CREATIVE IGNORANCE: NĀGĀRJUNA ON THE ONTOLOGICAL SIGNIFICANCE OF CONSCIOUSNESS

by EVIATAR SHULMAN

It is generally accepted that Nāgārjuna’s dialectic is aimed at exposing, or proving, the lack of self-nature2 (svabhāva) of all phenomena, all things whatsoever. The fact that this paper, for example, is dependent on the material conditions for its production (my computer, electricity, paper, my fi ngers, etc.), on my intention to write it, on its audience and/or readers, and so forth, suggests it has no true nature of its own. What the refutation of svabhāva actually means, both philosophically and experientially, is hotly debated; numerous views haven been suggested. Some believe the lack of svabhāva implies Nihilism,3 others see it as pointing to the decep

This paper is an elaborated version of the one I read at the XVth IABS conference, held in Atlanta, June 2008. I wish to thank Akira Saito and Ernst Steinkellner for their valuable comments following my presentation. I also wish to thank Jonathan Silk and Paul Harrison for their remarks on an earlier version of this paper. 2 I will hereby be translating svabhāva, most literally “self-existence,” “existence of/by/in/for/as itself,” or “own being,” and commonly translated as “intrinsic nature” or “inherent existence,” as “self-nature.” Svabhāva most simply means “nature,” and Nāgārjuna at times will use the term in such a non-technical sense as well (e.g. YṢ 55). Specifi cally, svabhāva refers to a quality of being attributed to something that has it own private nature that it possesses of itself. MMK 15.2cd is commonly viewed as a defi nition of svabhāva (see for example Saito [2007: 157]): akṛtrimaḥ svabhāvo hi nirapekṣaḥ paratra ca (Self-nature is unmade and does not depend on another). 3 Recent examples are Wood (1994), Tola & Dragonetti (1995) and Burton (1999: ch. 4).

tive nature of language, or even to a fundamental error which characterizes any form of knowledge.4 Still others view the refutation of svabhāva as being conducted from the point of view of ultimate reality, and thus as directing the mind toward the realization of an absolute truth.5 There are still more who doubt that Nāgārjuna had any positive philosophical message.6 Finally, there are those who believe the realization of the lack of self-nature to be an end in itself.7 Although these positions can be elaborated, and many others could be listed, I believe the views just mentioned are the major readings Nāgārjuna has received in modern scholarship. They are

4 This is possibly the most widely accepted view of Nāgārjuna today, and it consists of a number of separate but nonetheless related positions. Sprung (1977) and Ganeri (2001) are among the scholars who defi ne Nāgārjuna as a skeptic. Siderits (1988) believes him to be arguing against the correspondence theory of knowledge. The view of Nāgārjuna and the Madhyamaka as discussing the nature of language, often inspired by Wittgenstein, has achieved great popularity, and is advocated in such works as Thurman (1980), Huntington

(1983, 1989, 2007) and Loizzo (2001). 5 A clear defi nition of the “absolutistic” reading of Nāgārjuna is given by de Jong (1972: 5): “There is no doubt that paramārtha, being the ‘supreme goal’ of the believer, may be called ‘the absolute.’ But this absolute by its very nature is inaccessible to philosophical thought. One might try to approach it by indirect means, but all one could say or think about it would of necessity be false. It cannot be thought of as being or as nothingness. For the Mādhyamikas it is ‘the silence of the saints’.” More than vestiges of this position can be identifi ed in many works on Nāgārjuna and the Madhyamaka, such as Gómez (1976), Seyfort Ruegg (1977: 6, 12, 1981: 34–41), Harris (1994) and Lindtner (1997). 6 Schroeder (2000, 2001) is a prominent example, and he represents a common Zen-Buddhist approach to Emptiness. 7 The classic case in this regard is the understanding of the Madhyamaka developed in the Tibetan dGe-lugs-pa sect, which has infl uenced a great number of modern discussions on the subject. For prominent examples see Napper (1989) and Williams (1989: ch. 3).

all rooted in understandings of Nāgārjuna’s thought which were developed in the diff erent Buddhist philosophical traditions.8 But although all of these various teachings of emptiness do relate to certain aspects of Nāgārjuna’s writings, I will argue that they also suff er from fundamental errors, in regard both to what the texts reliably attributed to Nāgārjuna actually say, and to philosophical consistency. Not only do these readings misrepresent Nāgārjuna’s original message; they also fail to come to terms with the full implications of his thought. In fact, all these presentations of Madhyamaka ignore a central aspect of Nāgārjuna’s insight which concerns his understanding of the relation between consciousness and reality.

I. Genre sensitivity

In this paper I will attempt a faithful reconstruction of Nāgārjuna’s teaching, based on a careful reading of his Mūlamadhyamakakārikā (“The Core Verses of the Middle Path,” MMK) 9 in light of his Yuk ti ṣaṣṭikākārikā (“Sixty Verses of Reasoning,” YṢ)10 and Śūnya tāsaptati (“Seventy Verses of Emptiness,” ŚS).11 Naturally, my

Some of these views are characteristic of non-Buddhist Indian traditions’ take of Nāgārjuna as well. In the VV Nāgārjuna argues against a rival who is understood to represent the Nyāya school and who accuses Nāgārjuna of Nihilism. The “absolutistic” reading of Nāgārjuna was developed in the Hindu Advaita-Vedānta school, most explicitly by

Gauḍapāda. On Gauḍapāda’s adoption of Nāgārjuna’s rhetoric see Radhakrishnan (1956: 456), Whaling (1979), Darling (1987: I.G) and King (1989). 9 For the text of the MMK I am relying on the edition of de Jong (1977 [1958]), together with the emendations made by MacDonald (2007). 10 For the text of the YṢ I am relying on the excellent critical edition prepared by Scherrer-Schaub (1991). 11 The text of the ŚS presents many philological and interpretive problems, the greatest of which are the signifi cant divergences which exist between the version of the kārikās alone and the version accompanied by a svavṛtti attributed to Nāgārjuna. For the text of the ŚS itself I prefer the version of the kārikās over the one embedded in the svavṛtti. The verses quoted here are based on an edition of the text I have prepared, which I hope to publish in the

methodological position determines much of the reading of Nāgārjuna I will suggest. I argue that in order to achieve a clear picture of Nāgārjuna’s understanding of emptiness we must regard his four extant analytical treatises – the MMK, YṢ, ŚS, and his VigrahaVyā vartanī (“A Refutation of Objections,” VV)12 – as an integral unit of meaning. When Nāgārjuna is read in light of the MMK and VV alone, as commonly happens, a limited picture of his thought emerges. The MMK’s power lies in its unrelenting critical force, which

precludes the possibility of off ering a positive description of existence. The VV is a polemical, one could say a defensive treatise, in which, in a “user-friendly” fashion, Nāgārjuna attempts to blur the severe consequences of his theory and method. Alternatively, if the MMK and VV are read in light of texts belonging to distinct literary genres, such as the Ratnāvalī (“The Precious Garland”), the picture becomes rather hazy,13 since Nāgārjuna’s four analytical texts do not discuss the more practical aspects of the Mahāyāna Buddhist path, such as compassion and the path of the Bodhisattva.14 If we wish to reach a reliable understanding of what śūnyatā (“emptiness”) meant to Nāgārjuna, we must fi rst defi ne the message expressed in the texts he devoted to this subject it

near future. There I will also elaborate on my preference for the kārikā version of the text. My edition is based on the one presented by Lindtner (1986), who relied on the Narthang and Peking canons, which I compared to the Derge edition. I have also compared these versions of the text to those found in the svavṛtti, as well as to those found in Parahita and Candrakīrti’s commentaries. For Candrakīrti’s commentary, Erb (1997) has prepared a critical edition of his discussion of verses 1–14. For a discussion of the diff erent versions of the ŚS see Komito (1987: section 3). 12 For the text of the VV I am using the edition found in Bhattacarya, Johnston and Kunst (1978). 13 Examples of such a presentation which views Nāgārjuna as a traditional Mahā yāna teacher are Lindtner (1982, 1986), and Williams (1984). 14 A rare exception to this rule would be MMK 24.32 which speaks of “the practice of the Bodhisattva” (bodhisattvacaryā). See also the concluding verses of the MMK and the YṢ.

self. We should better fi rst achieve a clear defi nition of Nāgārjuna’s philosophical message, before we ask how emptiness relates to Bodhisattvas, their vehicles and the like. It should be emphasized that the four texts I wish to examine are regarded as authentic to Nāgārjuna by nearly all the knowledgeable authorities both among modern scholars and within the Buddhist tradition.15 Regarding all other texts attributed to Nāgārjuna in the Chinese and the Tibetan traditions, serious doubts have been raised regarding their authorship. Moreover, the Tibetan tradition has grouped these four texts as a distinct genre within Nāgārjuna’s writings, that is his “analytical corpus” (rigs tshogs).16 Although

