of anything at all. So what need is there to mention establishment by mere appearance, which is not even a means of reliable knowledge? If everything is just nonexistent, then what is it that is designated, and in dependence upon what?285 Candrakīrti’s position is that there is no special need to engage in elaborate refutations of appearances, since phenomena are already illusory and unreal. The relative truth is their mere (“illusory”) appearance, while the ultimate truth is their absence of essence-svabhāva or inherent existence. Ratnākaraśānti’s point is that mere argumentation cannot establish this to be the case; as he says, if phenomena are proven to be nonexistent by a pramāṇa, then there is nothing that can establish them to be existent. In other words, if Candrakīrti’s assertions were really means of reliable knowledge about phenomena, then they would actually be nonexistent, and therefore could not appear. But they do appear, and this is exactly why the type of analysis Ratnākaraśānti 285 Ratnākaraśānti, Pith Instructions, ff. 229.b.1-2. gzhan yang dbu ma pa chos thams cad sgyu ma lta bu nyid dbu ma'i lam yin no zhes zer te/ 'di ltar gal te 'di dag tshad mas med par ma grub na de'i phyir 'gas kyang yod pa bsgrub par mi nus so/ tshad ma ma yin pa la snang ba tsam gyis lta smos kyang ci dgos/ thams cad med pa nyid yin na ci zhig nye bar blangs nas ci zhig gdags/ 197 engages in is necessary. The “mere appearance” of phenomena is insufficient to establish their reality in any sense, particularly on an account (such as Ratnākaraśānti’s) where appearances as such are epistemically unreliable. On a related note, Ratnākaraśānti also critiques Candrakīrti’s anti-pramāṇa rhetorical stance: [Since] all consciousnesses are also deceptive, in the way that everything is false, thus not even the slightest direct perception is established. Therefore, since there is no reliable ascertainment via the three-fold syllogism, all evidence is only spurious evidence.286 Furthermore, all the refutations of something or proofs of something else would also not be established—because these rely on the means of reliable knowledge, while you assert that there are no means of reliable knowledge! If you refute through mere words, it is said that you yourself, among the cows, have the view of a heretical Far-Thrower. Others deny that you are a man or a follower of the Middle Way. The intention of the statement that all phenomena lack an inherent essence, and so on, is reliably determined by the Sūtra Unraveling the Intent and the Ornament of the Great Vehicle Sūtras, and also by Maitreya, Asaṇga, and Nāgārjuna. If you accept that there are means of reliable knowledge, then they are established in the manner of the three natures. So abandon the contention that everything is false! And if you say, “But we don’t assert that there are means of reliable knowledge!”, well, people who debate without any means of reliable knowledge are just clowns.287 286 287 See Appendix, note 343 concerning “evidence” (Skt. hetu, Tib. gtan tshigs). Ratnākaraśānti, Pith Instructions, ff. 229.b.3-6. yang thams cad brdzun pa'i tshul gyis shes pa thams cad kyang brdzun pas mngon sum cung zad kyang ma grub pa’o/ de'i phyir tshul gsum nges pa med pas gtan tshigs thams cad kyang gtan tshigs ltar snang bar 'gyur ro/ gzhan yang rang dang gzhan la dgag pa dang sgrub pa thams cad ma grub ste/ de dag thams cad ni tshad ma la rag lus pa'i phyir dang/ tshad ma yang khas mi len pa'i phyir ro/ tshig tsam gyis 'gog na ni khyod nyid ba lang rnams las mu stegs rgyang pan gyi lta ba can no zhes gzhan gyis smras pa na mi dang dbu ma bkag pa 'gyur ro/ chos thams cad ngo bo nyid med pa la sogs par gsungs pa'i dgongs pa yang mdo sde dgongs 'grel du gsungs pas (P/N/S: pa) nges te/ de nyid byams pa dang thogs med dang klu sgrub kyis kyang ngo/ tshad mar khas len na ni rang bzhin gsum gyi tshul du grub pas thams cad brdzun du smra ba thong zhig/ gal te bdag cag ni tshad (C: mtshan) ma mi 'dod pa'i phyir ro zhe na/ tshad ma med pa'i phyir rtsod par smra ba ni bzhad gad pa nyid yin no/ 198 In this manner, Ratnākaraśānti turns the argument from “mere appearance” and “general renown” on its head. By committing himself to the position that an adherent of the Middle Way tradition should not study the discourse of pramāṇa, and attacking the very idea of means of reliable knowledge, Candrakīrti defeats his own arguments—if there are no true means of reliable knowledge, how can one’s argument to this effect serve as a means of reliable knowledge that there are no such means of reliable knowledge? The larger point at stake here, however, concerns the ultimate efficacy of pramāṇas. In pramāṇa theory, the “object of reliable knowledge” or prameya is real by definition. 288 The point is that a pramāṇa, qua pramāṇa, guarantees the reliability of a cognition of its attendant prameya. Therefore, if something is “established” to exist by means of such an instrument of reliable knowledge, then as Ratnākaraśānti says above it cannot be refuted “even by a hundred” other instruments. Mutatis mutandis, if something is established by a pramāṇa to not exist, then its mere phenomenal appearance is not “instrumental” (i.e. not a pramāṇa) for establishing its true existence. This is the insight behind Dharmakīrti’s discussion of phenomenal distortions (bhrānta) such as hairs floating in the eye; phenomenal appearance in and of itself does not guarantee the epistemic reliability of information concerning what appears. Since in the philosophy of Candrakīrti all phenomena are merely apparent, and their relative existence is identical to their ultimate lack of self-nature, there is no way within this tradition to distinguish between the “substantially existent” causally efficacious phenomena of the dependent nature, and the “imputed” phenomena of the constructed nature. This is, needless to say, a major problem, 288 Cf. Dunne (2004), pp. 35-45 199 tantamount to the claim that floating hairs in the sky are no more or less illusory than the hair on a human being’s head. Thus, according to Ratnākaraśānti, mere appearance or the illusoriness of phenomena is not the proper way to understand the philosophy of the Middle Way. Following Śāntarakṣita, he argues that the Middle Way is best understood in terms of the pramāṇa discourse. For Ratnākaraśānti, as discussed above, the reflexivity of awareness is the only ultimately reliable pramāṇa. Together with his aforementioned views, this would seem to indicate that Ratnākaraśānti holds that the reflexive nature of awareness is ultimately real, at least insofar as he also follows Dignāga and Dharmakīrti in maintaining that the object, means, and result of a knowledge-act are all ultimately inseparable. This view diverges from Śāntarakṣita’s, who argues near the end of his autocommentary to the Ornament of the Middle Way that the reflexive nature of awareness is only conventionally real: “Reflexive awareness is classified as conventional truth because it cannot bear an analysis which looks for a singular or a manifold nature.” 289 However, it should be remembered that earlier in the Ornament of the Middle Way, while examining the Sautrāntika epistemology, Śāntarakṣita argues that reflexive awareness is a necessary element of the perceptual process, and goes so far as to define reflexive awareness as the quintessential feature of consciousness as opposed to insentience: Consciousness arises in the opposite manner as that which has the nature of insentience. Anything which has the nature of not being insentient, has the essential quality of [reflexive] awareness290 [MA 16] 289 290 Trans. Blumenthal, Op. Cit. rnam shes bem po’i rang bzhin las/ bzlog pa rab tu skye ba ste/ bem min rang bzhin gang yin pa/ de ‘di bdag nyid shes pa 200 It is evident that Ratnākaraśānti’s presentation of reflexive awareness diverges in certain key respects from Śāntarakṣita’s presentation in the Ornament. However, in this verse, even Śāntarakṣita acknowledges that the reflexivity of awareness is an essential feature of consciousness as such. And, as has already been noted, Ratnākaraśānti acknowledges the ultimate emptiness of reflexive awareness, as in the Fourth Yoga even the luminosity of reflexive awareness no longer appears. Thus the key difference between Śāntarakṣita and Ratnākaraśānti may well come down to the eleutheriological and epistemological significance (as opposed to the ontological status) of reflexivity in their respective systems of philosophy. For Śāntarakṣita, Middle Way ontology necessitates a denial that even the reflexive nature of awareness is ultimately real, as part of a general emphasis on the negation of any type of svabhāva. For Ratnākaraśānti, Yogic Practice epistemology requires that the reflexive nature of awareness, and perhaps only the reflexive nature of awareness, is ultimately valid as a means of reliable knowledge, and therefore (in some sense) ultimately real. But both are committed to a synthesis of the Middle Way and Yogic Practice schools via the methodology of Buddhist pramāṇa theory, and in any event it is not prima facie evident that these views are in direct conflict. Certainly, in equating the luminosity of reflexive awareness with thusness, but placing the cognition of thusness at the level of the Third Yoga beneath the Fourth Yoga of non-appearance, Ratnākaraśānti appears to follow Śāntarakṣita in maintaining that reflexive awareness itself is, finally, empty. The only question to be asked, then, is whether the emptiness of ordinary phenomena is different from the emptiness of luminosity. 201 IV. Conclusion Ratnākaraśānti’s Pith Instructions is a complex and multifaceted text drawing on several major traditions of Buddhist scholastic philosophy. Although the metaphor of a “sliding scale” of analysis has so far only been applied to the works of Dharmakīrti and Śāntarakṣita, Ratnākaraśānti appears to embrace a very similar shift in perspective; thus the proposition that phenomena such as blue have the nature of consciousness can alternately be described as false (insofar as this claim is susceptible to neither-one-nor-many style analysis) and true (insofar as Ratnākaraśānti asserts that all phenomena have the nature of luminosity). The question, of course, is which level is “higher” on the scale, and whether or not Ratnākaraśānti’s final position is in accord with Śāntarakṣita’s. On a related note, there is the interesting—and perhaps irresoluble—issue of the relationship between the Fourth Yoga of “non-appearance” and Śāntarakṣita’s Middle Way ontology of emptiness. As previously discussed, Śāntarakṣita insists that the idealistic ontology of the Yogic Practice school is only suitable for describing the relative truth, and that the reflexive nature of awareness is only relatively or conventionally existent. This position was adopted by Mi pham and integrated into his articulation of the “Consequentialist” style of Middle Way exegesis. However, the dominant “Consequentialist” tradition followed Tsong kha pa in denying even the conventional existence of reflexive awareness. Tsong kha pa’s account of how consciousness can function in the absence of reflexive awareness is both unnecessarily complicated and hermeneutically (not to mention logically) suspect; accordingly, it will not be considered here. But by way of conclusion, it is worth examining Garfield’s critique of Williams’ Reflexive Nature of Awareness. Although Williams misconstrues many of the technical details of 202 reflexive awareness, particularly in relation to pramāṇa theory, his overview of the ontological and epistemological issues at stake in Śāntarakṣita’s account of reflexive awareness is generally accurate. Williams frames his presentation as a “Tibetan Madhyamaka” defense of reflexive awareness, and presents the views of Ju Mi pham. Mi pham, in his commentary on Śāntarakṣita’s Ornament of the Middle Way, follows Śāntarakṣita in arguing that the reflexively or immediately-present nature of what is presented to consciousness—what Williams terms “selfawareness (ii)”—does exist conventionally, though not ultimately. Garfield’s critique of Williams is noteworthy, not because it faithfully reproduces either Tsong kha pa’s arguments against the conventional existence of reflexive awareness or the standard dGe lugs interpretations thereof—Garfield glibly dismisses the typical dGe lugs critiques of reflexive awareness as “terrible arguments”291 despite the fact that they “[crop] up in discussion with dGe lugs scholars with disturbing regularity”292—but precisely on account of its ahistorical timbre. Before turning to Garfield, however, it is important to recapitulate the most salient differences between Ratnākaraśānti and Śāntarakṣita. In Buddhist doxographical terms, Śāntarakṣita stands firmly within the Middle Way tradition, while Ratnākaraśānti ultimately lands in the Yogic Practice camp. Both attempt a synthesis of Madhyamaka, Yogācāra, and Pramāṇa theory, and are thus operating in similar ontological, phenomenological, and epistemological territory. However, their eleutheriological approaches are markedly different. Śāntarakṣita evidently considered the tendency toward ontological reification as the most significant obstacle to progress along the path. Accordingly, Śāntarakṣita’s main concern throughout the Ornament of the Middle Way (and other works) is to deconstruct the bases of any 291 Jay Garfield, “The Conventional Status of Reflexive Awareness: What’s at Stake in a Tibetan Debate?”, Philosophy East and West 56, no. 2 (2006): 218-219. 292 Ibid., 219. 203 type of reification by means of the neither-one-nor-many argument. While Śāntarakṣita saw Yogācāra as the best description of relative truth, and therefore as something worth studying, he also maintained argued that Yogācāra theory failed to withstand the neither-one-nor-many argument. Even the reflexive nature of awareness, which Śāntarakṣita acknowledged to be constitutive of consciousness, could not in his view withstand this analysis. Therefore, according to Śāntarakṣita, reflexive awareness is only relatively or conventionally existent. By contrast, Ratnākaraśānti clearly and repeatedly argues that the reflexive nature of awareness is ultimately existent. However, this view is predicated on his “False Imagist” ontologico-epistemological paradigm, essentially a type of anti-realism. Both his affirmation of reflexive awareness and his refutation of cognitive images are elements of his eleutheriology, which emphasizes transcending phenomenological duality rather than ontological reification. Ratnākaraśānti was not a realist any more than Śāntarakṣita, and perhaps even less so, but he framed the issues more in terms of phenomenology and epistemology rather than ontology and metaphysics. For Ratnākaraśānti, the ultimately existent (or: ultimately epistemically reliable) character of the reflexive nature of awareness is diametrically opposed to the ultimately nonexistent, epistemically unreliable character of ordinary phenomenal appearances—which are, according to Ratnākaraśānti, nothing but mental representations (vijñapti, Tib. rnam rig). In other words, even though Ratnākaraśānti embraces an idealistic ontology, the phenomena described by that idealistic ontology are only illusory, the “images” they produce in consciousness only false. This tracks well with Śāntarakṣita's position, at least to the extent that it implies a disavowal of the notion that phenomenal appearances of any type exist ultimately. 