***PLEASE NOTE THIS PAPER REPRESENTS WORK IN PROGRESS***

Thank you all for coming.

My goal here is to sketch, in rough outline, the relationship between svasaṃvedana or

“reflexive awareness” as a facet of epistemological theory on the one hand, and the stages of

yogic meditative practice on the other, as these are found in the works of Ratnākaraśānti, with

whom we’re all very familiar now thanks to Dan McNamara’s paper. In particular I will focus on

the Pith Instructions for the Ornament of the Middle Way (*Madhyamaka Alaṃkāra Upadeśa,

dbu ma rgyan gyi man ngag) and the Pith Instructions for the Perfection of Wisdom

(Prajñāpāramitā Upadeśa, sher phyin man ngag), neither of which is known to be extant in

Sanskrit. My operating hypothesis is that these texts, together with the Commentary on the

Ornament of the Middle Way (Madhyamaka Alaṃkāra Vṛtti, dbu ma rgyan gyi ‘grel pa),

represent a more or less continuous and coherent perspective; accordingly, I will frequently refer

to one text in order to illuminate the other, and vice versa.

In any case, as I will demonstrate, Ratnākaraśānti appears to maintain that Yogācāra

meditation can—and perhaps even ought best—to be understood in the context of tantric

practice. Furthermore, he implicitly justifies this stance by appealing to the nondual nature of

reflexive awareness, also known as “luminosity” (prakāśa, gsal ba), which forms the bridge

between Yogācāra theory and the Guhyasamāja Tantra, which is the main tantra commented

upon by the Ārya school at Vikramaśīla (including Ratnākaraśānti) and one of the primary

sources for generation stage (utpattikrama, bskyed rim) practice in Tibetan Buddhism,

particularly among the gSar ma schools. But in order to understand how this is so, it is first

necessary to say a few words about the four stages of Yogācāra.

These four stages are traditionally traced back to the Laṃkāvatāra Sūtra [LAS IX.256-

257]. In brief, the four stages as found in this sūtra may be described as follows:

1) The meditator realizes that all phenomena are “mind only,” and passes beyond the

conceptualization of objects as existing externally to the mind.

2) The meditator apprehends suchness, the nature of reality, and passes beyond “mind

only.”

3) Passing beyond “mind only,” the meditator abides in “non-appearance.”

4) Abiding in “non-appearance,” the meditator “sees the Mahāyāna,” whatever this

might mean.1

Ratnākaraśānti presents these four stages in a slightly different manner, closely following

(though not precisely identical to) Kamalaśīla’s commentary on these verses in the First

Bhāvanākrama.2 According to Ratnākaraśānti, the four stages are:

1) Apprehending the full range of phenomena.

2) Apprehending that the way in which phenomena exist is “mind only.”

3) Apprehending suchness, the nature of reality.

4) Seeing the Great Vehicle, which Ratnākaraśānti glosses as non-appearance.3

1 See Handout, number (1). 2 See Handout, number (2). Cf. Bentor (2000), 45ff. for an overview of the slight differences between Ratnākaraśānti and Kamalaśīla on this point. 3 This account is from the Madhyamakālaṃkāropadeśa. Ratnākaraśānti has the same account of the four stages in the Prajñāpāramitopadeśa, but explains them slightly differently (see below).

Leaving aside the differences among these accounts, there are two key points to understand.

First, there is a clear progression from the recognition of ontological idealism (or “mind only”) to

the realization of suchness. Second, the passage beyond “mind only,” concomitant with this

realization of suchness, necessitates the disappearance of phenomenological subject and object.

In his Pith Instructions for the Ornament of the Middle Way, Ratnākaraśānti glosses the

“full range of phenomena” (ji snyed yod pa) as the “two extremes of entities” (dngos po'i mtha'

gnyis). As an aside, it is worth noting that this is not how he glosses “the full range of

phenomena” in the sher phyin man ngag; in the latter text, he explains the first stage in terms of

the twelve āyatanas, that is, the six sensory bases and the six consciousnesses. The “two

extremes” may indicate an ontological valence in his presentation, i.e. existence and

nonexistence, that as we shall see is present in his commentary to the Guhyasamāja. In any case

this discrepancy may be reconciled if we consider the “two extremes” in terms of

phenomenological subject (the consciousness-āyatanas) and object (the sensory base-āyatana).

And this is, indeed, by far the most common manner in which Ratnākaraśānti glosses “duality.”

