（ 101 ） Dge lugs versus Sa skya* Yael Bentor Introduction The Guhyasamāja Tantra（1） is one of the earlier tantras of the highest tantra class dated to the 7th or 8th century, translated into Tibetan during the Early Propagation of Buddhism in Tibet and into Chinese by Dānapāla around 1002.（2） Whereas it seems that the tantra alone was translated into Chinese, the Tibetan Bstan ’gyur includes some hundred and thirty-four commentaries about the Guhyasamāja Tantra and its practices.（3） That there is such a large body of work on the Guhyasamāja Tantra found in the canon of works translated from Indian languages demonstrates the esteem in which the Guhyasamāja Tantra was held in India. This acclaim continued in Tibet. A Tibetan version of the Guhyasamāja Tantra was found in Dunhuang cave,（4） and another has been included in the set of eighteen great Tantras of the Rnying ma School.（5） Marpa Lotsāba（1012–1097）and his contemporary ’Gos Khug pa lhas btsas（6）―two famous translators from Sanskrit―were important teachers of this tantra, as * This research was supported by The Israel Science Foundation（grant no. 401/13）. （1）See Fremantle 1971, Matsunaga 1978 and Tanemura 2015. （2）Matsunaga 1977: 109. （3）Tōh. 1783–1917. （4）India Office Library, Mss. IOL Tib J 481 and IOL Tib. J 438. （5）Martin 1987: 180. （6）Roerich, Blue Annals, 1949: 359–360. — 70 — 密 教 文 化 第二三七号 平成二十八年十二月 Divergent Perspectives on the Guhyasamāja Sādhana in Tibet: （ 102 ） practiced by members of the Bka’ brgyud and Sa skya schools during the first centuries of the second millennium. This practice was important for Chos rgyal Divergent Perspectives on the Guhyasamāja Sādhana in Tibet ’Phags pa（1235–1280）, the fifth among the early Sa skya pa masters,（7） for Bu （8） ston Rin chen grub（1290–1364） , and for Red mda’ ba Gzhon nu blo gros （9） through whom it was transmitted to Tsong kha pa Blo bzang （1348–1412） （10） grags pa（1357–1419） . The importance of Guhyasamāja in Tibet stems from its remarkable hermeneutic system,（11） which was later applied to other tantras as well. Important Tibetan scholars, including Red mda’ ba and Rong ston Shes bya （12） kun rig（1367–1449） of the Sa skya and Tsong kha pa（13） of the Dge lugs school wrote lengthy commentaries on Candrakīrti’s Pradīpoddyotana, one of the main hermeneutic works of the Guhyasamāja cycle. One of the goals of the hermeneutic system is to bridge the text of the tantra and the actual practices of this system. This brings us to the sādhana, the manual for the practice of a tantra. Guhyasamāja sādhanas: Their authentication and authority While the Guhyasamāja Tantra itself is attributed to Buddha Vajradhara, sādhana manuals were written by Indian and Tibetan yogi-scholars. That multiple sādhanas existed for the practice of each tantra tells us that there is more than one way to engage in the practice of a given tantra. This fluidity （7）See works 107–110 in the Collected Works of the Sa skya, Tokyo, volumes 6 and 7. （8）Tōh. 5078 & 5169. （9）Yid kyi mun sel and Bla ma bsgrub pa dpal bas zhus pa’i lan. （10）Tōh. 5282, 5284, 5285, 5286, 5288, 5290, 5320, 5303, 5305, 5308. （11）On the hermeneutic system of the Guhyasamāja, see Matsunaga 1963 & 1964, Steinkellner 1978, Thurman 1988, Arénes 1998 & 2002. （12）Red mda’ ba, Yid kyi mun sel and Rong ston, Gsang ’dus rnam bshad. （13）Sgron gsal mchan, Tōh. 5282. — 69 — （ 103 ） of development was also signaled by the fact that sādhanas were written not of sādhanas leads us to wonder about how to determine the authenticity and 文 lama justify to his more educated followers his sādhana over that of Dge lugs’, 教 authority of a given sādhana manual. How, we may wonder, would a Sa skya 密 only by Indian gurus but also by Tibetan lamas. This variation in the content for example? in Tibet. In India as well there were several sādhanas for the Guhyasamāja Tantra, as we know from their Tibetan translations found in the Bstan ’gyur. We also know that the differences among them were acknowledged by the tradition itself since, at least in the Tibetan canon, the sādhanas, as well as other treatises on the Guhyasamāja Tantra, are arranged according to different schools of the Guhyasamāja Tantra. Though the total number of Indian schools of the Guhyasamāja is uncertain, two of the more famous among them are the Jñānapāda and Ārya Schools.（14） Earlier, we noted that the hermeneutic system seeks to bridge the text of the tantra and the practices of this system as they are found in its sādhanas. Such a bridge is necessary since the Guhyasamāja Tantra, much like other tantras, does not offer clear and comprehensive instructions for its practice.（15） In fact the relation between the Guhyasamāja Tantra and its practices is far from obvious. Moreover both the Guhyasamāja Tantra and its sādhanas continued to evolve in India. Already fifty years ago the great Japanese scholar Matsunaga Yukei,（16） who also edited the Tantra, suggested that the explanatory tantras of the Guhyasamāja were composed by members of the Ārya school in order to （14）See Tanemura 2015: 328. （15）See Bentor, 2009. （16）Matsunaga 1964: 25. — 68 — 化 Yet the question of sādhana authenticity did not arise for the first time （ 104 ） provide canonical authority to their own sādhanas.（17） The gap between the tantra and its practice generates a number Divergent Perspectives on the Guhyasamāja Sādhana in Tibet of questions: First, on what ground were the early Indian sādhanas of the Guhyasamāja Tantra composed? Second, which of these numerous sādhanas is authentic? In this paper we will limit our inquiry to the Ārya school. This school carries the name of its founder, Ārya Nāgārjuna, who composed two sādhanas: the Piṇḍīkrama-sādhana and the Sūtra-melāpaka sādhana.（18） Though these two works, attributed to the same author, might be expected to prescribe one and the same practice, this is not the case. Tibetan writings offer different ways of dealing with such variance. For example, Tsong kha pa, the founder of the Dge lugs school, who aspired to create a consistent and harmonious system, mostly ignored the disparities between the two works of Nāgārjuna.（19） The Sa skya scholar Ngor chen Kun dga’ bzang po（1382–1456）, for his part, devoted a short work to what he calls ‘slight differences＇ between Nāgārjuna’s two sādhanas.（20） With the disparities between Indian sādhanas such that even sādhanas written by the same master are incompatible, it is not surprising that different traditions for the practice of the Guhyasamāja developed in Tibet. Candrakīrti’s Vajrasattva-sādhana was another important Guhyasamāja （17）More recently Tomabechi Toru and Tanaka Kimiaki offered us important studies of the Guhyasamāja Tantra tradition in India, and in my own studies I constantly refer to their impressive publications. （18）It is to be noted that different versions of their names exist. The Piṇḍīkrama-sādhana or Piṇḍīkṛta-sādhana, Mdor byas, Tōh. 1796 and the Sūtra-melāpaka sādhana, Mdo bsre, Tōh. 1797. （19）See Bentor, 2015a. （20）Shin tu rnal ’byor gyi khyad par sgrub thabs kyi yan lag tu bris pa, folio 204b, p. 102.1-2, appended to his sādhana of the Guhyasamāja, Dpal gsang ba ’dus pa’i dkyil ’khor gyi sgrub pa’i thabs dngos grub rgya mtsho. This work was written in 1423, after the death of both Red mda’ ba in 1412 and Tsong kha pa in 1419. — 67 — （ 105 ） Sādhana of the Ārya tradition [edited by Luo Hong and Tomabechi].