204 Moreover, Ratnākaraśānti clearly did not consider the luminosity of reflexive awareness to be an ordinary phenomenon like a blue patch, and his refusal to subject the reflexivity of awareness to neither-one-nor-many argumentation must have stemmed from his understanding of a qualitative difference between these two different types of phenomena (if, indeed, the reflexive nature of awareness can even be properly described as a “phenomenon,” or for that matter an “entity,” at all). This appears to be a point of departure from Śāntarakṣita, but the extent to which this difference forms an outright contradiction, as opposed to a shift in emphasis, is not at all obvious—particularly since, as we have seen, Ratnākaraśānti holds “non-appearance” to be the final stage of Yogic Practice. This lack of obviousness points toward a larger problem in the academic study of Buddhist philosophy in general, and the study of the Ornament of the Middle Way in particular. Aside from the ambiguity inherent in such a dense and elliptical work, which is real and should not be downplayed as a significant source of difficulty, our ignorance of the relationship between Śāntarakṣita and his commentators is in no small part due to a continuing lack of substantive academic engagement with the tradition of exegesis stemming from the Ornament of the Middle Way, despite the fact that this seminal work was the primary touchstone for subsequent Indian Madhyamaka (and Yogācāra) commentaries in both India and Tibet over the centuries following its composition. It is easy to understand the reason why: the primary mode of Indian philosophical investigation, both internal to Buddhist traditions as well as between Buddhist and non-Buddhist interlocutors, was the discourse of pramāṇa, the means of reliable knowledge. Śāntarakṣita's synthesis provides an incomparable framework for articulating the precise manner in which Middle Way and Yogic Practice philosophy relate to the pramāṇa discourse, and thus to the wider sphere of intra- and inter-doctrinal polemics. Yet, apart from Blumenthal (2006), and 205 two translations of Mi pham's later (though valuable) Tibetan commentary, 293 there is no booklength treatment of the Ornament at all. Both Śāntarakṣita's autocommentary294 and Kamalaśīla’s subcommentary295 remain untranslated. Meanwhile, translations and Western-language commentaries on Candrakīrti's Entry into the Middle Way and Clear Words abound; there is even a monograph296 entirely devoted to the sixth chapter of the Entry into the Middle Way alone, independently of the rest of the text. This situation is perplexing from an historical-critical perspective, since as far as the textual record is concerned, there were no commentaries on the Entry into the Middle Way for the first five centuries of its existence, and only one Indian commentary in total. Later Tibetan authors would occasionally refer to a continuous lineage stretching from Candrakīrti (ca. 600 CE) to Jayānanda (ca. 1100 CE), who wrote the first known commentary to the Entry to the Middle Way. However, Vose (2008) dismisses this notion as a hopeful fantasy or literary conceit, citing the complete lack of any evidence that either the texts or the authors purported to constitute this lineage ever existed. Again, the motivation behind Indian authors' lack of engagement is clear: the Entry into the Middle Way attacks the legitimacy of the pramāṇa discourse, and in so doing sets itself apart from the entire edifice of Indian (not just Buddhist) scholastic philosophy. From the perspective of an Indian commentator, there was quite literally 293 These are trans. Padmakara Translation Group, The Adornment of the Middle Way: Shantarakshita’s Madhyamakalankara with commentary by Jamgön Mipham (Boston: Shambhala, 2005) and trans. Thomas Doctor, Speech of Delight. 294 Śāntarakṣita, Commentary to the Ornament of the Middle Way (Madhyamakālaṃkāravṛtti, dbU ma rgyan gyi rang ‘grel pa). P5285. sDe dge, dbu ma, vol. Sa: ff. 56.b.4-84.a.1. 295 Kamalaśīla, Explanation of the Ornament of the Middle Way (Madhyamakālaṃkārapañjikā, dbU ma rgyan gyi dka’ ‘grel). P5286. 296 Ari Goldfield, Jules Levinson, Jim Scott, and Birgit Scott, trans., The Moon of Wisdom: Chapter Six of Chandrakirti’s Entering the Middle Way with commentary from the Eighth Karmapa Mikyö Dorje’s Chariot of the Dagpo Kagyü Siddhas (Ithaca: Snow Lion Publications, 2005). 206 nothing to comment upon in his work. This may be usefully contrasted to the place occupied by Dharmakīrti, either a contemporary or an immediate successor of Candrakīrti, in the Indian scholastic discourse; both Buddhist and non-Buddhist commentators were effectively forced to respond to Dharmakīrti, precisely owing to Dharmakīrti's comprehensive engagement with pramāṇa theory. More broadly, the importance of pramāṇa theory in the history of Middle Way exegesis has, up until the present, been almost entirely overlooked. In the context of Western engagement with Tibetan scholasticism, the intellectual history of Madhyamaka has usually been presented as a story about the innate supremacy of Candrakīrti’s ideas concerning Nāgārjuna, particularly as articulated and defended by his dGe lugs interpreters. In broad outline, the dGe lugs system is a synthesis of Candrakīrti and Dharmakīrti that proceeds by glossing over Candrakīrti’s wholesale rejection of pramāṇa theory and Dharmakīrti’s embrace of Yogācāra ontology and epistemology. There are important questions, which are only now starting to be asked, about just how accurately dGe lugs exegesis treats Candrakīrti and Dharmakīrti. But the more interesting questions concern the extent to which—or even whether or not—Indian Madhyamaka ever existed independently of pramāṇa theory. Biographical information about the Nāgārjuna who composed the Root Verses of the Middle Way is even less forthcoming than biographical information about Ratnākaraśānti. Earlier, he was placed at some point in the second century, but this chronology is, in the words of Ian Mabbett, “…not properly justified; it is a self-validating majority vote, or a median possibility, rather than a demonstrable probability.”297 The only thing that is absolutely clear 297 Ian Mabbett, “The Problem of the Historical Nāgārjuna Revisited,” Journal of the American Oriental Society 118, no. 3 (1998): 333. 207 when it comes to Nāgārjuna is that there were no commentaries on his works for several hundred years—not until Buddhapālita (470-550). The only reason we are aware of a continuous Madhyamaka tradition in between is on account of Kumārajīva (344-413), who is reputed to have converted to Mahāyāna Buddhism after studying the works of Nāgārjuna. But Kumārajīva was a Kuchean monk, of Kashmiri lineage, whose greatest impact was on Chinese Buddhism. And given that Kumārajīva himself is one of the earliest (and only) sources for biographical information about Nāgārjuna, there are legitimate, if unanswerable, questions about just how well-known the works of Nāgārjuna were outside of the narrow slice of Central Asia where Kumārajīva happened to encounter them. The standard narrative about Nāgārjuna is that the Root Verses and other works attributed to him were always regarded with the high esteem in which later Indian, Chinese, and Tibetan Buddhists held them. But there is simply no evidence for this view. It must be remembered that, if a second-century (or earlier) chronology for Nāgārjuna is in fact correct, this places him squarely in the period during which the Mahāyāna was still being formed, and was not even necessarily well-defined. More to the point, at this stage the so-called “Great Vehicle” was still very much a minority movement, as is well-attested. And even if Nāgārjuna lived somewhat later (perhaps in the third century), it is a matter of historical record that Kumārajīva, like his younger contemporary Vasubandhu, had to be converted to the Mahāyāna from the Sarvāstivāda; even in the fourth century, the Mahāyāna had not yet achieved its later unquestioned dominance. Therefore, to the extent that he was articulating an unequivocally Mahāyāna philosophy, as well as openly critiquing the Sarvāstivāda Abhidharma, Nāgārjuna was in all likelihood simply unintelligible to his contemporaries. 208 This is not, of course, to say that Nāgārjuna’s works were not eventually revolutionary in their impact. But it is important not to “self-validate” the prevailing consensus nor read that consensus back into the textual record. In terms of the wider Indian scholastic world, both Buddhist and non-Buddhist, the works of Nāgārjuna seem to have been relatively obscure until several centuries after their first appearance. The reasons for this are, again, relatively easy to understand. From a Sarvāstivāda or Abhidharmika perspective, Nāgārjuna was speaking nonsense, making the ridiculous claim that dharmas were prajñaptisat (“designatedly existent”) and their svalakṣaṇas (“defining characteristics”) empty or nonexistent. And from a nonBuddhist perspective, there was no common basis for argumentation, as in the Resolving Disputes (Vigrahavyāvartanī) and elsewhere, Nāgārjuna had critiqued the early versions of pramāṇa theory that were already circulating in his time. This point is absolutely crucial, for as many Western scholars have noted, pramāṇa theory was not the exclusive province of any one Indian scholastic tradition. It was, rather, a set of shared presuppositions or a common ground for argumentation, a Wittgensteinian “language game” that everyone could play. Insofar as Nāgārjuna did not frame his arguments according to the rules of this “game,” there was no way for anyone to engage with his philosophy. This was the substance of Bhāvaviveka’s (500-570) critique of Buddhapālita, who had written the very first commentary on the Root Verses not long before. Buddhapālita had only employed “unacceptable consequence” (Skt. prasaṇga, Tib. thal ‘gyur) arguments in his exposition of Nāgārjuna’s text. Bhāvaviveka’s problem was that this made it impossible for Nāgārjuna’s arguments to reach a wider audience. Therefore, Bhāvaviveka prescribed the use of “autonomous” (Skt. svātantra, Tib. rang rgyud) syllogisms—logical arguments that are formally 209 valid according to the criteria of pramāṇa theory—and rigorously applied this framework in his own commentary to the Root Verses. Owing to the later Tibetan fascination with Candrakīrti, who savagely critiqued Bhāvaviveka precisely on account of the latter’s pro-pramāṇa advocacy, it is sometimes assumed that Middle Way exegesis proceeded in a straight line from Nāgārjuna to Candrakīrti and thereafter to Tibet. In fact, this assumption is demonstrably incorrect. There was only one Indian commentary to Candrakīrti, written by Jayānanda sometime in the eleventh century, several hundred years after Śāntarakṣita had introduced Buddhist philosophy—particularly, Madhyamaka—to Tibet. Moreover, the failure of Candrakīrti’s commentaries to attract any commentarial attention of their own highlights an important fact: the most important and influential (that is, commented-upon) Indian Madhyamaka works were composed by Bhāvaviveka and Śāntarakṣita, and employed the basic theoretical tools of the pramāṇa discourse. There is, accordingly, simply no reason to assume that some kind of “pure” Madhyamaka existed in India as an exegetical tradition independent of pramāṇa theory. There are no known subcommentaries on Buddhapālita’s pramāṇa-less exegesis of the Root Verses, and the only Indian work known to follow Buddhapālita in this regard is Candrakīrti’s ignored Entry into the Middle Way. Yogācāra certainly existed independently of pramāṇa, as evinced both in its earliest strata—the works of Asaṇga and Vasubandhu—and in the commentaries of Sthiramati (ca. 500). But by far the most influential articulations of Yogācāra theory were by way of pramāṇa discourse, first in the works of Dignāga and later in the works of Dharmakīrti. Regardless, neither Asaṇga, Vasubandhu, Dignāga, nor Dharmakīrti made any direct reference to Nāgārjuna or the Root Verses. To the extent that “Madhyamaka” as such still existed in ninth-century India 210 as a viable exegetical tradition at all, it appears to have done so only due to Śāntarakṣita’s synthesis of Madhyamaka with Yogācāra and pramāṇa theory. The constant here, of course, is pramāṇa theory, which was of central importance to the Buddhist philosophical tradition, as well as the main subject of academic inquiry at the great Buddhist monastic universities. This makes the Tibetan reception of Candrakīrti all the more baffling. In fact, it would not be overstating the point to note that there was a massive shift over time in the Buddhist commentariat's appraisal of Candrakīrti. The exact sequence of events in this shift is, as yet, not understood, and owing to the paucity of source material it may never be completely understood. Clearly, however, judging from the fact that Ratnākaraśānti felt the need to respond to arguments from the Entry into the Middle Way, Candrakīrti's reputation was experiencing a renaissance by the end of the tenth or the beginning of the eleventh century. The question is: Why? Kevin Vose hypothesizes that the resurgence of interest in the Entry into the Middle Way was due at least in part to the emergence of the Radiant Light (Pradīpoddyotana, Tib. sgron gsal) and other influential commentaries on the Gathering of Secrets (Guhyasamāja) Tantra.298 These texts were ascribed to a certain “Candrakīrti,” who was a member of the lineage of tantric teachings stemming from Saraha (also known as Rāhulabhadra), teacher of the “Nāgārjuna” who in turn taught the author of the Radiant Light. Vose (2008) notes that reputation of the “Noble” lineage of Saraha → Nāgārjuna II → Aryadeva II/Candrakirti II/etc. must have had a salutary effect on the reputation of the Entry into the Middle Way, as it is only around this time that its one and only subcommentary appears. Clearly, Ratnākaraśānti's dismissive treatment of the arguments from the Entry into the Middle Way serves as further evidence that, despite having 298 Kevin Vose, Resurrecting Candrakīrti: Disputes in the Tibetan Creation of Prāsaṇgika (Boston: Wisdom Publications, 2009), 30. 211 been ignored for centuries after its composition, within two generations after the appearance of the Radiant Light, the Entry into the Middle Way was the object of new or renewed scholarly interest. The later history of commentaries on the Entry into the Middle Way in Tibet has been the subject of endless discussion, and will not be considered here in any detail. What is most interesting with respect to this thesis is the dismissive attitude which not only Ratnākaraśānti, but also his Tibetan translators, held toward the content of the Entry into the Middle Way. In the unusually long colophon299 written by Śāntibhadra and Śākya 'Od, these two New Period (gsar ma) translators explain their motivation for translating the Pith Instructions: The monk Candrakīrti—who had deviated from the intent of Nāgārjuna— thenceforth, in his later life, completely gave up nihilism and wrote commentaries on the profound tantras.300 In their final words, they explicitly state that their purpose is to secure the place of pramāṇa study, and to refute those (like Candrakīrti) who say there is no point to studying the means of reliable knowledge: This was translated by the Indian abbot Śāntibhadra and the ordained translator Śākya ‘od. We descended on the main point in order to annihilate the deceitful. Logic is the central pole on the victory banner of omniscient wisdom! This is the Pith Instructions on the Ornament of the Great VehicleConsciousness-Middle Way, 299 Ruegg (1981) attributes this statement to Ratnākaraśānti himself, however this is extremely unlikely, since it follows both the thematic conclusion of the Pith Instructions and an account of the deeds and reputation of Ratnākaraśānti, which he simply would not have written about himself (far less at the end of a commentary). Most likely Ruegg was thrown off by the length of the colophon, which is nearly a full folio page as opposed to the standard one or two lines. 