This is evident in the extended discussion of the second stage from his Pith Instructions for the

Perfection of Wisdom; as Bentor summarizes,

In the second stage the yogis reflect on the perception of all phenomena as products of mental-processes-only, which appear due to habitual tendencies of clinging to objects. Since objects grasped as external to the mind do not exist as they are conceptualized, their grasper cannot exist in that way either.4

As is well known, this argument—to the effect that, the realization that external objects

cannot exist apart from the internal subject, necessitates in turn the realization that the internal

4 See Handout, number (3). Cf. Bentor (2000), 42-43.

subject does not exist as an independent entity, either—is traceable to Vasubandhu. Less well

known is that this reasoning also appears in the Pramāṇavārttika of Dharmakīrti (PV 3.212-213),

who expands the argument to include a point about cognitive error (translations by John Dunne):

This part of awareness—namely, the one that is established such that it seems external—is different from the internal determination [which is the part of awareness that apprehends that apparently external part]. Awareness is not differentiated, but its appearance is differentiated into two. This being so, that dualistic appearance must be cognitive confusion. The nonexistence of one of the two in awareness eliminates the existence of both. Therefore, the emptiness of duality is the suchness of the awareness.5

I will return to this passage, and in particular to its definition of “the suchness of awareness” as

“the emptiness of duality.” But first, I would like to say a little more about the role of pramāṇa

theory in Ratnākaraśānti’s exegetical project.

* * *

In general, Ratnākaraśānti rhetorically positions himself as an exponent of Śāntarakṣita

and Kamalaśīla’s synthesis of Yogācāra and Madhyamaka analysis with pramāṇa theory, though

he is perhaps more comfortable than them with the Yogācāra project, on its own terms—and any

precise doxographical categorization is problematic, for the reasons Dan just outlined. In the dbu

ma rgyan gyi man ngag, he terms his approach the “Middle Way of the Three Natures” (rang

bzhin gsum gyi dbu ma), includes homages to both Asaṅga and Nāgārjuna, and spends much of

his time making pramāṇa-theoretical arguments. As noted above, he also clearly uses

5 See Handout, number (4). Cf. Dunne (2004), 406-408.

Kamalaśīla’s commentary on the four stages from the Laṃkāvatāra, from the First

Bhāvanākrama, as a touchstone.

In any case, as far as pramāṇa is concerned, Ratnākaraśānti generally follows the first few

generations of commentators on the Pramāṇavārttika, especially Devendrabuddhi and

Śākyabuddhi. The primary manner in which Ratnākaraśānti departs from Dharmakīrti is in terms

of his emphasis: Dharmakīrti spends most of his time arguing from a bāhyārthavāda or “External

Realist” perspective, but not at any great length from the highest, antarjñeyavāda (also called

vijñānavāda) or “Epistemic Idealist” level of analysis. Ratnākaraśānti, on the other hand—owing

perhaps to his affinity for Yogācāra, although again this point should not be overemphasized—is

primarily concerned with the Epistemic Idealist perspective.

This is a subtle yet important point, since it defines a shift in the terms of the discussion.

As we all know, Buddhist pramāṇa theorists accept only two instruments of knowledge:

perception and inference. At the External Realist level of analysis, the perceptual instrument

(pratyakṣa-pramāṇa) is identified with the objective phenomenal form or “apprehended aspect”

of a cognition (grāhyākāra). That is to say, we have knowledge of the world by means of the

mental representations that sensory contact produces in consciousness. However, at the

Epistemic Idealist level of analysis, the intentional structure of phenomenolgical duality itself is

understood to be nothing more than a type of cognitive error or distortion; therefore, any

intentional cognition is by definition unable to serve as a reliable epistemological instrument

(pramāṇa). As such, the only manner in which we have ultimately reliable information is by

means of nondual reflexive awareness (svasamvitti or svasamvedana).

To step back for a moment, it is not necessarily true that the phenomenal object cannot

serve as a perceptual instrument at the Epistemic Idealist level. In principle, what distinguishes

Epistemic Idealist discourse is its rejection of any extra-mental causes for sensory cognition. In

other words, it is at least theoretically conceivable that an Epistemic Idealist could maintain that

there are, in fact, no extra-mental entities, but, at the same time, that we are only afforded

reliable information about reality by means of an intentional, dualistic form of cognition.

(Something like this seems to be the interpretation of Epistemic Idealism made by Dan Arnold

and Christian Coseru, among others).