The most important discussions of Nāgārjuna’s corpus have appeared in the following sources: Seyfort Ruegg (1981), Lindtner (1982, 1986) and Williams (1984) mainly discuss Tibetan views on the subject. Ramanan (2002 [1966]) and Robinson (1967) discuss Chinese positions. Further important studies of specifi c works by Nāgārjuna are Dragonetti (1978, 1986), Vetter (1992), Huntington (1995), Tola & Dragonetti (1995, 1995a, 1998) and Jamieson (2000). Tola & Dragonetti (1998) have argued against the authenticity of the VV, in a manner I fi nd unconvincing. They raise the plausible suggestion that the fi rst 20 verses of the text, which express the position of a pūrvapakṣin, were originally an

independent text. Aside from this argument, the authors off er no claims that should seriously cause us to doubt the traditional attribution of the VV to Nāgārjuna, as the majority of their arguments are answered by taking into consideration the diff erent perspectives from which Nāgārjuna may have written diff erent statements. In Tola & Dragonetti (1995: 54–57) the same authors have suggested that a number of verses from the ŚS may not be authentic, since the title of the text speaks of 70 verses while the text actually includes 73. Again, I do not believe such a technical argument to be persuasive, especially since such discrepancies are common to the genre (e.g. Vasubandhu’s Viṃśatikā). In this respect, see the comments made by Prebish (1974: 176). Regarding the SŚ, the Chinese tradition seems to be unfamiliar with the text, although the Dvadaśamukhaśāstra, a central text of the Chinese Madhyamaka, quotes ŚS 8 and 19. 16 The common Tibetan classifi cations of the rigs tshogs normally include

this category is clearly a retrospective classifi cation, it is not without its merits. We can safely assume that Nāgārjuna was aware of the diff erences which exist between writing a philosophical text and composing a devotional hymn or a “friendly letter.” In short, based on these four texts we can hope to achieve a clear defi nition of emptiness, or this is at least where we should begin. The YṢ and the ŚS expand on the analysis conducted in the MMK, and allow a fuller understanding of Nāgārjuna’s philosophical thought. They demonstrate that the MMK has a special place in the Nāgārjunian corpus, but that Nāgārjuna’s philosophical insight is not exhausted by the text. An attentive reading of the YṢ and the ŚS will lead us not only to a better understanding of the way Nāgārjuna viewed the world, but to a fuller comprehension of the MMK’s thought as well. 17

II. The object of refutation

What is Nāgārjuna actually refuting? A quick but bold look at the texts tells us that Nāgārjuna was troubled not by “self-existence” – svabhāva – but by existence in general – bhāva, or astitvam. Nāgār juna attempted to pave the middle path between existence

5 or 6 texts. The additional texts included are the (1) Vyavahārasiddhi, no longer extant (the fi rst 6 verses are quoted in Lindtner [1982: 96–99, 1986: 120–123], identifi ed by Lindtner in Śantarakṣita’s Madhymakālaṃkāravṛtti). (2) Vaidalyaprakaraṇa, a polemical text dedicated to a refutation of the 16 basic categories of Nyāya thought. Serious doubt regarding the text’s authenticity have been raised by Tola & Dragonetti (1995a) and Pind (2001). (3) Ratnāvalī, at times listed in the rigs tshogs, but generally assigned to the gtam tshogs (“The religious narrative corpus”), where it does in fact belong. The major bulk of this text is actually about Buddhist practice and belief, and discusses Buddhist concerns on a much wider scale. For a discussion of the Tibetan classifi cation of Nāgārjuna’s works see Seyfort Ruegg (1981: 7–9) and Williams (1984). 17 The YṢ and the ŚS off er rich insight in regard to Nāgārjuna’s soteriological views as well. This issue will not be discussed in the present context.

and non-existence: he believed all notions of existence to be rooted in ignorance. As he states in MMK 15.10: “Exists” is a grasping at eternalism. “Does not exist” is a view of annihilation. Therefore the wise should not base themselves on existence or non-existence. astīti śāśvatagrāho nāstīty ucchedadarśanam / tasmād astitvanāstitve nāśriyeta vichakṣaṇaḥ / This verse supplies an important defi nition of the middle path which avoids both existence and non-existence.18 Nāgārjuna is here extending the meaning of the traditional Buddhist defi nition of the middle as the path that avoids eternalism and annihilation (śāśvata and uccheda). These terms, which originally referred primarily to the nature of the self,19 now make an ontological statement about the nature of reality. A similar position is expressed in MMK 5.8: The slow-witted who see existence and non-existence of things do not see the auspicious quieting of objects. astitvaṃ ye tu paśyanti nāstitvaṃ cālpabuddhayaḥ / bhāvānām te na paśyanti draṣṭavyopaśamaṃ śivam // Again Nāgārjuna makes it more than clear that he believes any view, any actual seeing of existence or non-existence, to be mistaken. These verses alone should rule out the interpretations of Nāgār

A similar statement is given in ŚS 21: yod pa nyid na rtag nyid dang // med na nges par chad nyid yin // dngos po yod na de gnyis ’gyur // de phyir dngos po khas blang min // If there is existence there is eternalism, and if there is non-existence there is surely annihilation. When there are existent things, both occur, and therefore one should not accept existent things. 19 See my discussion of this issue in Shulman (2008: section III). In this context, it is of primary importance to notice that the terms used by the Kaccānagottasutta (SN II. 16–17) for the extremes, attitha and natthita, do not relate to abstract notions of existence but to particular ways of understanding the nature of the Self. In fact, for the Kaccānagottasutta, attitha and natthita represent sassata and uccheda (eternalism and annihilation). See also note 31 in the same article.

juna’s thought delineated at the outset of our discussion: Nāgārjuna denies non-existence and therefore cannot be a nihilist.20 He must not be expounding a vision of an absolute truth, since such a truth must exist.21 He is also making a defi nite philosophical statement regarding the nature of reality, which must not exist as it appears, and therefore his verses cannot be only of pragmatic (“upāyic”) value. Finally, the focus of the discussion must not be only language or knowledge, since that would imply an existent reality misrepresented by thought. If words or concepts are invalidated, surely the objects they refer to are unreal as well.22 In order for these verses to mean anything, they must be a description of reality itself, which is characterized as neither existent nor non-existent, neither absolutely true nor wholly false. Both of the verses quoted deny astitvam and nāstitvam, existence and non-existence, or better “is-ness” and “non-ness.” In other places Nāgārjuna prefers to target a more general notion of existence – bhāva. A most important example is the opening verse of the MMK (1.1):

Burton (1999: 90) has argued that Nāgārjuna’s thought is nihilistic even though Nāgārjuna did not believe so himself. See note 54 below for my response to such a position. 21 The discussion of Nāgārjuna’s view of absolute truth, or “the absolute,” cannot be fully developed in this paper. Clearly there are verses which suggest that Nāgārjuna accepted an ultimate and unconditioned reality, such as MMK 18.9 and 25.9. But the maṅgalaślokas of the MMK, as well verses such as MMK 7.32, 18.10, 25.19, 20, YṢ 5–6, and ŚS 30–32, strongly suggest that Nāgārjuna did not believe in ultimate truth in any absolute sense. These verses imply that Nāgārjuna used “absolutistic” terms such as tattva, dharmatā, and nirvāṇa, not as a description of an actual state, but rather as a poetic description of a truth that exists only in the realms of the imagination. 22 See also MMK 12.10, which explicitly states that the impossibility of suff ering arising from itself, from another, from both or without a cause, is true also in regard to external things (bāhyānām bhāvānām).