（21） Once grub,（23） accept the canonical authority of this sādhana, while most Sa skya 密 more we see a disagreement. Tsong kha pa,（22） as well as Bu ston Rin chen school authorities do not.（24） Those in the latter group maintain that the author 教 of this sādhana is not the renowned Candrakīrti, but another person of the same 文 name. sādhana do so on the basis of its incompatibility with Nāgārjuna’s Sūtramelāpaka. This returns us to the issue of the relation between the Root Tantra and its Sādhanas. The Sūtra-melāpaka, as its name suggests, relates each step of the practice to lines found in the Guhyasamāja Tantra. It divides the sādhana into twenty-four different acts, and cites lines from the Guhyasamāja Tantra as the scriptural authority for each of these steps. In other words, the Sūtra-melāpaka links the Guhyasamāja Tantra and its practice. Candrakīrti’s Vajrasattva-sādhana not only offers instruction on the practice but also points out specific lines of the Tantra that serve as scriptural authority for each step. The Sa skya scholar A myes zhabs（25） finds many incompatibilities between the Vajrasattva-sādhana and the Sūtra-melāpaka concerning scriptural attribution for the meditations in the Guhyasamāja Tantra. （21）Rdo rje sems dpa’i sgrub thabs, Tōh. 1814. （22）Kilty 2013: 66. （23）Bu ston cites Candrakīrti’s Vajrasattva-sādhana throughout his Mdor byas ’grel chen, Tōh. 5078. （24）Among them are Red mda’ ba, Bla ma bsgrub pa dpal bas zhus pa’i lan, TBRC 306b, Kathmandu, folio 51a, p. 281.4; Ngor chen Kun dga’ bzang po, Shin tu rnal ’byor gyi khyad par sgrub thabs kyi yan lag tu bris pa, folio 206a, p. 102.4.3; and A myes zhabs, Gsang ’dus chos ’byung, folio 18a, p. 35.2-4. （25）’Jam mgon A myes zhabs Ngag dbang kun dga’ bsod nams, Gsang ’dus chos byung, folio 18a, p. 35.2-5. — 66 — 化 Tibetan scholars who reject the authority of Candrakīrti’s Vajrasattva- （ 106 ） According to Red mda’ ba, another member of the Sa skya school:（26）“While Divergent Perspectives on the Guhyasamāja Sādhana in Tibet Ārya Nāgārjuna himself wrote a very clear sādhana such as the Piṇḍīkramasādhana, how is it possible that Glorious Candrakīrti did nothing other than putting it into prose? ” Red mda’ ba is here referring to the fact that whereas the Piṇḍīkrama-sādhana is written in verses conducive to recitation, the Vajrasattva-sādhana prescriptions are written in prose. Tsong kha pa,（27） for his part, accepts the authority of Candrakīrti’s Vajrasattva-sādhana. Its Indian provenance, he claims, is demonstrated by the fact that two well-known Indian masters, Tathāgatarakṣita and Līlāvajra,（28） wrote commentaries on it. Moreover, Tsong kha pa points out, Muniśrībhadra,（29） in his Pañca-krama-ṭippaṇī, edited by Zhongxin Jiang and Tomabechi,（30） acknowledged its authorship.（31） Further research is required to determine whether the rejection of （26）Bla ma bsgrub pa dpal bas Gsang ba ’dus pa’i bsgrub thabs mdor byas dang Bsgrub thabs rnam gzhag gi rim pa rnam gnyis kyi mi ’dra ba’i khyad par zhus pa’i lan, TBRC folio 306b. （27）Mtha’ gcod, folio 36a, p. 134.4-5. （28）Tōh. 1835 and 1815. （29）Tōh. 1813, D. folio 152b, p. 304.3, and Jiang and Tomabechi 1996: 11. nd （30）As Toru Tomabechi pointed out to me on April 22 2016 at Taisho University, in his commentary on the Pradīpoddyotana, Bhavyakīrti explains that the sādhana referred to in the line sādhanopayikāyāṃ kṛta eva, sgrub pa’i thabs su byas pa nyid（Tōh. 1785, D. folio 21b, Chakravarti 1984: 37）could be either the Piṇḍīkrama or Candrakīrti’s own work（presumably the Vajrasattva-sādhana）, Tōh. 1793, D. folio 139a3-4. Hence there is an indication by another Indian master that Candrakīrti could be the author of the Vajrasattva-sādhana. （31）Further supporting this line of thought, Tsong kha pa’s disciple Mkhas grub rje Dge legs dpal bzang（1385–1438）objects to the position of Red mda’ ba that nothing of value can be found in the Vajrasattva-sādhana that is not already present in Ārya Nāgārjuna’s Sādhana. According to Mkhas grub rje the Vajrasattva-sādhana also presents ways to relate the steps of the meditation to the scriptures that are not found in the Sūtramelāpaka. Bskyed rim dngos grub rgya mtsho, folio 14a, p. 29.2-5. — 65 — （ 107 ） the validity of Candrakīrti’s Vajrasattva-sādhana by Sa skya scholars together contributed to the predominance of the Jñānapāda school of the Guhyasamāja devoted much attention to the Ārya school.（34） Still from these discussions we meditation by identifying the lines in the Guhyasamāja Tantra that are their true scriptural authority. The hermeneutic system of the Guhyasamāja As mentioned, the Guhyasamāja literature developed an especially sophisticated hermeneutic system to link the root text of the tantra to the practice. The Candrakīrti who composed the most important commentary on the Guhyasamāja Tantra, entitled Pradīpoddyotana, formulated a set of rules for interpretation known as the seven ornaments.（35） The word‘rule’implies limits, in this case limits on opportunities to explain the tantra. Yet the very possibility of textual interpretation and the linking of the tantra with its practices is of utmost importance. The hermeneutic tradition allows authors of the sādhanas and commentaries a great deal of interpretive freedom with regard to identifying （32）Go rams pa Bsod nams seng ge, Gsang ’dus sgrub thabs [kun bzang] don ’grel. （33）Stag tshang Lo tsā ba Shes rab rin chen, ’Jam dpal rdo rje’i mngon par rtogs pa ’dod dgu rgya mtsho. （34）’Jam mgon A myes zhabs Ngag dbang kun dga’ bsod nams, Gsang ’dus rnam bshad. （35）Skt. saptālaṃkāra, Tib. rgyan bdun. See Matsunaga 1963 & 1964, Steinkellner 1978, Thurman 1988, Arénes 1998 & 2002. Similar tantric hermeneutical methods are found also in the Vajrajñānasamuccayatantra, Tōh. 447, one of the explanatory tantras of the Guhyasamājatantra（D. folio 285b2-3）. Matsunaga（1964）maintains that the second part of the Vajrajñānasamuccayatantra, which lists these seven methods of interpretation with all their subdivisions, was written after Candrakīrti’s Pradīpoddyotana. — 64 — 化 can conclude that one recurrent theme is how to authenticate the steps of the 文 during the 15th century, though no doubt in the 17th century A myes zhabs 教 among Sa skya scholars such as Go rams pa（32） and Stag tshang Lo tsā ba（33） 密 with their awareness of discrepancies between the two Sādhanas of Nāgārjuna （ 108 ） the scriptural authorities for the various stages of the meditation in the Divergent Perspectives on the Guhyasamāja Sādhana in Tibet Guhyasamāja Tantra. Furthermore, the tantric hermeneutics provide them with the methods for doing so. Thus a diversity of explanations in both Indian and Tibetan traditions about the‘right’way to practice is far from surprising. Shortly, we will examine the kinds of choices Tibetan scholars make with their interpretive freedom and on what grounds they make their decisions. First, however, we will take a moment to consider another Indian source for the Guhyasamāja sādhana. This is the Samāja-sādhana-vyavastholi. Macro-cosmic and micro-cosmic