Knowing Nothing: Candrakīrti and Yogic Perception

1. INTRODUCTION

Individuals who have reached advanced stages on the Buddhist path are renowned for being able to apprehend things beyond the ken of ordinary persons. A plethora of anecdotes, narratives and expository material in Indian Buddhist works, beginning with the earliest suttas and extending through the compositions of the Conservative (the so-called Hīnayāna) schools to the Mahāyāna scriptures and śāstras, depict and describe practitioners who have gained

perceptual and cognitive access to remote objects and otherwise inaccessible information, who know distant environments, the hidden or invisible in their immediate surroundings, and/or the fundamental nature of the world. The ability of these adepts to experience distinct phenomena, states of affairs, dimensions and supreme realities concealed to others is often attributed to their mastery of concentrative states and meditative techniques, and the

acquisition of refined levels of consciousness generated on their basis. Given the specifically Buddhist focus of these persons’ striving, their efforts tend to be ultimately aimed at the direct cognition of or immediate insight into their tradition’s conception of the final truth, this truth being

presented in early and Conservative Buddhism as, e.g., the four noble truths, and in the Madhyamaka school of Mahāyāna Buddhism as emptiness. Nāgārjuna (2nd/3rd c. CE), the founder of the Madhyamaka school, although without doubt convinced of an ultimate state of affairs, has little specific to say about

perception of the out-of-the-ordinary in

1 Research for this article was supported by the Austrian Science Foundation in the context of the FWF-Project S9805-G08. I am most grateful to Prof. Eli Franco, Dr. Dorji Wangchuk and Terry Chantler for carefully reading the present paper and for offering insightful comments and suggestions. I also thank Dr. Mudagamuwe Maithrimurthi for sharing with me his knowledge of the abhijñās.

his Mūlamadhyamakakārikā (henceforth MMK), and focuses its 447 verses primarily on refuting the existence of the phenomena known to the world, or, more particularly, on disproving the entities, categories and concepts accepted and taught by the Conservative Buddhist schools.2 With the exception of the very general reference in MMK 24.24 to the cultivation (bhāvanā) of the path (mārga),3 Nāgārjuna also does not mention or discuss in the MMK the means, such as

meditation techniques, for arriving at apprehension of the parokṣa, the imperceptible, or, as described by later scholars, the atyāntaparokṣa, the radically inaccessible. This dearth of references to methods and processes and his limited delineation of the result leave his stance on exactly what those who dare to appropriate and internalize his radical critique might in the end perceive, achieve or experience open to interpretation, and contribute to it

remaining a topic of debate among scholars. Although the complementary scrutiny of other writings attributed to him contributes to the illumination of his views, for more explicit and detailed statements about yogic perception and the objects of yogic perception in Madhyamaka it is necessary to examine the

works of later authors and commentators. The present paper will mainly concentrate on statements by Candrakīrti (600–650 CE) that address, and allude and relate to the topic of yogic perception. These can be found scattered throughout his works; I rely here on his commentary on the MMK, i.e., the Prasannapadā, his commentaries on Nāgārjuna’s Yuktiṣaṣṭikā and Śūnyatāsaptati,4 and on his independent work the Madhyamakāvatāra, together with its bhāṣya.

2 See Vetter 1982: 96, n. 21, where he considers MMK 7.4 to represent the view of a Sarvāstivāda opponent; MMK 9.1-2 and 9.6 that of a Pudgalavādin, possibly a Sāṃmitīya; MMK 17.1-11 to represent the view of an opponent who would at least later be termed a Sautrāntika; and MMK 17.12-20 to possibly be

that of a Sāṃmitīya. See also Kragh 2006, Chapter 3 for more detailed discussion concerning references to the opponents dealt with in MMK 17.1-20. Nāgārjuna’s approach in the MMK is apophatic, but he does refer to and even characterize (primarily negatively) the ultimate state (see, e.g., MMK 18.9);

important references to the highest truth and nirvāṇa in the MMK have been noted and discussed in Vetter 1982. 3 MMK 24.24: svābhāvye sati mārgasya bhāvanā nopapadyate | athāsau bhāvyate mārgaḥ svābhāvyaṃ te na vidyate ||. The mārga is also referred to in MMK 24.25 and 40, bhāvanā in 24.27. 4 Candrakīrti’s authorship of the Śūnyatāsaptativr̥tti is not completely beyond doubt. The work is not mentioned in Indian literature, and only the colophon of the Tibetan translation of the work (in all four Canonical editions of the Tanjur) and later Ti

2. CANDRAKĪRTI ON EXTRASENSORY PERCEPTION

Before proceeding to a presentation of the Madhyamaka understanding of the ultimate and an investigation of Candrakīrti’s views on its perception, it might be noted that Candrakīrti also makes allowance for more general types of extrasensory knowledge. One interesting reference to the wonder of yogic

perception in its wider sense is encountered at Madhyamakāvatāra 3.11, where Candrakīrti recites some of the attainments gained by the practitioner who has reached and dwells on the third Bodhisattva level of awakening, the bhūmi called prabhākarī, the “Illuminating.” 5 He states that the Bodhisattva who

abides on this level, in addition to completely destroying his craving and hatred and perfecting the dhyānas,6 attains supernormal knowledge, or, as it is sometimes interpreted, “direct knowledge” (mngon shes, *abhijñā). In the commentary on his verse, it is made clear that with his mention of supernormal knowledge he intends a traditional five-fold group of abhijñās, four of which might broadly be seen as types of clairvoyance.7 These five types of supranormal capability are generally said to be produced on the basis of the practitioner having reached the fourth dhyāna, an intensified concentrative state characterized by one-pointedness of mind and emotional equanimity. The first of the five abhijñās referred to in Candrakīrti’s commentary consists in

the ability to perform various types of paranor betan historians name him as its author. For internal criteria that nonetheless appear to support attribution of the work to him, see Erb 1997: 1-10. 5 MA 3.11 (MABhed 53.17-20): sa der rgyal sras bsam gtan mngon shes dang || ’dod chags zhe sdang yongs su zad par ’gyur || des kyang rtag tu ’jig rten pa yi ni || ’dod pa’i ’dod chags ’joms par nus par ’gyur ||. 6 The word bsam gtan (*dhyāna) is explained

in the bhāṣya as intending the four dhyānas, the four samāpattis, and the four apramāṇas. 7 For references to the group of five abhijñās, cf. Lamotte 1976: 1814; on the six abhijñās, cf. 1809ff. Lamotte notes (p. 1809) that the first five are usually given in the order r̥ddhividhijñāna (Pāli: iddhividha) / r̥ddhiviṣayajñāna; divyaśrotra (Pāli: dibbasota); cetaḥparyāyajñāna (Pāli: cetopariyañāṇa) / paracittajñāna; pūrvanirvāsānusmr̥tijñāna (Pāli: pubbe nivāsānussatiñāṇa); cyutyupapādajñāna (Pāli: sattānaṃ cutūpapātañāṇa), also known as divyacakṣus (Pāli: dibbacakkhu). Cf. also de La Vallée Poussin 1931; Lindquist 1935; Ñāṇamoli 1995: 37 (with references to Majjhima Nikāya suttas 6, 73, 77 and 108); AK 7.42-56 and AKBh thereon; Dayal 1932: 106ff.; Gethin 1998: 185f.; Gethin 2001: 84. On methods for developing the iddhis and the abhijñās, see Visuddhimagga chapters 12 and 13 (iddhividhāniddeso and abhiññāniddeso) and AKBh on 7.43d; see also Gethin 2001: 101f.

mal feats (r̥ddhi), such as being able to manifest mind-made bodies, to pass through physical matter such as walls and mountains, to fly, to walk on water and dive into the earth, to blaze like fire and shower down rain from oneself, and to touch the sun and the moon. The second abhijñā mentioned is the divine ear (divyaśrotra), by way of which the yogin is able to hear any sounds, divine or human, that he wishes to listen to. The third abhijñā enables him to know the state of mind of other beings (paracittajñāna), the fourth, to recollect millions of his previous lives in great detail (pūrvanivāsānusmr̥tijñāna). With the fifth supranormal achievement, that of the divine eye (divyacakṣus), he is able to see beings dying and being reborn,

and knows the wholesome or unwholesome karma that takes them to their respective good or difficult destinations. As astounding and fascinating as these powers and supernormal perceptions might be, Candrakīrti has nothing special to say about them himself, choosing instead to elaborate on them in his bhāṣya

by citing verbatim the Daśabhūmikasūtra’s brief but detailed account of the five abhijñās.8 His interest in them is exhausted in this account. As is obvious from Candrakīrti’s reliance on a Mahāyāna sūtra for their description, these five abhijñās are not unique to the Madhyamaka school; we are, in fact, familiar with presentations of them in Canonical and Abhidharma works, and two of them, the recollection of past lives and the divine eye, figure in

a number of Canonical portrayals of the Buddha’s own awakening.9 Individual abhijñās, explained as resulting automatically upon attainment of the fourth dhyāna (as in the case of the Buddha or persons who trained in them in previous lives) or as needing to be developed through effort by the yogin on the basis of this dhyāna,10 were viewed early on as extraordinary but mundane types of knowledge because they did not on their own accomplish release from saṃsāra for the practitioner, even though they might be conducive

8 De La Vallée Poussin presents the Sanskrit text of the Daśabhūmikasūtra citation in an appendix to his translation of MA chapter 3; see MABhtr 1907: 305-307. For the section cited, see also Rahder 1926: 34-36 (= section M). 9 For suttas in which these two abhijñās do not appear, see Schmithausen 1981: 221, n. 75. 10 Cf. Gethin 2001: 102. Cp. AKBh on 7.44b. On the dissociation of liberation from attainment of the dhyānas in some Canonical texts, see Schmithausen 1981: 219222.

to it.11 Canonical descriptions of the abhijñās in the context of the liberation process therefore usually included a further item, termed “knowledge of the destruction of the taints” (Pāli: āsavakkhayañāṇa, Sanskrit: āsravakṣayajñāna), the “taints” being [[[craving]] for] sensual pleasures (kāma), [[[craving]] for] existence (bhava), and ignorance (avidyā).12 This abhijñā came to be known as the supramundane abhijñā, for it informed of one’s attainment of freedom

from birth and death, of one’s nirvāṇa, and in the stereotypical account of the attainment of liberation has as a main component the insight that effects liberation.13 It is probable that two of the abhijñās included in the Canonical liberation accounts, viz., the recalling of former lifetimes and the witnessing of beings propelled by their earlier actions to new existences, were considered to provide experiential confirmation of soteriologically relevant truths, especially the truths of suffering and the origin of suffering, and in this way to contribute to the liberation process. Both Canonical

and post-Canonical authors also acknowledged the usefulness of other abhijñās, such as the ability to read others’ minds and the capacity to perform miracles, for benefitting ordinary persons, especially for converting them to Buddhism.14 Transferred to the Mahāyānist Bodhisattva context, the first five abhijñās – the sixth either reserved for Buddhahood or revised inasmuch as the end of the taints would deliver the Bodhisattva to a premature nirvāṇa and thus abruptly end his career15 – must have been interpreted as serving to deepen the adept’s 11 See Ñāṇamoli 1995: 37; cp. the

discussion and classifications in AKBh chapter 7 ad verse 42. See also Schmithausen 1981: 221f., where he suggests that the abhijñās may have been considered especially necessary in the case of the Buddha’s original discovery of the Four Noble Truths. 12 De La Vallée Poussin (1931: 338) remarks: “À ces cinq savoirs, fut ajouté un sixième: la connaissance que prend le saint de sa sainteté. Le caractère scolastique de cette invention n’est pas douteux.” 13 For the “stereotypical account,” see Schmithausen 1981: 203-205. On āsavakkhayañāṇa in the context of the stereotypical account, see Schmithausen 1981:

204, n. 16. 14 On Canonical views regarding the performance of miraculous feats, see, e.g., Gethin 2001: 97-101. Cf. AKBh on AK 7.47 regarding the value of miracles and mindreading for conversion; note also Granoff 1996 for problems connected with the performance of miracles. 15 The Mahāprajñāpāramitāśāstra, for example, distinguishes between an incomplete and complete āsravakṣayajñāna in order to explain statements in Prajñāpāramitā literature that connect Bodhisattvas with the sixth abhijñā. In the case of the incomplete form, the kleśas are stopped, but the vāsanās are not; see Lamotte 1976:

experience and strengthen his dedication to reach the final goal of Buddhahood, and as being of use in augmenting both his desire and ability to inspire and aid others. It is difficult to know whether Candrakīrti’s single-word reference to the abhijñās in his Madhyamakāvatāra verse and his uncommented citing of the Daśabhūmikasūtra indicate much more than a tipping of his hat to tradition; their mention may demonstrate his acknowledgement of the view

that the acquisition and employment of miraculous powers serve the Bodhisattva’s programme of helping other beings. However, his disinterest in further elaborating the five abhijñās signals that regardless of their value as useful sideeffects of the Bodhisattva’s endeavour, for him they are of minor importance owing to their negligible soteriological value, both individually and collectively having little, if anything, to contribute to the actual

of liberation. Yet like the early authors whose inclusion of a sixth abhijñā was inspired by a primary concern with knowledge with soteriological function, Candrakīrti’s main interest is in a type of knowledge that can be classified as supramundane and that provides the insight which breaks one out of, as the texts have it, the “jail of saṃsāra.” Writing nearly a millennium after the Canonical authors composed their accounts of the Buddha’s and his disciples’

liberation process, Candrakīrti, however, does not assert that the escape from repeated birth and death is effected through meditative stabilization in the fourth dhyāna and subsequent profound insight into the four noble truths accompanied by vanquishment of the taints. He declares rather that it is brought about by profound insight into the emptiness (śūnyatā) of things.

3. THE MĀDHYAMIKA’S NIRVĀṆA: AN ILLUSION?

In brief, Candrakīrti propounds the view that the world, including the subject perceiving and experiencing the world, is of an illusionary na 1816ff. Candrakīrti explicitly refers to āsravakṣayajñāna when he recites and comments on the ten powers (bala) of a buddha in MA 12.21 (cf. MABhed 369.13:

zag rnam zad pa mkhyen stobs); in MA 12.31 he declares that it informs the newly accomplished buddha of the destruction of the kleśas together with their vāsanās. Included in the stock list of the ten powers in MA 12.21 is the ability to recall past lives (sngon gnas dran pa mkhyen pa, *pūrvanivāsānusmr̥tijñāna) and the knowledge of the birth and death of beings (’chi ’pho skye blo, *cyutyupapādajñāna).

ture; it may be appropriate to refer to his view in this respect as one of metaphysical illusionism.16 According to him, the phenomena of the world, or universe, which appear and seem to be real, in actuality do not exist. He and others of the Madhyamaka tradition do admit that the things of the world

appear to ordinary, unawakened persons, but they deny that these things truly are as they appear to be, i.e., real as opposed to unreal. The Mādhyamikas maintain that the things of the world are empty of a real nature that would support or justify any claim to their being ontologically existent. Phenomena must be empty of a real nature, of an own-being (svabhāva), the Mādhyamika argues, because they arise in dependence (pratītyasamutpāda) upon other things;

whatever arises in dependence, in being reliant on something else and thus not capable of existing without the other’s support, obviously does not exist of its own accord, by its own nature. Would things exist on their own, i.e., be real, they could as a consequence neither arise nor perish, for a real thing, a thing with its “own” being (sva-bhāva), would on account of this not require causes for it to come into being or to pass out of being; it would not arise

in dependence on something else nor decay or vanish due to the influence of some other factor. Such an entity would exist forever, and change would be impossible. That the phenomena experienced by the unawakened are indeed apprehended to arise in dependence, and to change, reveals that they are empty of an own-being, and thus bereft of true existence. Their arising in dependence translates into not truly existing, to not actually arising in dependence. The merely apparent existence of the things of the world therefore inspires the Mādhyamikas to compare them to, among other illusory phenomena, the objects apprehended in dreams and mirages, or conjured by a magician. Even though such objects appear and seem to be real during the dream, on a hot day in the desert, or, in the case of a magical illusion, while one beholds the magic show, the elephants in the dream, the water in the mirage and the beautiful

damsel produced by the magician are empty of real existence and do not actually exist. Upon awakening from the dream, approaching the mirage for water, or seeing the magician dissolve the damsel, one relinquishes − even though their reality had been taken for granted until then − all ideas of the existence of these objects. Like these illusory objects, the dependently-arisen phenomena

16 On Nāgārjuna as a propounder of metaphysical illusionism, see Oetke 2007: 16ff.

of the world that are unquestioningly believed to be genuine by the unenlightened have only an apparent reality, a semblance of, a superficial, “fake” realness. The teaching of the emptiness of things thus discloses the deceptive nature of worldly phenomena: they are mere fictions, unreal appearances

masquerading, so to speak, as real things. As fictions they are actually no things, ontologically nothing, and thus in the final analysis, inexistent. According to the Mādhyamikas, no thing has ever really existed and no thing will ever come into existence. The cycling through repeated births and deaths that constitute the saṃsāric wandering − this too has never really occurred. Thus the question arises: If, according to the Mādhyamikas, saṃsāra is

actually a fiction, what, then, of nirvāṇa? Can one escape from something that never was? Does nirvāṇa, unlike saṃsāra, exist? Or is liberation also a fiction, and the counsel to strive for it, a Mahāyānist joke? Aid for answering these questions can be found in MMK chapter 25, the “Examination of

nirvāṇa,” and in Candrakīrti’s commentary on its individual kārikās. It should be mentioned that large circles within early Buddhism and some of the Conservative schools did indeed maintain a positively characterized nirvāṇa. A number of passages in early Buddhist works present nirvāṇa as an

unconditioned and enduring state or sphere, and as such as similar to the higher spheres of yogic concentration but radically transcending them; nirvāṇa appears in these specific cases to have been conceived as a metaphysical, or rather, meta-physical, world-transcending dimension into which the liberated mind/self would enter.17 The Theravādins, in spite of their dogmatic

17 For references, see Frauwallner 1953: 226f. [= Frauwallner 1984: 178f.]; Schmithausen 1969: 158f. Schmithausen (1969: 159) remarks that the occurrence of a far greater number of passages negatively characterizing nirvāṇa as the process or state of the termination of suffering derives from the fact that the positive nature of nirvāṇa, as it is in itself, was beyond the reach of thought and speech and experienceable only in a meditative state; positive

statements might indicate its not being nothing, but detailed speculation, given the nature of language, was dubious. He adds that such speculation on the nature of nirvāṇa was superfluous for the goal of Buddhism: inasmuch as all of worldly existence, on account of its impermanence, was considered to be

suffering, liberation from this suffering sufficed as the goal, regardless of whether it might be characterized as a positive state or as pure annihilation. The tradition thus recognized the existence of a positive though indescribable dimension as the “place” of liberation, but particularly emphasized its negative function as the ending of suffering.

denial of the existence of a self that might enter or experience nirvāṇa, postulated it as a positive, unconditioned, and enduring − and to that extent joyful − entity. Nirvāṇa in the Sarvāstivāda school has the unique characteristic of being a hypostatized elimination or stopping of the defilements and

suffering, an existing “non-being,” and was thus considered a real, unconditioned and permanent entity.18 N āgārjuna addresses the issue of an existent nirvāṇa in the fourth, fifth and sixth kārikās of MMK 25,19 commencing by unhesitatingly rejecting the possibility. He argues that if nirvāṇa would be an

existent thing, it a) would have to be characterized by aging and death (jarāmaraṇa), b) would have to be conditioned (saṃskr̥ta) and c) would have to be reliant on something else (upādāya), since all existent things have the characteristics of aging and death, are conditioned and are reliant. No Buddhist

would accept a nirvāṇa so characterized. Nāgārjuna likewise rejects the view that nirvāṇa is non-existence (abhāva).20 The equating of nirvāṇa and non-existence was, however, not completely foreign to Buddhism, for the Sautrāntika school did assert a nirvāṇa − at least an ontological nirvāṇa − that is mere non-existence. The Sautrāntika nirvāṇa is exhausted in its designation: it is solely the name for the fact that the emotional and intellectual defilements, and suffering, no longer arise, onto

18 Even though the Sarvāstivādins’ presuppositions that individual existence ends with the death of the liberated person and that an ātman (which might continue) does not exist relegated the spiritual experience of nirvāṇa without remainder to mere annihilation, the school did make room for the liberative

effects of nirvāṇa prior to death. These occurred in the form of a consecutive separation from the defilements brought about by religious praxis and by pratisaṃkhyānirodha (cessation resulting from consideration/insight, equated by the Sarvāstivādins with nirvāṇa), which of necessity was viewed as a succession of real, existent pratisaṃkhyānirodhas, or “nirvāṇas,” equivalent in number to the number of defilements removed. See, e.g., Schmithausen 1969: 161f.; Cox 1994; Cox 1995: 87f., 90f., 323 n. 72. 19 MMK 25.4: bhāvas tāvan na nirvāṇaṃ jarāmaraṇalakṣaṇam | prasajyetāsti bhāvo hi na jarāmaraṇaṃ vinā ||. MMK 25.5: bhāvaś ca yadi nirvāṇaṃ nirvāṇaṃ saṃskr̥taṃ bhavet | nāsaṃskr̥to vidyate hi bhāvaḥ kvacana kaścana || (pāda c emended following MacDonald 2007: 40f.). MMK 25.6: bhāvaś ca yadi nirvāṇam anupādāya tat katham | nirvāṇaṃ nānupādāya kaścid bhāvo hi vidyate ||. 20 He rejects this possibility in MMK 25.7-8. 25.7: bhāvo yadi na nirvāṇam abhāvaḥ kiṃ bhaviṣyati | nirvāṇaṃ yatra bhāvo na nābhāvas tatra vidyate ||. 25.8: yady abhāvaś ca nirvāṇam anupādāya tat katham | nirvāṇaṃ na hy abhāvo ’sti yo ’nupādāya vidyate ||.