Not from themselves, not from another, not from both or without a cause, are arisen entities ever found, anywhere. na svato nāpi parato na dvābhyāṃ nāpy ahetutaḥ / utpannā jātu vidyante bhāvāḥ kva cana ke cana // Here Nāgārjuna argues against the truth of bhāvāḥ in the plural, and hence we must translate “entities” or “things.” Such a translation could lead us to believe that Nāgārjuna is arguing against “thingness,” against the diff erentiation of entities into distinct phenomena with clear-cut boundaries.23 Obviously, “thingness” is part of what Nāgār juna is targeting here, but it cannot contain all of his purpose. If things do not have any true boundary, any well-defi ned state of existence, any bhāva, they can not really be understood to exist. This point is expressed more clearly when Nāgārjuna refutes bhāva in the singular, as in YṢ 46: When one accepts existence, there are the arising of passion and hatred, the holding of bad and violent views, and the strife which comes from them. dngos por khas len yod na ni // ’dod chags zhe sdang ’byung ba yi // lta ba mi bzad ma rungs ’dzin // de las byung ba’i rtsod par ’gyur // rāgadveṣodbhavas tīvraduṣṭadṛṣṭiparigrahaḥ / vivādās tatsamutthāś ca bhāvābhyupagame sati //24 The YṢ continues to discuss the great misfortunes caused by believing existence to be true. This verse tells us that the refutation of “things” in the plural, is related to the refutation of “existence” in the singular.25 In fact, the Sanskrit allows a meaning unavailable in

Such a reading of Nāgārjuna, based on the MMK and the VV, was articulated by Streng (1967). 24 In quoting from the YṢ I will provide the Tibetan text, accompanied by the Sanskrit verse in the few cases in which it has been identifi ed in later sources. 25 The compound bhāvābhyupagame could obviously be read as referring to bhāva in the plural (probably bhāvānām abhyupagame), and thus the verse would relate to “things” and not to “existence.” MMK 21.14–15 tells us what Nāgārjuna probably means by this phrase, speaking of bhāvaṃ abhyupapan

English – “existences,” that is bhāvāḥ in the plural. We should note that Nāgārjuna’s argument against “things” is better understood to be a refutation of “states of existence.” Nāgārjuna denies the reality normally attributed to all that is, saying it does not exist in any true fashion. Some readers may have noticed that the lack of self-nature has yet to appear in any of the verses I have quoted thus far.26 There is no need to amend the message of these verses so as to deliver a meaning not theirs – they are denying existence, not self-nature.27

nasya in the singular. 26 It is true that following MMK 15.10, MMK 15.11 explains the relation between astitva/nāstitva and śāśvata/uccheda in terms of existence by way of svabhāva. But 15.11 should not cause us to read 15.10 as denying existence/non-existence only by way of svabhāva. Rather, Nāgārjuna is explaining that in order for something to exist, it would have to have svabhāva, and thus it would be “eternal.” 27 There are a number of additional considerations that should cause us to doubt whether svabhāva is so central a concept for the thought of the MMK. First, it should be noted that svabhāva is a rather rare concept in the MMK, especially if we do not over-emphasize the importance of chapters 17 and 24. In these two chapters, Nāgārjuna uses the notion of svabhāva in order to defend himself against his pūrvapakṣins. I suggest that these two chapters, much like the VV, should not be understood as a positive articulation of Nāgārjuna’s position, but rather as a defensive strategy in which Nāgārjuna exposes the problems that arise

once one accepts svabhāva. Aside from chapters 17 and 24, and chapter 15 which is an analysis of svabhāva, the term svabhāva appears only in verses 1.3, 7.16, 13.3–4, 20.21, 21.17, 22.2–4, 9, 14, 16 and 23.2, 6, 24–25. This means that the term sva bhā va is absent from 17 of the MMK’s 27 chapters (which amounts to no less than 63%)! Clearly, the notion of svabhāva is central to the MMK’s thought, but this observation alone should cast a doubt on the idea that the MMK as a whole is a refutation of svabhāva. Furthermore, if the text’s main objective was a refutation of svabhāva, surely the term should appear in its concluding chapter. But MMK 27 makes no mention of svabhāva. Also, the fact that Nāgārjuna devotes a separate chapter to svabhāva, just as he does to the skandhas and āyatanas or the terms saṃskṛta and kāla, implies that sva bhā va is part of what the MMK is analyzing, rather than being the focus of the discussion.

In fact, the YṢ makes it clear that the refutation of svabhāva is not an end in itself (contra dGe-lugs-pa exegesis), but rather the means by which existence is refuted. Once things are proven to lack a true nature of their own, there is nothing left to lack self-nature. What can be said to lack self-nature? When self-nature is refuted, nothing is left. As Nāgārjuna states in YṢ 19: What appears dependent on this and that does not arise by way of self-nature. What does not arise by way of self-nature – how can it be called ‘arisen’? de dang de brten gang byung de // rang gi dngos por skyes ma yin // rang gi dngos por gang ma skyes // de ni skyes zhes ji ltar bya // tat tat prāpya yad

utpannaṃ notpannaṃ tat svabhāvataḥ / svabhāvena yan notpannam utpannaṃ nāma tat katham // The fact that things arise in dependence proves they do not arise “svabhāvically.” But if they have not really arisen in any true way, how can they be said to have arisen? If there is no svabhāvic arising there is, in fact, no arising at all.28 And again, more bluntly: What appears together with causes does not abide without conditions, and is destroyed as a result of their absence – how can it be understood that ‘it exists’? gang zhig rgyu dang bcas ’byung zhing // rkyen med par ni gnas pa med // rkyen med phyir yang ’jig ’gyur ba // de ni yod ces ji ltar rtogs // hetutaḥ saṃbhavo yasya sthitir na pratyayair vinā / vigamaḥ pratyayābhāvāt so ’stīty avagataḥ katham // YṢ 39 Verse 39 re-states what verse 19 said about arising in terms of existence. What exists in dependence cannot exist! In this verse

is followed by a verse that makes a similar statement regarding extinction, a verse that hints at the implausibility of understanding nirvāṇa as “cessation.” rgyu zad nyid las zhi ba ni // zad ces bya bar rtogs pa ste // rang bzhin gyis ni gang ma zad // de la zad ces ji ltar brjod // YṢ 20 The calm (which results) from an extinction of a cause is understood as ‘extinction.’ What (exists) by way of self-nature does not become extinct. How can it be understood to be extinct?

example of the principle that there can be no existence without svabhāva is MMK 13.3:29 There is no self-nature of things, since change is perceived. The emptiness of things (is understood) from the fact that there are no things devoid of self-nature. bhāvānāṃ niḥsvabhāvatvam anyathābhāvadarśanāt / asvabhāvo bhāvo nāsti bhāvānāṃ śūnyatā yataḥ //30 What changes has no svabhāva. What has no svabhāva is empty, it does not exist. There is no such a thing that lacks svabhāva. This verse summarizes the stages we have seen so far by which Nāgārjuna’s dialectic proceeds: Because of (1) change (or dependence), things are understood to have (2) no self-nature. But nothing can exist without a true nature, and hence (3) things are empty, they do

Both Candrakīrti and Bhāviveka understand MMK 13.3 (and 13.4ab) to be voiced by a pūrvapakṣin (see Nietupski [1994]). It is obviously not easy to argue against such authorities. Candrakīrti’s and Bhāviveka’s reading is adopted also by Streng (1967) and Inada (1993). For Buddhapālita as well, the verse articulates the position of a Buddhist who equates emptiness with impermanence (Derge, dBu ma, vol. 1, 3842, 217.2–218.2). It is obviously not easy to argue against such authorities. Nevertheless, in light of the many other places in which Nāgārjuna makes arguments similar to the ones I am outlining here, such as the ones I have quoted and the ones I will quote below (most importantly MMK 13.7 which appears in the same chapter), I believe my reading is more than plausible. Even if this may not be “what Nāgārjuna intended” by the verse, it is fully consistent with his overall system. In any case, the third pāda of the verse – asvabhāvo bhāvo nāsti (“There is no thing devoid of self-nature”) – emphasizes the point I am making regarding the meaning of svabhāva even if it is understood as a pūrvapakṣin’s claim: Once there is no self-nature, no existent thing remains. 30 Another translation that could be off ered here (substituting the referent of yataḥ) is “there is no thing lacking self-nature, because of the emptiness of things.” This translation would not change the meaning of the lack of selfnature I am discussing here.

not exist. We see in this verse that there is a qualitative diff erence between lacking self-nature and being empty. Because things lack self-nature, they are empty.31 This same point is made in the Vṛtti to VV 1: Since there is no self-nature anywhere (in any of its conditions), the sprout lacks self-nature. Because it lacks self-nature it is void. yasmād atra sarvatra svabhāvo nāsti tasmān niḥsvabhāvo ’ṅkuraḥ / yasmān niḥsvabhāvas tasmāc chūnyaḥ / If emptiness is equal to the lack of self-nature, the second sentence of this passage would be both tautological and meaningless. We see that emptiness results from the lack of self-nature, a statement quite distinct from the one which says that emptiness is emptiness of self-nature.32 This same point is made again in the commentary to VV 57, where Nāgārjuna adds that if something is empty, in this case a name, it is unreal:33 And also, because of the non-existence of the self-nature of things, the name lacks self-nature. Therefore it is empty. Because of its emptiness it is unreal. tad api hi bhāvasvabhāvasyābhāvān nāma niḥsvabhāvaṃ tasmāc chūn yaṃ śūnyatvād asadbhūtam /

See also Nagao (1991: p. 191). Nagao believes that the formulation “because it is devoid of self-being it is empty” was produced by Nāgārjuna’s Indian commentators in order to explain why the dependently originated is said to be empty and non-existent. As we see here, this formulation had already been introduced by Nāgārjuna himself. 32 There are clearly many places where Nāgārjuna speaks of emptiness of self-nature, such as ŚS 67. This is also the general drift of MMK 24. Nonetheless, the existence of such passages does not mean that lack of svabhāva is the sole meaning of emptiness. The refutation of svabhāva is, in fact, the main avenue by which the fuller meaning of emptiness is reached, and thus it comes as no surprise that Nāgārjuna speaks of the emptiness of self-nature as well. 33 The statement that what lacks svabhāva is unreal (asadbhūta) is made not only in the commentary but in the verse as well.

