Nagarjuna and Logic

It will be our purpose in the following chapter to investigate the doctrines contained in Nâgârjuna’s major works. We will examine the interpretation of some important scholars and attempt to show their various drawbacks. This will point the way to our own position with regard to his work, a position in

which a specific solution with respect to pratityasamutpada becomes the key concept in the understanding of reality. Pratityasamutpada will be shown to be as positive a description of reality as is possible, given Nâgârjuna’s, and the general Buddhist tradition’s, stance on the role of language. It will provide the rationale for the appearance of the enlightened and the unenlightened states. However before this exegesis is possible let us examine the

contemporary views on those texts which are indisputably claimed, by all, to be authentically written by Nagarjuna himself. It has been customary among scholars of the past to read Nagarjuna with the aid of a commentary. Indeed since the MMK itself was abstracted in the first place, and in totality, from the commentary (Prasannapadà) of Candraklrti,1 it is hardly surprising that the views expressed in that commentary are strongly associated with the

doctrines of the MMK. We are left then with a tradition of scholarship, initiated by Stcherbatsky, and in the present day represented by Murti, that attempts an exposition of Nâgârjuna’s doctrines based on commentarial literature written approximately four centuries after the event. One would suppose, though here information is very sketchy, that after significant developments in the use of logic in religio-philosophical debate, and a general interchange

of ideas, a somewhat modified world picture would have developed during this period. This view is certainly upheld by Kalupahana, in his recent translation of MMK,2 who admits to having become rather uncomfortable with Candrakirti’s interpretation of Nagarjuna upon hearing that contemporary Vedântists hold Candraklrti in high regard.3 Nevertheless scholars like Murti retain their position. They claim that the Mâdhyamika:

... uses only one weapon. By drawing out the implications of any view he shows its self contradictory character. The dialectic is a series of REDUCTIO AD ABSUR- DUM arguments (prasahgâpâdanam). Every thesis is turned against itself. The Mâdhyamika is a prâsangika or vaitandika, a dialectician or free-lance debater.

1 Müla Madhyamakâkarikâ de Nâgârjuna avec la Prasannapadà Commentaire de Candrakirti par L. de la Vallée Poussin (1903-1913) Bibliotheca Buddhica IV republished by Biblio Verlag, Osnabrück (1970). 2 Kalupahana (1986). 3 ibid. p. xiii.

The Madhyamika disproves the opponent’s thesis, and does not prove any thesis of his own

In fact, as we shall see in due course, not even Candraklrti himself can realistically claim to simply turn an opponent’s thesis upon itself and reduce it to absurd conclusions, without introducing positions that the opponent does not hold himself. More importantly he cannot disprove the opponent’s thesis without proving a thesis of his own. When we turn to the case of Nâgârjuna we shall see that such a description of his method is impossible to uphold. Robinson5 has attempted to demonstrate that in some instances, Nâgârjuna seems to be explicitly using at least two of the three traditional Western laws of thought as axiomatic to his system, though there is little evidence that this has been agreed by his opponent. Thus we have a number of explicit statements of the principle of contradiction in the kârikâs:

In truth, the cessation of a real existing entity is not possible. For indeed, it is not possible to have the nature of both existence and non-existence at the same time.6

or:

A completed-incompleted doer cannot create a completed-incompleted deed. For how could the mutually conflicting completed and incompleted states co-exist as one?7

These statements would seem to mirror the purport of the third position of the catu$koti, or tetralemma, employed by the Buddhists, that a thing cannot be both existent and non-existent, and in this general sense the third kofi appears to conform to the principle of contradiction. Now, although the law of identity is nowhere found in any of the works we have ascribed to Nâgârjuna, Robinson certainly believes that the law of the excluded middle is held. In support of his contention he cites:

Indeed, a passing entity does not come to pass, and neither does a non-passing entity. Apart from these, how could there be a third (type of) entity coming to pass?8and

