The Convergence of Theoretical and Practical Concerns in a Single Verse of the Guhyasamaja Tantra

by YAEL BENTOR

One of the most well-known verses of the Guhyasamaja Tantra a verse incorporated into many practices of the Guhyasamaja—is found in its second chapter, the chapter on the “mind directed at enlightenment” ([[[bodhicitta]]]], byang chub sems). Here the Tathagata, whose name is “Vajra-Body, Vajra-Speech and Vajra-Mind of All Tathagatas,” dwelt in absorption,1 and uttered the following verse.

abhave bhavanahhavo bhavana naiva bhavana |

iti bhavo na bhdvah syad bhavana nopalabhyate ||5

This chapter first examines the role of this verse in the practice of the Guhyasamaja, and then explores how it was understood. We then fol¬low the shifts in philosophical affiliation this verse underwent over time

The Role of Our Verse in the Practice of the Guhyasamaja

In the most important sadhana manual for the practice of the Guhyasamaja according to the Arya tradition, the Pindi-krama-sadhana (Mdor byas sgrub thabs), by Nagarjuna,4 our verse appears, with small variations, at the very beginning of the generation of the mandala and the deities dwelling in it. In

introducing it, Nagarjuna explains:

“[The yogis] meditate that in the ultimate truth the three realms are devoid of intrin¬sic nature (nihsvabhava)He then

concludes:6 “With this verse, [the yogis] meditate that the nature of the animate and inanimate [[[world]]] is empty (sunya, stongpa), and with this ritual

method, the animate and inanimate are blessed as the ground of pristine wisdom (jnana-bhumi, ye shes kyi sa).”

In another central manual on the practice of the Guhyasamaja of the Arya tradition, the Samaja-sadhana-vyavasthali (Rnam gzhag rim pa),7 Nagabuddhi instructs the practitioners to meditate, while reciting this verse, on everything as having the nature of the space that remains after the destruction of the three realms at the end of the eon.

Hence, the meditation here is a meditation on emptiness. In a type of ritual death, practitioners dissolve themselves and their entire world into emptiness. The new pure rebirth of the practitioners as deities in the celestial mansion of the mandala then arises from emptiness. Emptiness here

corresponds to the empty eon in between the previous and the later worlds in a cosmological cycle, which is understood not as nothingness, but as something that has the potential for the recreation of the new world. And for this reason, emptiness here is called the ground of pristine wisdom—it is the ground

for all phenomena. During the prac¬tice of the creation stage, the kyerim (bskyed rim), the elimination of all appear¬ances of the world and all its inhabitants within the practitioner’s own mind is the ground for all the visualizations during the meditation that follows. And this initial meditation on emptiness is practiced while our verse is recited.

This verse is obviously mantra-like, alliterating (anuprasa) the sounds bha, va, and na. Moreover, it puns on the meanings derived from the root Rhhu. Bhava is being, existing, that which exists, an entity, an existing thing, and all earthly objects. Thus, bhava indicates both a thing and a state of

existence. In the first sense it can be translated as an entity or a thing; and as a state of existence, bhava can mean existing, and abhava not existing. As for bhavana, it is usually translated as meditation. This noun is in the causative form, and carries the meanings of causing to be, bringing into existence, creating, and producing.

This meaning of meditation is indeed the foundation of the creation stage. Our verse is recited immediately after practitioners visualize away ordinary appearances, and right before they begin to visualize themselves as enlightened beings at the center of the celestial mansion of the mandala. The pun on the meaning of the nature of existence (bhu, bhava) and of meditation (bhavana)—in the sense of “causing to be”—is very germane at this point of the

practice. The practitioners may reflect here: “Into what would the ordi¬nary world disappear?” “How would the enlightened realm be created?” “Does the ordinary world exist?” “Is the ordinary world a meditation, that is to say, ‘caused to be,’ by the mind?” “Does the realm of the mandala exist? Is it more 

THE CONVERGENCE OF THEORETICAL AND PRACTICAL CONCERNS

real or less real than the ordinary world?” “Since it is obvious that this enlight¬ened realm is created by the mind, isn’t the ordinary world similarly a result of visualization or mental construction?” “On the other hand, the deities and the celestial mansion where they reside have arisen from emptiness, the true nature of all things, so they must be real.” And so on.

The pun on bhu and bhavana is lost when the Sanskrit verse is translated into other languages. The etymology of the Tibetan verb for “to meditate,” gompa (sgom pa), is not “to cause to be,” but rather “to habituate.” Still, in certain contexts the meaning of the Sanskrit word does carry through. The final

verb upa-jlabh also bears a variety of meanings. The difficulty in understanding the meaning of the Sanskrit verse itself, as well as the problem of translating it into Tibetan contributed to the great variety of interpretations the verse received.