logically nothing at all.21 Explicating Nāgārjuna’s rejection of nirvāṇa conceived as non-existence, Candrakīrti states that in the world a thing is termed non-existent when it gives up its own-being and becomes otherwise, i.e., becomes other than existent.22 But since nirvāṇa was never established as something that exists, it cannot relinquish its existence and become otherwise; that is, not having obtained the necessary prerequisite of having the state

of a thing, it is not in a position to abandon this state and become inexistent. Speaking to the Sautrāntika view of nirvāṇa as the cessation and thus end of the defilements, as their “having become otherwise,” Candrakīrti declares that if the nonexistence of the defilements23 is nirvāṇa, then the impermanence of the defilements (to be understood as their momentary perishing in the saṃsāric continuum) will have to be accepted as nirvāṇa. This is

definitely not accepted by the Sautrāntikas, asserts Candrakīrti, because it would entail that liberation is automatically achieved, without any effort on the part of the practitioner.24 Yet even though the Mādhyamikas reject a nirvāṇa conceived and classified either as an existent or as non-existence (as well as one characterized by both existence and non-existence, and by neither existence nor non-existence),25 they continue to speak of nirvāṇa. This is

confirmed, for instance, by Candrakīrti’s commentary on MMK 25.10. In 21 The Sautrāntika nirvāṇa as a spiritual event consists in liberation from the defilements and suffering existence; nirvāṇa without remainder thus expresses itself as the complete destruction, i.e., the end, of the body-mind continuum. 22 Candrakīrti’s statement here relates to MMK 25.7. See de Jong 1978: 245, entry for p. 527.6 (the sentences are missing from PsPed

527): iha hi bhāvaḥ svabhāvaparityāgād anyathā bhavann abhāva iti vyapadiśyate | yatra ca pakṣe nirvāṇaṃ bhāvo na bhavati vihitadoṣatvāt tatra pakṣe ’bhāvo ’pi nirvāṇaṃ na bhavati bhāvasvarūpeṇāsiddharūpasyābhāvarūpatānupapatter iti abhiprāyaḥ |. 23 Candrakīrti mentions birth (janman) along with the defilements (kleśa). 24 In his commentary on Yuktiṣaṣṭikā 4cd, Candrakīrti informs an opponent who holds that saṃsāra, specified as the skandhas, i.e.,

the body-mind continuum, exists and that its cessation is nirvāṇa understood as non-existence (abhāva), that such is indeed taught, but it is merely a strategy. The teaching is necessary because the unenlightened have been habituated since beginningless time to the belief that things really exist, and are not able to turn away from attachment to them without being told, as an antidote, that nirvāṇa is the cessation of saṃsāra. In coming to believe that the attainment of nirvāṇa involves great happiness, they are able to turn away from the pleasant things of saṃsāra, not to mention the disagreeable. 25 Cf. MMK 25.11-15 and 25.16-17 and Candrakīrti’s commentary thereon.

the kārikā, Nāgārjuna makes reference to a statement of the Buddha’s in which he has proclaimed that being (bhava) and non-being (vibhava) have to be abandoned, and from this Nāgārjuna concludes that nirvāṇa is appropriate neither as existence (bhāva) nor as non-existence (abhāva).26 Citing a sentence

from a Canonical work which negates that release from existence can be found by way of being or non-being,27 Candrakīrti declares that even though existence and non-existence are to be abandoned, the Buddha did not state that nirvāṇa is to be abandoned; he rather asserted that it is not to be abandoned.28 Following Candrakīrti’s interpretation of the MMK on nirvāṇa thus far, this would mean that the practitioner who has come to understand that

the world and even what was thought to be the escape from it are neither existing nor not existing (nor both nor neither) – this practitioner is nevertheless to continue to strive for liberation, for nirvāṇa. One might be inclined to interpret this call to continued effort from a subjective point of view, as meaning that even though the Mādhyamikas reject an ontologically existent nirvāṇa, and even though they equally reject nirvāṇa as the cessation of

an ontologically existent saṃsāra, they do accept nirvāṇa as a spiritual event. As an event it will belong to the conventional level, but as the paramount and decisive spiritual event it will effect the practitioner’s release from repeated birth and death, which are ultimately unreal but experienced as real until the event occurs. It is, as stated earlier, a profound insight, sharpened, deepened and solidified by meditative concentration, which is said to

effect the release. In Yuktiṣaṣṭikā 4cd, Nāgārjuna declares that the thorough knowledge (parijñā) of existence and non-existence is the liberating factor.29 In his commentary on this half-verse, Candrakīrti explains that because existence and non-existence are mutually dependent, they are not established by own-nature, i.e., they cannot exist in reality (for 26 MMK 25.10: prahāṇaṃ cābravīc chāstā bhavasya vibhavasya ca |

tasmān na bhāvo nābhāvo nirvāṇam iti yujyate ||. 27 PsPed 530.7: tatra sūtra uktam | ye kecid bhikṣavo bhavena bhavasya niḥsaraṇaṃ paryeṣante vibhavena vā ’parijñānaṃ tat teṣām iti |. De La Vallée Poussin (PsPed 530, n. 4) determines the text closest to the sūtra cited by Candrakīrti to be attested in the Udāna (p. 33, iii.10). 28 PsPed 530.8-9: na caitan nirvāṇaṃ prahātavyam uktaṃ bhagavatā kiṃ tarhy aprahātavyam |. 29 YṢ 4cd: dngos dang dngos med yongs

shes pas || bdag nyid chen po rnam par grol ||. I rely in this paper on Scherrer-Schaub’s edition of the YṢ as contained in the YṢV.

whatever is dependent cannot really exist), but the spiritually immature do not know this and, conceiving existence and non-existence and therewith

engendering desire and other defilements in regard to the two, they are bound and doomed to wander in saṃsāra.30 Awareness of the lack of real existence and non-existence, on the other hand, has the power to ultimately terminate the continuum of desire and other defilements because it jettisons the

objective basis onto which desire is projected. Thorough knowledge of the non-existence of both existence and non-existence is on account of this potent enough to break the bonds of the cycle of birth and death and deliver one from saṃsāra; it is thus suitable as a means of liberation. That thought and

conceptual activity have no part to play in this thorough knowing is indicated in Candrakīrti’s commentary to Yuktiṣaṣṭikā 4cd, where he describes it as having the nature of the nonimagining of an own-being of existence and non-existence. More epistemologically, he equates it with the non-perception of

existence and non-existence.31 It thus appears that for him thorough knowledge is the result of cultivation of the understanding that nothing exists, and involves, conceptually, the ceasing of all conceptualizing of and in regard to existence and its contingent opposite, and perceptually, the nonapprehension

of these two, i.e., of any thing or any lack of existence predicated upon a thing. When the yogin as “knower” is without the concept of, or apprehension of any of the things accepted as existing or not existing by the world or by the other Buddhist schools, the “object” of the thorough knowing must exclude all

possible phenomenal entities and non-entities. The object, conventionally speaking, is the truth behind the veil; the yogin’s thorough knowing characterized as the nonapprehension of existence or non-existence bespeaks a penetration of the world of fictions to its true nature, a nature which is untouched by conceptuality and stripped of the manifoldness of the illusion. It is a

30 Related text and French translation in Scherrer-Schaub 1991: 32.9-23 and 132-134 (I rely on Scherrer-Schaub’s edition of the YṢV in this paper). See alternatively Loizzo 2007: 259.6-260.6 and 140f. (Loizzo’s YṢ and YṢV translation is often unreliable). 31 … yongs su shes pa dngos po dang dngos po med pa’i rang gi ngo bo la sogs pa yongs su mi rtog pa’i rang bzhin … (Scherrer-Schaub 1991: 32.12-14; Loizzo 2007: 259.7-9); dngos po dang dngos po med pa mi dmigs pa ni ... (Scherrer-Schaub 1991: 32.23-24; Loizzo 2007: 260.6-7).

nature described in Madhyamaka works as peaceful (śānta) and the pacification of all objective manifoldness and all manifold conceptual and verbal activity (praprañcopaśama),32 and is such because nothing has ever arisen to disturb its calm: nothing has ever come into being, and nothing has ever ceased. Thus even though it is neither a “thing” nor a dependently conceived “non-thing,” the object of the thorough knowledge, viz., the true nature of the dependently originated and the dependently designated, is appropriate to be understood as the Mādhyamika’s ontological nirvāṇa. This true nature of the world coincides with nirvāṇa conceived as the removal of the defilements and the abandoning of all suffering existence because like all other things, the defilements and suffering, in never having arisen, have always been “abandoned.” Similar to the traditionally described nirvāṇa,33 the Madhyamaka nirvāṇa is set forth as

the pacification of all manifoldness, but in contradistinction to the previously mentioned interpretations of nirvāṇa, which envisioned it as an existent and enduring dimension or entity removed from the world, as an existent non-existent, or as the stopping of real defilements and a real personal continuum,

the Madhyamaka nirvāṇa is the world itself – in its innate and eternal state of peaceful non-arising. As the true nature of the world and the phenomena constituting it, it is not even, as the other schools’ nirvāṇa is, something to be attained through escape from the world, for it is already ontologically anticipated in things themselves and merely requires insight into this fact.34 The old opposition between nirvāṇa and saṃsāra is replaced in

32 Cf. MMK 7.16, 18.9, 25.24. 33 Cf. Aṅguttara Nikāya II.163 where nirvāṇa is characterized as papañcavūpasama. 34 Cf. Vetter 1982: 92f.: “Ich weise hier nur darauf hin, dass das von Nāgārjuna als Ziel genannte Nirvāṇa kein jenseitiger Ort ist, auch kein isolierter Zustand in der Welt, auch kein

Nichtmehrsein von etwas Besonderem, sondern die Welt selbst, insofern sie ihrer Bestimmtheiten und damit Bedingtheiten entkleidet und darum nicht mehr als solche wahrnehmbar ist.” (“I will here only point out that the nirvāṇa named by Nāgārjuna as the goal is not a place beyond, not an isolated state in the

world, also not the being no more of something particular; [it is] rather the world itself insofar as it is stripped of its determinacies and with that its conditionalities and therefore no longer perceptible as such.”). Cf. MMK 25.9: ya ājavaṃjavībhāva upādāya pratītya vā | so ’pratītyānupādāya nirvāṇam upadiśyate ||. See also Vetter 1982: 93, where he asserts that the Madhyamaka interpretation of nirvāṇa assures its definiteness: “... diese Endgültigkeit kann nur dadurch garantiert werden, dass es schon immer nur das Nirvāṇa gibt und dass die Welt nur eine falsche Vorstellung

Madhyamaka with an identification of nirvāṇa and saṃsāra, or rather with an identification of nirvāṇa and the true nature of saṃsāra. nirvāṇa as a spiritual event involves seeing through the world, the manifoldness of existence, such that its true nature is experienced. It is against this larger background that the seemingly paradoxical statements found in Madhyamaka texts as well as in Prajñāpāramitā and Mahāyāna literature in general which state

that the yogin sees the ultimate by not-seeing, or that “non-seeing is seeing” are to be understood.35 The knowing of the true nature of things, of the ultimate peacefulness of existence that has always been at its heart, or as the texts sometimes refer to it, of “thusness” (tattva), is a knowing that is without objects or appearances, one in which the yogin does not apprehend any thing. To dwell in a meditative state in which nothing appears is to see

reality. In his commentary on Yuktiṣaṣṭikā 6cd,36 Candrakīrti also defines nirvāṇa via the seeming paradox: he asserts that the thorough knowing of the non-arising of a real nature of existence which occurs by way of non thorough-knowing, is said to be nirvāṇa.37