4 Murti (1960) p. 131. 5 Robinson (1957) p. 295. 6 MMK. vii. 30. sataÉca tâvabhâvasya nirodha nopapadyate ekatve na hi bhâvas ca nâbhâvas copapadyate 1 MMK. viii. 7 kârakah sadasadbhütah sadasatkurate na tat parasparaviruddham hi saccâsaccaikatah kutah 8 MMK. ii. 8 ganta na gacchati tâvadagantâ naiva gacchati anyo ganturagantnàca kastrtïyo hi gacchati

One who admits existence will necessarily perceive permanence and destruction. For, it necessarily follows that such an existence must either be permanent or impermanent.9

We may simply comment at this stage that statements such as these seem to support the view that a law of the excluded middle is invoked on occasions by Nagarjuna. Robinson’s conclusions, with regard to Nagaijuna’s putative adherence to such laws, are suitably vague. This is obviously advisable, particularly since at no point in his writings does Nagarjuna exactly state the laws of thought as such. It has been suggested more than once that Indian thought forms need not precisely mirror those adopted in the West. Robinson seems to bear this in mind as he does not press Nagarjuna’s adherence to the laws very far, contending in his summing up merely that:

Since Nagarjuna’s argument relies on numerous dichotomies, the principle of contradiction is necessary to most of his inferences.10 In another article Robinson11 questions how far the view that Nagarjuna adopted the prasahga method with his opponents can be upheld. He concludes that, in fact, it is possible to tease out a number of positions that are Nagarjuna’s alone. They do not belong to an identifiable opponent. Using such a method

Robinson is able to show that six positive positions are axiomatically held solely by Nagarjuna in his MMK. These are as follows.12 (i) Whatever has extension is divisible, hence is composite and is therefore neither permanent nor real. In consequence an indivisible, infinitesimal thing cannot possess extension. Now all the schools of Buddhism, together with the Mahayanasutras, do in fact accept a category of non-composite, non-conditioned things

(asamskrtadharma). For instance, in the Sarvastivada, space {aka$a), as a dharma of this category, is considered to have infinite extension while at the same time being incapable of division. The Acintyastava of the Catuhstava agrees with this definition since we are told: That which arises not, disappears not, is not to be annihilated and is not permanent, that is (tattva) which is like space (akafa) (and) not within the range of words (or) knowledge (aksarajnana).13

9 MMK. xxi. 14 bhavamabhyupapannasya Saivatocchedadarfanam prasajyate sa bhavo hi nitya ’nityo 'tha va bhavet 10 op. cit. p.296. 11 Robinson (1968). 12 cf. ibid. p.4. 13 CS. III. 39. yan nodeti na ca vyeti nocchedi na ca f aSvatam tad akaSapratikaSam naksarajhanagocaram

It seems strange that, in MMK ch. 5, Nagarjuna should concentrate his attack on the notion of space (ákása) by picking the relation between ákaáa and its characteristics (lak$ana) as a weak link, when it is clear that his opponents, by regarding ákáéa as asamskrta, accept that it is devoid of attributes or characteristics (lak$ana). If Nagarjuna accepts his opponents’ position, space would be “not within the range of words or knowledge (aksarajñána)”, and

consequently would not be a legitimate target for his argument. (ii) To exist means to be arisen and consequently existence is synonymous with manifestation. There can be no unmanifested existence. This axiom seems to contradict the doctrines of other Buddhists who hold that the real is that which has never arisen, has no beginning, no end and is permanent. This seems to be the meaning of the Udána where we are told: There is that sphere wherein is neither earth nor water nor fire nor air, wherein is neither the sphere of infinite space nor of infinite consciousness,

nor of nothingness, nor of either ideation nor non-ideation; where there is neither this world nor a world beyond nor both together nor moon and sun; this I say is free from coming and going, from duration and decay; there is no beginning and no establishment, no result and no cause; this indeed is the end of suffering.14

In other places Nagarjuna holds to such a position. Hence:

Where the functional realm of the mind ceases, the realm of words also ceases. For indeed, the essence of existence (dharmata) is like nirvana, without origination and destruction.15

and such a view is echoed in the mahgalaÜoka oí MMK. We may therefore conclude with Robinson that:

Nagarjuna is not alone among the thinkers of classical India in promiscuously adhering now to one and now to another of these (two) axioms.16 (iii) A real thing would have to be an utterly simple individual which contains no diversity. If it had diversity, it would have extension and so would not be indivisible and real. This is a corollary of axiom (i). (iv) The perception of arising and ceasing is illusory. Nagarjuna makes such a point in the kárikás:

14 U. 80 Atthi bhikkhave tad áyatanam, yattha ríeva pathavl na apo no tejo na vdyo na dkásdnaňcáyatanam na viňňánánaňcáyatanam na ákiňcaňňáyatanam na neva- saňňdnasaňňdyatanam n’áyam loko naparaloko ubho candimasuriyá, tad amham bhikkave n ’eva ágatim vadami na gatim na thitim na cutim na upapattim ap- patittham appavattam anarammanam eva tam es’ ev’anto dukkhassa’tij 15 MMK. xviii. 7 nivrttam abhidhatavyam nivrttaš cittagocarah anutpannániruddhá hi nirvánám iva dharmatá 16 op. cit. p.5 .

You may think that both occurrence and dissolution can be perceived but such a perception only comes about from a deluded mind.17 Very often the perception of origination and duration are compared with a dream, an illusion or a city of the gandharvas Like an illusion, a dream or an illusory city in the sky. In such a way has origination, duration and cessation been described.18

Robinson shows that Nagarjuna’s attempt to demonstrate all phenomena as illusory (maya) is not arrived at by aprasahga treatment of an opponent’s position. Neither is it arrived at by resort to an empirical examination of perception which shows that the senses generate distorted information. On the contrary all that Nagarjuna is doing here is dogmatically asserting that perception is always distorted by false thought constructions (yikalpa, prapanca etc.). (v) Only transitive actions and relations are allowed. A good case in point is MMK. vii. 7-8:

(opponent’s contention) As light illuminates both itself and other entities, so does origination give rise both to itself and others. (Nagarjuna’s reply) There is no darkness in light or in its abode. What then does light illumine when, indeed, it destroys darkness?19 When Nagarjuna denies that a lamp can illuminate itself, he is merely disallowing the making of reflexive statements. Nagarjuna will claim that the statement “Light illuminates itself’ is incoherent even though that same statement may be reformulated as “Light is inherently bright” which is perfectly

coherent from a commonsense point of view. It seems that axiom five becomes a special case of axiom three where a real thing is defined as being utterly simple and hence without attributes. As we have already shown, axiom three is corollary of axiom one, and no one except Nagarjuna takes this axiom seriously. One is left feeling that Nagarjuna’s method is on occasions specious to say the least. (vi) It is claimed that the Buddhas teach:

... that the dharma is based on two truths; namely the relative (samvrti) truth and the ultimate (paramartha) trJth.20 17 M M K.xxi.U drfyate sambhavaS caiva vibhavaS caiva te bhavet drfyate sambhavai caiva mohad vibhava eva ca 18 MMK. vii. 34 yatha may a yatha svapno gandharvanagaram yatha tathotpadas tathd sthanam tathd bhanga udahrtam 19 MMK. vii. 8-9 pradipah svaparatmanau samprakdSayitd yatha utpadah svapardtmanav ubhdv utpadayet tathd pradipe nandhakdro ’sti yatra casau pratisthitah kim prakafayati dipah prakaso hi tamovadhah

However in chapter 24 of the kaňkáš Nagarjuna’s putative opponent is a Hinayánist who argues that Nágárjuna is denying the Buddha's teaching as contained in the Tňpitaka. Nágárjuna is not really in a position to invoke the Buddha's teaching on the two truths as contained in the Maháyánas- utras since his antagonist will not accept such texts as authoritative. We are now in a position to briefly summarise Nágárjuna's method in the káňkás. As Robinson puts it:

It consists (a) of reading into the opponent’s views a few terms which one defines for him in a contradictory way, and (b) insisting on a small set of axioms which are at variance with common sense and not accepted in their entirety by any known philosophy.21

This is most definitely not theprasahga method as defined by Murti. Other authors have noted the inconsistencies between the reductio ad absurdum method extolled by Candraklrti and Nágárjuna's own particular methodology. Lamotte is a major scholar who, in his introduction to a translation of the Vimalalártinirdešasutra, is prepared to put down a further six positions or theses which he considers are held in a positive sense by the early Madhymaka at least from the point of view of the conventional truth (samvrtisatya). These are:

(i) All dharmas are without own-being (nihsvabháva), i.e. empty of self-being (svabhávašunya). (ii) All dharmas are non-produced (anutpanna) and non-destroyed (aniruddha). (iii) All dharmas are originally quiet (ádisánta) and by nature in complete nirvana (prakrtipaňnirvrta). (iv) The dharmas are without a character (alakjana) and are consequently unutterable (anirvacaniya, anabhilápya) and inconceivable (acintya). (v) All dharmas are equal (sama) and non-dual (advaya). (vi) Emptiness (šiinyatd) is not an entity (bháva, dharma, padártha).

Although we may object to statement (vi), preferring to say that from the ultimate point ofVxewšunyatá neither exists (asti) nor does not exist (násti), nevertheless, here again, we have a respectable authority on Mahayana Buddhism admitting the fact that Nágárjuna, far from following the prasahga method, is quite ready to hold a number of views which appear axiomatic for his own system though they are not held by any known opponent. The whole debate has been recently summarised with great clarity by Ruegg.22 He holds that:

20 MMK. xxiv. 8 21 op. cit. p. 8-9. 22 Ruegg (1986)

What the Madhyamaka has disowned, then, is any thesis, assertion or view (drsti) that posits the existence of some kind of bhava or dharma possessing a svabháva, and not all philosophical statements, doctrines and theories (dariana) without distinction.23 While it is true that Nágáijuna avoids all positive assertion of entities, one should not see him as someone single-mindedly intent on the rejection of all philosophical and ethical thought. Rather:

The Madhyamaka philsophy is... a non-speculative and non-constructive discourse relating to non-substantial factors (dharma) originating in the structured conditionship ofprañtyasamutpáda 24

The pivotal point of the whole Madhyamaka system seems to be the term Éünyatá. Nágárjuna’s statement in MMK. xxiv. 11, that a wrongly grasped Éünyatá is like a badly seized snake appears to imply that an ontological existence value cannot easily be predicated of it. That it cannot be either an existent or a non-existent seems clear since the Madhyamaka would be guilty of the charge of etemalism (ÉáÉvataváda) if he endorsed the former position, and by condoning the second would be accused of nihilism (iucchedaváda). Since all Buddhist schools, and the Mádhyamika is no exception here, stress an avoidance of adopting any extreme position, and in consequence tread a Middle Path (madhyama pratipad), there is no difficulty in accepting an idea of Éünyatá which

avoids these two extremes. P. J. Raju25 makes an interesting point in his association of the term Éünya with the mathematical zero of Indian scientific thought. Zero is defined as a mathematically indeterminate number, being neither positive nor negative. This seems a reasonable interpretation and the only objection to Raju’s position here is that of Ruegg, whose argument seems more a quibble than anything else, since as we shall see, he is wholeheartedly opposed to any attempt to place a value on the notion of Éünyatá. He says:

... there is no evidence in the basic texts of the Madhyamaka school that a mathematical model (and place-value) had any immediate bearing on their theory oí Éünyatá. In the Madhyamaka the term Éünya refers to the fact that any dharma is empty of own being (svabhávaÉünya) in which notion there is no mathematical connotation.26

It seems to me that Raju has not been attempting to expand the whole of the Madhyamaka philosophy using a mathematical model as Ruegg seems to suggest. He is simply saying that the concept of zero, as a term referring to an entity, or entities, which cannot be determined with regard to being or non-being, and which consequently have a problematic ontological

23 ibid. p. 233. 24 ibid. p. 235. 25 Raju, P.T. The Principle of Four-Cornered Negation in Indian Philosophy ’, Review of Metaphysics 7 (1954) p. 694-713. 26 Ruegg (1977-8) p. 40 n. 154.