correlations of the Guhyasamāja sādhanas The Samāja-sādhana-vyavastholi,（36） was written by Nāgabuddhi, a tantric disciple of Nāgārjuna [edited by Tanaka Kimiaki]. By and large, Tibetan authors of works on the practice of the Guhyasamāja according to the Ārya tradition accept the authenticity of the Vyavastholi. Their interpretations on this work, however, differ. Like the three sādhanas already mentioned, the Vyavastholi, too, portrays parallels to steps of the meditation. This time, however, the parallels are not to the text of the Guhyasamāja Tantra, but to events that occur periodically in both the cosmos and the individual. In other words, the creations and destructions of the world as well as birth, death and the intermediate state during the life of sentient beings are referenced. These macro- and micro-cosmic parallels imbue the steps of the sādhana with new meanings in relation to the cosmic whole. Still in this case as well the parallels are not as straightforward as we would wish and thus open the door to different readings. For this reason, once more, Tibetan masters differ in their understanding of （36）Klu’i blo, Samāja-sādhana-vyavastholi, ’Dus pa’i sgrub pa'i thabs rnam par gzhag pa’i rim pa, Tōh. 1809. — 63 — （ 109 ） the new meanings the individual steps of the sādhana acquire by being related relates the destruction of the universe by the seven suns that burn the entire 密 to specific macro- and micro-cosmic events. For example, the Vyavastholi（37） three realms to the meditation on emptiness at the beginning of the main part of 教 the Piṇḍīkrama-sādhana.（38） And then it correlates the cyclical creation of the 文 universe to the visualization of the celestial mansion of the maṇḍala. How are The workings of the Guhyasamāja sādhanas—macrocosmic aspects The 11th century Tibetan scholar and translator ’Gos Khug pa lhas btsas（39） uses the terms“ground of purification”and“its purifier.” By meditating on emptiness, the meditators purify the destruction of the world. Likewise, by visualizing the celestial mansion of the maṇḍala they purify the creation of the world. The Sa skya scholar Red mda’ ba Gzhon nu blo gros closely follows the explanations of ’Gos Khug pa lhas btsas in this matter.（40） The assumption here is that by means of meditating on the sādhana, the meditators destroy their ordinary environment in order to create the pure world of the celestial mansion of the maṇḍala, where they will be enlightened. Hence in a type of ritual death, by meditating on emptiness, they dissolve their world, including themselves, so that they will be able to arise as deities in the pure celestial mansion of the maṇḍala. For this reason the terminology of ground of purification and purifiers is used—by means of the sādhana, the meditators purify their impure reality, the ground of purification. What are the implications of the position that by meditating on the （37）Tōh. 1809, folio 121b, p. 242.4-5. （38）Tōh. 1796, vv. 16cd–18, folio 2b, p. 4.3-5. （39）Gsang ’dus stong thun, folios 7a–14a, pp. 13.2–27.4. （40）Dpal gsang ba ’dus pa’i ’grel pa sgron ma gsal ba dang bcas pa’i bshad sbyar yid kyi mun sel, pp. 232–236. — 62 — 化 we to understand these correspondences? （ 110 ） sādhana, the yogis purify cosmological events? The idea that the mind can Divergent Perspectives on the Guhyasamāja Sādhana in Tibet alter the world of the individual, found in various Buddhist scriptures, lies at the core of different theories on the workings of the sādhana. How else can we understand that the sādhana can purify the meditator’s ordinary existence? Such a notion also seems to be at the basis of the relations the Vyavastholi draws between the sādhana and cosmic events. We will engage with theories of mind-only, citta-mātra, below, but here my purpose is to consider more generally individual and shared karma. Mkhas grub rje accepts the idea that yogis are capable of dissolving their own world or environment. Yet he maintains that they cannot affect the environment of others. Therefore he objects to the position that by means of the meditation, the impure world is purified:（41） What is the meaning of purifying the impure world by meditating on the celestial mansion? It is not that through your meditation on the celestial mansion, you transform this present impure world into a pure celestial mansion, but that you purify your own capacity to partake in the impure world in the future. Thus Mkhas grub rje asserts that, in fact, meditating on the sādhana does not enable yogis bring about a true transformation in their outer world. Their accomplishments can affect only themselves, for no matter how much they meditate, the world will remain impure. However, they can purify themselves and transform their own abilities. Hence, by meditating on the celestial mansion of the maṇḍala here, yogis can transform the way they will partake in the impure world in the future. （41）Mkhas grub rje Dge legs dpal bzang po, Rgyud thams cad kyi rgyal po dpal gsang ba ’dus pa’i bskyed rim dngos grub rgya mtsho, folio 50a, p. 101.1-2. — 61 — （ 111 ） Meditation on emptiness in tantric context 密 A further difficulty that arises from linking the destruction of the world with meditating on emptiness concerns the correlation between nothing 文 that burn the entire three realms, turn them into the nature of space.（43） The 教 whatsoever and emptiness.（42） According to the Vyavastholi, the seven suns Abhidharmakośa—on which the Vyavastholi relies—also explains that through top of Mt. Meru.（44） Emptiness here then is correlated to the empty eon between the previous and the later worlds in the cosmological cycle. Meditating on sheer nothingness, rejected by the Mādhyamika school, is highly in line with the sādhana. This is so because the liminal phase, wherein meditators‘visualize away’ ordinary reality, allowing a pure enlightened reality to emerge is at the foundation of the sādhana.（45） We might then ask: Is the emptiness meditated upon in the tantric sādhana different from that of the Mādhyamika school? The apparent incongruity between these two systems gave rise to numerous disputes. If yogis meditate at this point of the sādhana on emptiness as defined by the Mādhyamika school, their meditation is on the ordinary world as devoid of intrinsic nature [svabhāva]. In doing so, they would not totally destroy their present world in order to create the pure world of enlightenment. If on the other hand, the yogis completely eliminate ordinary reality, as the logic of the sādhana requires, they will be meditating on nihilistic emptiness. It is in this context that Tsong kha pa states:（46） （42）See Bentor 2015b. （43）Tib. nam mkha’i rang bzhin, Skt. ākāśa-maya. Tōh. 1809, folio 121b, p. 242.4-5. （44）Kośa ch. 3, v. 90, and its auto-commentary. The Kośa continues with a discussion of how much of the Form Realm is also burnt up. The destructions through water and wind extend even higher. （45）And the scriptural authority of the Vyavastholi instructs the meditators to do so. （46）Tsong kha pa, Bskyed rim zin bris, folio 15b, p. 181.1-2. — 60 — 化 destruction by fire, the world is destroyed and is entirely consumed up to the （ 112 ） The Perfection Vehicle does not specify that while meditating on emptiDivergent Perspectives on the Guhyasamāja Sādhana in Tibet ness one must definitely dissolve appearances, but the Mantra Vehicle specifies that one must definitely do so.