300 Ratnākaraśānti, Pith Instructions, ff.231.a.3-5. brtsun pa zla grags la sogs pa klu sgrub kyi dgongs pa las ‘chal bar gyur na rang slad kyis tshe smad la med par smra ba dor nas rgyud zab mo’i ‘grel pa byed par ‘gyur la/ ‘on kyang des log par bris pa de’i rjes su ‘jug cing g.yo sgyus spyod pa yang ‘ga’ ‘byung bar mdo sde dad pa cher ‘byung dang/ dgongs pa nges par ‘grel pa sogs par bcom ldan ‘das nyid kyis gsungs nas/ 212 Logical arguments which most eloquently refute Candrakīrti. Later, the Kashmiri Amogha and the translator ‘O-ru settled any questions.301 The colophon is lengthy, and can be found translated in its entirety in the Appendix. This passage, however, is particularly noteworthy for several reasons, not the least of which is the allegation that Candrakīrti “perverted” ('chal) the intent of Nāgārjuna. Considering that, for most Tibetan sources, particularly the so-called “Consequentialists” (thal ‘gyur pa), Candrakīrti's exegesis is simply coextensive with the intent of Nāgārjuna, this is nothing short of scandalous. It is especially significant because the proponents of the “Great Middle Way of Extrinsic Emptiness” (gzhan stong gi dbu ma chen po), the perennial rhetorical whipping-boys of the Consequentialists, were routinely caricatured as having deviated from the authoritative Indian Madhyamaka tradition of Nāgārjuna and Candrakīrti; although, from their perspective—as from Ratnākaraśānti's perspective—the authoritative Madhyamaka tradition passed from Nāgārjuna to Asaṇga and Vasubandhu. This passage is also noteworthy for its explicit distinction between two radically divergent phases in the philosophy of Candrakīrti. The Tibetan is ambiguous as to whether these phases occurred within a single life or across multiple lives, but it is clear that the translators considered the author of the Entry into the Middle Way and the author of the Radiant Light to be, in some sense, the same person, who later came to see the errors in his earlier work. Most importantly, however, this passage identifies the most basic problem that the minority of later commentators who dared to cross the Consequentialists would have with Candrakīrti: that the presentation of emptiness in the Entry into the Middle Way was, in effect, a species of nihilism. In their view, by radically negating not only the essences (i.e. essence-svabhāva) of objects, but also the means of reliable knowledge and the realization of the Buddhas, Candrakīrti committed himself to an untenable reading of emptiness. Ironically, the champions of Candrakīrti, primarily the Consequentialists, would 301 Ratnākaraśānti, Pith Instructions, ff. 231.a.6-7. lo tsa ba dge slong shAkya 'od kyis bsgyur ba/ g.yo sgyus spyod pa tshar gcod bsnyon 'debs gnad la 'bebs/ rtog ge'i srog shing kun mkhyen ye shes rgyal mtshan can/ theg chen rnam rig dbu ma'i rgyan gyi man ngag bcas/ legs mdzad gang de zla grags 'joms pa'i rtog ge yin/ slad nas kha (P/N/S: a) che a mo gha dang 'o ru lo tsa bas zhus te gtan la phab pa'o// 213 rehabilitate Candrakīrti's version of emptiness by insisting that, in spite of all the obvious evidence to the contrary, Candrakīrti's philosophy was actually compatible with Dharmakīrti's pramāṇa theory, particularly at the Sautrāntika or “External Realist” level of analysis. Of course, this required a distortion or outright denial of Dharmakīrti's shift to a purely idealistic epistemology, again in spite of all the evidence to the contrary, but that is a topic for another time. The point here is that both Ratnākaraśānti and his Tibetan translators identified critical flaws in the Entry into the Middle Way, several centuries before it became the accepted orthodoxy among most Tibetan commentators—and, later, their Western counterparts—that the Entry into the Middle Way and Consequentialism was the definitive statement of Nāgārjuna's intent. This is clear evidence that, contrary to the charges leveled by the Consequentialists, the proponents of “Extrinsic Emptiness” (gzhan stong), primarily Dol po pa Shes rab rGyal mtshan (12921361), stood on firm precedent in the Indian commentarial literature. Dol po pa’s descriptions of the ultimate truth302 bear more than a passing resemblance to Ratnākaraśānti’s, and there is no reason to assume that Ratnākaraśānti’s description of ultimate reality as nondual luminosity was in any way unique among Indian scholastics—indeed, Ratnākaraśānti’s primary rhetorical opponents, the “True Imagists” such as Jñānasrīmitra, appear to have accepted the basic paradigm of the relationship between pramāṇa theory and Middle Way exegesis, differing only insofar as the existence (or epistemic reliability) of cognitive images at the level of Buddhahood was concerned. And again, concerning the evidence of the textual record, Śāntarakṣita’s decision to place Madhyamaka above Yogācāra on the “sliding scale” was by no means unquestioningly followed by successive generations of commentators—even, and especially, commentators continuing the tradition of synthesizing Madhyamaka, Yogācāra, and pramāṇa theory. Regardless of the reasons, it is a historical fact that, within two centuries of Ratnākaraśānti’s tenure at Vikramaśīla, Consequentialism had become the dominant strain of Tibetan scholastic 302 Dol po pa Shes rab rGyal mtshan, trans. Jeffrey Hopkins, Mountain Doctrine (ri chos nges don rgya mtsho), 222. “That which has all five characteristics of non-duality is the definition of the ultimate.” See also pp. 9-21 and 93102. 214 methodology. While there were many Consequentialists” of note, for example the Sa skya scholar Sa skya Paṇḍita (1182-1251), Tibetan Consequentialism was in general closely associated with the dGe lugs tradition founded by Tsong kha pa (1357-1419). In broad outline, the touchstone of Tsong kha pa’s philosophy was the valorization of a Consequentialist reading of the Root Verses, and the refutation of “Mind Only” philosophy in all its forms, including the Yogācāra-Madhyamaka-Pramāṇa synthesis of Śāntarakṣita. As is well-known, Tsong kha pa considered Extrinsic Emptiness an untenable and heretical deviation from orthodox Buddhist doctrine; Tsong kha pa vigorously critiqued Dol po pa from a Consequentialist perspective, maintaining that the “Great Middle Way of Extrinsic Emptiness” was not in any sense “Middle Way” philosophy. Less known is the historical fact that, as the dGe lugs became ascendant in Central Tibet (dbU and rTsang) in the seventeenth century, the Great Fifth Dalai Lama, bLo bzang rGya mtsho (1617-1682), consolidated power by, among other actions, seizing land and assets belonging to the Jo nang school of Dol po pa: Due to conflicting territorial interests and struggles for power among the local rulers of Tsang in Central Tibet, the Jonang tradition found itself in opposition to the political ambitions of the 5th Dalai Lama and the Mongols. Consequently, after the passing away of Tāranātha, the 5th Dalai Lama forcefully converted Jonang Damcho Ling Monastery (Phuntsok Ling) and other Jonang monasteries into Geluk monasteries, put a ban on zhentong [Extrinsic Emptiness] teachings, sealed Jonang libraries, burnt texts on zhentong, and the Jonang were forced to relocate away from their establishments in Central and Western Tibet.303 As it turns out, dGe lugs dominance was so complete that only in the year 2000304 did Westerners realize the Jo nang pas still exist as a lineage of Tibetan Buddhism. Both the Jo nang pas and the current, 14 th 303 304 Jonang Foundation, “Frequently Asked Questions,” http://www.jonangfoundation.org/faq Andreas Gruschke, “Der Jonang-Orden: Gründe für seinen Niedergang, Voraussetzungen für das Überdauern und aktuelle Lage,” Tibet, Past and Present: Tibetan Studies I: Proceedings of the Ninth Seminar of the International Association of Tibetan Studies, ed. Hank Blezer (Leiden: Brill Academic Publishers, 2002): 183-214. 215 Dalai Lama bsTan ‘dzin rGya mtsho agree305 that the reasons behind the 5th Dalai Lama’s actions were essentially political, as opposed to doctrinal. This may well have been the case. It is difficult, however, to understand the decision to burn philosophical texts and ban religious teachings on purely political grounds. And it may well be impossible to separate the dGe lugs suppression of Extrinsic Emptiness from Tsong kha pa’s Consequentialist critique of that doctrine. Owing to these historical circumstances, from the time of the very first contact between Westerners and Tibetan Buddhism, it has frequently been assumed that Consequentialism is universally considered among Tibetan scholars to be the most authentic and direct continuation of Indian Madhyamaka. As a result, Extrinsic Emptiness has only very recently been the subject of any serious academic inquiry. And in general the forms of Middle Way philosophy from the Sanskrit tradition that were historically the most important and influential in India—most notably the Ornament of the Middle Way, and the many other presentations of Madhyamaka that were articulated in terms of pramāṇa theory—have as yet received only sparing attention. But, by way of a final conclusion, I submit that apart from the exigencies of history, the realist and anti-idealist ontology at the heart of the dGe lugs Consequentialist project has proved irresistible to Western academicians, precisely because it confirms the physicalist and materialist biases of the Western academy. For example, Mackenzie, in his article on the various forms of reflexive awareness, claims that one potential problem with reflexive awareness theorems is their tendency to lead toward idealism, since idealism is contrary to physicalism and physicalism is currently, and in his opinion “rightly,”306 hegemonic with respect to academic discourse. Arnold wonders aloud whether idealism is simply a “nonstarter”307 and recently published a monograph308 critiquing Dharmakīrti’s philosophy of mind on physicalist grounds. 305 Mullin, Glenn. The Fourteen Dalai Lamas (Santa Fe: Clear Light Publishers,2002), 207. 306 Mackenzie, “The Illumination of Consciousness,” 60. 307 Dan Arnold, “Self-Awareness and Related Doctrines,” 356. 216 In the Introduction, it was noted how Garfield's attempt to harmonize Nāgārjuna with “ordinary language” philosophy effaces the real and important difference between relative and ultimate truth. However, the two-truths distinction is not the only casualty of Garfield’s exegeses. In his response to Williams, Garfield brushes aside the traditional dGe lugs critiques of reflexive awareness, in favor of a distinctively modern tack: Reflexive awareness has no place in conventional reality, and is indeed incoherent. Reflexive awareness… involves a commitment to a view that intentionality is an intrinsic, rather than a relational aspect of cognition; to a view that we have a special kind of immediate, non-deceptive access to our own minds and to their states; and to the view that we specify an essence of the mental.309 First, it is necessary to reiterate the point that, for Yogācāra proponents of reflexive awareness, intentionality is in no wise “intrinsic” to consciousness. Like Arnold, Garfield confuses the argument that consciousness is intrinsically reflexive with the argument that consciousness is intrinsically intentional. Garfield may be unable to conceive of non-intentional or nondual consciousness, and he may have support for this view in any number of Western philosophical traditions, but this only serves to highlight perhaps the single most important difference between Western and Buddhist philosophies of mind. In any case, as exegesis of the relevant point, this is inaccurate on its face: reflexive awareness does not involve any kind of commitment to the view that intentionality is “intrinsic” to consciousness. Second, Garfield misconstrues the nature of the “immediate” and “non-deceptive” access to cognitive states afforded by reflexive awareness. First, like Arnold, he confuses reflexive awareness with reflective awareness: 308 Dan Arnold, Brain s, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind. New York: Columbia University Press, 2012. 309 Garfield, “The Conventional Status of Reflexive Awareness,” 202. 217 I certainly can be aware of the pleasure of a strawberry or the pain in my back without being aware that I am aware of it. Perceptual contact guides my behavior. Full stop. If I am then aware that I am aware, that is a further cognitive state, distinct from the first, and directed upon it. I can keep climbing the hierarchy of meta-awareness as long as I like, but that is only a potential regress, and hence is not vicious.310 The description of a second-order awareness of being aware, a “further cognitive state, distinct from the first, and directed upon it,” is a succinct definition of reflective awareness. This has, of course, nothing whatsoever to do with reflexive awareness, which is something else entirely. Indeed, for Ratnākaraśānti, as for Śāntarakṣita and Mi pham, the issue at stake is precisely that the immediate awareness of awareness is not the same thing as taking subjective mental contents for an intentional object. The former is reflexive awareness; the latter is not. Furthermore, this passage is clear evidence that Garfield has not studied pramāṇa theory in any kind of depth, for as Dreyfus, Dunne, McClintock, and others have clearly explained, the fact that sensory perceptions occur without there being any second-order (that is, conceptual) awareness of the sensory content of those perceptions is perhaps the most basic aspect of Dharmakīrti’s entire epistemology. Definite perceptual judgments (niścaya) are secondary, conceptual, and formed in strong dependence upon what the subject is interested in and capable of perceiving. Not everything in e.g. the visual field will be intentionally apprehended as a visual object; but this lack of determination has absolutely nothing to do with reflexive awareness. The point is that the subject is reflexively aware of those perceptual judgments which are formed. And yes, at some pre-conscious (or, more accurately, subliminal) level there is also reflexive awareness of those aspects of the sensory-cognitive image which do not form a perceptual judgment. However, this does not in any way imply that unprocessed sensory content is epistemically available, since only definite ascertainments grant epistemic access to the contents of senseperception. This is one of the most fundamental issues at stake in Dharmakīrti’s perceptual theory, and it 310 Ibid., 221. 218 is simply inconceivable that an ostensibly serious critique of reflexive awareness would overlook such a basic point. More broadly, in standard Yogācāra and pramāṇa theory, as for example the philosophy of Dharmakīrti, every moment of consciousness is qualified by reflexive awareness, but every moment of an ordinary, unenlightened being’s consciousness is also qualified by ignorance and dualistic fixation. It is just this ignorance which prevents reflexive awareness from affording sentient beings unimpeded access even to the contents of their own minds, to say nothing of ultimate truth. So Garfield is simply wrong when he asserts that reflexive awareness implies “that we cannot be in error about the nature of our own minds or cognitive activity.”311 Proponents of reflexive awareness such as Dharmakīrti and Śāntarakṣita proceed by first asserting that we are in error about the nature of our own minds and cognitive activity. This is just the problem that reflexive awareness, following long years of contemplative practice and inner development, solves. But perhaps the most interesting way in which Garfield reads his own ideas into the discussion concerns his claim that reflexive awareness entails a commitment to the view “that we specify an essence of the mental.” This may or may not be true, but the fact that Garfield apparently considers it problematic is telling. Because the real issue here is not that Arnold or Garfield (or, for that matter, Williams) have made errors of comprehension. Some degree of error is inevitable in every exegetical effort; it is the cost for playing this particular language-game, and I do not labor under the illusion that my interpretation—far less my translation—of Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way is beyond reproach. No, the fundamental problem is that physicalism has prejudiced Buddhist scholarship in favor of external realist interpretations of Madhyamaka, as though Middle Way philosophy were nothing more than a prefiguration of the postmodern nexus of physicalism, anti-essentialism, and “ordinary language” philosophy that forms the contemporary academy’s effectively unquestioned consensus view. Given the 311 Ibid., 223. 