This is how this verse appears in the Tibetan translation of the Guhyasamaja Tantra:

dngos po med pas sgom pa med/bsgom par bya ba bsgom pa min / de Itar dngos po dngos med pas/sgom pa dmigs su med pa'o // The differences between this Tibetan translation and the Sanskrit go beyond those that are the result of the grammar of these two languages. In the Tibetan translation, we have in the first line (pada), dngos po med pas (because things do not exist) for abhave, which is the usual translation of abhave, but as noted already, does not have exactly the same meaning; in the second pada we have bsgom par bya ba (one ought to meditate) for bhavana,8 and in the third

line again the verbal noun med pas (because they do not exist), while the Sanskrit has a negation of an optative of a verb of existence, na bhavah syad. The Interpretation of Our Verse in the Pradipoddyotana in Sanskrit

As we observed earlier, this verse may be rendered into English in more than one way. Let us now examine how this “mantra” was understood and inter¬preted. In the most famous commentary on the Guhyasamaja Tantra according to the Arya tradition, the Pradipoddyotana, Candrakirti interprets our verse by means of the tantric hermeneutical method called the tsulshi (tshul bzhi) or “the four ways,” which consists of the literal, common, hidden, and ultimate levels of interpretation.10

Even though Candrakirti does not explicitly say so, the literal level of interpretation here is clearly based on Nagarjuna’s tetralemma. The four lines are explained in correspondence with the four possibilities: existing, 

92 tibetan ritual not existing, both existing and not existing, and neither existing nor not exist¬ing. Still, we should not rush to the conclusion that since this work is writ¬ten by a Candrakirti, a Madhyamika explanation is what we must expect here.

Candrakirti explains

[I summarize and interpret]:11

1. [If there are no things], there can be no meditation (bhavana = causing to be) because if there are no things, there cannot be causing to be.

2. [If there are things], then meditation [causing to be] is not a meditation, because even without meditation [causing to be], there are existing things.

3. [If there are both things and no things]: that which is both a thing and a no thing would not exist, therefore, that thing [which is both] would not be a thing.

4. [If there are neither things nor no things], then, there cannot be medita¬tion [causing to be]. Therefore, no meditation is to be perceived. So far, this is the literal level of interpretation. If we look at all four levels of interpretation, then what we find here is not the usual tantric hermeneutic by means of the tsulshi.12 Instead, it is the fourfold meditation common in Yogacara writings that is applied here to explain our verse.

The stages of the fourfold meditation that are found in some of the Five Works of Maitreya (Byams chos sde lnga) and in Vasubandhu's commentaries on them are:13

1. Apprehending things to the extent they exist.

2. Apprehending mind-only or mental-events-only (cittamatra, sems tsam).

3. Apprehending that there is also no mind-only.

4. Realizing suchness.

How is this fourfold meditation applied in the Pradipoddyotana in explaining our verse?14

1. The first level is “apprehending things to the extent they exist”—here, according to the four possibilities of Nagarjuna.

2. The second stage is “apprehending mind-only” or “mental-events-only” (cittamatra, sems tsam) by realizing that external things are creations of the mind (cittamaya).

3. In the third stage, Candrakirti maintains that given the absence of things, neither is there mind-only, and the two truths are indivisible.

4. On the ultimate level, for those who realize the stage of union15—and here Candrakirti does use tantric terminology—there is no more cling¬ing to meditator, meditation, and object of meditation. 

The sädhanas use the recitation and meditation on our verse as a part of medita-tion on emptiness, and according to Candraklrti’s Pradipoddyotana, this medi-tation is the fourfold meditation typical of treatises of the Yogäcära School. As often pointed out,16 Yogäcära works offer more dynamic processes, especially meditative processes of transformation, and such processes are also the foun-dation of tantric practices, such as the creation stage. The fourfold meditation is a process that matches the creation stage well.

During the practice of the creation stage, at first the practitioners visualize away their ordinary world and reflect on the extent it exists. In the second stage, they create, in their minds, their enlightened realm—with themselves as dei¬ties and with their environment as the celestial mansion of the

mandala—and they meditate on mind-only. In the third stage, they realize that this creation, much like their ordinary world, is not real; and by understanding that the true nature of all phenomena is not mental-event only, they understand that neither is there mind-only. Finally, after dissolving their visualization into emptiness, they realize the suchness of all things, and the nonduality of emptiness and appearances.

Two Different Interpretations of Our Verse in Tibetan Translation

There are two different explanations of our verse in works translated into Tibetan: one in the Tibetan translation of the Pradipoddyotana,17 and the other in Santipa’s commentary on the Pindi-krama-sadhana, the Ratndvali,“ which, as we saw, contains this verse as well. Here is Santipa’s commentary on the first part of the verse (abhave bhavanabhavo, dngos po med pas sgom pa med) in the literal level of interpretation, which seems to have survived only in its Tibetan translation.19

brtan pa dang g.yo ba’i dngos po thams cad med na sgom20 pa ni med de/ bsgom par bya ba med pa’i phyir ro /

This seems to be a good translation of the Sanskrit of Candraklrti’s Pradipoddyotana:21

sthira-cala-sarva-padarthanam abhave sati bhavanaya abhavah bhavyabhavat.

Santipa’s explanation can be rendered into English as: “When there are not any animate and inanimate things, there is no meditation [causing to be], because there is nothing to meditate upon [to cause to be].”