value, may quite feasibly have been borrowed from mathematics, because of its symbolic sense. Now MMK. xxiv. 13 holds that emptiness (Éünyata) may not be an object of refutation. This stands to reason. Something may only be refuted or affirmed if it is capable of being understood in terms of being or non-being. Sünyatá is clearly not capable of being understood in such a way, which is why it is reported to be like a snake wrongly grasped {MMK. xxiv. 11). One can easily fall into the trap of assigning a definite value to it. This is what Lamotte is saying in his thesis (vi), i.e. that Éünyatá is not an entity. It does not follow from this that Éünyatü does not exist. It is not in a null class, along with mirages, etc., as Nakamura27 would have us believe. Nagarjuna’s statement that:

Whatever is in correspondence with emptiness (.Sünyatá) all is in correspondence (i.e. possible). Again whatever is not in correspondence with emptiness (,Éünyatá), all is not in correspondence.25

shows how Éünyatd is to be properly interpreted. When things are not understood as being empty, substantiality or own-being {svabháva) is imputed to them. Nagarjuna shows in MMK. xv. that the concept of svabháva, when imposed on things, renders them incapable of cooperating in dependent origination

(pratityasamutpáda). An ignorant world-view then destroys the essentially causal characteristic of things. Emptiness {áünyatá) is the abandonment of such a world-view. One comes to see how things actually cooperate. Robinson confirms our supposition, while at the same time repudiating the position of Nakamura: (All [[[sarvam]]]) means all mundane and transmundane dharmas (inMMK. xxiv. 14), that is all true predicables in the Buddhist domain of discourse. It

manifestly does not mean predications about rabbit horns and tortoise hairs... Dependent co-aris- ing is emptiness and therefore it is cogent. Emptiness is by definition ‘absence of own being’ {svabháva). The entire point of Nagarjuna’s argument is that the class of entities that possess own-being is null. Thus the class of empty phenomena (pratltyasamutpada) is the complement of the own-being or null class ... Thus the emptiness class is not null, but is co-extensive with the universal class.29

Things are not totally non-existence but simply falsely imputed to have own-being {svabháva). In fact these dharmas are svabhávaÉünya and cannot be confused with any null class from a logical point of view. Actually,

27 Nakamura, H. ‘Buddhist Logic expounded by Means of Symbolic Logic’/./.B.St. 7/1 (1958) p. 1-21. First published in Japanese 1954. cf. p. 14-15. 28 MMK. xxiv. 14. sarvam ca yujyate tasya iunyata yasya yujyate sarvam na yujyate tasya iunyam yasya na yujyate 29 Robinson (1957) p. 306.

this second, non-null or universal class has cogency simply because it is linked to pratityasamutpada. In another part of the karikds we find that: Dependent Origination (pratityasamutpada) we call emptiness (.§unyata). This is a provisional name and indeed it is the middle path.30 In other words ¿unyata. is a provisional name or metaphorical designation (upadaya prajhapti) for dependent origination (pratityasamutpada). It has already

been noted that the concept of pratityasamutpada occupied an important place in Nagarjuna’ system. Now we can see why. Pratityasa- mutpada and iunyata are synonymous. Whatever is in correspondence with these is ultimately true. Nagaijuna’s method then is to show that any of the alternatives supplied by discursive thought to characterise things, may be conventionally valid, but from the ultimate point of view they do not apply. In presenting the

conventional options he clearly, as Ruegg suggests, uses a logical method based on Aristotelean “two-valued logic founded on the dichotomously structured binary nature of discursive thinking in terms of alternatives”.31 Or again:

... the exclusion of the middle, as an onto-logical principle ... is ... one of the very foundations of Madhyamaka thought. And if the logical principle of excluded middle ... is not accepted in the Madhyamika’s procedure based on the use of the prasahga, this is because he considers that the subject of such sentences is in fact null.32

However, since he rejects all alternatives from the ultimate point of view, one will be wary in applying Western logical concepts to interpret his system in toto. Ruegg again sums this up by stating:

That the principle involved in the TERTIUMNONDATUR is indeed fundamental in Madhyamika thought follows from the consideration that, if a third position or value really existed, the mind would cling to it as some kind of thing, albeit one beyond the two values of “classical” logic. But if this were to happen there could be no “stillness” or “tranquility” on the level of paramartha, i.e. no absence of vikalpa and prapahca. And this would be radically opposed to Madhyamaka theory.33