219 historicism inherent in any such approach to Nāgārjuna, this is poor enough as exegesis, but the real scandal is how this prejudice has systematically precluded any extensive consideration of texts such as Śāntarakṣita’s Ornament of the Middle Way and Ratnākaraśānti’s Pith Instructions for the same which, judging by the available evidence, are far more representative of Buddhist philosophy as it actually exists—even and especially with respect to dGe lugs scholasticism, which perhaps more than any other Tibetan Buddhist tradition combines Madhyamaka with pramāṇa theory. Thus when Garfield claims that reflexive awareness necessitates a commitment to specifying an essence of “the mental,” this is not so much a well-formed argument against reflexive awareness, nor a hermeneutically-justified interpretation of Middle Way philosophy, as it is an indication of Garfield’s own commitment to physicalist reductionism and his desire to square away that commitment with (what he understands of) Tsong kha pa’s philosophy. But as this thesis draws to a close, it is worth pausing to consider whether physicalism is, in fact, a coherent statement of the academic discipline of physics as currently understood and practiced; whether, that is, the philosophical view that phenomena are real and composed of really- and objectively-existing localized point-masses all bouncing off of each other like so many atomic billiard balls, and that all minds and mental processes are reducible to the same, is justifiable on the basis of the scientific research so beloved of those who defend this view. The most curious fact about physicalism is that, while the atomic billiard ball picture follows naturally from Newton's mechanics, and was the dominant manner in which the physical world was understood through the early part of the twentieth century, subsequent developments in atomic physics proved conclusively, beyond the shadow of a doubt, that the billiard ball model of atomic particles was strictly incorrect. The famous “double slit” experiment demonstrated that, in effect, even a single particle possesses extension, as even a single particle is capable of interfering with itself. After this experiment, as well as subsequent investigations by Niels Bohr and others, it became clear that a light-wave extended in space was not truly a mechanical system of light-particles—that even a single particle of light behaved in 220 a wave-like fashion, exerting a causal influence upon the world as though it were extended across the entire space occupied by the experimental apparatus. The difficulty of maintaining any sort of realism on the basis of physics is only compounded by the experimental results of de Broglie, who demonstrated that “wave-like” behavior is not limited to single particles, but is in principle observable for any agglomeration of matter; in addition to photon-waves and electron-waves, there are table-waves and chair-waves. In general, as Bohr writes, “The special situation in quantum physics is above all... that the information gained about atomic objects cannot be comprehended along the lines of an approach typical of the mechanical conception of nature.”312 According to Bohr, mechanistic conceptions fail because “in quantum physics we are presented... with the inability of the classical frame of concepts to comprise the peculiar feature of indivisibility, or 'individuality,' characterizing the elementary processes”313 of causal interactions. But by far the most interesting result in quantum physics has been the evident necessity of rejecting a strictly “objective” conception of phenomena, owing to experimental refutation of the notion that causal processes occur independently of their observation by some phenomenal subject, or that the objects of knowledge (prameya) exist independently of the means of knowledge, “measuring instruments” (pramāṇa). Bohr called this principle “complementarity,” referring to “our position as observers in a domain of experience where unambiguous application of the concepts used in the description of phenomena depends essentially on the conditions of observation.”314 Because unambiguous descriptions depend on the conditions of observation, “Only in a complementary description transcending the scope of the mechanical conception of nature is it possible to find room for the fundamental 312 Niels Bohr, Atomic Physics and Human Knowledge (Oxford: Wiley, 1958), 99. 313 Ibid., 34. 314 Ibid., p. 99. 221 regularities responsible for the properties of the substances of which our tools and our bodies are composed.”315 In sum, as Bohr so eloquently wrote, How radical a change in our attitude towards the description of nature this development of atomic physics has brought about is perhaps most clearly illustrated by the fact that even the principle of causality, so far regarded as the unquestioned foundation for all interpretation of natural phenomena, has proved too narrow a frame to embrace the peculiar regularities governing individual atomic processes. …. While in relativity theory the decisive point was the recognition of the essentially different ways in which observers moving relatively to each other will describe the behavior of given objects, the elucidation of the paradoxes of atomic physics has disclosed the fact that the unavoidable interaction between the objects and the measuring instruments sets an absolute limit to the possibility of speaking of a behaviour of atomic objects which is independent of the means of observation. We are here faced with an epistemological problem quite new in natural philosophy, where all description of experiences has so far been based upon the assumption, already inherent in ordinary conventions of language, that it is possible to distinguish sharply between the behavior of objects and the means of observation. This assumption is not only fully justified by all everyday experience but even constitutes the whole basis of classical physics, which, just through the theory of relativity, has received such a wonderful completion. As soon as we are dealing, however, with phenomena like individual atomic processes which, due to their very nature, are essentially determined by the interaction between the objects in question and the measuring instruments necessary for the definition of the experimental arrangements, we are, therefore, forced to examine more closely the question of what kind of knowledge can be obtained concerning the objects. In this respect we must, on the one hand, realize that the aim of every physical experiment—to gain knowledge under reproducible and communicable conditions—leaves us no choice but to use everyday concepts, perhaps refined by the terminology of classical 315 Ibid, pp. 99-100. 222 physics, not only in all accounts of the construction and manipulation of the measuring instruments but also in the description of the actual experimental results. On the other hand, it is equally important to understand that just this circumstance implies that no result of an experiment concerning a phenomenon which, in principle, lies outside the range of classical physics can be interpreted as giving information about independent properties of the objects, but is inherently connected with a definite situation in the description of which the measuring instruments interacting with the objects also enter essentially.316 Far from being a mere historical curiosity, Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way has an enduring place, not merely in the history of Buddhist philosophy, but in the history of human knowledge. 316 Ibid., pp. 25-26. 223 Appendix: Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way Homage to noble Tara! Constant homage to Buddha Vajradhāra, Who is the very Body of Great Bliss Pervading the abode of the two Buddha-bodies Like the sun, moon, planets and stars! As Maitreya and Asaṇga Explained, And as Nāgārjuna also Taught, By means of reliable knowledge, endowed with scripture, The two truths shall be explained here. The “two truths” refers to the ultimate truth, and the relative. The ultimate has three aspects: (1) The ultimate itself is suchness, (2) The ultimate means of accomplishment is authentic wisdom, and (3) The ultimate to be attained is Nirvana. As it is said: Ultimate, attained, and accomplishment— The ultimate is propounded in three aspects. [MV III.11a] Similarly, the relative has three aspects: (1) The constructed [brtags] relative is the nature of all phenomena which are entirely imagined [kun brtags]; (2) Relative consciousness is all mistaken cognitions, and (3) The expressive relative is that which indicates the ultimate through sounds and thoughts. Extensively: imaginary construction, consciousness, and Expression in accordance [with the ultimate]. [MV III.10b] 224 The expressive relative is neither conceptual nor nonconceptual. Why is it nonconceptual? Because it is conducive to the production317 of nonconceptuality. Why is it not nonconceptual? Because it is a conceptualization of the authentic ultimate. Nonconceptuality does not arise, even from nonconceptuality itself, as it does not exist in spiritually immature beings. But nonconceptuality does not arise from ordinary concepts, either, because these are precisely that which does not accord with nonconceptuality. On the other hand, the nonconceptual itself is not unarisen. Following the cultivation of understanding through the correct placement of the intellect, the mind effortlessly enters into the Expanse of Phenomena free from any conceptual structuring. This very lack of conceptual structure is “nonconceptual wisdom.” As it is said in the Recitation for Engaging the Non-Conceptual, “This follows from genuine engagement, cultivation, intensification, and proper attention.” Others also say that, following the cultivation of the wisdom which arises from contemplation, the stainless and indestructible ultimate wisdom devoid of cognitive distortion will be made directly manifest. Therefore, those who wish to realize the nonconceptual awakening318 should rely on, and cultivate, an authentic ascertainment of the ultimate. This is our view. What is it ultimate truth? It is the object of noble beings. What is the relative truth? It is the object of spiritually-immature beings.319 317 bskyed (P/N/S: skyes) pa. 318 rnam par mi rtog pa’i byang chub rtogs (P: byung ba tog, N/S: ‘byung ba rtog). 319 yang don dam pa'i bden pa gang zhe na/ 'phags pa rnams kyi yul du gyur pa'i phyir ro/ kun rdzob kyi bden pa gang zhe na/ byis pa rnams kyi yul du gyur pa'i phyir ro/ This verse appears to be original. 225 The ultimate is posited in three modes: as the object of sublime wisdom, as the fruit of sublime wisdom, and as that which exists ultimately. Likewise, the relative is posited in terms of the three natures: the constructed, the dependent, and the perfected. This is the “Middle Way of the Three Natures.” All dualistic entities do not exist, But dualistic distortion, itself, does exist; It has the nature of emptiness. This is what is called the Path of the Middle Way.320 So it is said, and furthermore: The entirely-constructed, the other-dependent, And the completely-perfected [TSN 1a] Are imagined, arisen from causes, and Unchanging; thus are they arrayed. Our topic of discussion is the characteristics of the three natures, briefly summarized above. The defining characteristic of the constructed nature is that it does not exist in the way that it appears, namely as the duality of apprehended and apprehender. Although duality does not exist, dualistic appearances arise through the force of prior psychological imprints. The defining characteristic of the other-dependent nature is that it depends on causes and conditions. Alternatively, it is “other” in terms of being different from authentic wisdom, which 320 gnyis kyi bdag nyid thams cad med/ gnyis su ‘khrul pa'i bdag nyid yod/ de ni stong pa'i bdag nyid de/ dbu ma'i lam du de bzhed do/ This verse also appears to be original; Moriyama (2011) attributes this verse to Ratnākaraśānti. 226 is not other-dependent.321 The dependent nature also has three aspects: the storehouse consciousness, afflicted mind, and active cognition. The first is called “storehouse” because it is the basis for the “seeds” of all the phenomena of emotional affliction, and “consciousness” because it takes itself as its support.322 Afflicted mind is “afflicted” since it is always concomitant with the four emotional afflictions, and “mind” because it constantly thinks of itself: “I am.” Finally, active cognition continually arises, over and over again, through engagement with various objects. It takes these objects as its support, thus it is a consciousness. At all times, the dependent nature is empty of, isolated from, and without the constructed nature. This is the defining characteristic of the perfected nature. All consciousnesses are always empty of even the slightest trace of duality. And this emptiness never changes: it is of one taste everywhere, like space. This is how the perfected nature is explained. As the noble Maitreya taught, The non-existence of non-existent duality Is the characteristic of emptiness, which exists. [MV I.13b] One might ask: “Since this emptiness of duality is beginningless, doesn’t that mean that we should have been enlightened from that time, at the very beginning?” 321 yang na gzhan zhes bya ba ni rnam pa la bya la/ yang dag pa’i ye shes ni tshu rol te/ gzhan gyi dbang ma yin no/ Here, Ratnākaraśānti is playing with the etymology of the Sanskrit term paratantra. The Tibetan translation of this term, gzhan dbang (literally “other-powered”), emphasizes the relationship between the other-dependent nature and its causes and conditions; it is “powered” by those causes and conditions. However, unlike the Tibetan word dbang, the Sanskrit tantra can also simply mean “connected.” In this passage, then, Ratnākaraśānti draws attention to the literal sense of para as “other,” i.e. to the paratantra as something “other than” authentic wisdom, and thus not “connected” to it. 322 rang gi ngo bor (D: bos C, bo la P/N/S) nye bar dmigs pas rnam par shes pa’o. The sense in which the storehouse consciousness “takes itself as its focus” or indeed can even be considered a proper “consciousness” at all, proved a fertile ground for debate in India and beyond. Cf. William Waldron, The Buddhist Unconscious, 91-127. 