The Tibetan translation of the Pradîpoddyotana itself is somewhat different:

brtan pa dangg.yo ba’i dngos po thams cad kyi ngo bo nyid ni med pa yin na ni sgom pa med ste bsgom par bya ba med pa’i phyir ro / This may be translated as: “When ‘there is no’ essence to all the animate and inanimate ‘things,’ ‘there is no meditation,’ because there is nothing to medi¬tate upon.”

In the Sanskrit there is no equivalent to the word “essence” (ngo bo nyid) found in the Tibetan. In terms of the “view,” the difference between these two Tibetan translations is considerable. We can conclude then that the Tibetan translation of the Pradîpoddyotana is also a transition toward a more standard Mâdhyamika view. Still, it is not clear when this philosophical shift took place. According to its colophon in the Bstan, ‘gyur, the Pradîpoddyotana was trans-lated and revised in the eleventh century.

Tibetan Commentaries

There was a short commentary on the Guhyasamaja Tantra written by Chag Lotsawa Choje Pel (Chag lo tsa ba Chos rje dpal), who lived in the thirteenth century,23 but at present it is unavailable to me. The version that Buton Rinchen Drub (Bu ston Rin chen grub, 1290-1364) comments upon is very similar to

that of the Dronsel (Sgron gsal), the Tibetan translation of the Pradipoddyotana.24 Buton more or less reproduces the fourfold meditation of the Pradipoddyotana without commenting on it. Apparently for Buton, the question as to which school (Yogacara or Madhyamika) this meditation belongs was not an issue. Furthermore, in his explanation of another step in the creation stage of the Guhyasamaja,25 Buton explicitly advocates Mind Only: “this is so that you will understand all [[[phenomena]]] as Mind Only (sems tsam).”

Buton was one of the last commentators on the Guhyasamaja Tantra who actually knew Sanskrit, and who could see that the meaning of our verse in its Tibetan rendering was different from the meaning of the Sanskrit. He could also see that, in its most important commentary, the Praipoddyotana in its Tibetan translation, the meaning of our verse was further altered, at least since the fourteenth century, if not before. But Buton does not comment on this. Among the Gelugpas, the Arya School of the Guhyasamaja is considered to hold the philosophical positions of the Madhyamika School, and the empti¬ness meditated upon in practices of the “Path of Mantra” (i.e., in the Tantra) is

considered no different from the emptiness of the Madhyamika School. As we would expect, the portion of the commentary of Tsongkhapa (1357-1419) that explains our verse26 accords with his views on Madhyamika and emptiness. In Tsongkhapa’s interpretation, “a thing” must be glossed as “an inherently existing thing”; “no thing” therefore refers to “absence of own essence,” and “meditation” (sgompa) is meditation on suchness.

Furthermore, Tsongkhapa does not enter into the subject of the four¬fold meditation. For him, the explanation of the Pradipoddyotana here is, as Candrakirti names it, an explanation by means of the tantric hermeneutical method called “the four ways” (tshul bzhi). In the section on the creation stage in his commentary on the Namzhag Rimpa (Rnam gzhag rim pa),27 Tsongkhapa says that from among the “four ways” in which the Pradipoddyotana explains our

verse, it is the literal and common levels of interpretations which are rel¬evant to the creation stage. The common level of interpretation is common to both the creation and completion stages. But the hidden and ultimate levels pertain to practices with the consort, to the subtle body, and to the complet¬ing stage alone. Hence, in his discussion of the creation stage, Tsongkhapa discusses only the first two levels of interpretations. Thus, in the context of the creation stage, his explanations do not go into the fourfold meditation and the problems that this poses for Madhyamika.

Though Tsongkhapa did not concern himself with the fourfold medita¬tion, he did address the question of the nature of the external world. Tsong¬khapa28 explains the meaning of “external appearance” or “external aspect” (bahyakara, phyi rol gyi mam pa) in the Pradipoddyotana22 by specifying that this refers to external objects which “exist by their own essence (rang gi ngo bo nyid kyis grub pa),’’ and by adding that things have no existence “apart from being

merely imputed by the mind (sems kyis btags pa tsam las).” Simi¬larly, Tsongkhapa50 glosses the phrase “created by the mind” (cittamaya) or “of the nature of the mind” (sems kyi rang bzhin) in the Pradipoddyotana21 with “of the nature of being merely imputed by the mind (sems kyis btags pa tsam gyi rang bzhin).”22 Thus, rather than taking the explanation of the Pradipoddyotana at its (Yogacara) face value, Tsongkhapa gives it a Prasangika Madhyamika spin. In his commentary on difficult points in the Pradipoddyotana, entitled the Tacho Rinchen Nyugu (Mtha’good rin chen myugu),22 Tsongkhapa elaborates on the subject of external objects and Mind Only:

When [the Pradipoddyotana] explains [the verse that] begins with dngos po med pa on the common level of interpretation, there appears something like a refutation of external objects and an establishment

[of them] as mind-only; and there are similar occurrences also in other cases. It seems that [some people], unable to examine this very thor¬oughly, did not understand that the position of the Noble Father [[[Arya Nagarjuna]]] and his Spiritual Sons [[[Candrakirti]] and so on] in general, and the position of the

commentator [[[Candrakirti]]] in particular, which accept external objects as conventional designations. Therefore, they say that the system of the Pradipoddyotana does not accept external objects. However, since I already extensively explained elsewhere why this is unacceptable and how to eradicate the extreme views of eternal- ism and nihilism, I do not elaborate here.