（47） At the same time Tsong kha pa emphasizes that the meditation on emptiness during the sādhana must accord with the Madhyamaka view:（48） While meditating on the meaning of the mantra śūnyatā, one should recollect the view of Madhyamaka, and visualize all phenomena in the world and its inhabitants dissolving into clear light. Both are necessary, for it is meaningless to visualize the latter without meditating on the former. Likewise the Sa skya scholar Go rams pa Bsod nams seng ge（1429–1489） , ridicules the position of the opponents who remove only the true existence of appearances in their ordinary reality, but not the appearances themselves. He specifies （47）Tsong kha pa explains that dissolving appearances does not mean actually annihilating them, but preventing them from appearing in the yogi’s mind, ibid., folio 15a, p. 180.6. Tsong kha pa stresses that these are not appearances that appear to the sense consciousness, but appearances that appear to the mental consciousness as the ordinary world and its inhabitants, Sngags rim chen mo, pp. 463–464. This is just as śamatha meditation is a faculty of the mental consciousness, not sense consciousness. According to Tsong kha pa by meditating on emptiness and on the pure environment, the consciousness of the eye and the other senses do not arise for a while, but meditation on the sādhana cannot stop appearances completely. Tsong kha pa does state, however: “It is absurd to suppose that merely by applying one’s mind to their cessation, all appearances become imperceptible and thereby one abides in absorption in the three doors of liberation.” See Rnam gzhag rim pa’i rnam bshad, Tōh. 5290, folio 9a, p. 296.4. （48）Tsong kha pa, ’Dod ’jo, Tōh. 5320, folio 60a, p. 190.3-4. This is a commentary on the sādhana of Bde mchog by Lūyīpā, the Dpal bcom ldan ’das mngon par rtogs pa, Bhagavad-abhisamaya, Tōh. 1427. — 59 — （ 113 ） that if aspects of ordinary reality are not sufficiently eliminated, they might rethe maṇḍala:（49） 密 appear when after meditating on emptiness, the yogis visualize the pure realm of 教 maṇḍala wheel of the celestial mansion and its deities, is this ‘small spread above the seat of the lotus, the lunar disk and so on, or is it below it? Do the clothes you wear cover the many faces and arms you meditate on or tightly enfold them? Please think this over carefully and then tell us. Other Sa skya scholars, such as ’Jam mgon A myes zhabs [in 1636], emphasize that if ordinary appearances are not removed during the meditation on emptiness, then in arising from within the continuum of emptiness, the appearance of the deity will not dawn.（50） At the same time, Go rams pa as well stresses that yogis ought to meditate on emptiness during the sādhana using Madhyamaka reasoning, such as that phenomena are devoid of ‘one and many’.（51） Thus, both Tsong kha pa and Go rams pa tread a subtle line between the two ways to meditate on emptiness in this context, in fact instructing yogis to meditate on both. We could go on in this vein for some time. Tibetan lamas certainly had more to say about the implications of the notion that by meditating on the sādhana, yogis purify cosmological events. But we will turn now to parallels the Vyavastholi makes between steps of the meditation and events in the lives of sen（49）Gsang ’dus sgrub thabs don ’grel, folio 19a, p. 10.1.3-5. （50）See Gsang ’dus rnam bshad, folio 63a, p. 317.5-6. （51）Gsang ’dus sgrub thabs don ’grel, folio 24a-b, p. 12.3.5-4.6. — 58 — 化 room’ outside the celestial mansion or inside it? Is the cushion you have 文 When you are a yogi sitting inside a small room meditating on a huge （ 114 ） tient beings, such as birth and death. In this context the ambiguities in the text of the Vyavastholi are especially notable. Divergent Perspectives on the Guhyasamāja Sādhana in Tibet The workings of the Guhyasamāja sādhanas—microcosmic aspects Following its description of the empty eon and the evolution of the subsequent world that correspond to the meditation on the celestial mansion,（52） the Vyavastholi describes the evolution of the beings in the world and the four modes of possible birth, known also in the Vajracchedikā Prajñāpāramitā:（53） Birth from egg, birth from womb, birth from heat plus moisture, and miraculous birth. Tibetan lamas of all schools delineated four kinds of meditation for purifying these four modes of birth.（54） These authors include Grags pa rgyal mtshan（1147–1216）, Gser sdings pa（born in the 12th c.）, Bu ston Rin chen grub , Rong ston Shes bya kun rig（1367–1449） , ’Ba’ ra ba Rgyal mt（1290–1364） shan dpal bzang（1310–1391）, ’Jigs med gling pa（1729/1730–1798）, the 3rd Kar ma pa Rang byung rdo rje（1284–1339）, ’Jam dbyangs Mkhyen brtse dbang po （1820–1892）, and Kong sprul Blo gros mtha’ yas（1813–1899/1890）. There are no two identical descriptions among them. Tsong kha pa,（55） for his part, objects to each and every system that identifies specific steps in the sādhana as purifiers of the four modes of birth. He does this because the path of the unexcelled mantra is intended for human beings （52）Tōh. 1809, D. folio 122a, p. 243.2-4. （53）Harrison and Watanabe 2006: 114. （54）See Bentor, 2006: 193–194. To give an example, Grags pa rgyal mtshan maintains that instantaneous visualization of the deity purifies miraculous birth; generation through the five abhisaṃbodhis—that include solar and lunar disks—purifies birth from warmth and moisture, since the sun warms and the moon moistens; the visualization of the fruitional deity from a heap of drops purifies birth from egg; and sending forth the deities from the womb of the Mother purifies birth from womb. See Rgyud kyi mngon par rtogs pa, folio 69a, p. 35.1.3-5. （55）In his Rnam gzhag rim pa'i rnam bshad, Tōh. 5290, folios 13b–20a, pp. 305.1–318.4. — 57 — （ 115 ） endowed with characteristics essential for the practice of the sādhana, who will to Tsong kha pa, if from the very beginning of their practice, the yogis engage in world, who were endowed with wondrous qualities, turned into ordinary people and died. We know this from the Abhidharmakośa as well.（58）According to ’Gos Khug pa lhas btsas,（59）the meditation on gathering the first deities in the maṇḍala into clear light purifies both cosmological and individual events. On the cosmological level, it purifies the death of the first beings in the world, while on the individual level it purifies the clear light of the death of the yogi’s previous life. Tsong kha pa disagrees with the position that by meditating on the sādhana at present, the yogis can have a retroactive effect on events during the first kalpa or during previous lives. He says: “This is pointless since it is similar to the position that the fire that burns this year destroys the fire-wood of last year.”