Again – what can lack svabhāva? Once there is no self-nature, there is nothing left to lack its own nature, an insight Nāgārjuna expresses in what may be the acme of the MMK: If there were anything non-empty, there could be something empty too. And there is no non-empty thing – how will there be something empty? yady aśūnyaṃ bhavet kiṃ cit syāc chūnyam api kiṃ cana / na kiṃ cid asty aśūnyaṃ ca kutaḥ śūnyaṃ bhaviṣyati // MMK 13.7 This verse is followed by the famous denial of the possibility of taking emptiness itself as a true view of reality (13.8). MMK 13.7 tells us that once the thing is empty, there is nothing left that is empty. No existence remains after the Madhyamaka dialectic penetrates its object of scrutiny. Not only does the object lack self-nature, it is unreal and has no true existence. This is probably the right moment to re-affi rm that I do not believe that Nāgārjuna was a nihilist, intentionally or by default. Nāgārjuna denied the validity of notions of non-existence,34 and found non-existence to be morally dangerous.35 But before we ask ourselves how Nāgārjuna escapes nihilism, and before I present a positive defi nition of Nāgārjuna’s vision of the middle, I would like to push my point a little further and discuss two common intuitions about Nāgārjuna which I believe are rooted in error. I am referring to the role the traditional Buddhist insights of impermanence and dependence, as well as the notion of the two truths, play in Nāgārjuna’s thought.

MMK 5.6: When there is no existence, whose non-existence will there be? (avidyamāne bhāve ca kasyābhāvo bhaviṣyati). See also MMK 15.5 and 25.7. 35 YṢ 2ab: First refute non-existence, the source of all faults. (re zhig nye kun ’byung ba’i gnas // med nyid rnam par bzlog zin gyis /) CREATIVE IGNORANCE 153

III. Nāgārjuna’s innovations

In modern interpretations of Nāgārjuna, one often encounters the idea that Nāgārjuna attempted to retrieve the Buddha’s original message in response to dogmatic tendencies which prevailed in the Buddhism of his day, primarily in Abhidharma traditions. Such a position generally argues that for Nāgārjuna, emptiness is a diff erent way of saying impermanence and/or dependence.37 There are many problems with such an interpretation, among them the fact that the Buddha did not characterize all things as dependent,38 and that Nāgārjuna has much in common with Abhidharma traditions.39 In the present context I wish to concentrate

It is clear that Nāgārjuna is not responsible on his own for the ideational developments I will discuss in this section. Many Mahāyānasūtras (e.g. the Śūraṃgamasamādhisūtra and the sūtras of the Prajñāpāramitā genre) express these same doctrinal shifts. Nonetheless, Nāgārjuna is a clear exponent of these new Mahāyāna visions, and he off ers a full logical exposition of them. In this sense he is an emblematic fi gure who represents the innovations produced by thinkers and meditators in the earlier stages of the Mahāyāna. 37 The classic proponent of such a thesis may be Kalupahana (1986), although this view of Nāgārjuna is endorsed by many scholars, such as Gombrich (1996: 32) and Ronkin (2005:

200). Understanding emptiness as a synonym of dependence and or impermanence is also deeply rooted in the Zen reading of Nāgārjuna, and is central to dGe-lug-pa presentations of Madhyamaka as well (e.g. Garfi eld [1994, 1995]). 38 See my discussion of this issue in Shulman (2008). See also in Cox (1993). 39 Nāgārjuna shares with the Ābhidharmikas the conviction that a thorough and detailed analytic inquiry into the nature of reality can bring one to a vivid vision of truth. Moreover, I suggest that Nāgārjuna accepted traditional Abhi dharma insight and method, but wished to specify how this tradition’s basic concepts should be understood. I believe this is a better option than saying Nāgārjuna rejected Abhidharma thought and viewed it as a corruption of the Buddha’s message. These remarks are clearly not intended as a fi nal statement on the matter, and serve only as initial observations to be explored in the future. See also

only on the fact that Nāgārjuna directly refuted both impermanence and dependence, since both imply existence.40 When all existence is empty, there is nothing there to be impermanent, as he says in MMK 25.22–23: All phenomena being empty – what is endless, what has an end? What has and doesn’t have an end? What does not have nor not have an end? What is the same? What diff erent? What eternal? What ephemeral? What both eternal and ephemeral? What neither? śūnyeṣu sarvadharmeṣu kim anantaṃ kim antavat / kim anantam antavac ca nānantaṃ nāntavac ca kiṃ // MMK 25.22 kiṃ tad eva kim anyat kiṃ śāśvataṃ kiṃ aśāśvataṃ / aśāśvataṃ śāśvataṃ ca kiṃ vā nobhayam apy ataḥ // MMK 25.23 Or again, more

cogently: If everything is impermanent, and impermanence is also not permanent, how will there be permanent or impermanent things? thams cad mi rtag yang na ni // mi rtag pa yang rtag pa med // dngos po rtag dang mi rtag nyid // ’gyur na de lta ga la yod // ŚS 58 The same problem that Nāgārjuna identifi es in regard to the lack of svabhāva, applies to impermanence as well: Just as there must be something existent to be characterized as devoid of self-nature, there must be something permanent to be characterized as impermanent, or something independent to be characterized as dependent. This point is made explicit in the YṢ in regard to dependence: Those who are attached to the self and the world (and see them as) non-dependent – Oh! They are confused by views of permanence and impermanence.

Saito’s (2007: 158) reference to Nāgārjuna as “the founder or originator of the Mahāyāna-Abhidharma movement,” as well as the discussion in Williams (1988: 8), Hayes (1994: 361) and Walser (2005: chapters 6 and 7). 40 For a fuller discussion of this issue see Shulman (2008a).

Those who accept that being dependent, things are established in reality – how will the faults of permanence and the like not appear for them as well?! Those who accept that being dependent, things are like the moon on the water, neither true nor false, are not confused by views. gang dag gis ni ma brten par // bdag gam ’jig rten mngon zhen pa // de dag kye ma rtag mi rtag // la sogs lta bas ’phrogs pa yin // YṢ 43 gang dag brten nas dngos po rnams // de nyid du ni grub ’dod pa // de dag la yang rtag sogs skyon // de dag ji ltar ’byung mi ’gyur // YṢ 44 gang dag brten nas dngos po rnams // chu yi zla ba lta bur ni // yang dag ma yin log min par // ’dod pa de dag ltas mi ’phrogs // YṢ 45 Verse 43 attacks the non-Buddhist position which denies that all exists in dependence. The key verse is the following one (44), which attacks Buddhists who believe dependent things really to exist. Prior to these three verses, a similar claim was made regarding impermanence, where Nāgārjuna again attacks Buddhists who do not realize that impermanence denies the possibility of existence.41 Later on he again says that: What is born in dependence is unborn, said the best among knowers of reality. brten nas skye ba ma skyes par // de nyid mkhyen pa mchog gis gsungs // pratītya jātaṃ cājātam āha tattvavidāṃ varaḥ // YṢ 48cd Many more examples can be supplied in order to further substantiate the position that Nāgārjuna believed that emptiness empties

41 gal te yod par smra ba rnams // dngos mchog zhen nas gnas pa ni // lam de nyid la gnas pa ste // de la ngo mtshar cung zad med // YṢ 40 sangs rgyas lam la brten nas ni // kun la mi rtag smra ba rnams // rtsod pas dngos rnams mchog bzung bas // gnas pa gang yin de rmad do // YṢ 41 It is not at all surprising that those who speak of existence (astitvavādin) abide grasping at things as they abide on their path. But it is quite amazing that those who rely on the path of the Buddha, who speak of impermanence with regard to everything, abide as they grasp at things through strife.

impermanence and dependence, which both can only be viewed from the extreme of existence.42 The fact that there is nothing there to be impermanent or dependent should cause us to be very cautious with regard to the way we understand Nāgārjuna’s use of the theory of the two truths. Most often, this theory is used in order to re-affi rm the validity of the phenomenal world, in an attempt to balance the intensity of Nāgārjuna’s dialectic of emptiness. It seems that such a reading of Nāgārjuna may be no more than a futile eff ort to avoid the deep and thorough refutation of existence he conducts. We may be convinced by now that according to Nāgārjuna there really are no true phenomena that exist “conventionally” and are “ultimately empty.” The concept of the two truths is valuable as a reminder that Nāgārjuna is not affi rming non-existence, but should not be seen as a positive description of reality. Rather, what MMK 24.8–10, the locus classicus for the discussion of the two truths, actually say is that the Buddha’s teachings are useful in order to facilitate realization.43 This statement is corroborated by YṢ 21–22 and 30–33, that explain that basic Buddhist concepts amount to useful fi ctions.44