227 This does not follow, because even though one has had some experience, still [emptiness] is not definitively ascertained.323 Enlightenment results from ascertaining the completelyperfected nature of the other-dependent; cyclic existence results from ascertaining the constructed imagined nature. The defining characteristic of the perfected nature is that very emptiness which defines the other-dependent nature.324 Of what is it empty? It is empty of duality. Because it is emptiness, it is not an entity; Because it is real, it is not a non-entity. Since it exists as something, it is not other; In terms of its characteristics, it is not non-other.325 If it were not distinct from the bundles of mental conditioning, then emptiness would be definitively ascertained when these bundles are perceived; in this way, [perceiving] anything would also be perceiving the truth. Moreover, since mental conditioning is equivalent in meaning to the emotional afflictions which are the source of all suffering, emptiness would be an emotional affliction; and, since emptiness is permanent, the emotional afflictions would also be 323 nyams su myong ba yod du zin kyang nges pa med pa’i phyir ro/ What makes something a pramāṇa is that it is capable of producing a nges pa (Skt. niścaya) or “definitive ascertainment”—the term could perhaps also be translated as “correct perceptual judgment”—concerning what one is trying to know. For example, the visual perception of a red cup is a pramāṇa for knowing that the cup is red, since it is correctly perceived to be red; but this perception is not a pramāṇa for knowing that the cup is impermanent (i.e. composed of momentarily-existing particles), since it does not phenomenally appear to be momentary. In the same way, Ratnākaraśānti argues, every consciousness is indeed empty of duality; but consciousness is not ordinarily “reliably ascertained” to be empty of duality. That is to say, just as the cup is impermanent, but not phenomenally ascertained as momentary, in ordinary beings a moment of consciousness will never produce the correct perceptual judgment that “mind is empty of duality.” 324 325 yongs su grub pa'i ngo bo gang yin pa de ni gzhan dbang gi stong pa nyid kyi ngo bo yin no/ gang gis stong na gnyis kyis so/ stong pa nyid phyir dngos po min/ de dngos yin phyir dngos med min/ lus dngos yin phyir gzhan ma yin/ mtshan nyid kyis ni gzhan min min/ zhes gsungs so/ Ratnākaraśānti is apparently quoting someone, but I was unable to locate the source for this verse. Perhaps this is evidence for Moriyama’s “other” Ornament to the Middle Way (see note 19)? However, also see MV I.20. 228 permanent! And, since emptiness is insentient, mental conditioning would also be insentient. Therefore, emptiness is not, and cannot be, exactly the same as the bundles of mental conditioning. Nor are they completely different. If this were the case, then even those who do directly perceive the ultimate truth could not subdue dualistic marks [mtshan ma], nor could there arise any direct experience of emptiness; the characteristics of thorough affliction and complete purification would become both the same and different. Thus emptiness is not utterly distinct from them, either. As it is said, Not an entity, not a non-entity No characteristic difference or sameness [MV I.13b] Furthermore: As for its synonyms, these are: Suchness, the authentic limit, Without marks, the ultimate, The expanse of phenomena, and the nature of reality. [MV I.14] The meaning of these synonyms is as follows. Because of not changing, it is “suchness.” Because it is the limit of thusness, it is “the authentic limit.” Because the dualistic marks of entities do not exist within it, it is “without marks.” Because the ultimate truth is the object of sublime transcendent wisdom, it is “the ultimate.” Because it is the cause of the qualities [Skt. dharmas, Tib. chos] of the Noble Ones, it is the “Expanse of Reality” [Skt. dharmadhātu, Tib. chos kyi dbyings]. Because it is the nature of everything, it is the “nature of reality” [Skt. dharmatā, Tib. chos nyid]. Early on, it arises as the emotional afflictions, together with the adventitious defilements that are not its true identity; later, once these have been abandoned, there comes to be complete purification without any defilement. As it is said, 229 Together with stains, [or] stainless, [it is] Emotional affliction [or] complete purity; Like elemental water, gold, and space, When it is pure, it is asserted to be pure. [MV I.16] If the emotional afflictions were not manifest, All sentient beings would be liberated. If complete purity were not manifest, Even with effort, there would be no result. [MV I.21] No affliction, no non-affliction, No purity and no impurity; The nature of mind is clear light, Since emotional affliction is temporary. [MV I.22] So it is said, in a slightly expanded fashion. In this way, there is entirely-imagined form, form that is constructed as phenomenal images, and the form of the nature of reality. It is also said that there is designated existence, substantial existence, and ultimate existence. Concerning the statement that this is the “Path of the Middle Way which Possesses the Three Natures,” there is no existence in terms of the constructed nature, and there is no nonexistence in terms of the dependent and perfected natures. Therefore, it is free from the two extremes. Moreover, The imagination of the unreal exists. In it, duality does not exist. The emptiness of that [duality] does exist; It exists in the [imagination]. Thus, everything is explained as Not empty and not non-empty, By being existent, non-existent, and existent. This is the path of the Middle Way. [MV I.1-2] 230 Even though, for example, a conceptualized blue-patch may exist, the characteristic “blue” does not exist, because it cannot withstand logical analysis. There is cognitive distortion due to the contaminating force of the “blue” psychological imprint. Since it arises in this way, the experience is distorted, and it is experienced as though one were experiencing something else.326 By contrast, there is nothing that can refute the luminous nature of reflexive awareness. Reflexive awareness is direct [mngon sum, *pratyakṣa], and it is an authentic experience, because there are no means of reliable knowledge [tshad ma, pramāṇa] apart from it. Thus it is established as a means of reliable knowledge. Indeed, this means of reliable knowledge cannot be refuted, even by a hundred means of reliable knowledge—to say nothing of mere extreme assertions, which are not means of reliable knowledge. Thus the reflexivity of awareness is proven by the two means of reliable knowledge, as well as through refutation. Since they appear dualistically, [phenomena] are said to be the “mark of duality.” However, they come to cessation in transcendent wisdom. This wisdom is authentic and undistorted, thus it is the unmistaken perfected nature. The suchness of that wisdom is the unchanging perfected nature. From the mouth of Nāgārjuna, it was also said: The arising of entities does not exist, Nor does any cessation whatsoever exist. These are only conceptual notions, Of change as arising and cessation. What are called “the great elements” Are contained within consciousness; 326 That is to say, “something else” as opposed to one’s own psychological imprints, or one’s own mind. Ratnākaraśānti adopts the standard Yogic Practice stance that, due to dualistic cognitive distortion, blue—as, indeed, any and every phenomenal appearance—is experienced as though it were the characteristic of something external to the mind, as though it were itself external to the mind, while in fact it is nothing other than the mind. 231 As authentic wisdom is devoid of those, How would they not be mistaken constructions? [YS 34]327 Everything is contained in this explanation. Therefore the Yogic Practice tradition accords with the philosophy of the Middle Way. This is irrefutable. The unmistaken perfected nature is transcendent wisdom. At the time of the grounds of non-appearance, the grounds of completely pure superior intent, and the grounds of the especially exalted, as well as the adamantine samādhi, all phenomena are seen as being like the center of space. As it is said, for one with transcendent wisdom, all phenomena are seen as illusory, like in the eight examples.328 By contrast, pure worldly awareness is “pure” insofar as this awareness thoroughly resolves suchness, but it is “worldly” insofar as it is still distorted; by the force of necessity, even perfect enlightenment at the level of Buddhahood retains a tiny amount of cognitive distortion, because the purity is worldly in nature.329 Consciousnesses abiding on the Bodhisattva stages ripen dissimilarly, as they are of two types: worldly and transcendent. The worldly consciousness is also further subdivided, into the impure worldly consciousness and the pure worldly consciousness. This is the tradition which maintains the non-existence of cognitive images: In the worldly awareness, Once cognitive images disappear, The Lord of Non-Appearance, 327 Nāgārjuna, Sixty Verses (Yukti-ṣaṣṭā-kārikā), v. 33-34. “The Great Elements” are Earth, Water, Wind, Fire, and Space—i.e., according to Indian (meta)physics, all the material components of the cosmos. Ratnākaraśānti is citing a verse that maintains an explicitly idealistic ontology, attributed to Nāgārjuna. The fourth pada reads: log par rnam brtags (P/N/S: rnam dag) cis ma yin. 328 329 “Like a bubble, like a flower in the sky, like a flash of lightning,” and so on. de bzhin du sangs rgyas kyi sa la dgos (P/N/S: dgongs pa’i) pa'i dbang gis rdzogs pa'i byang chub cung zad 'khrul pa yin te/ dag pa 'jig rten pa'i bdag nyid yin pa'i phyir ro/ 232 Nondual suchness without conceptual structures, arises. Although worldly awareness has images, Those images are thoroughly resolved To be both false and unreal. Thus it is said to be “without images.”330 Other Buddhists have the following stances. Those who assert that the awareness of an object such as a blue-patch occurs without a cognitive image are the Traditionalists. The SūtraFollowers assert that objects are experienced through a cognitive image, a “reflection” that has been causally supported by the object. Appearing in the manner of the Middle Way is a third approach, and Yogic Practice is a fourth. This should be understood by those who are learned in the scriptures, their commentaries, and the means of reliable knowledge. How is there an awareness of blue-patches and so on? The Blessed One Himself said: External objects, as imagined by Immature beings, do not exist. Objects and appearances arise In a mind disturbed by psychological imprints. On the other hand, if one asks for the evidence for this, it should be understood similarly to how the dream-consciousness, appearing as various objects, no longer exists after one has awoken. Someone might object that a mere example does not prove this point. To this, it should be said that while phenomena themselves are established as luminous by the experience of their 330 Ratnākaraśānti does not specify that this is a quotation, and the verse appears to be original. 233 inherently luminous nature, that luminosity is spoken about in terms of supreme appearances, individual appearances, and luminous-like appearances.331 [Luminosity] is not insentient, and it is not epistemically remote, 332 as it has the nature of being utterly and completely luminous. If this luminosity were not established, then nothing would appear, and in consequence nothing at all could be established. And if it is established, then because it itself is consciousness, all phenomena are established as having the nature of consciousness. In verse: What is not luminous, is not illuminated; It cannot be made luminous by anything else. Blue-patches and so on are illuminated, Therefore awareness is non-dual.333 In the presentation of those like the Sūtra-Followers, what connects the experienced object to consciousness? They say, “The connection arises from that [object].” But then the eye would also be experienced by an eye-consciousness!334 Furthermore, although they do not exist 331 chos rnams rang nyid gsal bzhin par bdag nyid kyis gsal ba'i ngo bo nyams su myong bas grub pa yin la/ gsal ba de yang rab tu snang ba dang/ so sor snang ba dang/ gsal bzhin ba zhes bya’o/ Ratnākaraśānti’s point here is that, while all phenomenal appearances are illuminated in consciousness, there are three different degrees or types of illuminated appearances, presumably corresponding to the three natures. Thus even dream-appearances and distorted or imputed phenomena are in some sense luminous. Cf. Georges Dreyfus, Recognizing Reality, 339-40. “For Dharmakīrti… the validity of the self-cognizing factor of a mental episode is not reducible to that of the episode itself. For example, when I wrongly conceive of the moon as being made of green cheese, my thought is not valid. But I am aware that I have this thought. Hence, my apperception [i.e. reflexive awareness] is valid, though it is not a separate awareness.” In this passage, a “luminous-like” appearance would presumably refer to something like Dreyfus’ example of perceiving the moon to be made of green cheese; the appearance itself is thoroughly mistaken, but the observer is still reflexively aware of this mistaken perception. 332 lkog tu gyur pa ma yin. 333 gsal ba min pa gsal mi 'gyur/ gzhan gyis kyang ni gsal mi nus/ sngon po la sogs gsal ba ste/ de phyir gnyis med shes pa yin/ This verse also appears to be original to Ratnākaraśānti. 334 mdo sde pa ltar na yang/ shes pa de (P/N/S: des) 'brel pa gang gis don de myong bar 'gyur/ de las byung ba'i 'brel pas so zhe na/ mig kyang mig gi rnam par shes pas myong bar 'gyur ro/ 234 during a dream, their appearances nevertheless arise. Do they arise through the force of similarity? If they were similar in every regard, then even consciousness itself would be insentient!335 And if they are only similar in a certain regard, then all [dream-cognitions] would know everything.336 Additionally, the appearance of cloth and woven grass mats and so on must either be singular or manifold. If it is singular, then the beginning, middle, and end parts would not appear separately, and there would arise the contradiction of what is different being the same. And why is that? Because it is impossible that whatever possesses parts in this way is a single, individual entity. Not even “partless” particles are counted while remaining solitary, because they always appear clustered together; particles are not observed by themselves. Some claim that many particles are joined together. But if these particles are joined by their parts, then they are not “partless” particles. And if they are joined despite not having parts, then they would all be stacked in the same place, one on top of the other, and consequently even In other words, the assertion that the connection between sense-object and sensory consciousness arises from the object, itself, is tantamount to the claim that sensory experience arises simply as a result of the connection between sense-object and sensory consciousness. But, in classical Abhidharma perceptual theory, sense-object and sensory consciousness are only two of the three terms necessary for a sensory experience; classically, the interaction between a sense-faculty and an object of sensation is what produces a sensory consciousness. Therefore, the sensefaculty is no less “connected” to the sensory consciousness than the object of sensation, and if this “connection” is sufficient to establish an experience, then the sense-faculty itself—in this case, the eye or eye-faculty—should also be experienced as a result of an eye-consciousness (i.e. a visual consciousness). 