（60） Rather, for Tsong kha pa, by meditating on the sādhana in their present lives, yogis can induce a transformative effect on events that will take place in the future, and not those that already occurred in the past. Moreover, the practice can affect （56）In his Rnam gzhag rim pa’i rnam bshad, ibid., Tsong kha pa employs not only reason, but the scriptural authority of the Vyavastholi as well. Immediately after describing the four modes of birth, the Vyavastholi describes how in his final birth in the saṃsāric world, the Buddha was born in Jambūdvīpa as a human being, and as such attained awakening, and then concludes（folio 123a, p. 245.1）: “For this reason, the stages of birth into human existence are taught here.” （57）Tōh. 1809, D. folio 123a, p. 245.1-4. （58）Kośa, chapter 3, v. 90c-d and its auto-commentary. （59）Gsang ’dus stong thun, folio 10b, p. 20.1. （60）Tōh. 5290, Rnam gzhag rim pa’i rnam bshad, folio 45b, p. 369.5-6. — 56 — 化 The Vyavastholi（57） continues to set forth how the first beings in the 文 Why would such yogis need to purify their future birth from egg?（56） 教 the path of the unexcelled mantra, they can be enlightened within one lifetime. 密 attain awakening within several lifetimes as human beings. Moreover, according （ 116 ） the life cycle events of the yogis themselves, but not of others. Divergent Perspectives on the Guhyasamāja Sādhana in Tibet Further ramifications of differences among the Indian basic Guhyasamāja sādhanas Relying on the Vyavastholi, Tibetan scholars mostly agree that the Guhyasamāja sādhana serves the purification of the yogis’ birth, death and the intermediate state. However one finds much variance with regard to which step of the sādhana affects which of these events. The difficulty arises because of the differences in the practice between the three basic works of the Ārya tradition generally regarded as authoritative by Tibetan scholars: Nāgabuddhi’s Vyavastholi and the two works by Nāgārjuna. These sādhanas diverge, for instance, with regard to number of deities visualized. During Nāgārjuna’s Piṇḍīkrama-sādhana,（61） yogis visualize themselves as two deities. The first of these is the so-called First Lord,（62）and later this deity transforms into Vajrasattva. In the second sādhana by Nāgārjuna, the Sūtramelāpaka, from the start the yogi arises as Vajradhara. However we are told that since at first the nature of Vajradhara’s body is wisdom and as such he cannot act for the sake of sentient beings, later on during the sādhana he transforms into a visible body. The Vyavastholi, by contrast, offers no hint of two deities or of two aspects of the same deity. Returning to the question of steps of the sādhana that are purifiers and the identification of their grounds of purification, Tibetan scholars agree that meditation on emptiness prior to the visualization of the deity serves the purification of death. They disagree however about the steps involved in purifying the intermediate state and rebirth. For example, the Sa skya scholar Red mda’ ba maintains that meditation on the single deity Vajradhara, as in both the Vyavastholi and the （61）Tōh. 1796, D. folios 3b–4a, pp. 6.7–7.3, vv. 51–54. （62）Skt. ādinātha, Tib. dang po mgon po. — 55 — （ 117 ） Sūtra-melāpaka, serves to purify both the intermediate being and rebirth.（63） Bu in the Piṇḍīkrama-sādhana facilitates the purification of the intermediate being, namely, death, the intermediate state and rebirth, and the three major parts of the 化 sādhana. 文 Bu ston creates an elegant correlation between the three grounds of purification, 教 while the meditation on Vajrasattva serves the purification of rebirth. In this way 密 ston Rin chen grub（64）differently explains that the meditation on the First Lord as The fourth sādhana of the Ārya tradition, Candrakīrti’s Vajrasattvasādhana, [whose authenticity is not accepted by members of the Sa skya school such as Red mda’ ba] has no instructions on the First Lord, yet it divides the meditation into three distinct parts and links them respectively to the attainment of the three bodies of the Buddha. In other words, by meditating on emptiness, yogis attain the Dharmakāya,（65）by visualizing the principal deity, they attain the Sambhogakāya, and by placing the deities on their bodies and so forth, they attain the Nirmāṇakāya.（66） Bu ston,（67） who accepts the authenticity of Candrakīrti’s Vajrasattvasādhana, links the three parts of the sādhana not only to death, the intermediate state and rebirth, but also to the attainment of the three bodies of the Buddha. Cherishing coherent and comprehensive systems, Tsong kha pa（68） follows Bu ston. Tsong kha pa explains that the meditation on emptiness serves to purify （63）Yid kyi mun sel, pp. 234–235. Red mda’ ba follows here ’Gos Khug pa lhas btsas Gsang ’dus stong thun, folio 12b, p. 24.3-5. （64）Mdor byas ’grel chen, Tōh. 5078, folios 36a–37a, pp. 753.7–755.2 and folios 38b–39b, pp. 758.3–760.4. （65）Vajrasattva-sādhana, Tōh. 1814, D. folio 199a, p. 397.2-3, Hong and Tomabechi 2009: 14.8 & 47.4-5. （66）Vajrasattva-sādhana, Tōh. 1814, D. folio 199b, p. 398.2-3, Hong and Tomabechi 2009: 16.8 & 49.5-7. （67）Mdor byas ’grel chen, Tōh. 5078, folio 51a-b pp. 783.4–784.2. （68）Bung ba’i re skong, folios 15b–21b, pp. 367.2–379.5. — 54 — （ 118 ） the yogi’s future death, and its fruit is the Dharmakāya. The visualization of the Divergent Perspectives on the Guhyasamāja Sādhana in Tibet First Lord serves to purify the yogi’s future intermediate state, and its fruit is the Saṃbhogakāya. The transformation into Vajrasattva serves to purify the yogi’s future rebirth, and its fruit is the Nirmāṇakāya. This system influenced not only the Dge lugs school, but also the Sa skya tradition to which Red mda’ ba adhered. For example, the Sa skya scholar （69） who often followed Red mda’ ba, Rong ston Shes bya kun rig（1367–1449） , in this context rejects the position of Red mda’ ba on the very grounds offered by Bu ston. Red mda’ ba seems to have been familiar with the views of Bu ston.（70） Nonetheless, he chose to follow ’Gos Khug pa lhas btsas rather than Bu ston in this matter, while Tsong kha pa conformed to the position of Bu ston. Interestingly, although these Tibetan lamas belong to the same lineage of teachings on the Guhyasamāja Tantra Sādhana, and some were direct disciples of the others, they did not always make the same choices. Philosophical schools and the theoretical basis of the Guhyasamāja sādhana As already pointed out the philosophy of the Mind Only school can provide an excellent theoretical basis for the working of tantric sādhanas in which mental constructions such as visualizations play a central role. Indeed the term mind-only （cittamātra, sems tsam）is found in quite a few important Indian works of the Ārya school of the Guhyasamāja. For example, the commentaries on Nāgarjuna’s Piṇḍīkrama-sādhana by Muniśrībhadra（71） and Bhavyakīrti（72） instruct yogis to （69）Gsang ’dus rnam bshad, folios 6a–7a, pp. 11.3–13.4. （70）Red mda’ ba studied the Guhyasamāja according to the Ārya tradition with Bu ston, though Red mda’ ba was only sixteen years old when Bu ston passed away. See Roloff, 2009: 99 & 213. It was through Red mda’ ba that Tsong kha pa received teachings on Bu ston’s commentary on the Piṇḍīkrama-sādhana. See Tsong kha pa, Gsan yig, folio 5a, p. 241.1-2, and Mtha’ gcod, folio 40b, p. 143.4. （71）Muniśrībhadra [Thub pa dpal] in his Pañca-kramārtha-ṭippaṇī, instructs the yogi to — 53 — （ 119 ） meditate on the animate and inanimate worlds in their entirety as mind-only. The tion.