Regarding impermanence, see, for example, the discussion in chapters 7, 11 and 21 of the MMK, where Nāgārjuna makes it fully clear that he fi nds the ideas of origination and cessation, and thus of impermanence, to be unreasonable. The same idea is central to the YṢ (See, for example, the maṅgalaśloka and verse 18). I believe we should read these statements as they were phrased, rather than forcing them to say something else (“ultimately” or “conventionally”). I devote fuller attention to this issue in Shulman (2008a). Regarding dependence, see MMK 10.8–11, where Nāgārjuna clearly states that in order for the dependent to exist, it would have to be established prior to its dependence. Once two things depend on each other, neither of them is established. 43 A similar claim has been made by Wood (1994: ch. 5). 44 In YṢ 21–22 Nāgārjuna states that since there exist no arising or ceasing, the concept of impermanence was taught only for the practical purpose (kāryavaśāt, dgos pa’i don) of facilitating realization. YṢ 33 states again that the concepts of “I and mine” and of the skandhas, dhātus and āyatanas were

For brevity’s sake, I will quote only two verses from the ŚS. First, the opening verse of the text: Abiding, arising and ceasing, existence and non-existence, low, middle and superior – the Buddha spoke of these under the power of worldly convention, not under the power of truth. gnas pa’am skye ’jig yod med dam // dman pa’am mnyam pa’am khyad par can // sangs rgyas ’jig rten snyad dbang gis // gsung gis yang dag dbang gis min // ŚS 1 The basic concepts we employ in describing existence should not be understood to refl ect the truth, but only conventional agreement. Such conventions cannot be real, since we would need to specify an existent phenomenon that could be defi ned as empty.45 This would contradict the major thrust of Nāgārjuna’s argument and the explicit statements of all the verses we have examined. Near the end of the text, Nāgārjuna summarizes his discussion and defi nes his position regarding the two truths: The worldly principle “this arises in dependence on that” is not denied. (But) also – What is dependent has no self-nature, and hence – how could it exist? Understand this correctly! ’di la brten nas ’di ’byung zhes // ’jig rten tshul ’di mi ’gog cing // gang brten rang bzhin med pas de // ji ltar yod ’gyur de nyid nges // ŚS 71 There is truth in the way people see the world; thought is not totally mistaken in its analysis of experience. But once dependence is recognized, it should lead to the conclusion that nothing can exist. Again we encounter the three step procedure of Nāgārjuna’s dialectic: because of dependence, there is no svabhāva, and therefore there is no existence.

also taught for such practical purposes. 45 See in this regard ŚS 2ab: bdag med bdag med min bdag dang // bdag med min pas brjod ’ga’ ’ang med / There is no self, no non-self, no self and non-self. Therefore there is nothing which can be expressed.

We must now ask ourselves what this severe deconstruction of existence, that I awkwardly insist does not lead to non-existence, actually means.

IV. Creative ignorance

We have now reached the heart of our discussion, the attempt to come to terms with Nāgārjuna’s deep and total denial of existence. There is, according to this vision, nothing truly out there in the world. Nonetheless, we are not in a non-existent void but can actually discuss the meaning and value of our experience. How can a world that is not existent or non-existent (or both or neither) be described? In other words, how is it that a non-existent reality comes into being? Surprisingly enough, the YṢ and ŚS supply a rather straightforward answer to these questions, explaining that the world is created out of ignorance, as a result of processes of conceptualization. The clearest statement in this regard is YṢ 37: Since the buddhas have said that the world47 has ignorance for its condition, does it not follow that this world is a mental construction?

Initial articulations of the reading of Nāgārjuna presented in this section have previously appeared in de la Santina (1987: 174) and Tola & Dragonetti (1995: xxix), and more importantly in Scherrer-Schaub (1991: 252–259, n. 492). Regrettably, none of these sources developed these points much beyond the level of preliminary observations. Siderits (2004) has provided some of the logical basis for the reading of Nāgārjuna suggested here, although he bases his discussion on MMK 1 alone. Burton (1999: ch. 4) defi ned certain elements of Nāgārjuna’s thought in a similar way to my treatment of them, although I believe he has misunderstood Nāgārjuna’s basic position (see note 54 below). 47 Candrakīrti explains “the world” to be equal to “the fi ve aggregates of clinging” (’jig rten ni nye bar len pa’i phung po’i nga rnams so, ScherrerSchaub [1991:77]). The same defi nition is given also in chapter 12 of the Aṣṭa (see Conze [2006 (1973: 173)]). Although this statement seems to suggest that only the world of subjective experience is conditioned by vikalpa, we must note that there is, for Nāgārjuna, no “world” which exists beyond the fi ve ag

’jig rten ma rig rkyen can du // gang phyir sangs rgyas rnams gsungs pa // de yi phyir na ’jig rten ’di // rnam rtog yin zhes cis mi ’thad // YṢ 37 The world is a mental construction, an act of creative imagination, a vikalpa propelled by ignorance. The following verse strongly suggests that everything depends on ignorance: That which ceases when ignorance ceases, how can it not be clear that is an imagination constructed out of mis-knowledge? ma rig ’gags par gyur pa na // gang zhig ’gog par ’gyur ba de // mi shes pa las kun brtags par // ji lta bu na gsal mi ’gyur // YṢ 38 When ignorance will cease to be, it seems that the world will not be there either. What appears to exist is constructed by our own imagination, out of ignorance. Nāgārjuna, if I understand him correctly, is asking why we believe, given that our perception of the world is colored by

ignorance, that the world is true? How is it that our very knowledge of the world’s existence is not created by ignorance? Moreover, when we realize that the world is conditioned by ignorance, why is it that we don’t realize it to be an act of creative, ignorant imagination? What this means is not that our perception or ideation of things mistakenly constructs a mental image it replaces for a true object. Rather, the object itself is constructed by ignorance, since there is nothing objectively there independent of ignorant perception. Earlier in the YṢ Nāgārjuna has stated twice that the true vision of reality means seeing that things are born of ignorance. The fi rst instance is YṢ 10: When true knowledge sees the appearance conditioned by ignorance, no arising or ceasing is perceived. ma rig rkyen gyis byung ba la // yang dag ye shes kyis gzigs na // gregates. More importantly, the rūpa aggregate traditionally includes all that is material, and specifi cally the 6 perceptual objects. One would probably not want to argue that there is no relation between the “real” physical object and its representation in perception. Examples of the emphasis on the physical aspect of the rūpa aggregate can be found in Majjhima Nikāya i185–190, i88–90, i421–423. See also Gethin’s (1986) treatment of the fi ve aggregates in the nikāyas and early Abhidhamma.

skye ba dang ni ’gags pa’ang rung // ’ga’ yang dmigs par mi ’gyur ro // Nāgārjuna goes on to state that “this is nirvāṇa and the seeing of reality in this very life, what is to be done has been done” (YṢ 11ab: de nyid mthong chos mya ngan las / ’das shing bya ba byas pa’ang yin). YṢ 10 is based on a delicate play of meaning. In traditional Buddhist exegesis “appearance conditioned by ignorance” refers to the 12 links of conditioned arising, the descriptions of the process by which saṃsāric transmigration proceeds. The causational principle underlying this process is based on each link conditioning the arising of the following one, or, when it is absent, conditioning its ceasing.48 But Nāgārjuna envisions a very diff erent picture: When one rightly observes the conditioning of ignorance – he sees no arising and ceasing! This is because he understands that what seems to be real is actually not much more than a fantasy, and therefore that it does not truly arise or cease. Nāgārjuna is hereby articulating a fully new import for “appearance conditioned by ignorance.” Nothing whatsoever undergoes arising and ceasing, because all such things are not really there, they are fi ctions produced by ignorance. Nāgārjuna will make this point again in verse 26 where he asserts that “the knowers of things” (dngos po la mkhas pa rnams gyis, verse 25) know them to “appear caused by ignorance” (ma rig rgyu las shin tu byung). It is tempting to try to read these verses as describing the nature of experience, rather than characterizing existence in general. But in YṢ 34 Nāgārjuna declares he believes the physical-materialobjective reality to be dependent on consciousness: Things spoken of, the great elements and so forth,49 are enclosed in consciousness. When this is understood, they dissolve. Indeed, they are a mistaken construction.

For an exceptionally strong statement of this principle see SN ii105. 49 For an explanation of mahābhūtādi ( ’byung ba che la sogs pa), see Scherrer-Schaub (1991: n. 492, p. 256).