4. PERCEPTION OF EMPTINESS ACCORDING TO A POST-CANDRAKĪRTI MĀDHYAMIKA

As realization of the ultimate was deemed to be direct and unmediated, with the rise of the Buddhist epistemological-logical tradition Mādhya ist.” (“This definitiveness can only be guaranteed when there was always solely nirvāṇa and the world is only a wrong idea.”). 35 See, e.g., MABhed

229.18-20 (MABhtr 1911: 279): rnal ’byor pa ’phags pa’i lam mngon du mdzad par gyur pas ma gzigs pa’i tshul gyis de kho na nyid gzigs pa dag gis … . Cf. also PsPed 265.3-5. The author of the Tarkajvālā quotes the sentence mthong ba med pa ni de nyid mthong ba’o (similar to the oft-cited adarśanaṃ bhagavan

sarvadharmāṇāṃ darśanaṃ saṃyagdarśanam); see Heitmann 2004: 98f. and 99, n. 3. On interpretations of such statements, see Keira 2004: 99, n. 151. 36 YṢ 6cd: parijñānaṃ bhavasyaiva nirvāṇam iti kathyate || (Sanskrit cited in ScherrerSchaub 1991: 144, n. 125). Tibetan text and translation in Scherrer-Schaub

1991: 37.21-22 and 146; Loizzo 2007: 268.1-2 and 147. 37 Text of the entire relevant passage and translation in Scherrer-Schaub 1991: 37.2338.2 and 146-147; Loizzo 2007: 268.3-6 and 147. See also Scherrer-Schaub p. 146, n. 129, where she remarks that the aparijñāna defined as nirvāṇa may be best interpreted as a state of consciousness without subject or object. Note that Candrakīrti has defined existence (bhava, srid pa) as the five appropriated skandhas in his commentary to Yuktiṣaṣṭikā 6ab.

mikas felt increasingly behooved to explain how the soteriologically critical non-seeing occurred, and in doing so to situate it in an epistemologically suitable context.38 Candrakīrti’s presentation of the direct perception of the non-existence of phenomena was, as we shall see, in part a response to the

influence of the epistemologists. In order to highlight the distinctiveness of his presentation, it may be instructive to briefly describe, as a point of contrast, the theory of ultimate perception set forth by Kamalaśīla (740–795), a later Mādhyamika who dealt with the issue by appropriating and revamping

the ideas of the epistemologists. Heavily influenced by Dharmakīrti (c. 600–660), Kamalaśīla relied on his theory of non-perception (anupalabdhi) for the theoretical elucidation and traditional grounding of the vision of emptiness. Slightly modifying Dharmakīrti’s theory, which determines that the non-

perception of a specific thing X implies a perception other than that of thing X (anyopalabdhi) and indeed a perception of something other than X, viz., Y, (anyabhāva), Kamalaśīla maintains that the non-perception consisting in the non-seeing of any and all things is a perception other than that of X

(anyopalabdhi) because it is a perception that is different from the seeing of things, but rejects that the perception of something other than X (anyabhāva), that is, of some other thing, plays a role because no other thing truly exists which might serve as the object of perception.39 38 R. Keira (2004: 47-49) explains: “Now, since ordinary beings cannot perceive the ultimate nature of entities, it is also impossible that they would

perceive the voidness (śūnyatā) of entities, since that is what entities ultimately are. Here, however, the following problem arises: if nobody could understand the Mādhyamika thesis of the absence of real intrinsic nature by means of direct perception, the Mādhyamikas would not have a method for obtaining the nonconceptual wisdom of thusness. In that case the religious project of Mādhyamika philosophy would not be fulfilled: bodhisattvas would not

be able to progress spiritually on the path to buddhahood by directly realizing the ultimate thusness, i.e., the absence of real intrinsic nature. The Mādhyamika theory of meditation upon all dharmas as being without real intrinsic nature thus could not be established. Furthermore, if the Mādhyamikas could not prove the possibility of a direct perception which understands the absence of real intrinsic nature, they also could not prove the existence of

the Buddha’s wisdom directly understanding selflessness (nairātmya). Therefore, since the Buddha would not be established to be someone who can directly realize thusness, his authority would accordingly be lost, as it is an essential facet of the Buddha’s wisdom that it be nonconceptual and direct.” 39 On Dharmakīrti’s theory of non-perception, see Keira 2004: 52-64; on Kamalaśīla’s revision of this theory for Mādhyamika consumption, see pp. 64-86.

Put simply, knowing the ultimate involves a cognition which does not take any thing as its support. Kamalaśīla admits that like all other things, this cognition does not ultimately exist, but he unambiguously declares that it, like the yogin in possession of it, does exist conventionally.40 Even though the conventionally existing cognition which knows the true nature of things is a cognition devoid of content, it is proper to confer on it the status of valid direct perception because it is clear (spaṣṭa), that is, non-conceptual (kalpanāpoḍha), and non-belying (avisaṃvāda). Opponents who, in consideration

of the fact that cognition by definition requires an object, would argue that non-existing things are incapable of generating cognition are countered by Kamalaśīla’s assertion that the gnosis (ye shes) which arises from meditation clearly realizes the thusness (de kho na nyid) of the selflessness of things; by no means, he states, on occasions where this gnosis is said, e.g., in the Dharmasaṅgītisūtra,41 to involve “non-seeing” is a nonimplicative negation, i.e., no cognition at all, intended.42 Even so, it is challenging to imagine how yogic cognition, as a clear perception in which nothing appears, might

have as its “object” the state without appearances; as R. Keira has noted, Kamalaśīla could be criticized for assuming an anyopalabdhi which has non-existence (abhāva) as its object.43 Kamalaśīla deals with the problem by turning to reflexive cognition (svasaṃvedana) − the aspect of cognition which knows the content of cognition and makes memory possible. According to him, when the yogin reaches the stage in his meditation on the selflessness of phenomena in which nothing appears, the reflexive aspect of his cognition, here in the role of anyopalabdhi, takes the clear perception without appearances

as its object, first recognizing that it lacks any appearances and subsequently recognizing that the cognition itself does not truly exist. On the basis of this experience, the yogin is afterwards, upon emerging from the non-conceptual state, able to understand by way of a conceptual subsequent [judging] cognition (phyis rjes su thob pa’i shes pa) that

40 See Keira 2004: 105-110. For the Madhyamakāloka text containing Kamalaśīla’s affirmation of the conventional existence of yogic cognition, see ibid., pp. 226-228. 41 Dharmasaṅgītisūtra (as cited in the Śīkṣāsamuccaya): adarśanaṃ bhagavan sarvadharmāṇām darśanaṃ samyagdarśanam iti; see Keira 2004: 69-71 and 99. 42 See Keira 2004: 98-104; for the Madhyamakāloka Tibetan text, see ibid., p. 225f. 43 See ibid., p. 83f.

the cognition lacked appearances and that it also lacks true existence, and, as the upshot of this, that all things are without a real nature.44 5. CANDRAKĪRTI ON PERCEIVING NOTHING Candrakīrti, who is estimated to have been active approximately a century and a half before Kamalaśīla and who seems not to have known Dharmakīrti’s views on non-perception, would concur with Kamalaśīla that the yogin’s perception of reality occurs in the form of a direct perception. In a section of his Yuktiṣaṣṭikā commentary, to which I shall return shortly, Candrakīrti explicitly asserts that there is direct perception of reality. His understanding of the nature of the cognition that directly perceives the final nature of things is, however, quite different from Kamalaśīla’s. A passage relevant to Candrakīrti’s views on cognition of the ultimate, albeit occurring in another context, can be found in his commentary on the second kārikā of Nāgārjuna’s Śūnyatāsaptati.45 The discussion there, sparked by the kārikā’s reference to the self (bdag, *ātman), commences with Candrakīrti’s rebuttal of an opponent view that the words “I” and “mine,” although without an objective support for the Buddhas who have relinquished the belief in a self (ngar ’dzin pa, *ahaṃkāra) and the belief in “mine” (nga yir ’dzin pa, *mamakāra), do have an objective support when it comes to ordinary, unenlightened people because they still maintain the belief in a self; Candrakīrti argues that this is not the case because the self simply does not exist. The opponent responds that even if the self does not exist, the belief in a self nevertheless exists as a mind associate (sems las byung ba, *caitta) and therefore cannot be just a word. Candrakīrti inquires what the objective support (dmigs pa, *ālambana) for this mind associate might be, and when the opponent states that it is the self, Candrakīrti reiterates that the self does not exist, and points out that in the absence of an objective support, consciousness and its associates cannot arise. He then moves on to address the Yogācāra objection that consciousness and its associ 44 I rely on Keira for this explanation. Kamalaśīla’s assertions on this point from the Madhyamakālaṃkārapañjikā, etc., and R. Keira’s elucidation of them may be found in Keira 2004: 77-81. 45 For the text and a German translation of the entire relevant passage, see Erb 1997: 218.33-223.32 and 46-53. 150 ANNE MACDONALD ates arise and exist without an external object, as they do in dreams, etc.,46 and chides the Yogācāra opponent for not adhering to the wellestablished worldly convention that consciousness occurs together with an object, arguing that consciousness and its associates, which are in fact produced by an object, cannot exist when the object is missing. An extended discussion with another opponent follows in which a favourite non-existent of Indian philosophy is made topical, with the opponent contending that not all consciousnesses have an (existent) object because otherwise the consciousness which apprehends the son of a barren woman (mo gsham gyi bu, *vandyāputra) could not occur. Candrakīrti asserts that the designation (ming, *nāman/*abhidhāna) “son of a barren woman” serves as the consciousness’s object, and asks the opponent why he would then say that this consciousness is objectless. The opponent retorts that if the mere name would be the object when one hears “son of a barren woman,” then with the utterance of the sentence “The son of a barren woman does not exist,” the designation should also not exist, but since this is not the case, the name cannot be the object of consciousness. The opponent goes on to argue that non-existence (dngos po med pa, *abhāva) would constitute the object-support (dmigs pa, *ālambana), positioning himself in the well-known ākāra theory of perception attributed to the Sautrāntika school and recognized by epistemologists like Dignāga, according to which the object bestows its image to consciousness and is cognized by means of this image;47 for those admitting external things consciousness knows the outer object exclusively via the image of the object reflected in it. Consciousness thus appears with an image that conforms to its object; for example, consciousness of the colour blue arises assuming the aspect of its object, namely, blue. Consciousness of a barren woman’s son, in the view of Candrakīrti’s opponent, would therefore arise with non-existence as its image. Candrakīrti attacks this idea, likewise in reliance on the Sautrāntika theory that consciousness assumes the image of the object, focusing first on the idea that the image is not in its nature different from consciousness. Given that the opponent presumes that the son of a barren 46 Cf. also Candrakīrti’s rebuttal of the dream example for the Yogācāra argument that consciousness arises without an object in MABhed on MA 6.48-53; see MABhed 140.5-145.9 and MABhtr 1910: 328-333. 47 Cf. AKBh 62.6, 473.25-474.1; Cox 1988: 38-40; Hattori 1968: 98, 102; Erb 1997: 142, n. 400.