（73） avoids the term mind-only, but still instructs yogis to meditate on everything as cept of mind-only in the Ārya tradition of the Guhyasamāja. In his view all works that belong to the Ārya Nāgārjuna tradition of the Guhyasamāja were composed by Nāgārjuna, the Madhyamaka philosopher. Hence philosophies such as mindonly could not form a part of their core. Without a doubt, Tsong kha pa encountered the notion of mind-only in the commentaries on the Piṇḍīkrama-sādhana, which he cites in his own works. Yet he overlooked these ideas for the sake of philosophical harmony of both Sūtra and Tantra or both Pāramitā and Vajra Vehicles. Though Tsong kha pa provided a brilliant resolution to this issue,（76） he meditate on the full moon orb as mind-only in the nature of very radiant wisdom light, Tōh. 1813, D. folio 159a, p. 311.4-5, Jiang and Tomabechi 1996: 17.1. （72）Bhavyakīrti in his Pañca-krama-pañjikā instructs to meditate on the moon of mindonly, Tōh. 1838, folio 2b, D. p. 4.2-3. （73）Tōh. 1814, D. folio 199a, p. 397.5, Hong and Tomabechi 2009: 15.7-8 & 47.17. （74）Mdor byas ’grel chen, Tōh. 5078, folio 34a-b, pp. 749.4–750.1. （75）Gsang ’dus dngos grub rgya mtsho, folio 9b, p. 96.1.3-4. （76）Tsong kha pa glosses mind-only, sems tsam, as rlung sems tsam, as mere “wind-andmind” [Sādhana, folio 40a, p. 95.2]. For him the root of all phenomena here is not mindonly, but mere wind-and-mind. He applies this notion to the ground of purification of the meditation—the intermediate being, who has left behind the coarse physical body of its former life, and now is made of mere wind-and-mind. In this ingenious move, Tsong kha pa removes the philosophical school of mind-only from this context while building a system that works for the purifiers and their grounds of purifications of the tantric sādhana as well as for its Madhyamaka theoretical basis. In this way the comprehensive system Tsong kha pa designed was not limited to the Vajrayāna, but encompassed both Sūtra and Tantra. — 52 — 化 their own minds. Tsong kha pa however is troubled by the occurrence of the con- 文 produces the meditation in terms of mind-only. Ngor chen Kun dga’ bzang po（75） 教 In his commentary on the Piṇḍīkrama-sādhana, Bu ston（74） simply re- 密 Candrakīrti who composed the Vajrasattva-sādhana provides the same instruc- （ 120 ） diverged decisively from Indian works as well as from Bu ston, who as we saw, Divergent Perspectives on the Guhyasamāja Sādhana in Tibet made no attempt to present a unified philosophical approach to the Tantra and the Mādhyamika. We might mention a similar case of interpreting a practice along different theoretical positions.（77）In this instance a meditation on the body maṇḍala was interpreted on the basis of the theory of tathāgata-garbha. According to the Indian Siddha Vajraghaṇṭa:（78） “Sentient beings are naturally present non-dual maṇḍalas.” On this basis the early Sakya scholar, Bsod nams rtse mo（1142– （79） , as well as Bu ston（80）maintained that maṇḍalas are present naturally in 1182） yogis’ bodies. In this case, not only Tsong kha pa,（81）but the later Sa skya scholar Ngor chen Kun dga’ bzang po too,（82）rejected the more literal interpretation of the tathāgata-garbha theory. They were philosophically unable to accept Bsod nams rtse mo’s and Bu ston’s straightforward understanding of Vajraghaṇṭa’s line that there are maṇḍalas naturally present in sentient beings. Instead, they were obliged to resort to a non-literal and quite complicated reading.（83）Clearly for these lamas it was more important to adhere to the ‘proper’ view on tathāgata-garbha than （77）Though the dispute among Tibetan scholars on this matter originated in the context of the cycle of Cakrasaṃvara, it is primarily known as a consequence of its occurrence in Mkhas grub rje’s work on Guhyasamāja sādhana, Bskyed rim dngos grub rgya mtsho, folios 116b–118b, pp. 234.1–238.1. See Bentor 2017 and forthcoming. （78）Vajraghaṇṭa, Tōh. 1431, D. folio 219b, p. 438.5-6. （79）De’i dbang gi bya ba mdor bsdus, folio 118a1, p. 407.4.1-4. （80）Lus dkyil dbang chog, folio 1b, p. 386.3-5. （81）’Dod ’jo, Tōh. 5320, folio 122a-b, pp. 440.6–441.1. （82）Dril bu pa’i lus dkyil gyi bshad pa, folios 370a1–371a1. （83）They take this line to mean that the bases—out of which the body maṇḍalas are achieved—are naturally present in the bodies of sentient beings. For the body maṇḍala is achieved or generated from the channels and elements of the body that abide there naturally from the very moment one’s body was formed. On the other hand, such bases are not found in painted or colored powder maṇḍalas. Certainly an explanation that demands some pondering. — 51 — （ 121 ） to provide a literal reading of Vajraghaṇṭa. Moreover, for them adherence to the earlier authorities in their respective traditions. 密 ‘proper’ view on tathāgata-garbha trumped conformity with the perspectives of Let’s take one final related case. We can identify two distinct ap- 教 proaches to Vajraghaṇṭa’s instruction to meditate in a way not found in any simi- 文 lar sādhana.（84） Ngor chen Kun dga’ bzang po, for instance, follows the master tack, holds systematic practice to be of cardinal importance. Relying on reason and analogous practices, he privileges consistency of practice over the specific instructions of previous lineage masters. Conclusions: We have learned that Indian masters offered different instructions on the practice of Guhyasamāja Tantra. Tibetan masters, for their part, introduced even greater diversity. Such variation stemmed from the fact that the Root Tantra is not an instruction manual. This absence of specification in the Root Tantra did more than create difficulties, however. It also enabled the authors of the sādhanas to enjoy a certain degree of interpretive freedom. Tibetan scholars, attempting to establish the authenticity of their own sādhanas, were further challenged by the internal contradictions present in the two sādhanas composed by Nāgārjuna. The new meanings the sādhana acquired by being associated with specific macro- and micro-cosmic events in Nāgabuddhi’s authoritative Vyavastholi caused their own complications. And when Tibetan scholars tried to integrate the theoretical basis of the sādhana with notions of Buddhist philosophical schools, yet more discord was sown. The rationale of the practice did not always conform to foundational Buddhist thought, leading to controversy concerning meditating on nothingness （84）Cakrasaṃvara-sādhana, Dpal ’khor lo sdom pa’i sgrub pa’i thabs. Tōh. 1432. See Bentor, 2017. — 50 — 化 Vajraghaṇṭa, despite his instructional incongruity. Tsong kha pa, taking a different （ 122 ） vis-à-vis meditation on emptiness. Moreover, Tibetan scholars diverge on BudDivergent Perspectives on the Guhyasamāja Sādhana in Tibet dhist philosophy concerning mind-only and tathāgata-garbha. While some of these commentators accepted a different theoretical basis for Tantra Vehicle, others, such as Tsong kha pa, insisted on a single philosophical ground for both Sūtra and Tantra Vehicles. When the choices of a particular Tibetan lama contradicted those of his predecessors in the same transmission lineage of the Guhyasamāja, things