’byung ba che la sogs bshad pa // rnam par shes su yang dag ’du // de shes pas ni ’bral ’gyur na // log pas rnam brtags ma yin nam // mahābhūtādi vijñāne proktaṃ samavarudhyate / tajjñāne vigamaṃ yāti nanu mithyā vikalpitaṃ // YṢ 34 The elements are “checked by” or “enclosed in consciousness” (vijñā ne samavarudhyate, rnam par shes su yang dag ’du).50 They can be dissolved when this is understood, and hence are not objectively real but depend on consciousness for their being. They are further defi ned as a mistaken mental construction (mithyā vikalpitam, log pas rnam brtags). When we realize that Nāgārjuna understood things to rise out of ignorance, we can better understand his intention in describing them as being similar to illusions, dreams, phantoms, cities of gandharvas, and the like. This is a central feature of Nāgārjuna’s thought, which he expresses in diff erent verses and contexts.51 A good example is ŚS 66: Conditioned things52 are like a city of gandharvas, an illusion, a phantom, hairs (seen by a person suff ering from a cataract), a bubble in the stream, a magical display, a dream and a whirling fi re-brand.

he meaning of samavarudhyate is not fully clear. In this context the verb could imply either that the elements etc. are made of consciousness, or, more probably, that they depend on consciousness in order to be. Numerous translations for samavarudhyate have been presented, such as Lindtner (1986: 83): “made to cohere in consciousness,” Ichigo (1989: 155) and Loizzo (2001: 506) “reduced to consciousness,” Scherrer-Schaub (1991: 252) and Tola & Dragonetti (1995: 38): “contained in consciousness,” and Jinpa: “absorbed in consciousness.” See further Scherrer-Schaub (1991: 254–255). What is important to note about samavarudhyate in YṢ 34 is that it probably does not make the strong idealistic claim that reality consists only of mind. 51 See MMK 7.34, 17.31–33, 23.8; ŚS 14, 36, 40–42, 66; VV 65–67; YṢ 15–17, 27, 56. 52 This verse speaks of ’du byed (saṃskāra) in the sense of ’dus byas (saṃskṛta), a use familiar from diff erent Buddhist contexts (see, for instance, Boisvert [1995: 93–95]). One should note that the technical and generally subconscious meaning of saṃskāra makes little sense in this verse. It should

’du byed dri za’i grong khyer dang // sgyu ma smig rgyu skra shad dang // dbu ba chu bur sprul pa dang // rmi lam mgal me’i ’khor lo mtshungs // A similar idea is expressed in YṢ 17 as well: When one understands that existence53 is like a mirage and an illusion, one is not polluted by views of the extremes of a beginning or an end. srid pa smig rgyu sgyu ’dra bar // blo yis mthong bar gyur pa ni // sngon gyi mtha’ ’am phyi ma’i mtha’ // lta bas yongs su slad mi ’gyur // Things are unreal, but nonetheless appear. This appearance does in fact occur, but has no substantial reality to it. Moreover, as we have learned, the appearance is conditioned by ignorance and caused by conceptualization. This is why it is similar to an illusion, a dream or a mirage, phenomena which are created mentally without having any true objective support. This is, in fact, Nāgārjuna’s vision of the middle way, in which appearance is neither truly existent nor fully denied.

be clear as well that this verse is not intended as an affi rmation of an unconditioned (asaṃskṛta) truth, argued against earlier in verses 30–32 of the same text (ŚS 32): ’dus byas dang ni ’dus ma byas // du ma ma yin gcig ma yin // yod min med min yod med min // mtshams ’dir sna tshogs thams cad ’dus // Conditioned and unconditioned are neither one nor many. They do not exist, not-exist or (both) exist and not-exist. The whole variety (of things) is included in these boundaries. 53 In this verse, “existence” translates the more particular srid pa (bhava), rather than the more abstract bhāva (yod pa or dngos po). 54 This formulation of the middle way also answers Burton’s (1999: ch. 4) claim that the fact that Nāgārjuna believes reality to be a mental construction leads his view to Nihilism. I would argue the opposite view to Burton’s: the fact that Nāgārjuna understands reality to be conditioned by subjectivity demands a great degree of moral responsibility of people, since man naturally conditions and creates his own reality. According to this view, morality is not only validated, but enforced. The argument could be made that only in an empty world is morality understood to be not only a necessary, but even a constitutional force. On an ontological level, there is obviously a diff erence between

Thus far I have been quoting mainly from the YṢ. The ŚS discusses the creative capacity of the mind somewhat diff erently. First, it connects illusory existence to karma.55 Verses 33–43 are devoted to a discussion of karma, in which Nāgārjuna shows that karma lacks svabhāva. Of primary importance for our discussion are the conclusions the ŚS draws from showing karma to lack svabhāva. Just as the victorious Tathāgata creates a magical manifestation by way of his magical power, and that same magical manifestation in turn creates another magical manifestation, In such a case, the manifestation (created by) the Tathāgata is empty, and what need we say about the manifestation (created) by the manifestation? Both exist only as names,56 and are wholly conceptiononly. In just the same way the agent is like the manifestation, and his act like the manifestation created by the manifestation. What is empty of self-nature in every bit, is conception-only. ji ltar bcom ldan de bzhin gshegs // rdzu ’phrul gyis ni sprul pa sprul // sprul pa de yis slar yang ni // sprul pa gzhan zhig sprul gyur pa // ŚS 40 de la de bzhin gshegs sprul stong // sprul pas sprul pa smos ci dgos // gnyis po ming tsam yod pa yang // ci yang rung ste rtog pa tsam // ŚS 41 de bzhin byed po sprul dang mtshungs // las ni sprul pas sprul dang mtshungs // rang bzhin gyis stong gang cung zad // yod pa de dag rtog pa tsam non-existence and the statement that what exists is constructed by the mind. 55 las ni rkyen skyes yod min zhing // rkyen min las skyes cung zad med // ’du byed rnams ni sgyu ma dang // dri za’i grong khyer smig rgyu mtshungs // ŚS 36 Karma does not arise from conditions, and it does not arise at all from non-conditions. Conditioned things are like illusions, a city of gandharvas, and mirages. 56 The characterization of things as “only names” (ming tsam) is absent from the svavṛtti version of the text.

Karma is similar to a magical manifestation.57 Anything that appears due to karmic conditioning is “conception-only,” merely a name. The rationale of this insight is defi ned in verse 42: “What is empty of self-nature in every bit, is conception-only.” This is, in fact, exactly what I have been arguing that the lack of self-nature means – when there is no true existence of itself, reality proves to be a conceptualization. The ŚS informs us that this conceptualization is caused not only by ignorance, but also by karma. The discussion of karma in the ŚS concludes with a statement regarding the enigmatic nature of existence (verse 44), following an elaborate discussion of the problems Nāgārjuna

identifi es in defi ning the perceptual process (verses 45–57). The argument is too complex to be treated fairly in this context, since it rests on a very challenging and counter-intuitive assumption: Nāgārjuna seems to believe that if we cannot supply a coherent defi nition for the way perception functions, every experience, every act of knowledge and every object are proven to be unreal.58 I hope to give the intricate arguments of the ŚS fuller attention in another context. For now it will suffi ce if we note the intimate relation Nāgārjuna intuits between defi nition and reality. Of even greater importance in the present context are the formulations he provides at the end of this discussion, in which he defi nes the creative power of conceptualization. First he states that the kleśas lack self-nature, since they are conditioned by pleasant and unpleasant sensation.59 Next he states:

The discussion of karma in chapter 17 of the MMK reaches the same conclusion. In 17.31–33, Nāgārjuna uses the same image as in ŚS 40–42 in order to defi ne karma as being similar to an illusion. 58 The most signifi cant verse in this regard is ŚS 51: mig blo mig la yod min te // gzugs la yod min par na med // gzugs dang mig la brten nas de // yongs su rtog pa log pa yin // Eye-consciousness is not in the eye, in the object or between the two. What depends on the form and the eye is a mistaken conception. 59 ŚS 59: sdug dang mi sdug phyin ci log // rkyen las chags sdang gti mug dngos //

Because desire, anger and ignorance are directed toward one and the same thing, they create it through conceptuality. That conception, too, is unreal.60 The conceived object does not exist, and without it – how will there be conception? Therefore the conceived and the conception, because they arise from conditions, are truly empty.61 gang phyir de nyid la chags shing // de la zhe sdang de la rmongs // de phyir rnam par rtog pas bskyed // rtog de’ang yang dag nyid du med // ŚS 60 brtag bya gang de yod ma yin // brtag bya med rtog ga la yod // de phyir brtag bya rtog pa dag // rkyen las skyes phyir stong pa nyid // ŚS 61 Nāgārjuna understands the functioning of conceptuality in a surprising manner. Rather than conceptuality being an attempt to defi ne and understand reality, Nāgārjuna sees conceptuality as responsible for the creation of reality. Things are not objectively “out there,” but are brought into being by ideation.