K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 151 

woman as object transfers its image of non-existence onto consciousness, the consciousness, Candrakīrti points out, in conforming to the sheer non-existence of the image, will not be able to be existent. It is not logically possible, he asserts, for an existent consciousness to take on the image of that which is bereft of existence, because existence and non-existence are mutually exclusive and cannot occur simultaneously in a single phenomenon. If it would nevertheless be supposed that the consciousness would become both existent (to accommodate its own existence) and non-existent (to accommodate the image of the son of a barren woman), then the opponent will be forced to accept a double consciousness. When the opponent shifts the focus to the object and argues that the case of the apprehension of non-existence will exactly parallel the case of the apprehension of blue, i.e., the appearing image will reflect the object, he is informed that the image concerned (and the consciousness by implication), in conforming to the inexistence of the son of a barren woman, can only be non-existent, because otherwise the object and its image, the former non-existent, the latter existent, would contradict each other. In the same vein, Candrakīrti stresses a few lines later that inasmuch as consciousness does not have a nature different from the image, a consciousness that is produced through conforming to non-existence will have to be non-existent, since non-existence and an (existent) image are incompatible. He adds that consciousness lacks any nature prior to its arising, and not apprehending the image of an object, it simply does not arise.48 The debate does not stop here, but the main point has been made: for Candrakīrti, a consciousness of which the objective support is non-existence is a non-existent one. The lack of an object for consciousness precludes the arising of consciousness. While the above discussion from the Śūnyatāsaptativr̥tti deals with the Sautrāntika ākāra theory in the context of a response to opponents who rely on it to defend their own doctrinal theories, in other of his works Candrakīrti independently introduces it and adopts it for the sake of underpinning his own views; he appears to have accepted this doctrine on the conventional level.49 Perhaps his most interesting use of the ākāra theory occurs in the passages in which he describes and defends his interpretation of consciousness’s apprehension of ultimate

                     48 Cf. Erb 1997: 221.14-15 and 49.  49 For references, see Erb 1997: 142, n. 400. 

152 ANNE MACDONALD reality, ontological nirvāṇa. An epistemologically focused discussion on this topic can be found in his comments on Yuktiṣaṣṭikā kārikā 8,50 where he attacks fellow Buddhists for the sake of demonstrating that his view regarding consciousness and the ultimate is the sole logically and epistemologically viable one. Proceeding from the kārikā’s characterization of the Buddhist opponents’ nirvāṇa as (a real) cessation (’gog, *nirodha),51 Candrakīrti commences by exposing the inadequacy of the Vaibhāṣika and the Sautrāntika nirvāṇa conceived as the ceasing of the defilements and the psycho-physical continuum (in the case of the Vaibhāṣikas the conclusion of a series of hypostatized “stops,” and in the case of the Sautrāntikas nothing but the utter end of the continuum) when it comes to realization, i.e., direct perception, of this cessation.52 Among other arguments, he denies that cessation, and thus perception of it, could occur as long as the aggregates still exist and adverts to the fact that, as the kārikā has indicated, once the psycho-physical continuum has come to an end there is no subject left to apprehend the cessation.53 Candrakīrti then turns to the views of the logical-epistemological school on direct perception of the ultimate. Quoting from and paraphrasing Pramāṇasamuccaya I.6cd and its auto-commentary, he sets forth Dignāga’s definition of yogic perception, presenting it as the yogins’ seeing of the mere thing (don tsam, *arthamātra), a seeing that is 50 YṢ 8: rnam par ’jig pas ’gog ’gyur gyi || ’dus byas shes pas ma yin na || de ni su la mngon sum ’gyur || zhig ces pa de ji lta bu ||. 51 See also YṢ 7: dngos po skyes pa zhig pa la || ji ltar ’gog pa brtag pa bzhin || de bzhin sgyu ma byas pa ltar || mkhas pa dag gis ’gog par dgongs ||. 52 As Scherrer-Schaub (1991: 149f. [n. 141]) has already indicated, Candrakīrti exploits the traditional notion that nirvāṇa/nirodha must be “realized’ (sākṣātkr̥) to bring the discussion onto epistemological terrain. “La discussion qui s’ouvre avec la kār 8 et se poursuit jusqu’à la kār 12 et son commentaire … porte sur la nature de l’arrêt (nirodha) et de la connaissance de l’arrêt (nirodha-jñāna). Les sources scripturaires et les traités parlent de cette dernière comme d’une connaissance directe, un ‘vue devant les yeux’: ainsi de l’opération sur la troisième vérité, où l’arrêt doit être perçu directement (nirodha-sākṣātkāra). … La synonymie des expressions sākṣātKR̥- et pratyakṣī-KR̥-, de leur dérivés et expressions apparentées, autorise Candrakīrti à déplacer le centre de la discussion sur le terrain de l’épistémologie.” Cf., e.g., Saṃyutta Nikāya V, 422.19-22: Taṃ kho panidaṃ dukkhanirodham ariyasaccaṃ sacchikātabban ti me bhikkhave … āloko udapādi ||; further references in ScherrerSchaub 1991: 150 (n. 141). 53 For text and translation of these and other arguments, see Scherrer-Schaub 1991: 39.3-40.11 and 151-155; Loizzo 2007: 270.3-272.7 and 148-150.

K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 153 

without the superimposition of a unitary object and not mixed with conceptuality deriving from the guru’s teaching;54 he declares that such a view of yogic perception is not suitable when it is a question of direct perception of cessation. The opponent epistemologist responds by specifying that it is generally established (grags, *prasiddha) that when a real particular (rang gi mtshan nyid, *svalakṣaṇa) is meditated upon by way of its general characteristics (spyi’i mtshan nyid, *sāmānyalakṣaṇa), the gnosis arisen from meditation (bsgoms pa las byung ba’i ye shes, *bhāvanāmayam jñānam) gradually arises. That which is apprehended by this non-conceptual gnosis (rnam par mi rtog pa’i ye shes, *avikalpajñāna), he adds, being free of any conceptual superimposition, is nothing but the particular. Thus, when one realizes, e.g., impermanence, one knows the mere thing (dngos po tsam).55 The opponent concludes his argument by stating that since the object apprehended by gnosis is the particular, this object, like the object in the case of consciousness perceiving mere blue, etc., is directly perceived. The nonconceptual gnosis is thereby situated at the culmination of the epistemologist yogin’s meditation on, we may assume, the four Noble Truths, with the “mere thing” the real aspects such as impermanence, suffering, emptiness, selflessness, etc., connected with these Truths. The yogin envisaged by the opponent would thus initially meditate on a conceptual image or conceptual ascertainment of his object, such as impermanence or emptiness, and his intense concentration and repeated effort would effect a gradual refinement of the conceptualized object, with the end result that the meditation would issue in a direct, i.e., exclusively nonconceptual, perception of the object.56 In the view of certain later 54 YṢV: rnal ’byor pa rnams kyi bla mas bstan pa las skyes pa rnam par rtog pa dang ma ’dres pa gcig tu yul sgro btags pa med pa don tsam mthong ba gang yin pa de yang ’gog pa la mi srid do || (Scherrer-Schaub 1991: 40.12-14; Loizzo 2007: 272.79. Translation in Scherrer-Schaub p. 155f.; Loizzo p. 150). Pramāṇasamuccaya I.6cd: yogināṃ gurunirdeśāvyavakīrṇārthamātradr̥k; (see Steinkellner 2005: 3; the fragments presented in Hattori 1968: 94 read °āvyatibhinnā° for °āvyavakīrṇā°). Pramāṇasamuccayavr̥tti to I.6cd: yoginām apy āgamavikalpāvyavakīrṇam arthamātradarśanaṃ pratyakṣam (Steinkellner 2005: 3; see also Vincent Eltschinger’s article in the present volume, n. 93, as well as Eli Franco’s article in the present volume p. 122). 55 Cf. Scherrer-Schaub 1991: 40.12-21; Loizzo 2007: 272.10-273.2. 56 For a concise summary of the basic structure of the yogic path according to Dharmakīrti, see section 3.1 (under “The Path to Salvation”) of Vincent Eltschinger’s ar

154 ANNE MACDONALD Mādhyamikas like Kamalaśīla, who is known for having appropriated a number of the logical-epistemological school’s doctrines, the yogin involved in conceptual meditation on the emptiness of things would be able to evolve the meditation to the point that upon reaching the “ultimate limit of [[[Wikipedia:conceptual|conceptual]]] meditation” (bhāvanāprakarṣaparyanta), a non-conceptual perception of emptiness would arise,57 which, as previously explained, would take the form of a cognition without content and would be recognized and registered by the self-knowing aspect of cognition. The postulation of Dignāga’s,58 later elaborated by Dharmakīrti and his commentators, and tailored to fit Madhyamaka requirements by Kamalaśīla, that extended conceptual cultivation of an object would issue in direct perception of the object, is dismissed by Candrakīrti as preposterous. He asserts that when the idea being maintained is examined more closely, the epistemologists, given that they strictly maintain the distinctiveness, i.e., the mutual exclusivity, of the particular and of the general characteristic, respectively the object of direct perception and conceptual cognition, will have to admit that it is logically unacceptable to hold that the object used for meditation on the general characteristic could be the particular, since this would involve overextension (ha cang thal bar ’gyur ba, *atiprasaṅga) – I assume because the scope of the particular is thereby extended to include general char ticle in the present volume. For Dharmakīrti’s description of the cognitions and the meditative process the yogic path involves, see section 4 (“Yogijñāna as an Epistemological Topic”) of the same article. 57 Cf. Keira 2004: 50, 69ff. 58 Though I here attribute to Dignāga the idea that extended conceptual cultivation of an object issues in its direct perception, it should be noted that this theory is not recorded in any of his works available to us and is usually associated with Dharmakīrti. While it is of course possible that Dignāga set forth this view in one or more of his non-extant works and our YṢV passage provides documentation for this, its absence in the extant materials brings up the question of whether Candrakīrti might have known Dharmakīrti. Christian Lindtner, solely on the basis of text in the Catuḥśatakaṭīkā which appears to refer to the Pramāṇavārttika assertion pramāṇam avisaṃvādi jñānam, maintains that it “seems probable” that Candrakīrti did know Dharmakīrti (see Lindtner 1992: 57; the Catuḥśatakaṭīkā clause Lindtner cites, viz., mi bslu ba’i shes pa ni ’jig rten na tshad ma nyid du mthong na, can be found in Tillemans 1990: 67.11-12 [vol. 2]). The evidence is still too slim for definitive conclusions. I am grateful to Dr. Helmut Krasser for discussions on the matter and for providing me with valuable references.