It seems that the purpose of this passage is to explain how “some people” might come to the conclusion that the Pradipoddyotana refutes external objects and maintains that they are “mind-only.” Buton was an important teacher in the lineage that came down to Tsongkhapa, and it seems that Tsongkhapa had much reverence for Buton, although he did not always agree with him.34 For Tsongkhapa, there is a crucial difference between holding that external objects exist

as conventional designations, and holding that external objects do not exist at all. Tsongkhapa does not agree that the author of the Pradipoddyotana rejects external objects. According to Tsongkhapa, the Arya school of the Guhyasamaja maintains that external objects exist as conventional designations, and this causes him to offer an alternative gloss to the line of the Dronsel that refers to external objects—namely that while existing externally, inherently they do not exist, but are mere mental imputations.

In commenting on our verse, Tsongkhapa’s disciple Khedrubje (Mkhas grub rje, 1385-1438) follows his teacher. In the context in which Buton explains:35 “This is so that you will understand all [[[phenomena]]] as Mind Only, and realize the two truths as indivisible,” Khedrubje refers to this very passage, without naming names, by saying:36

Some lamas (bla ma kha cig) say that this is so that you will understand all [[[phenomena]]] as mind-only, and realize the two truths as indivisible. They do not understand that the author of the Pradipoddyotana accepts external things as conventional designations, and that this is also the intention of Arya Nagarjuna. The world and its inhabitants... are only conventional truth.... Hence this [statement] is just pointless.

For Tsongkhapa and Khedrubje then, there is no doubt that Candrakirti, the author of the Pradipoddyotana, as well as Nagarjuna, the author of the Pindi- krama-sadhana, do not accept the Mind-Only School, but hold the view of the Prasangika Madhyamika School.

Conclusions

We have analyzed the recitation of a verse, or rather “mantra,” as an important ritual event during the creation stage of the Guhyasamaja. As we saw, in his Pindi-krama-sadhana Nagarjuna calls this recitation of the mantra a “ritual” or “ritual method” (vidhi, cho ga). While rituals may remain almost unchanged, their interpretations are often adjusted in accordance with current theories. Indeed, the meaning of the Sanskrit mantra is fluid and enigmatic—

the more one reflects on it, the more implications one finds—and it is precisely this flu¬idity that serves as the basis for reflections during the meditation and allows for different interpretations throughout history. Still, this mantra is somewhat different from other mantras recited during this tantric practice, insofar as it has overt philosophical content—that is, since it resembles verses from Bud¬dhist philosophical treatises.

Our focus was the import given to this liminal point of the practice, just after the practitioners visualize away (mi dmigs) their ordinary identity and the ordinary appearances of themselves and their world, and just before they cre¬ate in their mind the mandala with the deities of their enlightened realm. On one level, this ritual may simply be taken as the erasure of one’s own ordinary existence by transforming it into utter nothingness so that a new

reality can arise. However, in a Buddhist philosophical context, this stage is understood as dissolution into emptiness. While in the Mula-madhyamaka-kdrikd (chapter 24) and in the Vigrahavyavartani-karika (v. 70),37 Nagarjuna does emphasize that emptiness is that which makes change possible, in Buddhist Tantric litera¬ture on the creation stage, such as the Pindi-krama-sadhana, emptiness is more explicitly understood as the ground or potential for all phenomena.

The commentators saw it as their task to explain the mantra, and since it lends itself to a number of interpretations, various commentators stepped up to the challenge, and most of them explained it by employing the theoretical frameworks they most favored.38 The Pradipoddyotana applies first Nagarjuna’s

tetralemma, but then it applies the fourfold meditation common in Yogacara treatises. The meaning Candraklrti, the author of the Pradipoddyotana, saw in our verse was modified twice. The first philosophical reorientation toward a so- called authentic Madhyamika or so-called authentic Prasangika Madhyamika was written into the Tibetan translation of the Pradipoddyotana. Unlike its Sanskrit version (and unlike the Tibetan translation of Santipa’s Ratnavali,

the Rin chen phreng ba), the Tibetan translation does not speak about the absence of things, but rather about the absence of their essence. Perhaps, when ear¬lier versions of the translations of the Tengyur texts and some of the former commentaries on the Pradîpoddyotana become available, we will be able to

determine with more precision when this modification occurred. The second transformation of the meaning of our verse took place in Tibetan composi¬tions, when especially among the Gelugpa, typical Yogâcâra practices and what came to be called Mâdhyamika-Yogâcâra fell from favor. Then, together with all the other authors of the Ârya School of the Guhyasamaja, Candraldrti, the author of the Pradîpoddyotana, came to be identified with “orthodox” Prâsangika Madhyamaka.