got even more interesting. Overall, we can characterize the Tibetan tradition of the Guhyasamāja Tantra as dynamic, lively and abounding with vigorous debate. Bibliography Kangyur: Guhyasamāja-Tantra or Sarva-tathāgata-kāya-vāk-citta-rahasya-guhyasamāja-nāmamahā-kalpa-rāja, Gsang ba ’dus pa or De bzhin gshegs pa thams cad kyi sku gsung thugs kyi gsang chen gsang ba ’dus pa zhes bya ba brtag pa’i rgyal po chen po, Dunhuang, IOL （India Office Library）Tib J 481 and IOL Tib. J 438; The Rnying ma rgyud ’bum（Thimbu: Dingo Khyentse Rimpoche, 1973）, vol. 17, folios 1b1–314a4; Stog Palace, vol. 96 [ca], folios 1b–82a, pp. 2.1–163.5; Tōh. 442, Derge vol. 81 [ca], folios 90a–148a, pp. 181.1– 295.6; Peking Bka’ ’gyur, Ōtani 81, vol. 65 [ca], folios 95b–167b, pp. 174.3.5–203.2.1; for Sanskrit editions see Fremantle 1971 and Matsunaga, 1978; for an English translation, see Fremantle 1971. Tengyur: Bhavyakīrti, Skal ldan grags pa, Pradīpodyotanābhisaṃdhi-prakāśikā-nāma-vyākhyā-ṭīkā, Sgron ma gsal bar byed pa’i dgongs pa rab gsal zhes bya ba bshad pa’i ṭī kā, Tōh. 1793, D. vol. 32 [ki], folios 1b–292a, pp. 2.1–583.7, continued in vol. 33 [khi], folios 1b–155a, pp. 2.1–309.5. ―― Pañca-krama-pañjikā, Rim pa lnga’i dka’ ’grel, Tōh. 1838, D. vol. 37 [chi], folios 1b–7b, pp. 2.2–14.7. Candrakīrti, Zla ba grags pa, Pradīpoddyotana-nāma-ṭīkā, Sgron gsal = Sgron ma gsal bar byed pa zhes bya ba’i rgya cher bshad pa, Tōh. 1785, D. vol. 30 [ha], folios 1b–201b, pp. 2.1–402.2; for an edition of the Sanskrit, see Chakravarti 1984. ―― Vajrasattvasādhana, Rdo rje sems dpa’i sgrub thabs, Tōh. 1814, D. vol. 35 [ngi], folios 195b–204b, pp. 390.6–408.6, Ōtani 2678, P. vol. 62 [gi], folios 168b–178a, pp. — 49 — （ 123 ） 19.5.3–23.4.2. For an edition of the Sanskrit, see Luo Hong and Tomabechi 2009. 1815, D. vol. 35 [ngi], folios 204b–209a, pp. 408.6–417.3. 密 lnga’i don mdor bshad pa rnal ’byor pa’i yid kyi ’phrog, Tōh. 1813, D. vol. 35 [ngi], 教 Nāgabuddhi, Klu’i blo, Samāja-sādhana-vyavastholi, ’Dus pa’i sgrub pa’i thabs rnam par 文 Ōtani 2674, P. vol. 62, folios 137b–149a, pp. 7.3.6–12.1.4. For a Sanskrit edition, see 化 Līlāvajra, Vajra-sattva-sādhana-nibandha, Rdo rje sems dpa’i sgrub thabs kyi ’grel pa, Tōh. Muniśrībhadra, Thub pa dpal bzang po, Pañca-kramārtha-yogi-manoharā-ṭippaṇī, Rim pa folios 148b–195b, pp. 296.4–390.6. For the Sanskrit see Jiang & Tomabechi 1996. gzhag pa’i rim pa, Tōh. 1809, Derge, vol. 35 [ngi], folios 121a–131a, pp. 241.6–261.5; Tanaka 2016. Nāgārjuna, Klu sgrub, Piṇḍīkrama-sādhana or Piṇḍīkṛta-sādhana, Mdor byas = Sgrub pa’i thabs mdor byas pa, Tōh. 1796, D. vol. 35 [ngi], folios 1b–11a, pp. 2.1–21.2; Ōtani 2661, P. vol. 61 [gi], folios 1a–12a, pp. 268.1.1–273.1.6. For Sanskrit editions, see de La Vallée Poussin 1896 and Tripathi, 2001. ―― Śrī-guhyasamāja-mahāyoga-tantra-utpādakrama-sādhana-sūtra-meśravaka[melāpaka], Mdo bsre = Rnal ’byor chen po’i rgyud dpal gsang ba ’dus pa’i bskyed pa’i rim pa bsgom pa’i thabs mdo dang bsres pa, Tōh. 1797, D. vol. 35 [ngi], folios 11a–15b, pp. 21.2–30.1; Ōtani 2662, P. vol. 61 [gi], folios 12a–17a, pp. 273.6–275.1.7. Tathāgatarakṣita, Vajra-sattva-sādhana-vyākhyā, Rdo rje sems dpa’i sgrub pa’i thabs kyi bshad pa, Tōh. 1835, D. vol. 36 [ci], folios 280b–285b, pp. 560.2–570.4. Vajraghaṇṭa, Rdo rje dril bu or Dril bu pa, Cakrasaṃvara-ṣeka-prakriyopadeśa, Dpal ’khor lo sdom pa’i dbang gi bya ba mdor bsdus pa, Tōh. 1431, D. vol. 21 [wa], folios 219b– 222b, pp. 438.3–444.5. ―― Cakrasaṃvara-sādhana, Dpal ’khor lo sdom pa’i sgrub pa’i thabs, Tōh. 1432, D. vol. 21 [wa], folios 222b–224b, pp. 444.5–448.5. Tibetan Works: A myes zhabs Ngag dbang kun dga’ bsod nams, ’Jam mgon（1597–1659/60）, Gsang ’dus chos byung = Dpal gsang ba ’dus pa dam pa’i chos byung ba’i tshul legs par bshad pa gsang ’dus chos kun gsal ba’i nyin byed, in History and Explanation of the Practice of the Guhyasamāja, Sakya Centre: Dehradun, 1985, 96 folios, pp. 1–191. ―― Gsang ’dus rnam bshad = Dpal gsang ba ’dus pa’i dkyil ’khor ’khor lo sgrub pa’i thabs rnam par bshad pa nges don phrin las rgya mtsho’i ’byung gnas, in History and Explanation of the Practice of the Guhyasamāja, Sakya Centre: Dehradun, 1985, 162 folios, pp. 193–515. Bsod nams rtse mo（1142–1182）, de’i dbang gi bya ba mdor bsdus, in Collected Works of — 48 — （ 124 ） the Sa skya, Tokyo: The Toyo Bunko, 1968, vol. 2, work 25, folios 117b–140a, pp. 407.3.1–418.4.6. Divergent Perspectives on the Guhyasamāja Sādhana in Tibet Bu ston Rin chen grub（1290–1364）, Mdor byas ’grel chen = Dpal gsang ba ’dus pa’i sgrub thabs mdor byas kyi rgya cher bshad pa bskyed rim gsal byed, Tōh. 5078, in Collected Works. New Delhi: International Academy of Indian Culture, 1967, vol. 9 [ta], 98 folios, pp. 683–878. ―― Rgyud sde spyi’i rnam gzhag = Rgyud sde spyi’i rnam par gzhag pa rgyud sde rin po che’i mdzes rgyan, Tōh. 5169, ibid., vol. 15 [ba], 305 folios, pp. 1–609. ―― Lus dkyil dbang chog = Dpal ’khor lo sdom pa’i rang bzhin gyis grub pa’i dkyil ’khor du dbang bskur ba’i cho ga zab don gsal ba, Tōh. 5053, ibid., vol. 7, [ja], 20 folios, pp. 385–424. Go rams pa Bsod nams seng ge（1429–1489） , Gsang ’dus sgrub thabs don ’grel = Dpal gsang ba ’dus pa’i sgrub thabs kun tu bzang po’i nyi ’od kyi don ’grel lam bzang gsal ba’i snang ba, in Collected Works of the Sa skya, Tokyo: The Toyo Bunko 1969, vol. 15, 71 folios, pp. 1.1.1–36.1.5. ’Gos Khug pa lhas btsas（11th cent.） , Gsang ’dus stong thun, New Delhi: Trayang, 1973. , New ’Gos Lo tsa ba Gzhon nu dpal（1392–1481） , The Blue Annals（Deb ther sngon po） Delhi: International Academy of Indian Culture, 1971. Grags pa rgyal mtshan（1147–1216） , Rgyud kyi mngon par rtogs pa rin po che’i ljon shing, in The Complete Works of the Great Masters of the Sa Skya Sect of the Tibetan Buddhism, Tokyo: The Toyo Bunko, 1968, vol. 3, 139 folios, pp. 1–70. Mkhas grub rje Dge legs dpal bzang po（1385–1438） , Bskyed rim dngos grub rgya mtsho = Rgyud thams cad kyi rgyal po dpal gsang ba ’dus pa’i bskyed rim dngos grub rgya mtsho, Tōh. 5481, in Collected Works, New Delhi: Gurudeva, 1982, vol. 7, 190 folios, pp. 3–381. Ngor chen Kun dga’ bzang po（1382–1456） , Gsang ’dus dngos grub rgya mtsho = Dpal gsang ba ’dus pa’i dkyil ’khor gyi sgrub pa’i thabs dngos grub rgya mtsho, in Collected Works of the Sa skya, Tokyo: Toyo Bunko, 1969, work 106, vol. 10, folios 184a–204b, pp. 91.4.1–102.1.2. ―― Shin tu rnal ’byor gyi khyad par sgrub thabs kyi yan lag tu bris pa, ibid., work 106, vol. 10, folios 204b–206a, pp. 102.1.2–4.5. ―― Dril bu pa’i lus dkyil gyi bshad pa, ibid., work 184, vol. 10, folios 117b–140a, pp. 398.1.1–405.4.1. Red mda’ ba Gzhon nu blo gros（1348–1412） , Bla ma bsgrub pa dpal bas zhus pa’i lan = Bla ma bsgrub pa dpal bas Gsang ba ’dus pa’i bsgrub thabs mdor byas dang Bsgrub thabs rnam gzhag gi rim pa rnam gnyis kyi mi ’dra ba’i khyad par zhus pa’i lan, in — 47 — （ 125 ） Spring yig gi tshogs, in Collected Works, Kathmandu: Sa skya rgyal yongs gsung rab slob gnyer khang, 2009, vol. 5, pp. 273.4–283.5; and in Gsung thor bu, TBRC 密 W1CZ1871, folios 302b–307a. bshad sbyar yid kyi mun sel, in Collected Works, Lhasa: Sa skya’i dpe rnying bsdu 教 Rong ston Shes bya kun rig（1367–1449）, Gsang ’dus rnam bshad = Dpal gsang ba ’dus 文 Jyekundo（Skye dgu mdo） , Gangs