K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 155 

acteristics. Candrakīrti’s rejection of the epistemologists’ theory thus focuses on the fact that if the particular free of all conceptual overlay is the actual object of meditation and, on account of this, the final object of yogic direct perception, then during the long and repeated course of the meditation, this particular cannot also be its opposite, the general characteristic constituted purely by conceptuality. The epistemologists, in maintaining that a yet unrealized particular can be conceptually contemplated to the point that it finally escapes the superimposition of general characteristics, contradict their fundamental differentiation of the objects of cognition with this, in his view, fogging of the distinction between the two and overlapping of their definitions. The mutually exclusive nature of the two objects and their respective cognitions otherwise posited by the epistemologists automatically prohibits any coinciding, intersection or reconciliation. As stated, Candrakīrti obviously intends to expose the faults of the views of his Buddhist colleagues in order to clear the way for his own position on perception of the ultimate. In the discussion on kārikā 8 preceding the altercation with the epistemologists, he is asked if direct perception takes place at the moment of seeing, and responds that because there is the making known (rnam par rig pa, *vijñapti) of the object by consciousness (rnam par shes pa, *vijñāna) even after seeing has ceased, this may be designated “direct perception.”59 He uses this as a lead-in to reference to the Sautrāntika theory of direct perception, according to which it is the image in consciousness, which conforms to the actual object, that is actually perceived, invoking here the theory’s stock example of the consciousness of blue. The discussion is taken in another direction by a Vaibhāṣika objection, but subsequent to his refutation of the epistemologists’ theory of yogic direct perception, there is another allusion to the Sautrāntika theory. Candrakīrti initiates the presentation of his Madhyamaka view by rhetorically asking how, even if the meditative process posited by the epistemologists would be correct, there could be the direct perception of the consciousness of cessation (’gog pa, *nirodha) when in cessation there does not exist even a trace of an entity having the form of the cessation of suffering. Next, in reliance on scriptural testimony which states that awareness of the nonarising of suffering is direct perception, he argues that it would, in fact,

                     59 Cf. Scherrer-Schaub 1991: 39.19-22; Loizzo 2007: 271.3-5. 

156 ANNE MACDONALD be impossible for consciousness to arise when its objective support (dmigs pa; *ālambana) has the form of non-arising; in such a case consciousness would definitely assume the mode of non-arising,60 that is, it would not arise at all. With this last brief statement Candrakīrti’s initially seemingly unusual take on direct perception of the ultimate is disclosed. For him, at the time of perception of the ultimate, of the emptiness of things that were never really there in the first place, inasmuch as there is nothing whatsoever to be perceived, that is, since an object for consciousness does not exist, consciousness will simply not come into being; Candrakīrti’s assertion that consciousness assumes the mode of non-arising translates into no consciousness at all. Yet in this way consciousness still fulfills the Sautrāntika demand that the consciousness resemble, conform to, its object: like its object, the non-arisen true nature of things, consciousness “takes,” so to speak, a non-arisen and nonexistent form. In Candrakīrti’s words: If consciousness, like its object, has the form of non-arising, it is proper to maintain that it has proceeded by way of the object just as it is.61 And given its proceeding by way of its object, its conforming to its object, it is proper to designate it direct perception. In the everyday world, too, he avers, situations occur in which one speaks of “direct perception” in regard to non-existent things. He provides the following example: A traveller sees an area off in the distance that appears to be abounding in clear water. He intends to cross the water but feels incompetent and nervous to do so, and therefore inquires of a local farmer just how much water might actually be out there. In response, the local, apparently taken aback by the question, asks where the water might be that the traveller claims to see, and then explains that what indeed looks like water off in the distance is actually only a mirage. He adds that if the traveller doesn’t believe him, he should go and look for himself; then he will directly perceive what he has just been told. It is the same in regular life, Candrakīrti points out, where things that do not exist and are not perceived are conventionally designated as directly perceived; therefore, from the point of view of worldly concealing truth, it is not contradictory to call a consciousness of non-perception (mi dmigs pa’i shes pa) − which for Candrakīrti is no

                     60 Cf. Scherrer-Schaub 1991: 40.28-41.3; Loizzo 2007: 273.7-11.  61 Cf. Scherrer-Schaub 1991: 41.3-5; Loizzo 2007: 273.11-12. 
K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 157 

consciousness at all – “direct perception.” He bolsters his position by referring to a scriptural statement which asserts that the determination (yongs su gcod pa, *pariccheda) of an object, corresponding to the way it is, by that which makes it known (shes pa byed pa, *jñāpaka) is direct perception. Candrakīrti considers this statement applicable to the present case because the consciousness which does not arise on account of the fact that its object is non-arisen accurately reflects, makes known, the fact that the object is non-arisen, i.e., that the object does not exist; therefore, inasmuch as the exact state of the object is accurately reflected through consciousness’s own inexistence, it is appropriate to term it direct perception. Candrakīrti had earlier presented basically the same view, although in another context, in the fourth verse and its commentary in the chapter on the level of a Buddha in the Madhyamakāvatāra.62 There he is replying to an opponent who contends that if the peaceful (zhi ba, *śānta), viz., the eternal calmedness of all non-arisen things, is reality (de nyid, *tattva), the mind will not proceed in regard to this, and when the mind does not proceed, it cannot thoroughly know its object; as a consequence, statements to the effect that precisely the non-existence of thorough knowledge (yong su shes pa med pa) constitutes thorough knowledge of reality, or complete non-knowing is knowing, are inappropriate. In his verse response, Candrakīrti admits that in this specific case of the mind relying on the aspect (rnam pa, *ākāra) of reality, it is only “as if” (lta bu, *iva) consciousness knows the ultimate, clarifying in his commentary, after reciting the Sautrāntika main requisite for perception, namely, that the consciousness be in conformity with the aspect of the object, and illustrating this with the example of blue, that it is metaphorically stated that the consciousness “arising” in conformity with the aspect of reality knows reality. It is owing to conceptuality that one establishes that this consciousness knows reality; in actuality there is not any consciousness of anything because neither consciousness nor its object come into being. Yet even with the qualification, Candrakīrti intends for the idea of a merely metaphorical apprehension of the ultimate to be taken seriously, indicating in the course of his explanation

                     62 Cf. the discussion in MABhed 356.18-358.20. See also the analysis of MA 12.3-4 in Dunne 1996: 546-548. 

158 ANNE MACDONALD that his reference to the Sautrāntika model of perception supplies a generally established example for the argument.63 6. KNOWING THE ULTIMATE So what exactly, we might ask, is the point of all this, besides the fact that Candrakīrti has displayed his agility in the performance of a very nice little pirouette with the Sautrāntika theory of perception? And why does he insist on describing the Mādhyamika yogin’s lack of consciousness as direct perception? One might initially conjecture that Nāgārjuna’s explicit mention of the realization of cessation, i.e., of nirvāṇa, in kārikā 8 of the Yuktiṣaṣṭikā inspired Candrakīrti, whose criticism of the other Buddhist schools demonstrates his awareness of the prevailing theories regarding nirvāṇa and the perception of it, to come up with his own specifically Madhyamaka view on the topic as he composed his commentary on the kārikā. But this is too simplistic, and we have just seen that he had already given a less developed explanation of the knowing of the ultimate in his earliest composition, the Madhyamakāvatāra. It is more probable that Candrakīrti took Nāgārjuna’s reference to realization of cessation primarily as an opportunity to confront Dignāga’s school, with its developed epistemological theories his main rival on the issue of direct perception of the ultimate, in order to both discredit its explanation of it and to set forth, in an argumentative and scripturally backed form, his own ingenious but typically pragmatic version. His intent, one assumes, would have been to enter the arena of Buddhist theories of perception of the ultimate and defeat the epistemologist on turf that was quite possibly already monopolized by him. It was certainly necessary that Candrakīrti find a way to acknowledge direct perception of nirvāṇa, for not to have done so would have left him open to attack regarding the Mādhyamika’s and even the Buddha’s direct realization of nirvāṇa, and as a consequence, easy prey when it came to questions of the value of Madhyamaka doctrines and to the issue of the Buddha’s establishment as an authority. At the very least, by securing his own rather unusual portrayal of the consciousness that “directly perceives” nirvāṇa as the ultimate state of things by means of scripture and a widely accepted theory of perception, Candrakīrti was

                     63 Cf. MABhed 358.13-14. 
K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 159 

able to retain his faction’s standing within the general Buddhist tradition on these issues of fundamental importance. His specifically unique presentation of the consciousness of the ultimate, on the other hand, spared him from having to compromise his integrity as a Mādhyamika. But what are the implications of the stance that when the final state of things is realized there is no object to be known, and no consciousness to know this? Would it mean that the ultimate is a sheer void, or a pure abstract nothing, and that the ultimate realization of this is in the end impossible since one can’t know nothing? I would contend that our author’s view is more sophisticated and suggestive than this. One of the main points, if not precisely the point that Candrakīrti intends to make in the above discussions is that consciousness (vijñāna) as conceived by his fellow philosophers and accepted by himself on the conventional level is fundamentally incapable of knowing the ultimate, because its functioning is restricted to occurring in relation to objects, and the ultimate is no object and has no objects in it. One can be quite certain that Candrakīrti would have rejected Kamalaśīla’s version of the highest awareness as clear perception not only because it is based on the epistemologist’s model but also because in this version the clear perception does not escape being described in terms that relate it to and therefore bind it to the conventional level; Kamalaśīla in fact allows this consciousness conventional existence. Candrakīrti’s non-acceptance of reflexive awareness would further have led him to repudiate the idea that non-existence is not the object of the consciousness and to charge that the positing of consciousness devoid of content, i.e., bereft of an object, would contradict the general Buddhist doctrine that consciousness occurs in tandem with an object. Candrakīrti’s underscoring that consciousness does not arise when the object is the ultimate is secondarily intended to point to the fact that for him all perceptual activity as we know it − as well as all conceptual and linguistic activity − ceases in the experience of the true nature of things, of ontological nirvāṇa.64 Toward the end of the Śūnyatāsaptati’s earlier referenced discussion of the perception of the son of a barren woman, Candrakīrti declares that the 64 Cf., e.g., Candrakīrti’s commentary on MMK 5.8, where he states that the pacification of all visibles (draṣṭavyopaśama) that is free of the net of all conceptuality (sarvakalpanājālarahita) has the nature of the ceasing of consciousness (here intended in the sense of conventional consciousness) and the object of consciousness (jñānajñeyanivr̥ttisvabhāva). 160 ANNE MACDONALD Mādhyamikas accomplish the clearing away, the elimination, of the factors of existence (chos, *dharma) in the sense that with the nonperception of any of these factors, all of which are of the nature of nonexistence, consciousness stops.65 This statement is followed by two supporting scriptural citations, one of which is Āryadeva’s famous verse that states that consciousness (rnam shes, *vijñāna) is the “seed of existence.”66 For Candrakīrti, the actual realization of the true nature of things is performed by a completely different category of awareness, if I may call it that, namely, by jñāna, “gnosis,” which does not belong or relate to the everyday level. I am aware that a number of Madhyamaka scholars construe the situation regarding ultimate knowledge and its object quite differently, in large part because they interpret the fundamental Madhyamaka stance on the possibility of existence – in my interpretation that it is impossible – as espousing it. I digress with this, but let it be noted that C.W. Huntington in his book “The Emptiness of Emptiness” describes the consciousness which knows the ultimate, i.e., jñāna, as a non-dualistic knowledge that is coterminous with the bodhisattva’s everyday experience “in both its conceptual and perceptual aspects.” He writes, “The Mādhyamika does not advocate any radically unconventional category of epistemic act, but rather a radically unconventional form of life, in which one is constantly and profoundly in touch with the holistic, contextual nature of all experience—with ‘the suchness of dependent origination.’”67 Huntington’s jñāna, albeit acknowledged to be 65 See Erb 1997: 221.40-222.2 and 50-51. It is to be noted that antecedent to this text passage, in the extended debate concerning perception of the non-existent Candrakīrti adverts to the absurd consequence entailed by acceptance of objectless consciousness in regard to nirvāṇa, namely, that (ordinary) consciousness would permanently continue, taking nirvāṇa as its objective support. He also briefly weaves in his view of the status of consciousness at the time of perception of the ultimate; see ibid., 221.31-222.12 and 50-51. Cf. also 223.7-16 and 52. 66 Cf. Catuḥśataka XIV.25 (Suzuki 1994: 360): srid pa’i sa bon rnam śes te || yul rnams de yi spyod yul lo || yul la bdag med mthong na ni || srid pa’i sa bon ’gag par ’gyur ||. The verse as cited in the ŚSV reads: srid pa’i sa bon rnam par śes || yul ni de yi spyod yul la || mthong ba’i yul rnams bdag med phyir || srid pa’i sa bon ’gag par ’gyur ||; see Erb 1997: 222.9-12 and p. 144, n. 421. 67 Huntington 1989: 119f. One notes also other comments in reference to jñāna: “Jñāna is the essential clarity and unerring sensibility of a mind that no longer clings to reified concepts of any kind. It is a direct and sustained awareness of the truth, for a bodhisattva, that meaning and existence are found only in the interface