335 yang rmi lam de med kyang der snang ba 'byung ba'i phyir ro/ 'on te 'dra ba'i dbang gis so zhe na/ de yang kun gyis mtshungs na ni/ shes pa yang bems po nyid du 'gyur ro/ That is to say, the objects experienced in a dream are strictly nonexistent. If it is asserted that these experiences arise due to some putative “similarity” between the objects perceived in the dream-state and their counterparts in waking life, it is necessary to strictly delimit the sense in which these are “similar.” Otherwise, the absurd consequence follows that the objects perceived during a dream are in some sense external or material. Effectively, this is a broad critique of the notion that there exist separate cognitive-perceptual apparati for processing visual images in dreams and in waking life. Ratnākaraśānti’s point here is that the appearances in a dream are just as “illuminated,” i.e. just as immediately accessible to sense-consciousness, as appearances in waking life. 336 'ga' zhig gis mtshungs na ni kun gyis (P/N/S: gyi) thams cad rig par 'gyur ro. 235 mountains would only measure as high as a single particle! Therefore woven cloth and so on are neither singular nor manifold. If they are joined by their six sides, Then “partless” particles would have six parts! If the six are together in one single place, Masses would be merely a single particle! [VK 12] So it was said. The flaws337 which follow from the claim that blue-patches and so on are external are the same for a blue-patch which has the nature of consciousness, because there is no difference in the unacceptable conclusions that follow. The distinction between [external] objects and consciousness simply does not amount to anything at all. Now, someone might say, “External particles are surrounded by six directions, but the individual moments of consciousness have four points of connection.” However, because they are the same in terms of having parts, these [differences] do not amount to anything at all, either. As there is no third kind of bundle apart from something which is either singular or manifold, blue-patches and so on are not established as either internal or external objects, and are therefore false. This [false image] also has the nature of mere luminosity, which is not disturbed by the force of prior psychological imprints. Every nature is illuminated. Successively, these natures are: 1) Not positioned with respect to anything else 2) Positioned, but real 337 skyon (P: rkyen) 236 3) Unreal Because luminosity is not positioned with respect to anything else, it is incorrect to say that it is deceptive.338 Because it is undifferentiated, and because there would arise the unacceptable consequence of an infinite regress, there can be no act of positioning it. Therefore, this luminosity is a direct means of reliable knowledge, because there is no distortion in its nature. As for positioned339 entities, these are distorted, similar to the appearance of an assortment of hairs floating in the sky. This being the case, do blue-patches and so on have an identity of one single luminosity? Characteristics such as “blue” and “yellow” are mutually exclusive; when appearing, they are not 338 gang zhig gi bdag nyid du gsal ba ste/ bdag nyid de dag ni go rims bzhin du gzhan gyis ma bzhag pa dang/ bzhag pas yang dag pa dang/ yang dag pa ma yin pa yin gyi/ gsal ba ni gzhan gyis bzhag (D/C: gzhag) pa ma yin pa'i phyir/ brdzun par mi rung ngo/ This passage hinges on the technical term vyavasthā, rendered here in the Tibetan translation as bzhag pa. Elsewhere in this work, bzhag pa is presumably rendering the Sanskrit sthāpan(īy)a, translated into English as “posited” in the sense of being laid out or fixed in place. Here, however, the term is being used in a different, more technical sense, which follows Dharmakīrti’s famous definition of a real entity in PV 1.40: sarve bhāvāḥ svabhāvena sva-svabhāvavyavasthiteḥ/ svabhāva-parabhāvābhyāṃ yasmād vyāvṛtti-bhāginaḥ// Tib. (sDe dge: tshad ma, ci, ff. 96.a.6-7) gang phyir dngos kun rang bzhin gyis/ rang rang ngo bo la gnas phyir/ mthun dngos gzhan gyi dngos dag las/ ldog pa la ni brten pa can// “All real entities are positioned with respect to their own self-nature; their opposites [i.e. nonentities] rely on the self-nature of some other entity.” Interestingly, the Tibetan translation of PV 1.40 renders vyavasthiteḥ as gnas, which has the literal sense of remaining in a particular place. The point is that, for Dharmakīrti, all entities are “positioned” or “located”—Dunne (2004) translates vyavasthiteḥ in PV 1.138 as “situated”—with respect to something else: the metaphorical zero-point of the coordinate system which rigorously defines their location. For Dharmakīrti, as the above verse illustrates, real entities (bhāva, Tib. dngos po) are positioned with respect to their own self-nature, which is to say that they are not located with respect to anything external to themselves; they are their own zero-point. Unreal or nonexistent entities (abhāva, Tib. dgnos med), by contrast, are located with respect to some other zero-point. For example, the phenomenal object (grāhyākāra, Tib. gzung rnam) is only located with respect to the phenomenal subject (grāhakākāra, Tib. ‘dzin rnam), and vice versa, thus both are unreal. Ratnākaraśānti adopts this basic framework, and argues that ultimately real entities such as the luminous nature of reflexive awareness are not located or positioned with respect to anything else [gzhan gyis ma bzhag pa]. That is to say, luminosity is not located by virtue of its relation to some phenomenal subject or object— or, for that matter, anything else at all. Ratnākaraśānti then distinguishes between two different classes of entities which are positioned with respect to an external locus: “real” and “unreal” entities. Although he does not specifically relate this distinction to the theory of the three natures, it clearly encompasses the difference between the dependent and the constructed natures. Likewise, the first “unpositioned” category maps closely on to the perfected nature. 339 bzhag (D/C: gzhan) pa ni nam mkha' la skra'i tshogs snang ba bzhin 'khrul pa yin no/ 237 singular. However, while it is mistaken to say that they have the identity of a single luminosity, they do not have a manifold nature, either. One might ask, “Is it not the case that blue-patches and so on are singular in terms of how they are experienced? After all, they are experienced simultaneously.” But this is not so; in such a case, what is different would be undifferentiated.340 Or: “Is luminosity also [experienced] differently, like those things?” Because each thing is experienced individually, and there are no “simultaneous” experiences, a single [experience] does not result in the experience of everything. Therefore blue-patches and so on are neither singular nor manifold, and are similar to hairs in the sky. Just as objects in the reflection of a mirror Are exclusive of “one” or “many,” Although appearing in this way, they have no existence; The essential nature of entities is like that. [LAS X.709] So it was said. Concerning what is meant by “exclusive,” [phenomena] are neither single nor manifold. Are blue-patches and so on the same as luminosity, or different from it? What is different from luminosity cannot appear. On the other hand, if they are not different from luminosity, then because they are this luminous nature, they would be one single thing. And it is impossible for the variety [of phenomena] to be singular, as was discussed above. In this context, “sameness” and “difference” are the same as “one” and “many” in the example of the reflection in a mirror. 340 ci ste sngon po la sogs pa 'di ni ji ltar nyams su myong ba de ltar gcig yin te/ lhan cig nyams su myong ba'i phyir ro zhe na/ ma yin te tha dad pa nyid tha dad pa ma yin no/ The interlocutor asserts that a multicolored patch is experienced as a single entity; Ratnakarasnati follows Dharmakirti [PV 3.200] in maintaining that a multicolored (i.e. phenomenally differentiated) patch cannot be a single (undifferentiated) entity. Cf. PV Dunne (2004) pp. 396ff. ad PV 3.194-224. 238 Some adherents of the Yogic Practice and Middle Way schools, who maintain that consciousness possesses [real] cognitive images, claim that it is unacceptable for luminous bluepatches to have a nature that is both deceptive and non-deceptive, as these qualities are contradictory. If [luminosity] were not their nature, then blue-patches and so on would not be luminous; however, blue-patches are indeed luminous. Since they are not something other than the luminosity which is real or non-imputed, blue-patches and so on are also real or non-imputed. Well, according to them, because all luminous [phenomena] are experienced in their nonerroneous nature, there would not be any cognitive distortion whatsoever. Therefore, all sentient beings would be eternally liberated, and would always already be completely perfect, genuine Buddhas! [Objection:] “While cognitive images are not joined341 to blue particulars, they are not false. But the external “blue” is false, because its appearance, despite not being an object, is nevertheless positioned as an object through an imaginative determination.”342 Well in that case, since the appearance is imaginatively determined, that [appearance] should be apprehended as just consciousness in every respect. It would not be [apprehended] as an object, because it is not an object. You say, “It is distorted for precisely that reason!” But this very distortion ought to be examined. 341 Ordinarily, mi rigs pa (Skt. ayukta) means “illogical.” Here, however, it is clear that rang gyi mtshan nyid (Skt. svalakṣaṇa) or “specifically characterized [phenomenon]” is functioning as the object of mi rigs pa, since la follows in every edition, but mi rigs pa in the sense of “illogical” does not take an object. We must therefore surmise that the original Sanskrit reads ayukta in the more literal sense of not being joined, connected, or associated with the svalakṣaṇa. 342 According to John Dunne (personal communication), mngon par zhen pa (Skt. abhiniveśa) or “imaginative determination” is a type of conceptual act that takes some information—usually, but not always, sensory information—and misconstrues it in some way. The implication is that the object is not being completely misconstrued, only that the conceptualizing mind engages in some type of imputation (sgro btags pa, *samāropa) with respect to the relevant information. 239 If an object is apprehended, then because it is in every respect the object of that [perceptual apprehension], there is no apprehension when the object does not exist. In such an event, the object is not apprehended, and it is also not cognized. In this manner, mental contamination is the positioning [of objects] as they are generally renowned in the world, which is not the genuine way things actually are. The evidence for the prior objection was that anything which is not different from luminosity—even a false image—is a real entity. But if [luminosity] is not different from the nature of being a real entity, this evidence does not apply. 343 And if the evidence is mere non-difference then it is inconclusive, since this evidence also holds for apprehended and apprehender and so on, even though they do not exist.344 Some “False Imagist” followers of Yogic Practice and the Middle Way say that conceptualized extended entities do not exist substantially.345 They say that the direct means of 343 gang yang gsal ba las tha mi dad pa'i phyir rnam pa brdzun pa yang dngos por 'gyur ro zhes pa'i gtan tshigs 'di [la] gal te dngos po'i rang bzhin du tha mi dad do/ zhe na gtan tshigs ma grub par 'gyur ro/ D/C is missing [la]. The argument here relies heavily on the technicalities of predicate logic in pramāṇa theory; cf. Dunne (2004) pp. 26ff. for an introduction to this topic. Basically, a subject (Skt. dharmin, Tib. chos can) is asserted to possess a predicate quality (dharma, Tib. chos), on the basis of some type of evidence (Skt. hetu, Tib. gtan tshigs). However, if it can be demonstrated that there is no “pervasion” (Skt. vyāpti, Tib. khyab pa) connecting the evidence to the predicate, the predication fails. In this case, the opponent’s position is that false images (the subject) are real (the predicate), because they are not separate from luminosity (the evidence). As Ratnākaraśānti points out, this argument turns on the assumption that whatever has the nature of being luminous is a real entity. But this assumption begs the question; if “being luminous” is the same as “being real,” then the interlocutor’s argument can be rephrased as a tautology: false images are real, because they are real. Therefore the supplied evidence is insufficient to establish the validity of the proposition [gtan tshigs ma grub], and there is no pervasion. 344 'on te tha mi dad pa tsam gtan tshigs yin na ni ma nges par 'gyur te/ gzung ba dang 'dzin pa nyid la sogs pa la de med par gtan tshigs yod pas so/ The interlocutor, a “True Imagist” Yogācārin, admits that the phenomenal subject and phenomenal object are not real. But, by virtue of appearing at all, they are illuminated. Thus the “mere non-difference” [tha mi dad pa tsam] between being luminous and being real is contradictory on its own terms, and cannot establish the reality of cognitive images (or anything else), since—as in the case of the phenomenal subject and object—there are instances of illuminated, yet unreal, phenomena. 345 rnal 'byor spyod pa pa dang/ dbu ma rnam pa med par smra ba rnams rnam par rtog pa'i lus rdzas su med par smra ste/ P/N/S has …rjes su med… 240 reliable knowledge are free from concepts at the very first moment, and that subsequent inferences are not, therefore it is not possible to refute any cognitive images. Some followers of the Middle Way tradition claim that whatever is generally renowned in the world is the relative truth. External objects are generally known in the world, as are the mind and mental functions; so these, too, exist in terms of the relative truth. [They say] mind and mental functions are distorted, and reflexive awareness is like a sword that cannot cut itself, therefore it is contradictory. [They say:] “Both346 of us assert that awareness of something else is merely contamination.347 Once cognitive distortion has been extinguished, the ultimate is realized. Therefore Buddhas are those in whom the mind and mental functions have been extinguished. At that point, the “arising” of a Buddha is posited as an experience in conformity with the ultimate, which is the non-arising of all phenomena. In order to accomplish the various aims of sentient beings within unlocated Nirvāṇa, imbues form with blessings.348 Having done that, genuine enlightenment is made manifest.” This is even more inconceivable than the inconceivable! Since there are no mind and mental functions, there is not any realization, either—so how can there be any ultimately For these “False Imagists,” the cognitive image consists of “conceptualized extended entities” [rnam par rtog pa’i lus], thus to say that these conceptual forms are “not substantially existent” [rdzas su med pa] is to say that cognitive images do not exist. This is considerably different from Ratnākaraśānti’s “False Imagism.” 346 i.e., both the interlocutor (who denies the existence of reflexive awareness) and his opponent (who maintains the existence of reflexive awareness). 347 rang gi rig pa 'gal ba'i phyir ral gri'i sos rang mi gcod pa bzhin no/ gzhan rig pa yang bslad pa tsam du 'u bu cag gnyi ga 'dod do/ 348 This contention is found in the last section of the Entry into the Middle Way, wherein Candrakirti argues that the Buddha, no longer having any mind or mental functions, merely appears to be acting for the welfare of sentient beings from their unenlightened perspective. In reality, according to Candrakirti, the Buddha is no more conscious of his beneficial activity than a jewel that has been “blessed” with the power to fulfill wishes—hence the metaphor. Cf. Dunne (1996). 