In theory, then, it is possible to frame Epistemic Idealist discourse entirely within the

“mind only” stage of Yogācāra, without any further reference to nonduality or the “suchness of

awareness.” In practice, however, this does not appear to be Dharmakīrti’s view, and it is

certainly not Ratnākaraśānti’s perspective. Recall the Pramāṇavārttika verse from above:

This part of awareness—namely, the one that is established such that it seems external—is different from the internal determination. Awareness is not differentiated, but its appearance is differentiated into two. This being so, that dualistic appearance must be cognitive confusion. The nonexistence of one of the two in awareness eliminates the existence of both (my emphasis). Therefore, the emptiness of duality is the suchness of the awareness.

For his part, concerning this very issue, Ratnākaraśānti writes:

The flaws which follow from the claim that blue-patches and so on are external are the same for a blue-patch which has the nature of consciousness, because there is no difference in the unacceptable conclusions that follow. The distinction between [external] objects and consciousness simply does not amount to anything at all.6

6 See Handout, number (5).

Time unfortunately prohibits a more thorough discussion of the issue, but the basic point is that

for Dharmakīrti, and, following him, Ratnākaraśānti, phenomenological duality—the bifurcation

of cognition into subjective and objective, or “apprehending” and “apprehended” aspects—is

nothing more than a type of cognitive error, that is produced, in the final analysis, by ignorance.

This is the “internal distortion” (antarupaplava) identified by Dharmakīrti at PV3.359-

362.7 In brief, the sensory cognition of e.g. a blue patch—or, more precisely, the appearance in

consciousness of a blue “apprehended aspect”—is distorted or mistaken, precisely insofar as the

mode of its appearance is dualistic; in other words, such a cognition is misleading to the extent

that it feels like something else, something outside the mind, is being “apprehended” by a first-

person or intentional subjectivity.

In any case, as Ratnākaraśānti writes:

The characteristic “blue” does not exist, because it cannot withstand logical analysis, but there is cognitive distortion due to the contaminating force of the “blue” psychological imprint. Since it arises in this way, the experience is distorted, and it is experienced as though one were experiencing something else [i.e. as something external to the mind].8

For this very reason, the objective or “apprehended aspect” cannot be considered an

ultimately reliable instrument of knowledge, defined as the type of pramāṇa that affords access

to reality as it really is, or enables the view of things just as they are (yathābhūtadarśana): in a

word, to suchness. As Dan McNamara has argued before elsewhere, the only ultimately reliable

instrument of knowledge is reflexive awareness. And why is that? Because reflexive awareness

7 Cf. Dunne (2004), 89n57 and 315-18. 8 See Handout, number (6).

is not structured by subject-object duality, and is therefore not contaminated by the “internal

distortion.”

This fact is amply attested in the literature, particularly in the commentaries of

Devendrabuddhi and Śākyabuddhi. For our purposes here today, though, their views on the

matter are less important than that of Ratnākaraśānti, who writes (this section immediately

follows the previous citation):

By contrast, there is nothing that can refute the luminous nature of reflexive awareness. Reflexive awareness is direct [mngon sum, *pratyakṣa], and it is an authentic experience, because there are no means of reliable knowledge apart from it. Thus it is established as a means of reliable knowledge.9

And:

Although their nature is luminosity, blue-patches and so on are false since they are harmed by analysis. But that luminosity is established as being real, because as an awareness that is free from distortion, it is a direct means of reliable knowledge.10

And, most simply:

Luminosity is a direct means of reliable knowledge, because there is no distortion in its nature.11

It is amply clear, then, that access to suchness, or the ultimate nature of reality, cannot in

principle be granted by a dualistic cognition—precisely insofar as such a cognition is, by

definition, distorted or mistaken, and, additionally, is prima facie evidence of the cognizer’s

ignorance. The state of “mind only” meditation might in theory be attainable within or by means

9 See Handout, number (7), first item. 10 See Handout, number (7), second item. 11 See Handout, number (7), third item.

of a dualistic cognition, but the realization of suchness is in principle impossible. To put it

slightly differently: reflexive awareness, or luminosity, is the only candidate for a pramāṇa by

means of which a meditator may have access to suchness (insofar as this may be spoken of as the

“object” of a particular type of meditation), precisely because it is nondual. Therefore,

interpretations of reflexive awareness that seek to cast it as the subjective aspect’s (dualistic)

apprehension of “itself,” or as dualistic intentional structure itself, obscure its most important

feature. Additionally, such interpretations render Ratnākaraśānti’s account of meditative practice,

or indeed any theory of nondual meditation, unintelligible.