K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 161 

meditatively cultivated, is essentially a rational insight into a profound interconnectedness inherent in the truly existing world, and thus merely a worldly, lived awareness of a state of affairs, and one that is involved not only with perception but with conceptual thought. Dan Arnold does not refer to jñāna per se in his book “Buddhists, Brahmins and Belief,” but he does speak of a realization, which he qualifies, at least parenthetically, as “radically transformative.”68 He clarifies that the subject of such a realization would be “a Buddha.” Arnold states that the object of the realization would be ultimate truth, but rejecting Madhyamaka argumentation as world-denying, he contends “that, for Candrakīrti, the only ultimate truth is that there is no ultimate truth—that the ‘ultimate truth,’ in other words, is the abstract state of affairs of there being no set of ‘ultimately existent’ (paramārthasat) ontological primitives like the dharmas of Abhidharma.”69 “‘[U]ltimate truth’ (nirvāṇa),” he writes, attempting to explain MMK 25.19ab, “does not consist in something fundamentally different in kind from ‘conventional’ reality (saṃsāra); rather, what is ‘ultimately true’ is simply the fact that there is nothing fundamentally different from the world as conventionally described.”70 According to him, the content of a Buddha’s realization would therefore consist in knowing that there is not something more real than the dependently arisen, but truly existing, world;71 similar to Huntington’s, Arnold’s ultimate insight is a real rational insight into the way the de between the components of an unstable and constantly shifting web of relationships, which is everyday life” (ibid., p. 104), and in reference to prajñā (Huntington states that it is “difficult to draw a distinction” between prajñā and jñāna): “Perfect wisdom graphically reveals the holistic, contextual nature of all forms of existence and knowledge and allows the bodhisattva to adjust his attitude so that it accords with the ‘suchness’ of all experience, with the self, and with the world, as they are in the context of the moment. In this way, he is invested with the ability to act effectively and in harmony with the demands of every situation as it presents itself in the web of interrelated events (ibid., p. 88).” 68 Arnold 2005: 204. 69 Ibid., p. 184. 70 Ibid., p. 172. Two sentences before the one quoted, he states, “Thus, the point of insisting on the ‘emptiness of emptiness’ is to throw us back into the world and to compel the recognition that, although events are dependent, contingent, and conventional, they are, for all that, real.” 71 Arnold (2005: 204) writes, “That claim [= Candrakīrti’s] is that the ‘ultimate’ consists not in some radically ‘other’ state of affairs but in the realization (radically transformative, to be sure) that there is nothing more real than this.” 162 ANNE MACDONALD pendent but real world, inclusive of its concepts, exists. Taking up John Dunne’s claim (a claim based on the previously mentioned Madhyamakāvatāra verse which states that suchness is only metaphorically known by the non-arisen consciousness) that a buddha would know neither ultimate reality nor the ordinary world because nothing occurs in his mind, Arnold contests that such a being would indeed perceive something, though certainly not an ultimate reality: “It seems that the ordinary world is all that such a Buddha would see.”72 On my reading of Candrakīrti’s works, however, neither Huntington nor Arnold come near to capturing what our author intends to convey with his references and allusions to the knowing of reality or his more general pronouncements on the nature of things. To return to my understanding of Candrakīrti’s perspective on the ultimate and its awareness: The cessation of all consciousness in the face of no objects is pivotal for the realization of emptiness, the true nature of things; it does not, however, fully define it. The coming to rest of consciousness merely serves as the necessary condition for the experience of the ultimate. As stated, it is a completely different type of awareness, viz., “gnosis” (jñāna), that knows it. Unlike ordinary consciousness, gnosis does not take a thing, or as Candrakīrti sometimes terms it, a mark (nimitta), as its object.73 The fact that its object-support is conventionally described by Candrakīrti to be emptiness74 does not, however, necessarily mean that the ultimate realization is a realization of nothing, or that it involves acquiescing to absolute nothingness, to a sheer, abstract void. Emptiness elucidated as the pacification of all manifoldness (prapañcopaśama) implies that no concept or linguistic designation applies to the true nature of the world; nothing can be predicated of it, not even “non-existence.” The notion of emptiness has in this context a spiritual function; as Lambert Schmithausen has stated,

                     72 Arnold 2005: 204. Arnold is referring to Dunne 1996: 548.  73 Cf. Candrakīrti’s commentary on MMK 25.16 (PsPed 533.11-15), where he argues that nirvāṇa, in this case conceived by the opponent as both “existent and nonexistent,” cannot be ascertained by vijñāna because there is no nimitta in nirvāṇa: kenaitad itthaṃvidhaṃ nirvāṇam astīti paricchidyate | saṃsārāvasthitaḥ paricchinattīti cet | yadi saṃsārāvasthitaḥ paricchinatti sa kiṃ vijñānena paricchinatti uta jñānena | yadi vijñāneneti parikalpyate tan na yujyate | kiṃ k āraṇam | yasmān nimittālambanaṃ vijñānaṃ na ca nirvāṇe kiṃcin nimittam asti |.  74 See PsPed 533.16: yasmāj jñānena hi śūnyatālambanena bhavitavyam |. 
K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 163 

“The concept ‘emptiness’ is also not intended to make a positive statement about this reality; it is rather merely a call to thought to deliver itself to its negation, to think itself away, to shake off all manifold conceptuality and thereby enable the manifestation of the inconceivable reality that was always there.”75 The ceasing of all conceptual and perceptual activity in the Madhyamaka yogin would thus act as a catalyst for experience of the concept-, designation- and percept-transcending ultimate that is neither an existent entity nor pure nothingness. Consciousness’ coming to rest would create, so to speak, a vacuum in which emptiness as thusness (tattva), the true nature of the world free of any appearances or conceptual content, could reveal itself. Inasmuch as the pacification of manifoldness coincides in meaning with the cessation of all things worldly, to which belong karma and the defilements, emptiness as the true nature of the world is equivalent to nirvāṇa;76 the experience of emptiness, then, would translate into an experience of nirvāṇa, and the gnosis that has, conventionally speaking, emptiness as its object-support (ālambana) would convey this experience of nirvāṇa. But if gnosis is not a real consciousness that takes the nonexistent as its object or a conventionally existing clear, contentless consciousness, just how does Candrakīrti envision it? Given the general Madhyamaka focus on demonstrating the ontological impossibility of known or postulated phenomena, and Candrakīrti’s hesitancy to attempt to describe an ultimate that could mistakenly be construed as existent or non-existent, details regarding the nature of gnosis are extremely rare in his works. There is, however, one interesting passage in which he does dare to sketch its features; it occurs in his commentary on MMK 25.16, 75 Schmithausen 1969: 166: “Auch der Begriff ‘Leerheit’ soll keine positive Aussage über diese Wirklichkeit machen; es ist vielmehr lediglich eine Aufforderung an das Denken, sich zur Negation seiner selbst zu vermitteln, sich zu zerdenken, alle vielfältige Vorstellung abzuschütteln und dadurch die Manifestation der immer schon vorhandenen unbegreiflichen Wirklichkeit zu ermöglichen.” 76 Schmithausen (1969: 166) describes the intersection of nirvāṇa and saṃsāra: “Das Nirvāṇa (als metaphysische Größe) ist also im Madhyamaka kein Jenseits; es ist nicht außerhalb der Welt, es ist vielmehr in ihr; ja, Nirvāṇa und Welt sind überhaupt nicht verschieden, sofern man nur die Welt nicht in ihrer unwahren Endlichkeit, sondern in ihrem wahren Wesen nimmt.” (“Nirvāṇa (as a metaphysical dimension) is therefore in Madhyamaka not a ‘beyond’; it is not outside the world, it is rather in it; indeed, nirvāṇa and the world are not at all different, as long as one takes the world not in its unreal finitude, but in its true nature.”) 164 ANNE MACDONALD in which he contrasts consciousness (vijñāna) and gnosis (jñāna) and, as already noted, asserts that gnosis’ object-support is emptiness. Immediately after this reference to its object-support he qualifies gnosis by way of two adjectives, namely, “having the form of non-arising” (anutpādarūpa) and “having a non-existing own-form” (avidyamānasvarūpa), both of which could be applied to the consciousness which in Candrakīrti’s pirouette “directly perceives” the ultimate, and both of which could also be taken to support the idea that there is no experience of reality, or that its non-experience is experience of it. It is the third adjective given − although at first glance seemingly insignificant − that provides perhaps one of the most telling references to his take on it. The modifier is “having a form that transcends all manifoldness” (sarvaprapañcātītarūpa), and in contrast to the previous two, it clearly does not intend a purely negative characterization. With it, there is allusion to an awareness that surmounts all manifold conceptualization and designation, one which neither exists nor does not exist, and is as unfathomable as its so-called object, the thusness that is true reality, ontological nirvāṇa. Of course as an awareness that is diametrically opposed to ordinary consciousness, it will not be configured in a subject-object relationship with emptiness, expressed as its focus for conventional convenience; its functioning would rather be non-dual. Intimated by this and the third adjective is the idea that gnosis consists in a radical mystical experience. Elsewhere, Candrakīrti states that the Buddhas abide in the objectless gnosis, far beyond the spiritually immature.77 It will not be irrelevant to mention, in this connection, that Nāgārjuna, in his chapter on the Tathāgata in the MMK, describes the Buddha, here understood not as a distinct person but as the true reality that is his nature, in the same way that Candrakīrti describes gnosis, namely, as “transcending all manifoldness” (prapañcātīta).78 Just as striking is Candrakīrti’s comment in the same chapter where he declares that the Mādhyamikas do not teach that the Tathāgatas are inexistent inasmuch as they are

                     77 Cf. YṢV on YṢ 4cd: de’i phyir de dag skye bo byis pa rnams las shin tu ’das pa dmigs pa med pa’i ye shes la gnas pas de dag nyid che ba’i phyir bdag nyid chen po zhes bya ste | (Scherrer-Schaub 1991: 32.16-18; Loizzo 2006: 260. 1-3).  78 MMK 22.15: prapañcayanti ye buddhaṃ prapañcātītam avyayam | te prapañcahatāḥ sarve na paśyanti tathāgatam ||. Cf. also PsPed 446.5: sarvās tv etāḥ kalpanā niṣprapañce tathāgate na saṃbhavanti |. 
K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 165 

“completely outside [the domain of] manifoldness.”79 These descriptions of the awakened beings and their gnosis, limited to being made by way of modifiers indicating indescribability and inconceivability, merely point to the unfathomable state beyond the nothingness of worldly phenomena. It is probably not inappropriate to state that for the Mādhyamika as yogin the final goal, and the final state, is not nothingness, but transcendence. Although he is more often occupied with and thus associated with rigorously arguing an uncompromising denial of the world, it is in passages such as the ones examined here that we encounter Candrakīrti, as he moves on from this to allude to the outcome and purpose of that denial, as a conveyer of spiritual, mystical experience.