NOTES

This research was supported by The Israel Science Foundation (grant no. 874/02-1).

1. In the concentration called “the vajra-mode of awakening into manifestation of all Tathâgatas” (sarva-tathâgatâbhisambodhi-naya-vajra, de hzhin gshegs pa thams cad kyi mngonpar rdzogs par hyang chuh pa 'i tshul rdo ije); for references, see note 2.

2. This reading is found in the editions of Francesca Fremantle, “A Critical Study of the Guhyasamâja-tantra: (Ph.D. diss., London: School of Oriental and African Studies, 1971), 190; Yukei Matsunaga, ed., The Guhyasamaja Tantra: A New Critical Edition (Osaka: Toho Shuppan, 1978), 9; Benoytosh Bhattacharyya, ed., Guhyasamaja Tantra or Tathâgataguhyaka (Baroda: Oriental Institute, 1931), n; S. Bagchi, ed., Guhyasamaja Tantra or Tathâgataguhyaka (Darbhanga: The Mithila Institute, 1965), 8. Fremantle notes a variant reading of abhâvi for abhâve in her manuscripts C and P and comments that a

substitution of‘i’ for ‘e’ occurs several times; Matsunaga notes the same variant reading in his manuscripts A and T5. The Pradîpoddyotana [Chintaharan Chakravarti, ed., Guhya-samâja-tantra-Pradïpoddyotana-tïkâ-sat-kotï-vyâkhyâ (Patna: Kashi Prasad Jayaswal Research Institute, 1984), 31] explains the

literal meaning with a locative absolute: abhâve sati, and most commentaries take the abhâve in the begin¬ning of our verse to mean abhâve sati. In his edition of the Pindï-krama-sâdhana, which cites our verse, Louis de La Vallée Poussin [Etudes et textes tantriques: Pancakrama (Gand: H. Engelcke, 1896),

2] has abhâvabhâvanâ bhâvo for abhâve bhâvanâbhâvo. According to David L. Snellgove, [[[Hevajra Tantra]]: A Critical Study (London: Oxford University Press, 1959), 77] the Sekoddesatïkâ also has this reading. Giuseppe Tucci [“Some Glosses upon the Guhyasamaja,” Mélanges Chinois et Bouddhiques, no. 3 (193d 1935): 352] “corrects” the Sanskrit text in light of its Tibetan translation, by changing abhâve to abhâvena, although he admits that this makes the firs I line hypermetric.

3. As we shall see, the meaning of this verse is purposely enigmatic, and indeed it was interpreted in various ways. For some translations of this verse, see Benoytosh Bhattacharyya, Guhyasamaja Tantra, xx; Tucci, “Some Glosses,” 353-53; Snellgrove, Hevajra Tantra, part 1, 77; Fremantle, “A Critical Study,” 34 and 143-4, n.i; Pio Filippani-Ronconi, “La formulazi- one liturgica della dottrina del Bodhicitta nel 2 Capitolo de Guhyasamâjatantra," Annali (Istituto Universitario Orientale di Napoli) vol. 32, no. 2, n.s. XXII (1972): 190; Kenneth Eastman, “Mahâyoga Texts at Tun-huang” (Master’s thesis, Stanford University, 1983),

18-19; Raniero Gnoli, “Guhyasamâjatantra (chapters 1, 2, & 5),” Testi Buddhist'. (Turin: Unione Tipografico-editrice Torinese, 1983), 628; Peter Gang, Das Tantra der Verhorgenen Vereinigung: Guhyasamâja-Tantra (München: Eugen Diederichs Verlag, 1988), 123.1 would like to thank Jake Dalton for providing me with a copy of Eastman’s Thesis. Here is justone possible translation: “When there are no [[[existing]]] things, there is no meditation [causing to be]. Meditation indeed is no meditation. Thus, a tiling would be no thing. No meditation is to be perceived [or, there is no object to the meditation].” As we shall see, there are vari¬ous other alternative translations.

4. Pindïkrama-sâdhana (Pindïkrta-sâdhana); the Sanskrit was edited by La Vallée Poussin, Pancakrama, 1-14; also, Ram Shankar Tripathi, Pindïkrama and Pancakrama of Âcârya Nâgârjuna (Sarnath: Central Institute of Higher Tibetan Studies, 2001), 1-32. For the Tibetan translation, see Sgrubpa’i thabs mdor byaspa, Sde dge Bstan ‘gyur, Toh. no. 1796, Rgyud ngi, folios ib-na; Peking Bstan ‘gyur, Ôtani 2661, vol. 61, 268.1.1-273.1.6.

5. La Vallée Poussin, Pancakrama, v. i6cd; Tripathi, Pindïkrama and Pancakrama, v. ijcd; Sde dge, Sgrub pa’i thabs, folio 2U3 /g Peking, Sgrub pa’i thabs, 269.3.2. While La Vallée Poussin (v. i6d) and Tripathi (v. 15c!) have bhavatrayam, all versions of the Bstan ‘gyur available to me have dngos po mams/ srid gsum. For nihsvabhâva, the Bstan ‘gyur has dngos po med pa.