We can give a qualified support for such a view, the qualification being that at the level of paramartha, i.e. that state devoid of thought construction {[nirvikalpajhana), “stillness” does not imply the complete obliteration of mental processes. As we have seen vijhana is transformed into jhana, 30 MMK. xxiv. 18 yah pratityasamutpddah funyatam tarn pracaksmahe sa prajhaptirupadaya pratipat saiva madhyama 31 op. cit. p.51 32 ibid. p. 50 33 ibid. p.

and the jnana of a Buddha has an object. This object paradoxically has no objectivity since at such a level of spiritual attainment objectivity and subjectivity have been transcended. Of importance in connection with Nagarjuna’s method is the question of where in his writings the two-valued logic,

which he generally employs, breaks down. It seems, from what has already been observed, that it would most probably do so when discussion turns away from the conventional and towards the ultimate nature of things. Now we know that aprasahgika is supposed to avoid the characterisation of things from the ultimate point of view, but is this actually the case in the writings of Nagarjuna? Ruegg certainly believes that it is:

... there appears to be no doubt that Nagarjuna, and his successors in the Madhyamaka school, founded many of their analyses of concepts and entities and their arguments based on reasoning by undesired consequences (prasahga) on the twin principles of non-contradiction and the excluded middle, before going on to show that in fact none of the members of a conceptual pair or tetralemma can in fact apply in reality.34

Staal35 disagrees. In his examination of the logical structure of the catuskofi he allows an interpretation of the fourth koti in which adherence to the law of the excluded middle is rejected:

When the Madhyamika philosopher negates a proposition, it does not follow that he himself accepts the negation of that proposition. Accordingly, there are other alternatives than A and not-A, and the principle of the excluded middle does not hold.36

The point at issue here seems to be the aspect of Nagarjuna’s doctrine which most disturbs his opponents. We will agree with Fenner’s characterisation of Nagarjuna’s overall approach to the conventional world such that:

The assumptions that undergird the Madhyamika analysis are these (1) that conceptuality depends on the consistent ascription of predicates to an entity,

(2) that predicates arise in the context of their logical opposites, which in its strong interpretation, as is required by the Madhyamikas, means that the presence of predicate implies its absence (and vice versa). This principle assumes a status equal to the aristotelean principles and its significance is

that analysis is effective to the extent that this principle is structurally formative (in its strong interpretation) for conceptuality. (3) the logical validity and formative influence and role of the three aristotelean principles of thought in structuring the development of conceptuality.37 34 ibid. p. 50 35 Staal, J.F. ‘Negation and the Law of Contradiction in Indian Thought’ Bulletin of the School of Oriental and African Studies 25 (1962) p. 52-71. and Staal, J.F. Exploring Mysticism Berkeley, University of California Press, 1975. 36 ibid. p. 44 37 Fenner, P. ‘A Study of the Relationship between Analysis (vicâra) and Insight

However by totally negating the predicates which arise in the context of their logical opposites, is Nagarjuna not opening himself to the charge of nihilism by appearing to suggest that such predicates in fact refer to nothing at all? All Buddhists, including Nagarjuna, are quick to reject the charge of nihilism. In fact the Vigrahavydvartani was written specifically with such a purpose in mind. Ruegg himself conducts such a defense when he comments:

... a thing may be said, followingMahayanist theory to be like a magical projection (maya) (not in a nihilistic sense but in the sense that it is imagined to be otherwise than it is in its true nature of dependent origination and emptiness).38 If this is so Nagarjuna must surely wish to negate the predicates without at the same time negating the ground to which they have been incorrectly applied. This may be the purport of Staal’s aforementioned statement. Let us now turn to an associated problem. Of central importance in our study of Nagarjuna’s

thought is the specific form of negation he employs. The Buddhist tradition accepts two alternative forms of negation and we are now in the position to examine which of the two is most appropriate to Nagarjuna’s work, acknowledging beforehand that nowhere in those texts ascribed to him does he explicitly make the distinction himself. The two forms of negation of interest are the total negation (prasajyapratisedha) and the limited or partial negation (paryudasa- pratisedha). Put briefly the prasajyapratisedha is a total negation because it negates a thesis without at the same time affirming any