Above all, this interpretation of reflexive awareness/luminosity is incompatible with his

understanding tantric meditation. But in order to understand why, it is necessary to return to the

discussion of the four stages from the sher phyin man ngag. As Yael Bentor has pointed out,

following the discussion of the four stages—wherein, it should be noted, Ratnākaraśānti

explicitly describes the third stage as “luminosity, empty of duality,” free from “all the false

conceptual marks of object and subject,” and the fourth stage as “non-dual, free of appearances

and apprehension”—Ratnākaraśānti relates this fourfold presentation, first to the Laṃkāvatāra,

and then, immediately afterward, to a verse from the Guhyasamāja:

The Gathering of Secrets teaches this exact thing [i.e. the same four stages] in one verse: When you investigate your own mind, All phenomena reside in the mind. Phenomena abide as the sky-vajra; There are no phenomena, and there is no nature of phenomena.//12

12 See Handout, item (8).

Ratnākaraśānti then explains that the stages of meditation can be described in the same

manner for both texts (noting that, on his reading, the first stage is only implied, rather

than directly stated, in both the Laṃkāvatāra and this verse from the Guhyasamāja):

The first stage is implicit in both [texts]: because, as long as yogis do not grasp that “This is all phenomena,” they are unable to apprehend their emptiness. Thus the first stage is the definitive resolution [yongs su gcod pa] that “This is all phenomena.” The second stage, which possesses appearances, is an extreme conviction [lhag par mos] in mind-only, which is empty of apprehended and apprehender. The third stage is the view as luminosity, in which the dualistic marks of phenomena do not appear, due to precisely that extreme conviction. The fourth stage is the vision, in that very state, [that arises] due to the utter non-appearance of all of the dualistic marks of both phenomena and the nature of the phenomena.13

He even goes so far as to gloss this verse from the Gathering of Secrets, word by word, in terms

of the fourfold meditation:

Moreover, “investigate the mind” refers to the second stage of yoga. “Reside” means that all phenomena are the appearance of one’s own mind, and “one’s own” is the achievement of the certainty: ‘My own mind, though nonexistent, appears.’ “Sky-vajra” refers to the two non-appearances. “Abiding as that” means abiding in the non-appearance of phenomena and the nature of phenomena, successively. Thus the non-appearance of the self-nature [rang bzhin] of phenomena is the third stage, and the non-appearance of the self-nature of the nature of phenomena is the fourth stage.14

Finally, as Bentor has noted, Ratnākaraśānti uses this fourfold schema—which, as I hope

to have convincingly demonstrated, is both intimately related and deeply indebted to his account

of pramāṇa theory—in his commentaries to the Guhyasamāja Tantra, itself. In addition to some

comments in the Kusumañjali, which there is unfortunately only time to mention in passing, 13 See Handout, number (8). 14 See Handout, number (9).

Ratnākaraśānti also uses the fourfold framework in his explanation of another famous verse from

the second chapter of the Guhyasamāja, which appears at the very beginning of the generation

phase in Ārya Nāgārjuna’s Piṇḍīkramasādhana:

abhāve bhāvanābhāvo bhāvanā naiva bhāvanā | iti bhāvo na bhāvaḥ syād bhāvanā nopalabhyate||

dngos po med pas sgom pa med/ bsgom par bya ba sgom pa min| de ltar dngos po dngos med pas/ sgom pa dmigs su med pa’o||15

Although this verse is (perhaps deliberately) obscure, one possible translation, following

Ārya Candrakīrti’s commentary, is: “When there are no entities (abhave), there is no meditation

(bhavanabhavo). Meditation is not even meditation. Thus, an entity could not be an entity;

meditation is not observed.” Ārya Candrakīrti glosses this verse in four ways, adopting the

common tantric hermeneutical technique of the “four methods” (tshul bzhi): literal, common,

secret, and ultimate. According to Bentor, however, Ārya Candrakīrti does something very

interesting: he relates these four levels of interpretation to the four stages of meditation from the

Laṃkāvatāra.

Commenting in turn upon Ārya Candrakīrti’s explanation of this verse, and focusing on

the second half of the first pada (“Meditation is not even meditation”), Ratnākaraśānti remarks

that, at the literal level, i.e. the first stage, there are no entities which exist or do not exist (please

note that the following are all highly provisional translations):

As all animate and inanimate entities do not exist, there is no meditation, since there is no object of meditation. “Meditation is not even meditation.” That which is meditation, is not even meditation, because of the nonexistence of meditation. Therefore there are no entities which exist or

15 See Handout, number (10).

do not exist, and no object of meditation, meditator, or meditation is observed. This is the literal meaning.16

As mentioned above, Ratnākaraśānti glosses the “full range of phenomena,” the object of

the first stage, as the “two extremes” of entities in the dbu ma rgyan gyi man ngag. Earlier I

pointed out that this is perhaps best understood as phenomenological subject and object, but

there is a clear ontological valence here—the gloss is in terms of existence and nonexistence. In

the absence of further research it is simply impossible to say anything more, but this does seem a

tantalizing clue as to the broader place of Madhyamaka analysis within the Ārya tradition of

tantric commentaries in general, and Ratnākaraśānti’s project in particular.