’byung phyir chags sdang gti mug dang // rang bzhin gyis ni yod ma yin // The phenomena of desire, anger and ignorance arise conditioned by the mistaken perceptions of pleasant and unpleasant. Therefore desire, anger and ignorance do not exist by way of self-nature. 60 This verse could be read as a statement that conception creates the kleśas rather than the kleśas creating the object. The reason I believe my translation – which stresses that the kleśas create the object – to be more convincing, is that the next

verse begins with the statement “The conceived does not exist…” (brtag bya gang de yod ma yin), a fact which must have been referred to in the previous verse. Also, in a Nāgārjunian world, the fact that something is experienced through the distorting lenses of the kleśas clearly implies that it is unreal. I wish to thank Prof. Ernst Steinkellner and Prof. Akira Saito for carefully reading this important verse, as well as the following verse, with me. 61 We could translate “emptiness itself” (stong pa nyid), but following the svavṛtti I translate “are empty” (stong pa yin). The nyid which ends the verse appears to be a translation of eva. It is diffi cult to decide whether the original Sanskrit reads śūnyatā-eva or śūnya/e/ā-eva.

What leads Nāgārjuna to conclude in these verses that objects are created in the manner they are envisioned by the mind? Nāgārjuna’s analysis leads him to the conviction that there is no true existence; the object is not real. Observing that experience is manifold, as objects take diff erent forms (in this case they are experienced through the threefold division of the kleśas), Nāgārjuna realizes that it is ideation which creates the object. There exists no unitary reality which conditions experience, and hence the objects of experience, which appear to be unitary, are created as part of the way they are envisioned by consciousness. They are not actually perceived, but rather, are

projected as part of the “perceptual” process. For Nāgārjuna, it is not the object which conditions experience, but experience which conditions the object. The logic Nāgārjuna is employing in this case rests on the well-known “one or many” argument: The object cannot have a unitary or a manifold nature. Once it appears in diff erent ways, the Mādhyamika views it as a result of the way it has been conceived.62 Moreover, once things are proven to be brought into being by the power of ideation, that ideation itself is realized to be unreal as-well, since it perceives objects which are not really there. Emptiness is said to be the play of unreal conceptualization perceiving unreal objects. The description of reality as “conception-only” in the ŚS is highly signifi cant. It may remind us of Vasubandhu’s statement at the opening of his Viṃśatikā: “In the Mahāyāna these three worlds are established as being mere fi gments of consciousness” (mahāyāne trai dhātukaṃ vijñaptimātraṃ vyavasthāpyate). ŚS 61 is also remarkably similar to Madhyāntavibhāga 1.3 and 1.6, and to Tri sva

bhāvanirdeśa 36.63 Some readers will possibly be worried that Nāgārjuna has turned into a Yogācārin.

62 The non-unitary nature of the object, which serves as a proof of its logical and ontological impossibility, is central to the logic employed by Nāgārjuna in the ŚS. See verses 46 and 50 for examples of this principle. 63 I quote here only the last of these 3 verses, Trisvabhāvanirdeśa 36:

I believe that to a great extent such an understanding is true.64 In a future publication I wish to provide a complementary discussion, which will show that Vasubandhu was a sort of a Mādhyamika. In my mind, in the earlier stages of their evolution the Madhyamaka and Yogācāra shared a very similar intuition about reality, understanding it to be an empty presentation determined by conscious and unconscious processes of conceptualization.65 cittamātropalambhena syāc jñeyārthānupalambhatā / jñeyārthānupalambhena syāc cittānupalambhatā // From the perception of mind-only, there should be the non-perception of knowable things. From the non-perception of knowable things, there should be the non-perception of mind. See also ŚS 57 for another remarkably similar statement by Nāgārjuna. 64 For further discussion of the intimate relations between Madhyamaka and Yogācāra thought, albeit in a diff erent context than the one developed here, see Nagao (1991: ch. 13)

and King (1994). It is also interesting to note in this regard the way Saito (2007) refers to Nāgārjuna as “…the founder or originator of the Mahāyāna-Abhidarma movement, that was later developed by the so-called Yogācāras and Mādhyamikas” (p. 158). Saito refers also to the commentaries written by early Yogācāra masters on the MMK. His words suggest the possibility that there was a viable Yogācāra reading of Nāgārjuna, now forgotten due to the immense infl uence thinkers such as Buddhapālita, Candrakīrti and Tsong-kha-pa exert on our understanding of Madhyamaka. The possible, or even natural synthesis of Madhyamaka and Yogācāra is also attested by Śāntarakṣita’s so-called Yogācāra-Madhyamaka school. This school was particularly dominant during the early stages of Tibetan Madhyamaka (see Seyfort Ruegg [2000: section one]). 65 It is most important to note that when Nāgārjuna or Vasubandhu identify the creative role of mental forces, this should not be meant to imply they believed people can control the processes of creation propelled by their own minds. Karma, ignorance and mistaken forms of imagination and conceptuality (vikalpa, vijñapti, abhūtaparikalpa) are clearly not willful, and are to a large extent unconscious. This explains not only why we cannot create at will, but also why we make perceptual and ideational mistakes.

V. Conclusion

The basic argument developed in this paper was that for Nāgārjuna, the fact that phenomena lack svabhāva implies that they are created by ignorance through processes of conceptualization. When nothing exists, as it has no true nature, it cannot be independent of the way it is known or perceived. The dialectic of Emptiness shows things to be a sort of a “real illusion.” Phenomena are not really there in any objective or substantive sense. Nonetheless, they do appear, and hence are understood to be “like an illusion, like a dream, like a city of gandharvas.”66

Abbreviations

JAOS Journal of the American Oriental Society JIP Journal of Indian Philosophy MMK Mūlamadhyamakakārikā – cf. n. 9 PEW Philosophy East and West ŚS Śūnyatāsaptati – cf. n. 11 VV Vigrahavyāvartanī – s. Bhattacharya et. al 1978 WZKS Wiener Zeitschrift für die Kunde Südasiens und Archiv für indische Philosophie YṢ Yuktiṣaṣṭikākārikā – s. Scherrer-Schaub (1991)

References

Bhattacarya, K., Johnston, E. H., and Kunst, A. (1978). The Dialectical Method of Nāgārjuna: Vigrahavyāvartanī. Delhi: Motilal Banarsidass. Boisvert, M. (1997). The Five Aggregates: Understanding Theravāda Psychology and Soteriology. Delhi: Sri Satguru Publications.

MMK 7.34ab: yathā māyā yathā svapno gandharvanagaraṃ yathā.

Burton, D. F. (1999). Emptiness Appraised: A Critical Study of Nāgārjuna’s Philosophy. Great Britain: Curzon. Conze, E. (tr.) (2006 [1973]). The Perfection of Wisdom in Eight Thousand Lines & its Verse Summary. San Francisco: City Lights. Cox, C. (1993). “Dependent Origination: Its Elaboration in Early Sarvāstivādin Abhidharma Texts.” In: Sharma, R. K. (ed.). Researches in Indian and Buddhist Philosophy: Essays in Honour of Professor Alex Wayman. Delhi: Motilal Banarsidass: 119–142. Darling, G. J. (1987). An Evaluation of the Vedāntic Critique of Buddhism. Delhi: Motilal Banarsidass. de Jong, J. W. (1977 [1958]). Mūlamadhyamakakārikā of Nāgārjuna. Madras: Adyar Learning and Research Center. de Jong, J. W. (1972). “The Problem of the Absolute in the Madhyamaka School.” JIP 2: 1–6. de la Santina, P. (1987). “The Madhyamaka Philosophy.” JIP 15: 173–185. Dragonetti, C.

(1978). “The Pratītyasamutpādahṛdayakārikā and the Pratītyasam utpādahṛdayavyākhyāna of Śuddhamati.” WZKS 22: 87–93. Dragonetti, C. (1986). “On Śuddhamati’s Pratītyasamutpādahṛdayakārikā and on Bodhicittavivaraṇa.” WZKS 30: 109–122. Erb, F. (1997). Śūnyatāsaptativṛtti: Candrakīrtis Kommentar zu den “Siebzig Versen über die Leerheit” des Nāgārjuna (Kārikās 1–14): Einleitung, Übersetzung, textkritische Ausgabe des Tibetischen und Indizes, Tibetan and Indo-Tibetan Studies 6. Franz Steiner Verlag Stuttgart. Ganeri, J. (2001). Philosophy in Classical India: The Proper Work of Reason. London and New York: Routledge. Garfi eld, J. L. (1994). “Dependent Arising and the Emptiness of Emptiness: Why did Nāgārjuna Start with Causation?” PEW 44/2: 219–250. Garfi eld, J. L. (1995). The Fundamental Wisdom of the Middle Way: Nā gārjuna’s Mūlamadhyamakakārikā. New York and Oxford: Oxford Univ. Press. Gethin, R. (1986). “The Five Khandas: Their Treatment in the Nikāyas and Early Abhidhamma.” JIP 14: 35–53. Gombrich, R. (1996). How Buddhism Began. London & Atlantic highlands, N. J.: Athlone. Gómez, L. (1976). “Proto-Mādhyamika in the Pāli canon.” PEW 26/2: 137– 165.