241 authentic, perfect enlightenment at all?349 “Realization is just that [absence],” he says. But realization itself is mental, therefore it is not the case that mind and mental functions no longer exist. Furthermore, if the ultimate is experienced as something else, then that experience is distorted. And if it is not different, then it is reflexively known.350 At the time when mind and mental functions are distorted, what is the antidote? The coarse contrary of cognitive distortion is its opposite, while the subtle is the antidote.351 Since ordinary cognitions are not an antidote, they do not extinguish the mind and mental functions. Thus the experience of the innate essence, itself, by means of “mere cognitive representation” is a direct means of reliable knowledge, because it is not distorted.352 The experience of an imputed self-nature is distorted; therefore, all distorted dualistic marks should gradually be abandoned. Like endless space without a single stain, the experience of the luminous body of all phenomena, devoid of duality, is the authentic realization of the ultimate. However, the cessation of mental 349 sems dang sems las byung ba 'di dag med pa'i phyir rtogs (D/C: rtog) pa nyid kyang med na/ don dam par yang dag par rdzogs pa'i byang chub lta ga la yod/ 350 gzhan yang don dam pa de gzhan (D/C: bzhin) du gyur pa myong na ni 'khrul par 'gyur la/ tha mi dad na (D/C: pa) ni rang rig par 'gyur ro/ In other words, an experience of the ultimate as though the ultimate were some external object is necessarily distorted, as all dualistic cognitions are distorted by definition. And, if the ultimate itself is not different from the realization of the ultimate, then that realization is reflexive; it is not separate from its object. Thus the interlocutor’s rejection of reflexive awareness is misguided. 351 This is a play on words that is lost in Tibetan (and English) translation. Ratnākaraśānti distinguishes the “opposite” [mi mthun phyogs, *vipakṣa] from the “antidote” [gnyen po, *pratipakṣa]. 352 de bas na 'di ltar rnam par rig pa tsam gyis (D/C: gyi) gnyug ma rang gi ngo bo myong ba ni mgnon sum gyi tshad ma yin te/ ma 'khrul ba'i phyir ro/ Here, Ratnākaraśānti explicitly identifies “mere cognitive representation” [rnam par rig pa tsam, *vijñaptimātra] as a “direct means of reliable knowledge” [mngon sum gyi tshad ma, *pratyakṣa-pramāṇa]. Ratnākaraśānti also identifies “mere cognitive representation” as the second stage of Yogic Practice. This implies that Ratnākaraśānti considered contemplative practices such as the Four Yogas to be direct means of reliable knowledge. See above, §III.B.1.c. 242 functions is not logical.353 And while [mind and mental functions] are defiled, the “seeds” of all phenomena have not yet been extinguished. It is precisely for this reason that we speak of the “storehouse” consciousness. Additionally, they are brought to cessation in the three types of Arhants. 354 How is this brought about? Through the transformation of the basis355 into the expanse free from any defilement whatsoever or the Body of Liberation, which is like a pure crystal.356 The True Body [chos kyi sku, Skt. dharmakāya] is especially exalted over and above these, as it supports all the qualities [chos, Skt. dharmas] of Buddhahood. This is similar to how, in terms of mere purity, there is no difference between a sunstone and a moonstone as compared to the sun and the moon, but there is a difference in terms of the magnitude and excellence of their rays of light. How is it that all phenomena are within the transcendent wisdom of the Buddhas and the Bodhisattvas? They appear as one taste—suchness—like limitless space that is utterly without stains. And why is that? Because the marks of duality have disappeared. “But then, since transcendent wisdom appears as a single cognitive image, wouldn’t it not be authentic wisdom, as it is distorted?” Then everything is deceptive, and since they are conceptual constructions of 353 sems las byung ba ‘gag pa ni rigs (D/C/N: rig) pa ma yin no/ That is to say, it is illogical to maintain that the authentic realization of the ultimate is merely the cessation (‘gag pa, Skt. nirodha) of mental functions. This argument has precedent in some of the earliest strata of Buddhist praxeological literature, which identify a state of complete mental cessation (nirodha-samāpatti) apart from the ultimate fruition. Indeed, as Ratnākaraśānti points out, the postulation of the “storehouse consciousness” stems from the need to respond to theoretical problems surrounding the existence of just this state. Cf. Waldron (2003), pp. 78ff. 354 dgra bcom pa gsum, i.e. realized disciples (Pali savakabuddhas), solitary Buddhas (Pali paccekabuddha), and completely perfect Buddhas (Pali sammasambuddhas). 355 This is another play on words that is lost in translation; Ratnākaraśānti states that the mind and mental functions are brought to “cessation” [ldog pa, *vyāvṛtti] through the “transformation of the basis” [gnas gyur pa, *āśrayaparāvṛtti], a synonym for enlightenment especially in early Yogic Practice literature. The “basis” is the storehouse consciousness. 356 See Thirty Verses on Consciousness (TK 30) 243 unreal things, nothing more than the imputation of names and dualistic marks. Therefore it is illogical that the cessation of the mind and mental functions is authentic enlightenment. Some adherents of the Middle Way claim that the Path of the Middle Way is just the illusoriness of phenomena. But if these [phenomena] are proven to be nonexistent by a means of reliable knowledge, then it is impossible to establish their existence by means of anything at all. So what need is there to mention establishment by mere appearance, which is not even a means of reliable knowledge? If everything is just nonexistent, then what is it that is designated, and in dependence upon what? They say that everything is not existent, and that everything is also not nonexistent, and that this is the path of the Middle Way. But such a position is untenable, as existence and nonexistence are mutually-contradictory: each excludes the other. Thus, when one is refuted, the second’s characteristics are present. It is therefore untenable to refute both. Additionally, all consciousnesses are also deceptive, in the way that everything is false, thus not even the slightest direct perception is established. Therefore, since there is no reliable ascertainment via the three-fold syllogism, all evidence is only spurious evidence. Furthermore, all the refutations of something or proofs of something else would also not be established— because these rely on the means of reliable knowledge, while you assert that there are no means of reliable knowledge! If you refute through mere words, it is said that you yourself, among the cows, have the view of a heretical Far-Thrower.357 Others deny that you are a man or a follower of the Middle Way. The intention of the statement that all phenomena lack an inherent essence, and so on, is reliably determined by the Sūtra Unraveling the Intent and the Ornament of the Great Vehicle 357 mu stegs rgyang pan (em. phan). The Cārvākas or “Far-Throwers” maintained that only material phenomena, composed of the elements (numbering four or five in classical Indian physics), exist. They accept the epistemic reliability (pramāṇya) of perception, but not of inference. 244 Sūtras, and also by Maitreya, Asaṇga, and Nāgārjuna. If you accept that there are means of reliable knowledge, then they are established in the manner of the three natures. So abandon the contention that everything is false! And if you say, “But we don’t assert that there are means of reliable knowledge!”, well, people who debate without any means of reliable knowledge are just clowns. As for ourselves, we assert that the unreal dependent nature exists, and that the bundles and so on are designated in reliance upon it, because all proofs and refutations are performed by using the means of reliable knowledge. Whatever arises as interdependent origination, This is explained to be emptiness: That which is caused and imputed.358 This is the Path of the Middle Way. [MMK 24.18] So it is said. Although their nature is luminosity, blue-patches and so on are false since they are harmed by analysis. But that luminosity is established as being real, because as an awareness that is free from distortion, it is a direct means of reliable knowledge. Luminosity is the inherent selfnature which gives illumination, because it is not positioned by the contamination through which some experience would be designated as distorted. Something is asserted to be lacking in “self” if it continuously changes by becoming something else. This refutes phenomena which are permanent and possess parts. However, phenomena possessing the characteristic of being momentary, yet abiding uninterruptedly, cannot be refuted because they are established by a means of reliable knowledge. 358 The canonical rescension of this verse reads, rten cing ‘brel bar ‘byung ba gang/ de ni stong pa nyid du bshad/ de ni brten nas gdags pa ste/ de nyid dbu ma’i lam yin no/ However, Ratnākaraśānti’s version reads: gang zhig rten cing ‘brel bar ‘byung/ de ni stong pa nyid du bshad/ de ni gyur byas brtags (D: btags) pa ste/ de ni dbu ma’i lam yin no/ This alteration brings Nāgārjuna’s verse closer in line with the classical Yogācāra perspective. 245 Deceitful fools deny what even cowherds can see: that sometimes phenomena are pervaded in general, but there is no pervasion in a specific instance. They pervert the meaning of “Do not abandon the distinction between engaging and refraining,” and claim that darkness arises from tongues of flame.359 But this is like saying that, because beer is medicinally useful for those with nervous disorders, it should also be useful as a medicine for those with liver disease. Because this is an unacceptable consequence, it negates one’s own arising; the negation of one’s own arising is also—or just is—the unacceptable consequence.360 [You say”] “The pervasion is not established.” The same person361 later saw the vigorous criticisms of the one with this logical reasoning and even repudiated himself. So you must not say that!362 359 g.yo can blun po chos gang dang gang (C/S: em. gi) spyi la khyab pa dang ma khyab pa'i bye brag ba lang rdzis kyang mthong ba la'ang bsnyon cing kho nyid 'jug ldog gi khyad par mi 'dor ba sgyu thabs la spyod pa dag me lce las mun pa skye bar 'gyur ro zhes zer ba ni The grammatical and logical relationship between the “perversion of meaning” or deceptive argumentation (sgyu thabs, *chala-kṛta), i.e. the claim that darkness can arise from tongues of flame, and “not abandon[ing] the difference between engaging [in virtue] and refraining [from vice],” a common Buddhist maxim, is unclear. The specific example of darkness arising from “tongues of flame” (me lce), which is not found in the Root Verses of the Middle Way, the Entry to the Middle Way, or the Clear Words, appears to come from Jayānanda’s Commentary to the Entry of the Middle Way (D3870: 133.a.4-.7, 160.a.3, 289.a.5), the only known Sanskrit commentary to any of the works of Candrakīrti I. This phrase occurs in the context of a refutation of causality—tongues of flame and darkness being compared to a seed and a sprout, respectively—although the relationship between light and darkness was earlier explored by Nāgārjuna in MMK VII.9-12. Certainly the reference, in the next line, to “unacceptable consequences” implicates Candrakīrti in this discussion. 360 ci de lta na de nyid kyang thal bar 'gyur ba yin pas rang skye 'gog byed du 'gyur te rang skye 'gog pa de yang de ste thal 'gyur yin no/ The “unacceptable consequence” (Skt. prāsaṇga, Tib. thal ‘gyur ba) is something of a loaded word. For more detail on the technical meaning of this term, as well as the long history of polemics it inspired, cf. Dreyfus and McClintock, ed., The Svātantrika-Prāsaṇgika Distinction. Most basically, a prāsaṇga is a type of logical argument, wherein the premises are demonstrated to contradict the conclusion, or else to lead inexorably to an obviously false or absurd result; as such, it encompasses, but is not strictly limited to, a species of reductio ad absurdum. Ratnākaraśānti is pointing out here that exclusive reliance on “unacceptable consequences,” particularly in the context of an insistence that all phenomena are strictly false, is ultimately self-defeating. 361 362 i.e., Candrakīrti khyab pa ma grub ce na ni mtshungs pas slan chad de'i rigs can la btsan thabs su bsnyon pa mthong la rang dma’ (D/C: sma) 'bebs de skad kyang ma zer cig/ 246 Moreover, blue-patches and so on are contingent, therefore they are conditioned by contamination. Since they are like that, the experience of them is also distorted. Thus there are contexts in which phenomena such as blue are harmed by analysis; but this is not so for luminosity. From the Large Sūtra of the Great Dharma Mirror-Like Great Vehicle: Twenty-eight inauthentic views appear. These are: 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17) 18) 19) 20) 21) 22) 23) 24) 25) 26) 27) 28) The view of nonexistent dualistic marks The view which denigrates the constructed nature The view which denigrates the imagination of the unreal The view which denigrates suchness The view of thorough grasping The view of complete transformation The view that there are no evil actions The view of renunciation Contemptuous views Incendiary views Misguided views The view of production The view of non-assertion Deceitful views The view of paying respect Views which teach stupidity Basic views The view of “not having a view” The view which blocks application The view of not having renunciation The view of accumulating obscurations The view that [liberation] arises without accumulating merit The view that there is no result The view of clinging to reliable ascertainments Untrue views The view of not speaking The great view The view of those with manifest pride Lay vow-holders who desire to realize nonconceptual awakening, the thorough investigation of reality, should attend and meditate as follows. There are two ways in which this 247 meditation may be undertaken: intensively, and continuously. “Meditation” here means the unity of Calm-Abiding and Clear-Seeing. The focal object of Calm-Abiding meditation is a nonconceptual image of both the full range of phenomena and the way that phenomena are, just as they are. The focal object of Clear-Seeing meditation is the same image, but together with concepts.363 When there is an uninterrupted experience of both of these, that is called the “Path which is the Unity of Calm-Abiding and Clear-Seeing.” These are the ordinary divisions of yoga. As for the special divisions of yoga, these are: the observation of the two extremes of entities, the observation of “mind only,” the observation of suchness, and no observation. The first stage of yoga takes as its support the full range of phenomena. The second stage of yoga takes as its support the way that phenomena are, which is “mind only.” The third stage of yoga takes as its support the suchness of all phenomena, just as it is. The fourth stage of yoga, in which one sees the Great Vehicle, is non-appearance. From the Descent to Lanka: In dependence on “mind only,” External objects become unintelligible. Having focused on suchness, One should also pass beyond “mind only.” Having passed beyond “mere mind,” One should pass into non-appearance. The yogi who abides in non-appearance Sees the Great Vehicle. [LAS IX.256-257] So it was said. 363 de la ji snyed yod pa dang/ ji lta ba bzhin du gyur pa la rnam par mi rtog pa'i gzugs brnyan du gyur pa ni zhi gnas kyi dmigs pa'o/ de la rtog pa dang bcas pa'i gzugs brnyan du gyur pa ni lhag mthong gi dmigs pa'o/ 248 This was written by Ratnākaraśānti To please those learned masters Who discern the stages of the Buddha-Vehicle And correctly rely on the methods of reliable knowledge! Whatever virtue there has been in this endeavor, I dedicate the entirety of it to all sentient beings. *** These were the Pith Instructions to the Ornament of the Middle Way, composed by the great master Ratnākaraśānti, who was praised by all at that time as being supreme and unrivaled among the scholars. He was invited by the wisdom ḍākinīs to Oḍḍiyāna, and even among the forty whose accomplishments are celebrated in stories he was especially renowned for his commentaries to the sūtras and tantras. He was also the chief among the four Gatekeepers [at Vikramaśīla] of his time, because of his unmistaken realization of the intent of Ārya Asaṇga and Nāgārjuna Garbha, and his most excellent clarifications of their teachings. The monk Candrakīrti—who had deviated from the intent of Nāgārjuna—thenceforth, in his later life, completely gave up nihilism and wrote commentaries on the profound tantras. As the Blessed One said, for example in the Sūtra Unraveling the Intent and the Increasing Faith in the Sūtras, it came to pass that some behaved deceitfully, following after those mistaken writings: “Those people will not realize my intention. Later, although they misunderstood, there will come one who unerringly explains my intention. From the first, there have arisen many who followed after my words.” Among the prophecies, this master in particular was referred to in the Root Tantra of Mañjuśrī: 249 One who is called Ratnākara by name Will arise to clarify the teachings of the Teacher. Such were the prophecies. This was translated by the Indian abbot Śāntibhadra and the ordained translator Śākya ‘od. We descended on the main point in order to annihilate the deceitful. Logic is the central pole on the victory banner of omniscient wisdom! This is the Pith Instructions on the Ornament of the Great Vehicle-Consciousness-Middle Way, Logical arguments which most eloquently refute Candrakīrti. Later, the Kashmiri Amogha and the translator ‘O-ru settled any questions. 