6. La Vallée Poussin, Pancakrama, v. 18; Tripathi, Pindïkrama and Pancakrama, v. 17; Sde dge, Sgrubpa’i thabs, folio 2b/j5; Peking, Sgrubpa’i thabs, 269.3.3-4.

7. Nâgabuddhi (Klu’i bio), Samâja-sâdhana-vyavasthâlï (’Dus pa’i sgrub pa’i thabs mam par gzhag pa’i rimpa), Sde dge Bstan ‘gyur, Toh. no. 1809, Rgyud ngi, folio 12^4-5; Peking Bstan ‘gyur, Ôtani 2674, vol. 62, 7.4.4-6.

8. De bzhin gshegs pa thams cad kyi sku gsung thugs kyi gsang chen gsang ha ‘dus pa zhes bya ba brtag pa’i rgyal po chen po. The Tantra is found in a number of recensions: Dunhuang, IOL (India Office Library) Tib J 481 and IOLTib. J 438; The Rnying ma rgyud ‘bum (Thimbu: Dingo Khyentse Rimpoche, 1973), vol. 17, folios ^1-31434; Sde dge Bka’ ‘gyur, Toh. no. 442, Rgyud ‘bum ca, folios goa-i48a (vol. 81, 181-295); Peking Bka’ ‘gyur, Ôtani 81, vol. 65,

174.3.5-203.2.1; Stag Palace, vol. 96, 2-190; also in Dpal gsang ba ‘duspa’i rtsa rgyud ‘grelpa bzhi sbrags dang bcaspa (Lhasa: Zliol Printing House, made from block-prints carved in 1890). The reading of the verse given here is found in both the Stog Palace edition (vol. 96, 17.5-6) and the Zhol edition of the ‘Grel pa bzhi sgrags (folio 6a2). The most significant variant reading is found in the Sde dge edition (vol. 81, 187.7-188.1) and the Peking (vol. 65:

176.3.2-3) which have bsgom pa bsgom pa ma yin nyid for bsgom par bya ba bsgom pa min in the second line (pada); and in the Dunhuang (IOLTib. J 438, folio 8b4) which has bsgompa’i dngos for sgompa med, at the end of the first pada. In the Hevajra Tantra (I.viii.44; Snellgrove, 30-31): bhâvanâ naiva bhâvanâ is

similarly translated as sgom pa nyid ni sgom pa min. As for other variant read¬ings in the first pâda, the Sde dge and the Peking have la and the Dunhuang has par for pas; the Rnying ma rgyud ‘bum edition (vol. 17, 15.3) has bsgom pa med for sgom pa med. In the second pâda, the Rnying ma rgyud ‘bum has sgom pa min for bsgom pa min. In the third pâda, the Sde dge and the Peking have de for pas. And in the fourth pâda, Dunhuang and the Peking have bsgom for sgom and the Dunhuang has do zhes for pa’o. The Sgron gsal (Sde dge Bstan ‘gyur, Toh. no. 1785, Rgyud ha, vol. 30, 47.2; Peking Bstan ‘gyur,

Ötani 2650, vol. 60, 35.3.4; The Golden Bstan ‘byur, vol. 30, 3235-6) has similar read¬ings to those of the Stog Palace and the Zhol editions; the variants are: sgom pa med at the end of the first päda, and bsgom par bya ba bsgom pa min in the second päda. The Mdor byas sgrub thabs (Sde dge, folio 2b4 and Peking, 269.3.3) has in the fast päda: dngos po med la bsgompa’i dngos. In his commentary on the Mdor byas sgrub thabs, the Rin chen phreng ba (Sde dge

Bstan ‘gyur, Toh. no. 1826, Rgyud ci, vol. 36, 50.1), Säntipa has dngos po med pas sgom pa med, as do the Zhol and the Stog Palace in the Root Tantra, and not bsgompa’i dngos. Bu ston Rin chen grub [Dpalgsangba ‘duspa’i sgrub thabs mdor byas kyi rgya eher bshad pa bskyed rim gsal byed {Mdor byas ‘grel chen), The Collected Works of Bu-Ston (New Delhi: International Academy of Indian Culture, 1967), 708.6-710.3] rejects the reading bsgom pa’i dngos, because it lacks a negative particle, while Santipa’s commentary explains it with a negation.

9. As Fremantle, “A Critical Study,” 143 points out, the Tibetan here is a transla¬tion of bhävya and not bhävanä.

10. The literal {aksarärtha, tshig gi don or yi ge’i don), the common {samastänga, spyi’i don), the hidden (garbhin, sbaspa) and the ultimate {kolika, mtharthugpa) levels of interpretations. This passage was also translated from Tibetan into Italian in Filippani- Ronconi. “La formulazione,” 194-95. Unfortunately, until I can see the Sanskrit manu¬script itself, everything I can say is rather tentative, since I must rely on Chakravarti’s edition, Guhya-samäja-tantra-tfkä, 31-32.

11. Chakravarti, Guhya-samäja-tantra-tikä, 31; the Sgron gsal (Sde dge, 47.2-3; Peking, 35.3.3-5; The Golden Bstan ‘gyur, 3235-32^).