Harris, I. C. (1994). The Continuity of Madhyamaka and Yogācāra in Indian Mahāyāna Buddhism. Leiden, New York, København, Köln: E. J. Brill. Hayes, R. F. (1994). “Nāgārjuna’s Appeal.” JIP 22: 299–378. Huntington Jr., C. W. (1983). “The System of the Two Truths in the Pra sannapadā and the Madhyamakāvatāra: A Study in Mādhyamika Soteriology.” JIP 11: 77–106. Huntington Jr., Ch. W. (1989). The Emptiness of Emptiness. Honolulu: University of Hawaii Press. Huntington Jr., C., W. (1995). “A Lost Text of Early Indian Madhyamaka.” Asiatische Studien 49: 693–768. Huntington Jr., C., W. (2007). “The Nature of the Mādhyamika Trick.” JIP 35: 103–131. Ichigo, M. (1989). “Śāntarakṣita’s Madhyamakālaṃkāra.” In: Gómez, Luis O, and Silk, Jonathan A. (eds.). The Great Vehicle: Three Mahāyāna Buddhist Texts. Ann Arbor: Collegiate: 141–240. Inada, K. K. (1993).

Nāgārjuna, a Translation of his Mūla madh ya ma kakārikā with an Introductory Essay. Delhi: Sri Satguru. Jamieson, R. C. (2000). Nāgārjuna’s Verses on the Great Vehicle and the Heart of Dependent Origination. New Delhi: D.K Printworld. Jinpa, Th., “Sixty Stanzas of Reasoning.” http://www.tibetanclassics.org/ pdfs/SixtyStanzas.pdf (last accessed: 14.07.2009) Kalupahana, D. J. (1986). Nāgārjuna: The Philosophy of the Middle Way. New York: SUNY. King, R. (1989). “Śūnyatā and Ajāti: Absolutism and the Philosophies of Nāgārjuna and Gauḍapāda.” JIP 17: 385–405. King, R. (1994). “Early Yogācāra and its Relationship with the Madhyamaka School.” PEW 44/4: 659–684. Komito, D. R. (1987). Nāgārjuna’s “Seventy Stanzas.” A Buddhist Psychology of Emptiness. Ithaca: Snow Lion. Lindtner, Ch. (1982). Nāgārjuniana: Studies in the Writings and Philosophy of Nāgārjuna. Delhi: Motilal Banarsidass. Lindtner, Ch. (1986). Master of Wisdom : Writings of the Buddhist Master Nāgārjuna. Oakland: Dharma Press. Lindtner, Ch. (1997). “The Problem of Precanonical Buddhism.” Buddhist Studies Review 14/2: 109–139.

Loizzo, J. J. (2001). Candrakīrti and the Moon-fl ower of Nālandā: Objectivity and Self-Correction in India’s Central Therapeutic Philosophy of Language. PhD Dissertation, Columbia University. MacDonald, A. (2007). “Revisiting the Mūlamadhyamakakārikā: Text-Critical Proposals and Problems.” Indo Tetsugaku Bukkyōgaku Kenkyū 14/3: 25–55. Nagao, G. (1991). Mādhyamika and Yogācāra. New York: SUNY. Napper, E. (1989). Dependent Arising and Emptiness. Boston, London, Sydney: Wisdom Publications. Nietupski, P. (1994). “The Examination of Conditioned Entities and the Examination of Reality: Nāgārjuna’s Prajñānāma Mūlamadhyamakakārikā XIII, Bhāvaviveka’s Prajñāpradīpa XIII, and Candrakīrti’s Prasannapadā XIII.” JIP 22: 103–143. Pind, O. H. (2001). “Why the Vaidalyaprakaraṇa cannot be an Authentic Work of Nāgārjuna.” WZKS 45: 149–172. Prebish, C. S. (1974). “The

Prātimokṣa Puzzle: Fact versus Fantasy.” JAOS 94/2: 168–176. Radhakrishnan, S. (1956). Indian Philosophy, vol. II. New York: Macmillan. Ronkin, N. (2005). Early Buddhist Metaphysics: The Making of a Philosophical Tradition. London and New York: RoutledgeCurzon. Saito, A. (2007). “Is Nāgārjuna a Mādhyamika?.” In: K. Mochizuki (ed.). Hokekyō to Daijōkyōten no Kenkyū (Studies in the Saddharma puṇ ḍa rīka sūtra and Mahāyāna Scriptures). Sankibo Press: 153–164. Scherrer-Schaub, C. A. (1991). Yuktiṣaṣṭikāvṛtti:

Commentaire à la Soixantaine sur le Raisonment ou Du Vrai Enseignment de la Causalité par le maître Candrakīrti. Bruxelles: Institut Belge des Hautes Études Chinoises (Mélanges Chinois et Bouddhiques). Schroeder, J. (2000). “Nāgārjuna and the Doctrine of ‘Skillful Means’.” PEW 50/4: 559–583. Schroeder, J. W. (2001). Skillful Means: The Heart of Buddhist Compassion. Honolulu: University of Hawaii Press. Seyfort Ruegg, D. (1977). “The Uses of the Four Positions of the Catuṣkoṭi and the Problem of the Description of Reality in Mahāyāna Buddhism.” JIP 5: 1–71.

Seyfort Ruegg, D. (1981). The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Harrassowitz. Seyfort Ruegg, D. (2000). Three Studies in the History of Indian and Tibetan Madhyamka Philosophy: Studies in Indian and Tibetan Madhyamaka Thought Part 1. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien. Shulman, E. (2008). “Early Meanings of Dependent-origination.” JIP 36: 297–317. Shulman, E. (2008a). “Nāgārjuna on Impermanence, the Buddha on Illusion.” In: Shulman, D. and Weil, Sh. (eds.). Karmic Passages: Israeli Scholarship on India. Delhi: Oxford University Press: 168–187. Siderits, M. (1988). “Nāgārjuna as Anti-Realist.” JIP 16: 311–325. Siderits, M. (2004). “Causation and Emptiness in Early Madhyamaka.” JIP 32: 393–419. Sprung, M. (1977). “Nietzsche and Nāgārjuna: The Origins and Issue of Skepticism.” In: Coward, H., and Silvaraman, K. (eds.). Revelation in Indian Thought: A Festschrift in Honor of Professor T. R. V . Murti. Emeryville, California: Dharma Publishing: 159–169. Streng, F. J

(1967). Emptiness: A Study in Religious Meaning. Nashville and New York: Abingdon Press. Thurman, R. (1980). “Philosophical Nonegocentrism in Wittgenstein and Can drakīrti in their Treatment of the Private Language Problem.” PEW 30/3: 321–337. Tola, F. & Dragonetti, C. (1995). On Voidness: A Study in Buddhist Nihilism. Delhi: Motilal Banarsidass. Tola, F. & Dragonetti, C. (1995a). Nāgārjuna’s Refutation of Logic (Nyāyā), Vaidalyaprakaraṇa, Źib Mo Rnam Par ḥthag Pa Źes Bya Baḥi Rab Tu Byed Pa: Edition of the Tibetan Text, English Translation and Commentary, with Introduction and Notes. Delhi: Motilal Banarsidass. Tola, F. & Dragonetti, C. (1998). “Against the Attribution of the Vigraha vyāvar tanī to Nāgārjuna.” WZKS 42: 151–166. Vetter, T. (1992). “On the Authenticity of the Ratnāvalī.” Asiatische Studien 46: 492–506. Walser, J. (2005). Nāgārjuna in Context: Mahāyāna Buddhism and Earlier Indian Culture. New York: Columbia University Press. Whaling, F. (1979). “Śaṅkara and Buddhism.” JIP 7: 1–42

Williams, P. (1984). “Review Article.” JIP 12: 73–104. Williams, P. (1988). The Refl exive Nature of Awareness: A Tibetan Madhyamaka Defence. Richmond, Surrey: Curzon. Williams, P. (1989). Mahāyāna Buddhism: The Doctrinal Foundations. New York: Routledge. Wood, Th. E. (1994). Nāgārjunian Disputations: A Philosophical Journey through an Indian Looking-Glass. Honolulu: University of Hawaii press.