250 Bibliography Abhayadhatta. Lives of the Eighty-Four Siddhas (Caturaśīti-siddha-pravṛtti). Translated into Tibetan by sMon grub Shes rab sMon grub Shes rab as grub thob brgyad cu rtsa bzhi’i lo rgyus. Translated into English by James B. Robinson as Buddha’s Lions: The Lives of the Eighty-Four Siddhas. Berkeley: Dharma Publishing, 1979. Anacker, Stefan. Seven Works of Vasaubandhu: The Buddhist Psychological Doctor. Delhi: Motilal Banarsidass, 1984. Arnold, Dan. Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind. New York: Columbia University Press, 2012. —. “Buddhist Idealism, Epistemic and Otherwise: Thoughts on the Alternating Perspectives of Dharmakīrti.” In Sophia 47 (2008): 3-28. —. “Is Svasaṃvitti Transcendental? A Tentative Reconstruction Following Śāntarakṣita.” In Asian Philosophy 15, no. 1 (2005): 77-11 —. “Self-Awareness (svasaṃvitti) and Related Doctrines of Buddhists Following Dignāga: Philosophical Characterizations of Some of the Main Issues.” In Journal of Indian Philosophy 38 (2010): 323-378. Asaṇga. Mahāyānasaṃgraha. Translated by Étienne Lamotte in Mahāyānasaṃgraha: La Somme du Grand Vehicule d’Asaṇga. Translated from the French by Gelongma Karma Migme Chodron. Batchelor, Stephen. Buddhism Without Beliefs: A Contemporary Guide to Awakening. New York: Riverhead Books, 1997. Bentor, Yael. “Fourfold Meditation: Outer, Inner, Secret, and Suchness.” In Religion and Secular Culture in Tibet: Tibetan Studies II: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, 41-58. Leiden: International Association of Tibetan Studies, 2000. Bilgrami, Akeel. Self-Knowledge and Resentment. Cambridge: Harvard University Press, 2006. Bohr, Niels. Atomic Physics and Human Knowledge. Oxford: Wiley, 1958. Boquist, Ake. “Trisvabhāva: A Study of the Development of the Three-nature-theory in Yogācāra Buddhism.” In Lund Studies in African and Asian Religions 8 (1993): 1-159. 251 Chatterjee, Ashok Kumar. The Yogācāra Idealism. Delhi: Motilal Banarsidass, 1975. Cheng, Hsueh-Li. Empty Logic: Mādhyamika Buddhism from Chinese Sources. Delhi: Motilal Banarsidass, 1991. Conze, Edward (translator). The Large Sutra on Perfect Wisdom: with the divisions of the Abhisamayālaṇkāra. Berkeley: University of California Press, 1975. Dignāga, Compendium of Pramāṇa (Pramāṇasamuccaya). Edited by Ernst Steinkellner. In “Dignāga’s Pramāṇasamuccaya, Chapter 1: A hypothetical reconstruction of the Sanskrit text with the help of the two Tibetan translations on the basis of the hitherto unknown Sanskrit fragments and the linguistic materials gained from Jinendrabuddhi’s Ṭīkā.” http://ikga.oeaw.ac.at/Mat/dignaga_PS_1.pdf. Doctor, Thomas (translator). Speech of Delight: Mipham’s Commentary on Śāntarakṣita’s Ornament of the Middle Way. Ithaca: Snow Lion Publications, 2004. Damasio, Antonio. The Feeling of What Happens. Boston: Harcourt, 1999. D’Amato, Mario. “Three Natures, Three Stages: An Interpretation of the Yogācāra TrisvabhāvaTheory.” In Journal of Indian Philosophy 33 (2005): 185-207. Dharmakīrti. Commentary on Pramāṇa (Pramāṇavarttika, tshad ma rnam ‘grel). Edited by John Dunne in Foundations of Dharmakīrti’s Philosophy. Dol po pa Shes rab rGyal mtshan, Mountain Doctrine: Ocean of Definitive Meaning (ri chos nges don rgya mtsho). Translated by Jeffrey Hopkins as Mountain Doctrine: Tibet’s Fundamental Treatise on Other-Emptiness and the Buddha Matrix. Ithaca: Snow Lion Publications, 2006. —, Shes rab kyi pha rol tu phyin pa man ngag gi bstan bcos mngon par rtogs pa’i rgyan gyi rnam bshad mdo’i don bde blag tu rtogs pa. In gsung ‘bum/ shes rab rgyal mtshan (‘dzam thang). TBRC W21208. Vol. 5 (ma): ff. 123.a-310.b. Dowman, Keith. Masters of Mahāmudrā. Albany: State University of New York Press, 1985. Dreyfus, Georges B. J. Recognizing Reality: Dharmakīrti’s Philosophy and its Tibetan Interpretations. Albany: State University of New York Press, 1997. Duckworth, Douglas. “Two Models of the Two Truths: Ontological and Phenomenological Approaches.” In Journal of Indian Philosophy 38 (2010): 519-527. Dunne, John. Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publications, 2004. 252 —. “Realizing the Unreal: Dharmakīrti’s Theory of Yogic Perception.” In Journal of Indian Philosophy 34 (2006): 497-519. —. “Thoughtless Buddha, Passionate Buddha.” In Journal of the American Academy of Religion 64, no. 3 (1996): 525-556. Fox, Alan. “Empty Logic: Mādhyamika Buddhism from Chinese Sources by Hsueh-li Cheng” (Review). In Journal of Chinese Philosophy 13 (1986): 361-365. Garfield, Jay. “The Conventional Status of Reflexive Awareness: What’s at Stake in a Tibetan Debate?” In Philosophy East and West 56, no. 2 (2006): 201-228. —. The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā. Oxford: Oxford University Press, 1995. ‘God Lo tsa ba gZhon nu dpal, The Blue Annals (deb ther sngon po). Translated by G. N. Roerich. Delhi: Motilal Banarsidass, 1976. Gold, Jonathan. “No Outside, No Inside: Duality, Reality and Vasubandhu’s Illusory Elephant,” Asian Philosophy 16, no. 1 (2006): 1-38. Goldfield, Ari, Jules Levinson, Jim Scott, and Birgit Scott (translators). The Moon of Wisdom: Chapter Six of Chandrakirti’s Entering the Middle Way with commentary from the Eighth Karmapa Mikyö Dorje’s Chariot of the Dagpo Kagyü Siddhas. Ithaca: Snow Lion Publications, 2005. Grimes, John. A Concise Dictionary of Indian Philosophy: Sanskrit Terms Defined in English. Albany: State University of New York Press, 1996. Gruschke, Andreas. “Der Jonang-Orden: Gründe für seinen Niedergang, Voraussetzungen für das Überdauern und aktuelle Lage.” In Tibet, Past and Present: Tibetan Studies I: Proceedings of the Ninth Seminar of the International Association of Tibetan Studies: 183-214. Edited by Henk Blezer. Leiden: Brill Academic Publishers, 2002. Gundmundsen, Chris. Wittgenstein and Buddhism. London: Macmillan, 1977. Hopkins, Jeffrey. Meditation on Emptiness. Boston: Wisdom Publications, 1996. Hume, David. A Treatise of Human Nature. 1739. Reprint, Project Gutenberg: http://www.gutenberg.org/files/4705/4705-h/4705-h.htm. Jayānanda, Commentary to the Entry of the Middle Way (dbu ma la ‘jug pa’i ‘grel bshad ces bya ba). D:3870, dbu ma vol. Ra: ff. 1.b.1-365.a.7. 253 Kajiyama, Yuichi. “Controversy between the Sākāra- and Nirākāra-vādins of the Yogācāra School—Some Materials.” In Journal of Indian Buddhist Studies 14, no. 1 (1965): 41829. Kellner, Birgit. “Self-Awareness (svasaṃvedana) in Dignāga’s Pramāṇasamuccaya and -vṛtti: A Close Reading.” In Journal of Indian Philosophy 38 (2010): 203-231. Kriegel, Uriah. “Consciousness as Intransitive Self-Consciousness: Two Views and an Argument.” In Canadian Journal of Philosophy 33, no. 1 (2003): 103-132. Lamotte, Etienne. The History of Indian Buddhism. Translated by Sara Webb-Boin. Louvain-LaNeuve: Publications de l’Institut Orientaliste de Louvain, 1988. Liu, Ming-Wood. “The Three-Nature Doctrine and Its Interpretation in Hua-Yen Buddhism.” In T’oung Pao 68, no. 4 (1982): 181-220. Lusthaus, Dan. Buddhist Phenomenology: A Philosophical Investigation of Yogācāra Buddhism and the Ch’eng Wei-shih lun. New York: RoutledgeCurzon, 2002. Mabbett, Ian. “Nāgārjuna and Deconstruction.” In Philosophy East and West 45 (1995): 203225. —. “The Problem of the Historical Nāgārjuna Revisited.” In Journal of the American Oriental Society 118, no. 3 (1998): 332-346. Moriyama, Shinya. “On Ratnākaraśānti’s Theory of Cognition with False Mental Images (Alīkākāravijñānavāda).” In Journal of the International Association of Buddhist Studies 33 (forthcoming). http://www.shinshuu.ac.jp/faculty/arts/prof/moriyama_1/docs/IABS2011Moriyama.pdf. Murti, T. R. V. The Central Philosophy of Buddhism: A Study of the Mādhyamika System. London: C. Tinling & Co., Ltd., 1955. Nāgārjuna, Root Verses of the Middle Way (Mūlamadhyamakakārikā). Edited by P.L. Vaidya. Darbhanga: Mithila Institute of Post-Graduate Studies and Research in Sanskrit Learning, 1960. McClintock, Sara. “The Role of the ‘Given’ in the Classification of Śāntarakṣita and Kamalaśīla as Svātantrika-Mādhyamikas.” In The Svātantrika-Prāsaṇgika Distinction: What Difference Does a Difference Make?, 125-171. Edited by Sara McClintock and Georges Dreyfus. Boston: Wisdom Publications, 2003. Maitreya, Distinguishing the Middle from the Extremes (Madhyāntavibhāga, dbu mtha’ rnam ‘byed). Edited by Gadjin M. Nagao. In Madhyāntavibhāga-bhāṣya: A Buddhist 254 Philosophical Treatise Edited for the First Time from a Sanskrit Manuscript. Tokyo: Suzuki Research Foundation, 1964. Edited by Ramchandra Pandeya in Madhyāntavibhāgaśāstra. Delhi: Motilal Banarsidass, 1971. Makransky, John. Buddhahood Embodied: Sources of Controversy in India and Tibet. Albany: State University of New York Press, 1997. Matilal, Bimal Krishna. Perception: An Essay on Classical Indian Theories of Knowledge. Oxford: Clarendon Press, 1986. Mackenzie, Matthew. “The Illumination of Consciousness: Approaches to Self-Awareness in the Indian and Western Traditions.” In Philosophy East and West 57, no. 1 (2007), 40-62. McCrea, Lawrence and Parimal Patil. Buddhist Philosophy of Language in India: Jñānaśrīmitra on Exclusion. New York: Columbia University Press, 2010. Mullin, Glenn. The Fourteen Dalai Lamas: A Sacred Legacy of Reincarnation. Santa Fe: Clear Light Publishers, 2000. Nadou, Jean. Buddhists of Kashmir. Delhi: Agam Kala Prakashan, 1980. Olcott, Henry Steel. The Buddhist Catechism, 44th ed. London: Theosophical Publishing Society, 1915. Padmakara Translation Group. The Adornment of the Middle Way: Shantarakshita’s Madhyamakalankara with commentary by Jamgön Mipham. Boston: Shambhala, 2005. Padmanabh S. Jaini, “The ‘Ālokā’ of Haribhadra and the ‘Sāratamā’ of Ratnākaraśānti: A Comparative Study of the Two Commentaries of the ‘Aṣṭasāhasrikā.” In Bulletin of the School of Oriental and African Studies, University of London 35, no. 2 (1972): 271-284. Park, Jin Y. (editor). Buddhisms and Deconstructions. Lanham: Rowman and Littlefield, 2006. Ram-Prasad, Chakravarthi. Indian Philosophy and the Consequences of Knowledge: Themes in Ethics, Metaphysics, and Soteriology. Burlington: Ashgate Publishing Company, 2007. Ratnākaraśānti, trans. Śāntibhadra and Śākya ‘Od, Pith Instructions for the Ornament of the Middle Way (Madhyamakālaṃkāropadeśa, dbu ma rgyan gyi man ngag). D: 4085 (sDe dge hi: vol. 138, ff. 223.b-231.a), ff. 231.a.3. Śāntarakṣita. Commentary to the Ornament of the Middle Way (Madhyamakālaṃkāravṛtti, dbU ma rgyan gyi rang ‘grel pa). sDe dge, dbu ma, vol. Sa ff. 56.b.4-84.a.1. —. Ornament of the Middle Way (Madhyamakālaṃkāra, dbu ma rgyan gyi tshig le’ur byas pa). P5284. 255 Siderits, Mark. Personal Identity and Buddhist Philosophy: Empty Persons. Burlington: Ashgate Publishing Company, 2003. Siderits, Mark, Tom Tillemans, and Arindam Chakrabarti. Apoha: Buddhist Nominalism and Human Cognition. New York: Columbia University Press, 2011. Sinha, Jadunath. Indian Realism. Delhi: Motilal Banarsidass, 1999. Streng, Frederick. Emptiness: A Study in Religious Meaning. New York: Abingdon Press, 1967. Suzuki, D. T. (translator). The Laṇkāvatāra Sūtra: A Mahāyāna Text. 1932. Reprint, Delhi: Motilal Banarsidass, 1999. Tāranātha. History of Buddhism in India (rgya gar chos ‘byung, dgos ‘dod kun ‘byung). Translated by Lama Chimpa and Alaka Chattopadhyaya as Tāranātha’s History of Buddhism in India. Simla: Indian Institute of Advanced Study, 1970. Tatz, Mark. “Maitrī-pa and Atiśa.” In Tibetan Studies: Proceedings of the 4th Seminar of the International Association for Tibetan Studies, edited by Helga Uebach and Jampa L. Panglung (Munchen: Kommission für Zentralasiatische Studien Bayerische Akademie der Wissenschaften, 1998): 473-82. Thakchoe, Sonam. The Two Truths Debate: Tsongkhapa and Gorampa on the Middle Way. Boston: Wisdom Publications, 2007. Thanissaro Bhikku (translator). “Phenna Sutta: Foam” (SN 22.95). Access to Insight, 2010: http://www.accesstoinsight.org/tipitaka/sn/sn22/sn22.095.than.html. Tillemans, Tom and Guy Newland, “An Introduction to Conventional Truth.” In Moonshadows: Conventional Truth in Buddhist Philosophy by The Cowherds. Oxford: Oxford University Press, 2011. Tola, Fernando and Carmen Dragonetti. “Dignāga’s Ālambanaparīkṣāvṛtti.” In Journal of Indian Philosophy 10 (1982): 105-134. —. “The Trisvabhāvakārikā of Vasubandhu.” In Journal of Indian Philosophy 11 (1983): 225266. Vasubandhu. Thirty Verses on Consciousness (Triṃśikāvijñaptikārikāḥ). Edited by Stefan Anacker in Seven Works of Vasubandhu. —. Twenty Verses (Viṃśatikākārikā). Edited Ram Shankar Tripathi. Varanasi: Sampurnananda Sanskrit University, 1992. 256 —. Explaining the Three Natures (Trisvabhāvanirdeśakārikā). Edited by Fernando Tola and Carmen Dragonetti in “The Trisvabhāvakārikā of Vasubandhu,” JIP 11 (1983): 225-266. Venerable K.L. Dhammajoti, “Ākāra and Direct Perception (Pratyakṣa),” Pacific World: Journal of the Institute of Buddhist Studies 3, no. 9 (2007): 245-272. Vose, Kevin. Resurrecting Candrakīrti: Disputes in the Tibetan Creation of Prāsaṇgika. Boston: Wisdom Publications, 2009. Waldron, William. The Buddhist Unconscious: The ālaya-vijñāna in the Context of Indian Buddhist Thought. London: RoutledgeCurzon, 2003. Westerhoff, Jan. Nāgārjuna’s Madhyamaka: A Philosophical Introduction. Oxford: Oxford University Press, 2009. Whitney, William Dwight. The Roots, Verb-forms, and Primary Derivatives of the Sanskrit Language: A Supplement to his Sanskrit Grammar. 1885. Reprint, Delhi: Motilal Banarsidass, 2005. Williams, Paul. Mahāyāna Buddhism: The Doctrinal Foundations. New York: Routledge, 1989. —. The Reflexive Nature of Awareness: A Tibetan Madhyamaka Defense. London: Routledge, 1997. Yao, Zhihua. The Buddhist Theory of Self-Cognition. New York: Routledge, 2005.