12. In the usual tshul bzhi, the common level of interpretation is common to both creation or generation stage {bskyed rim) and completion stage {rdzogs rim), both Sütra and Tantra, and so on; the hidden level often refers to practices with the consort, the subtle body, and so on; and the ultimate level of interpretation applies to the rdzogs rim alone.

13. See Vasubandhu’s Trimsikä (w. 28-29) aric’ Trisvabhävanirdesa (w. 36-37); also Madhyäntavibhäga (ch. 1, v. 6), Mahäyäna-süträlamkära (ch. 6, v. 8 and ch. 14, w. 23-28), and Dharmadharmatävibhäga; sec Klaus-Dieter Matlies, Unterscheidung der Gegebenheiten von ihrem wahren Wesen [Dharmadharmatävibhäga] (Swisttal-Odendorf: Indica et Tibetica, 1996), 102-103, 64, 110, 139. Ronald Davidson [“Buddhist Systems of Transformation: Äsraya-parivrtti / -parävrtti

among the Yogäcära” (Ph.D. diss., University of California, Berkeley), 1985, 295—97], David Jackson [[[The Entrance Gate for the Wise]] [section III] (Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien, 1987), 348-51 and notes thereon] and Christian Lindtner [“Cittamätra in Indian

Mahäyäna until KamalasTla,”Wiener Zeitschrift für die Kunde Südasiens, 41 (z997): 159-206] have pointed to still other parallels. This fourfold meditation appears also in the works of Säntaraksita and KamalasTla as well as Säntipa (Ratnäkarasänti) and therefore came to be associated with what was later called

the Yogäcära-Mädhyamika. Chizuko Yoshimizu [“The Theoretical Basis of the bskyed rim as Reflected in the bskyed rim Practice of the Arya School,” Report of the Japanese Association for Tibetan Studies, 33 (1987): 25-28] who edited and translated Bu ston’s commentary on our verse as it appears in Nägärjuna’s Pindlkrama-sädhana (Bu ston’s Mdor byas ‘grel chen; see below), also consulted the texts of the Pradipoddyotana and the RatnävalT, and commented (27):

“[T]he Ärya school employed Yogäcära-Mädhyamika theory virtually from its starting point.” Her conclusion on this point is (28): “Most Tantric authors including Nägärjuna seem to lack any concrete understanding of Mahäyäna philosophies.” Since Katsumi Mimaki [“The Bio gsal grub mtha’ and the Mädhyamika

Classification in Tibetan Grub mtha’ Literature,” in Contributions on Tibetan and Buddhist Religion and Philosophy, ed. Ernst Steinkellner, 2 vols. (Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien, 1982), 2: 163] maintains: “We know today that the terms indicating the sub-schools of the Mädhyamika, such as the...Yogäcära- Mädhyamika, ... have been invented by Tibetan authors, and do not appear in Indian texts,” I use here the term Yogäcära and not Yogäcära-Mädhyamika. My conclusion are different from those of yoshimizu on this point.

14. Chakravarti, Guhya-samäja-tantra-tlkä, 31-32.

15. The stage of union (yuganaddha-krama, zung ‘jug gi rim pa) of the completion stage (rdzogs rim) is the fifth of the five stages in Nägärjuna’s text, the Five Stages (Panca- krama).

16. See, for example, Gadjin M. Nagao, “What Remains in Sünyatä: A Yogäcära Interpretation ofEmptiness,” in Mahäyäna Buddhist Meditation, ed. M. Kiyota (Honolulu, HI: University of Hawaii Press, 1978), 66-82, and David Malcolm Eckel, To See the Buddha (San Francisco: HarperSanFrancisco, 1992), 69-71.

. The Sgron gsal, Sde dge Bstan ‘gyur, Toh. no. 1785, Rgyud ha, vol. 30, 47.2- 48.4; Peking Bstan ‘gyur, Ötani 2650, vol. 60, 35.3-3-4.7; The Golden Bstan ‘gyur, vol. 30, 3235-3335; translated by Sraddhäkaravarman, Rin chen bzang po, SrTjnänäkara and ‘Gos Lhas btsas and revised by Nag po and ‘Gos Lhas btsas.

18. Pindl-krta-sädhana-vrtti-ratnävall (Mdor bsduspa’i sgrub thabs kyi ‘grelpa rin chen phreng ba = Rin chen phreng ba), Sde dge Bstan ‘gyur, Toh. no. 1826, Rgyud ci, vol. 36: 50.1-51.3; translated into Tibetan by Karmavajra. In his commentary on the Guhyasamäja Tantra entitled Kusumänjali (Sde dge Bstan ‘gyur, Toh. no. 1851, Rgyud ti, 463.5-465.4) Säntipa has quite a different discussion of this verse.

19. Säntipa, Rin chenphrengba (Ratnävali), Sde dge ed., 50.1-2.

20. Reading sgompa for sompa in the Sde dge edition.

21. Chakravarti, Guhya-samäja-tantra-tlkä, 31.

22. Candrakirti, Sgrongsal (Pradlpoddyotana), Sde dge ed., 47.2; Peking ed., 35.3.3-4; The Golden Bstan ‘gyur ed., 3235-6.