In any case, Ratnākaraśānti glosses the second, “common” level of interpretation as a

form of nondual, mind-only meditation. Even more interestingly, however, he appears to argue

that causality itself is no longer applicable at this level of analysis, presumably due to the

artificial and dualistic nature of the distinction between cause and effect:

Having cleared away all phenomena that are categorized as above, there is no meditation, “Conditioned and unconditioned phenomena are empty,” because such a view grasps at emptiness, and the marks of duality (tshan ma) are to be cleared away. “Meditation is not even meditation.” This means that a meditation which apprehends cause and effect is untenable, because there is neither cause nor effect; since intentions (smon pa) are to be eliminated, how much more so entities and so on? Thus there does not exist any intended effect which is an external phenomenal form, because there are no focal objects such as intentions. Therefore, meditation on external phenomenal forms is not observed, because they have the nature of mind.17

16 See Handout, number (11). 17 See Handout, number (12).

At the third level, luminosity—implicitly identified with ultimate truth—clears away the

last vestiges of relatively true, differentiated, phenomenal appearances. In this case, those

“appearances” form the body of the deity visualized during generation stage practice:

In terms of the nonexistence of entities, just as the bundles and so on do not exist, mind-only meditation does not exist, because it is incorrect and impermanent, and because the relative truth is to be cleared away. “Meditation is not even meditation.” [Here,] the meditation on the body of the deity—which has the nature of relative truth—is untenable, since it is to be purified by luminosity. Thus entities which are merely relative are not entities, because of the indivisibility of the two truths. Therefore, in terms of meditation, meditation on two truths is not observed, because of the impossibility of the observation. This is the secret [meaning].18

And at the fourth level, “non-observation” is glossed as the stage of “union” in which

neither ultimate nor relative truth is observed:

There is no meditation of ultimate truth in terms of the non-existence of entities, because they have been pure since the beginning. Nor is relative truth a fit object for meditation, because it is not [actually] true. Thus there are no entities to be meditated upon which have the nature of either ultimate or relative truth, since liberation is only by means of the vision of the two truths as nonexistent. Therefore, due to perfectly comprehending the stage of union, grasping at meditator, meditation, and object of meditation is not observed. This is the ultimate [meaning].19

Finally, following this discussion, but before continuing to his commentary on the next

verse, Ratnākaraśānti relates the three natures of Yogācāra theory to this verse: the imagined

nature to the nonexistence of entities, the dependent nature to the nonexistence of meditation,

and the perfected nature to the non-observation of (or in) meditation.20 But this will have to be a

topic for another occasion.

18 See Handout, number (13). 19 See Handout, number (14). 20 See Handout, number (15).

BIBLIOGRAPHY

Bentor, Yael. (2000) “Fourfold Meditation: Outer, Inner, Secret, and Suchness.” In Religion and Secular Culture in Tibet: Tibetan Studies II: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, 41-58. Leiden: International Association of Tibetan Studies.

————. (2010) “The Convergence of Theoretical and Practical Concerns in a Single Verse of

the Guhyasamāja Tantra.” In Tibetan Ritual (ed. José Cabezón), 89-102. New York: Oxford University Press.

Dharmakīrti. (1938) Pramāṇavārttikam by Ācārya Dharmakīrti. Edited by Rāhula Sāṅkṛtyāyana.

Patna: The Bihar and Orissa Research Society. Dunne, John. (2004) Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publications. Ratnākaraśānti, Madhyamakālaṃkāropadeśa (dbu ma rgyan gyi man ngag). sDe dge hi: vol.

138, pp. 446-461. ————, Prajñāpāramitopadeśa (shes rab kyi pha rol tu phyin pa’i man ngag = sher phyin

man ngag). sDe dge, mdo ‘grel hi: vol. 138, pp. 266-324. ————, Piṇḍīkṛtasādhanavṛttiratnāvalī (mdor bsdus pa’i sgrub thabs kyi ‘grel pa rin chen

phreng ba = rin chen phreng ba). sDe dge, rgyud ci: vol. 36, pp. 4-191.