ABBREVIATIONS AND BIBLIOGRAPHY

AK Abhidharmakośa. See AKBh. AKBh Abhidharmakośabhāṣyam of Vasubandhu. P. Pradhan, ed. Tibetan Sanskrit Works Series 8. 2nd ed. Patna, 1975. Arnold 2005 Dan Arnold, Buddhists, Brahmins, and Belief: Epistemology in South Asian Philosophy of Religion. New York: Columbia University Press, 2005. Cox 1988 Collett Cox, “On the Possibility of a Nonexistent Object of Consciousness: Sarvāstivādin and Dārṣṭāntika Theories.” In Journal of the International Association of Buddhist Studies 11 (1988): 63-105. Cox 1994 Collett Cox, “Attainment through Abandonment: The Sarvāstivāda Path of Removing Defilements.” In Paths to Liberation. The Mārga and its Transformations in Buddhist Thought. R.E. Buswell, Jr. & R.M. Gimello, ed. Delhi 1994: 31-87. Cox 1995 Collett Cox, Disputed Dharmas: Early Buddhist Theories on Existence. Studia Philologica Buddhica: Monograph Series, 11. Tokyo: The International Institute for Buddhist Studies, 1995. 79 PsPed 443.11: na ca vayaṃ sarvathaiva niṣprapañcānāṃ tathāgatānāṃ nāstitvaṃ brūmaḥ… |. 166 ANNE MACDONALD Dayal 1932 Har Dayal, The Bodhisattva Doctrine in Buddhist Sanskrit Literature. London: Routledge & Keagan Paul Ltd., 1932. Reprint Delhi: Motilal Banarsidass, 1978. Dunne 1996 John Dunne, “Thoughtless Buddha, Passionate Buddha.” In Journal of the American Academy of Religion LXIV/3 (1996): 525-556. Erb 1997 Felix Erb, Śūnyatāsaptativr̥tti: Candrakīrtis Kommentar zu den „Siebzig Versen über die Leerheit“ des Nāgārjuna [[[Kārikās]] 1-14]. Tibetan and Indo-Tibetan Studies 6, Institute for the Culture and History of Indian and Tibet at the University of Hamburg. Stuttgart: Franz Steiner Verlag, 1997. Frauwallner 1953 Erich Frauwallner, Geschichte der indischen Philosophie, I. Band. Salzburg: Otto Müller Verlag, 1953. Frauwallner 1984 Erich Frauwallner, History of Indian Philosophy, Vol. 1. Bedekar, V.M., tr. Delhi: Motilal Banarsidass, 1984. Gethin 1998 Rupert Gethin, The Foundations of Buddhism. Oxford: Oxford University Press, 1998. Gethin 2001 Rupert Gethin, The Buddhist Path to Awakening. Leiden: E.J. Brill, 1992; 2nd edition Oxford: Oneworld, 2001. Granoff 1996 Phyllis Granoff, “The Ambiguity of Miracles. Buddhist Understandings of Supernatural Power.” In East and West 46 (1996): 79-96. Hattori 1968 Masaaki Hattori. Dignāga, On Perception. Cambridge: Harvard University Press, 1968. Heitmann 2004 Annette L. Heitmann, Nektar der Erkenntnis. Buddhistische Philosophie des 6. Jh.: Bhavyas Tarkajvālā I-III.26. Aachen: Shaker Verlag, 2004. Huntington 1989 C.W. Huntington, Jr., The Emptiness of Emptiness. An Introduction to Early Indian Mādhyamika. Honolulu: University of Hawaii Press, 1989. de Jong 1978 J.W. de Jong, “Textcritical Notes on the Prasannapadā.” In Indo-Iranian Journal 20 (1978): 25-59; 217-252. Keira 2004 Ryusei Keira, Mādhyamika and Epistemology. A Study of Kamalaśīla's Method for Proving the Voidness of all Dharmas. Introduction, Annotated Translations and Tibetan Texts of Selected Sections of the Second Chapter of the Madhyamakāloka. Wiener Studien zur Tibetologie und Buddhismuskunde 59. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, 2004. Kragh 2006 Ulrich Timme Kragh, Early Buddhist Theories of Action and Result: A Study of Karmaphalasambandha. Candra

K NOWING NOTHING: CANDRAKĪRTI AND YOGIC PERCEPTION 167 

kīrti's Prasannapadā, Verses 17.1–20. Wiener Studien zur Tibetologie und Buddhismuskunde 64. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, 2006. Lamotte, 1976 Étienne Lamotte, Le traité de la grande vertu de sagesse de Nāgārjuna (Mahāprajñāpāramitāśāstra) avec une étude sur la Vacuité. Tome IV. Publications de l’Institut Orientaliste de Louvain, 12. Louvain: Institut Orientaliste, Université de Louvain, 1976. La Vallée Poussin 1931 Louis de La Vallée Poussin, “Le Bouddha et les Abhijñā.” In Le Muséon (1931): 335-342. Lindquist 1935 Sigurd Lindquist, Siddhi und Abhiññā. Eine Studie über die Klassischen Wunder des Yoga. Uppsala 1935. Lindtner 1992 Christian Lindtner, “On the Date of Dharmakīrti etc.” In The Adyar Library Bulletin (1992): 56-62. Loizzo 2007 Joseph Loizzo and the AIBS Translation Team, Nāgārjuna’s Reason Sixty with Candrakīrti’s Reason Sixty Commentary. New York: The American Institute of Buddhist Studies at Columbia University, 2007. MA Madhyamakāvatāra. See MABhed. MABhed Madhyamakāvatāra par Candrakīrti. Traduction Tibétain. Louis de La Vallée Poussin, ed. Bibliotheca Buddhica IX. St. Petersburg, 1907-1912. Reprint: Delhi: Motilal Banarsidass, 1992. MABhtr “Madhyamakāvatāra, Introduction au Traité du Milieu de l’Ācārya Candrakīrti, avec le commentaire de l’auteur, traduit d’aprés la version tibétaine.” Louis de La Vallée Poussin, tr. In Le Muséon VIII (1907): 249-317; XI (1910): 271358; XII (1911): 235-328. MacDonald 2007 Anne MacDonald, “Revisiting the Mūlamadhyamakakārikā: Text-Critical Proposals and Problems.” In Studies in Indian Philosophy and Buddhism 14 (2007): 2555. MMK Mūlamadhyamakakārikā. See PsPed and de Jong 1978. Ñāṇamoli 1995 Bhikkhu Ñāṇamoli and Bhikkhu Bodhi, tr., The Middle Length Discourses of the Buddha. A New Translation of the Majjhima Nikāya. Boston: Wisdom Publications, 1995. Oetke 2007 Claus Oetke, “On MMK 24.18.” In Journal of Indian Philosophy 35 (2007): 1-32. PsPed Mūlamadhyamakakārikās (Mādhyamikasūtras) de Nāgārjuna avec la Prasannapadā Commentaire de Candrakīrti. Louis de La Vallée Poussin, ed. Bibliotheca Buddhica 4. St. 168 ANNE MACDONALD Petersburg 1903-1913. Reprint Osnabrück: Biblio Verlag, 1970. Rahder 1926 Johannes Rahder, Daśabhūmikasūtra. Leuven, 1926. Saṃyutta Nikāya M. Leon Feer, ed. London: Pali Text Society, 1898. Reprint 1960. Scherrer-Schaub 1991 Christina Scherrer-Schaub, Yuktiṣaṣṭikāvr̥tti. Commentaire à la soixantaine sur le raisonnement ou Du vrai enseignement de la causalité par le Maître indien Candrakīrti. Bruxelles: Institut Belge Des Hautes Études Chinoises, 1991. Schmithausen 1969 Lambert Schmithausen, “Ich und Erlösung im Buddhismus.” In Zeitschrift für Missionswissenschaft und Religionswissenschaft 53 (1969): 157-170. Schmithausen 1973 Lambert Schmithausen, “Spirituelle Praxis und philosophische Theorie im Buddhismus.” In Zeitschrift für Missionswissenschaft und Religionswissenschaft 57 (1973): 161186. Schmithausen 1981 Lambert Schmithausen, “On some Aspects of Descriptions or Theories of ‘Liberating Insight’ and ‘Enlightenment’ in Early Buddhism.” In Studien zum Jainismus und Buddhismus. Gedenkschrift für Ludwig Alsdorf. Klaus Bruhn and Albrecht Wezler, ed. Wiesbaden 1981: 199-250. ŚSV Śūnyatāsaptativr̥tti. See Erb 1997. Steinkellner 2005 Ernst Steinkellner, Dignāga’s Pramāṇasamuccaya, Chapter I. A hypothetical reconstruction of the Sanskrit text. Online publication (http://ikga.oeaw.ac.at/Mat/dignaga_PS_1.pdf), 2005. Suzuki 1994 Kōshin Suzuki, ed., Sanskrit Fragments and Tibetan Translation of Candrakīrti's Bodhisattvayogācāracatuḥśatakaṭīkā. Tokyo: The Sankibo Press, 1994. Tillemans 1990 Tom J.F. Tillemans, Materials for the Study of Āryadeva, Dharmapāla and Candrakīrti. Wiener Studien zur Tibetologie und Buddhismuskunde 24. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, 1990. Vetter 1982 Tilmann Vetter, “Die Lehre Nāgārjunas in den MūlaMadhyamaka-kārikās.” In Epiphanie des Heils. Zur Heilsgegenwart in indischer und christlicher Religion. Arbeitsdokumentation eines Symposiums. Gerhard Oberhammer, ed. Publications of the de Nobili Research Library, Vol IX. Vienna 1982: 87-108. YṢ, YṢV Yuktiṣaṣṭikā, Yuktiṣaṣṭikāvr̥tti. See Scherrer-Schaub 1991.