23. It is entitled Gsang ‘dus sgron gsal gyi bsdus don [[[Tibetan Buddhist Resource Center]], source code W11238].

24. Bu ston wrote commentaries on both the Pradlpoddyotana [the Dpal gsang ba ‘dus pa’i tlkkä sgron ma rab tu gsal ba = Sgron gsal bshad sbyar, The Collected Works of Bu-ston (New Delhi: International Academy of Indian Culture, 1967), vol. 9, 141-682] and on the Pindl-krama-sädhana [the Mdor byas ‘grel chen, The Collected Works of Bu-ston (New Delhi: International Academy of Indian Culture, 1967), vol. 9, 683—979], but he comments on our verse only in the latter work (708.6-710.3). In this latter commentary, Bu ston often consults Säntipa’s commentary on the Pindl-krama-sädhana, the Ratnävali. However, in writing about our verse, Bu ston’s text corresponds to the Tibetan version, not of the Ratnävali, but of the Pradlpoddyotana (Sgron gsal). Like the Sgron gsal (Sde dge ed., 47.2), Bu ston (708.6-7) says, “there is no essence to all the animate and inanimate

things,” and not “there are no animate and inanimate things,” as in the Sanskrit edition (Chakravarti, Guhya-samâja-tantra-tïkâ, 31) and in Sântipa’s Rin chen phreng ba (Sde dge ed., 50.1-2). And also parallel to the Sgron gsal (Sde dge ed., 47.7), in the hidden level of interpretation, Bu ston (709.5) has dngos po med pa and not just med pa, as in the Rin chen phreng ba (Sde dge ed., 50.6). There are only minor differences between his expla¬nation and the explanation of the Sgron gsal. For an English translation of Bu ston com¬mentary here, see Yoshimizu, “The Theoretical Basis of the bskyed rim,” 25-27.

25. The dissolution into the moon in Mdor byas ‘grel chen, 749.5-6.

26. Rgyud thams cad kyi rgyal po dpal gsang ba ‘dus pa’i rgya cher bshad pa sgron me gsal ba’i tshig donji bzhin ‘byed pa’i mchan gyi yang ‘grel, in The Collected Works (Gsung ‘bum) ofRje Tsong-kha-pa Blo-bzang-grags-pa (New Delhi: Ngawang Gelek Demo, 1977), vol. 6:180.2-186.6. Also in Dpalgsang ba ‘duspa’i rtsa rgyud ‘grelpa bzhi sbrags dang bcas pa (Lhasa: Zhol Printing House, made from block-prints carved in 1890), 55b6-57b6.

27. Rnam gzhag rimpa’i mam bshad dpal gsang ba ‘duspa’i gnad kyi don gsal ba, The Collected Works (Gsung ‘bum) ofRje Tsong-kha-pa Blo-bzang-grags-pa, zy vols. (New Delhi: Ngawang Gelek Demo, 1975-79) v°l- 9: 296.2.

28. Tsong kha pa, Sgron gsal mchan, New Delhi edition, 182.3-4; Zhol edition, 56b2.

29. Chakravarti, Guhya-sanuya-tantra-Gka, 31; Sde dge ed., 47.5.

30. Tsong kha pa, Sgron gsal mchan, New Delhi edition, 182.6; Zhol edition, 5663.

31. Chakravarti, Guhya-samâja-tantra-Gkâ, 31; Sde dge ed., 47.6.

32. The term “mind-only” (cittamâtra, sems tsam) itself appears only in the hidden level of interpretation, which Tsong kha pa explains in terms of the subtle body, and not of the fourfold meditation (New Delhi, 183.4; Zhol, 56b6).

33. Rgyud kyi rgyalpo dpal gsang ba ‘duspa’i rgya cher bshadpa sgron ma gsal ba’i dka’ gnas kyi mtha’ gcod rin chen myu gu, The Collected Works (Gsung ‘bum) ofRje Tsong-kha-pa Blo-bzang-grags-pa (New Delhi: Ngawang Gelek Demo, 1977), vol. 8, 207.6-208.2. 34. See Yael Bentor, “Identifying the Unnamed Opponents of Tsong-kha-pa and Mkhas-grub-rje 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, ed. Ronald M. Davidson and Christian K. Wedemeyer (Leiden: Brill, 2006), 185-200.

35. Bu ston, Mdor byas ‘grel chen, 749.5-6.

36. 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, The Collected Works (Gsung 'bum) oj the Lord Mkhas-grub Rje Dge-legs-dpal-bzang-po (New Delhi: Gurudeva, 1982), vol. 7, 209.2-4.

37. I would like to thank José Cabezon for pointing this out to me.

38. While in Nâgârjuna’s sâdhana, the Pindï-krama-sâdhana, our verse is recited during meditation on emptiness, in the Guhyasamaja Tantra, this verse is spoken as an explanation of the arising of the mind for enlightenment (bodhicitta). The commentar¬ies discuss both of these contexts.