ljongs rig rgyan gsung rab par khang, 2004, vol. 2, 化 ―― Yid kyi mun sel = Dpal gsang ba ’dus pa’i ’grel pa sgron ma gsal ba dang bcas pa’i sgrig khang, vol. 1, pp. 225–391; vol. 2, pp. 1–497. pa’i rnam bshad byin rlabs kyi bdud rtsi rnam par rol pa’i gter, in Collected Works, 414 folios, 827p. Stag tshang Lo tsā ba Shes rab rin chen（1405–?）, ’Jam dpal rdo rje’i mngon par rtogs pa ’dod dgu rgya mtsho, Collected Works, Pe cin: Krung go’i bod rig pa dpe skrun khang, 2007, vol. 3, pp. 225–278. Tsong kha pa Blo bzang grags pa（1357–1419）, Collected Works, New Delhi: Ngawang Gelek Demo, 1975–1979, the works below are arranged according to their Tōh. numbers. ―― Gsan yig = Rje rin po che Blo bzang grags pa’i dpal gyi gsan yig, Tōh. 5267, vol. 1, 31 folios, pp. 233–293. ―― Sngags rim chen mo = Rgyal ba khyab bdag rdo rje ’chang chen po’i lam gyi rim pa gsang ba kun gyi gnad rnam par phye ba, Tōh. 5281, ibid., 512 folios, vol. 4, pp. 1–494 and vol. 5, pp. 1–530. My notes refers to the edition published in Xining: Mtsho sngon mi rigs dpe skrun khang, 1995. ―― Sgron gsal mchan = Rgyud thams cad kyi rgyal po dpal gsang ba ’dus pa’i rgya cher bshad pa sgron me gsal ba’i tshig don ji bzhin ’byed pa’i mchan gyi yang ’grel, Tōh. 5282, 521 folios, vol. 6, pp. 1–509, vol. 7, pp. 1–534. ―― Mtha’ gcod = Rgyud kyi rgyal po dpal gsang ba ’dus pa’i rgya cher bshad pa sgron ma gsal ba’i dka’ ba’i gnas kyi mtha’ gcod rin chen myu gu, Tōh. 5284, vol. 8, 143 folios, pp. 64–348. ―― Bskyed rim zin bris = Gsang ’dus bskyed rim gyi zin bris, Tōh. 5288, ibid., vol. 9, 40 folios, pp. 152–230. ―― Rnam gzhag rim pa’i rnam bshad = Rnam gzhag rim pa’i rnam bshad dpal gsang ba ’dus pa’i gnad kyi don gsal ba, [Don gsal], Tōh. 5290, vol. 9, 90 folios, pp. 280–459. ―― Bung ba’i re skong = Dpal gsang ba ’dus pa’i bskyed rim blo gsal bung ba’i re skong gnad don gsal ba, Tōh. 5305, vol. 10, 29 folios, pp. 338–394. ―― ’Dod ’jo = Bcom ldan ’das dpal ’khor lo bde mchog gi mngon par rtogs pa’i rgya cher bshad pa ’dod pa ’jo ba, Tōh. 5320, vol. 14, 195 folios, pp. 72–460. — 46 — （ 126 ） ―― Sādhana = Dpal gsang ba ’dus pa’i bla brgyud gsol ’debs dang bdag bskyed ngag ’don bkra shis lhun po rgyud pa grwa tshang gi ’don rgyud rje thams cad mkhyen pas Divergent Perspectives on the Guhyasamāja Sādhana in Tibet zhus dag mdzad pa, n.p., n.d., 82 folios, pp. 17–180 My notes refer to this publication. Partly translated into English by Thurman 1995. Published also in The Collected Ritual Texts of the Rnam rgyal grwa tshang Phan bde legs bshad gling, Dharamsala: Namgyel Dratsang 1977. Modern works: Arénes, P. 1998. Herméneutique des tantra: Étude de quelques usages du “sens cache,” Journal of the International Association of Buddhist Studies, vol. 21.2, pp. 173–226. ―― 2002. Herméneutique des Tantra: Le Ye shes rdo rje kun las btus pa’i rgyud las ’byung ba’i rgyan bdun rnam par ’grol ba de Śraddhākaravarman, in Henk Blezer, ed., Reli- gion an Secular Culture in Tibet, Tibetan Studies 2. Leiden: Brill, pp. 163–83. Bentor, Yael. 2006. Identifying the Unnamed Opponents of Tsong-kha-pa and Mkhas-grubrje Concerning the Transformation of Ordinary Birth, Death and the Intermediate State into the Three Bodies, in Tibetan Buddhist Literature and Praxis: Studies in Its Formative Period 900–1400 , eds., Ronald M. Davidson and Christian K. Wedemeyer, pp. 185–200, E. J. Brill, Leiden. ―― 2009. Do ‘The Tantras Embody What The Practitioners Actually Do’? in Contributions to Tibetan Buddhist Literature, ed., Orna Almogi, pp. 351–373, International Institute for Tibetan and Buddhist Studies, Beiträge zur Zentralasienforschung 14, Halle. ―― 2015a. Tsong-kha-pa’s Guhyasamāja Sādhana and the Ārya Tradition, in Vimalakīrti’s House: A Festschrift in Honor of Robert A.F. Thurman on the Occasion of his 70 th Birthday, eds., Christian K. Wedemeyer, John D. Dunne, and Thomas F. Yarnall, The American Institute of Buddhist Studies at Columbia University in New York with Columbia University’s Center for Buddhist Studies and Tibet House US, pp. 165–192. ―― 2015b. Meditation on Emptiness in the Context of Tantric Buddhism, The Journal of Buddhist Philosophy, 1, pp. 136–155. ―― 2017. Tibetan Interpretations of the Opening Verses of Vajraghaṇṭa on the Body Maṇḍala, in Chinese and Tibetan Esoteric Buddhism, eds., Yael Bentor and Meir Shahar, Leiden, Brill, pp. 230–259. ―― forthcoming. Did mKhas grub rje Challenge the Authenticity of the Sa skya Lam ’bras Tradition? In Towards a History of 15 th Century Tibet: Cultural Blossoming, Religious Fervour and Political Unrest, edited by Volker Caumanns and Marta Sernesi, Lumbini International Research Institute, Nepal. Chakravarti, Chintaharan. 1984. Guhya-samāja-tantra-Pradīpoddyotana-ṭīkā-ṣaṭ-koṭīvyākhyā. Patna: Kashi Prasad Jayaswal Research Institute. — 45 — （ 127 ） Fremantle, Francesca. 1971. A Critical Study of the Guhyasamāja-Tantra, Ph.D. Disserta密 Jiang, Zhongxin and Toru Tomabechi 1996, eds., The Pañcakramaṭippanī of Muniśrībhadra: 教 Kilty, Gavin. 2013. Tsong kha pa, A Lamp to Illuminate the Five Stages: Teachings on 文 tion, London: School of Oriental and African Studies. Harrison, Paul and Watanabe Shôgo 2006. Vajracchedikā Prajñāpāramitā, in Buddhist Manuscripts, vol. 3, pp. 89–132, ed. Jens Braarvig, Oslo: Hermes Publishing. Introduction and Romanized Sanskrit Text, Bern: Peter Lang. series, vol. 15. Luo Hong and Toru Tomabechi. 2009, eds., Candrakīrti’s Vajrasattva-niṣpādana-sūtra （ Vajrasattva-sādhana）Sanskrit and Tibetan Texts, Beijing & Vienna: China Tibetology Publishing House & Austrian Academy of Sciences Press. Martin, Dan. 1987. Illusion Web-Locating the Guhyagarbha Tantra in Buddhist Intellectual History, in Silver on Lapis: Tibetan Literary Culture and History, ed., C.I. Beckwith, Bloomington: Tibet Society, pp. 175–220. Matsunaga, Yukei. 1963. On the Saptālaṅkāra, Journal of Indian and Buddhist Studies [IBK], vol. 11, pp. 470–76 [in Japanese]. ―― 1964. A Doubt to Authority of the Guhyasamāja-Ākhyāna-Tantras, Journal of Indian and Buddhist Studies [IBK], vol. 13.2, pp. 16–25. ―― 1977. Some Problems of the Guhyasamāja-tantra, in Studies in Indo-Asian Art and Culture, eds., Lokesh Chandra and Perala Ratnam, vol. 5, pp. 109–119. ―― 1978. The Guhyasamāja Tantra: A New Critical Edition, Osaka: Toho Shuppan. Roerich, George N., tr., The Blue Annals, Calcutta: Asiatic Society, 1949; rept. Delhi: Motilal Banarsidass 1979. Roloff, Carola 2009. Red mda’ ba: Buddhist Yogi-Scholar of the Fourteenth Century, Wiesbaden: Ludwig Reichert Verlag. Steinkellner, Ernst. 1978. Remarks on Tantristic Hermeneutics, in Louis Ligeti, ed., Proceedings of the Csoma de Kőrös Memorial Symposium, Budapest: Akadémiai Kiadó, pp. 445–58. Tanaka, Kimiaki 2016. Samājasādhana-Vyavastholi: of Nāgabodhi/Nāgabuddhi: Introduction and Romanized Sanskrit and Tibetan Texts, Tokyo: Watanabe Publishing. Tanemura, Ryugen. 2015. Guhyasamāja, in Brill’s Encyclopedia of Buddhism, vol. 1, pp. 326–333. Thurman, Robert A.F. 1988. Vajra Hermeneutics, in Donald Lopez, ed., Buddhist Hermeneutics, Honolulu: Kuroda Institute, pp. 119–48. 〈キーワード〉Guhyasamāja, Tibet, Vajrayāna, Sādhana — 44 — 化 Guhyasamāja Tantra, Boston: Wisdom Publications, The Library of Tibetan Classics THE MIKKYO BIJNKA