contrapositive thesis. In other words the total negation signifies the total avoidance of any thesis formulation whatsoever. The paryudasapratisedha or parital negation however is one in which, although an original thesis may be refuted, there is no implication that the contrapositive thesis is also negated. As Kajiyama39 describes it theprasajya type negation is primarily negative as in the case of the sentence, “they do not look at the sun”. Here there is no suggestion that they are looking at anything else. On the other hand a paryudasapratisedha may contain the suggestion of an affirmation. A good

example here is the sentence, “He is a non-brahmin”. One naturally assumes from this that he has some other caste status. Reference to a typical neither ... nor (i.e. fourth kofi) statement from the karikas will establish what is meant. In MMK. xxv. 10 we have: The teacher (Buddha) has taught the abandonment of the concepts of being and non-being. Therefore nirvana is properly neither (in the realm) of existence nor non-existence.40

(prajhd) based on the Madhyamakdvatdra' J.I.P. 12 (1984) p. 139-197 cf. p. 164. 38 op. cit. p. 51. 39 Kajiyama, Y. An Introduction to Buddhist Philosophy. An Annotated translation of the Tarkabhdsa of Moksakaragupta Kyoto, 1966. p. 38-9. 40 MMK. xxv. 10

Now if we take this statement to be a prasajya type of negation then we are led to conclude that the twin ideas of being and non-being totally exhaust the ontological status of the concept which in this case is nirvana. In the prasajya negation of nirvana no further position can arise once the negation is

concluded, which would lead to any proposition being tendered concerning the status of nirvana. Theparyudasa or limited negation, works in a different way. The initial negation here does not exhaust all that may be held concerning the concept to be negated. In our example one would, on the surface, accept the idea that nirvana is neither being nor non-being. Nevertheless one would not wish to state that these two positions exhaust the modes in which nirvana may

be said to occur. On the contrary nirvana as we have already noted, is empty {šunya) rather than totally devoid of existence, as Fenner makes clear: ... an entity is shown to be empty rather than non-existent through the exclusion of all possible predicates as being inapplicable to an entity. The entity A is neither a P nor a -P where P and not P exhaust the universal set of modalities. The nihilistic conclusion for the non-existence of something

presupposes the applicability of predicates to an entity which are in actuality absent... If A goes uncharacterised because all predicates are inapplicable to it, its existence or non-existence is unascertainable as the entity itself would be unidentifiable.41

If we make A=nirvana, the total negation will indicate that P and -P completely exhaust all the modes in which A can be said to occur. This would not however be the case for Nágárjuna since in his writings he implicitly holds the view that, while A “goes uncharacterised because all predicates are inapplicable to it”, there is some indeterminate sense in which A may be said to exist. A useful way of indicating such indeterminacy will be to say that A

exists ultimately in its emptiness (šunyatá) mode. This will be the equivalent of saying that is ultimately uncontaminated by all attempts to define it existentially. This is what I mean when I talk about the ontologically indeterminate existence of an entity. Most scholars who have treated this subject are heavily in debt to Candrakirti. Because he insists on the prasajya type of negation as the characteristic negation of the Prásaňgika-Madhyamika it has been taken for granted that Nágárjuna himself, even though he makes no specific reference to either, avoided the use of the limited paryudasa type. There

is, in fact, a diversity of thought amongst more recent scholars on this particular issue. Fenner42 tells us that Candrakirti distinguishes his school from the Svátantrika school of Bhávaviveka on the basis that while Bháva- viveka and his followers adopt the paryudasa, the Prásaňgikas plump for prahanam cabravic chasta bhavasya vibhavasya ca tasman na bhavo nábhávo nirvánám itiyujyate 41 op. cit. p. 187 42 ibid. p. 188.

the prasajya. However, he fails to tell us precisely where Candriklrti says this. Ruegg is similarly vague and does not quote sources. Nevertheless he opts for a different interpretation. For him both the Prasahgikas and the Svatantrikas use only the prasajya negation. He claims that:

In this form of negation (i.e. prasajya) as used by the Madhyamika denial of a position does not necessarily involve commitment to any other position ... The Madhyamika is certainly not working towards tome ontolgical or logical third value between contradictories any more than he is seeking a dialectical synthesis. Indeed, if there really existed such a dialectical synthesis or third value, there would be something on which conceptual thinking could base itself and cling, and the whole purpose of the Madhyamaka method could then no longer be achieved.43

Although this statement may be said, with some reservations, to outline the position of an author such as Candraklrti there does not seem to be any good justification to extend it to include Nagarjuna and his earlier followers. Let us take as an example the eight (negated) epithets of pratityasamutpada in the mahgalaSloka of MMK anirodham anutpadam anucchedam aSaSvatam anekartham andnartham anagamam anirgamam

Ruegg asks the question, do these epithets commit the Madhyamika to a positive view concerning pratityasamutpada equivalent to the opposite of what is here negated? He answers “no”. However from what has already been said concerning the status of pratityasamutpada in Nagarjuna’s non- MMK works, and his general method which only follows logical principles up to the limit of the conventional, we must be more careful than to give such an unqualified “no”. Now Ruegg

is completely consistent here. He applies the total (prasajya) negation in the manner that he expects Nagarjuna would have done. Ultimately of course pratityasamutpada cannot be characterised and Ruegg is, in this sense, correct to say “no”. However this is only half of the truth for we have already seen that an entity may also exist in its emptiness mode even though an attempt at predication has failed. In other words it may exist in a state of ontological indeterminacy. Pratityasamutphda is exactly the type of thing we should expect to possess such indeterminacy. Being ontologically indeterminate

pratityasamutpada will survive the partial (paryudasa) negation, and this is the point that Ruegg’s “no” does not take into account. Pratityasamutpada is not non-existent. From the point of view of ultimate truth (paramartha- satya) it may not be presented as an object to consciousness. It is not the object of mundane consciousness (vijhana), though it may be conceived, in a transcendent emptiness mode, as self and other intimately united in gnosis (jhana). In other words there is such a thing as pratityasamutpada, 43 op. cit. p.4

though it may not be characterised in terms of the eight epithets mentioned. It may be said to be ontologically indeterminate. At another point in MMK we hear that the Buddha may not be determined with regard to existence or non-existence after having attained nirvana, and having died. This of course corresponds with the general unwillingness of the Buddha to ascribe an existence value to such a state in the unanswered questions of the Tripitaka. Nagarjuna simply expands on what the Buddha has already said:

That image of nirvana (in which) the Buddha either “is” or “is not”—by him who (so imagines nirvana) the notion is crudely grasped. Concerning that which is empty by its own-nature (svabhava), the thoughts do not arise that: the Buddha “exists” or “does not exist” after death.44 However he makes it perfectly clear that the Buddha, in his ultimate condition, does have an ontological value for:

Those who describe the Buddha in detail, who is unchanging and beyond all detailed description—Those, completely defeated by description, do not perceive the Tathagata. The self-existence of the Tathagata is the self-existence of the world. The Tathagata is devoid of self-existence and the world is likewise.45

It would be much easier for Nagarjuna, should he have so desired, to assert that neither the Buddha, nor the world exist, but this he pointedly refuses to do. We must assume that this is not the position he wishes to adopt. Such a position would, as far as our researches lead us to believe, be the consequence of a total negation (prasajyapratijedha) of the predicates. The position here taken with regard to the Buddha, since it assigns some indeterminate

ontological value to his ultimate existence, corresponds closely with the consequences of a partial negation (paryudasaprati^edha). Now, before turning to a textual analysis of MMK, let us briefly look at some of the logical aspects of the Vigrahavyavartani (W ). Our point here will be to decide whether, in this text, Nagarjuna applies the prasanga approach prescribed by Candraklrti. In other words, does he formulate propositions not held by his opponents, and utilise a logic at odds in many places from that adopted by the so-called prasanga method. In the first place the precise nature of his opponent in this text is an object of

44 MMK. xxii. 13-14 Yena graho grhitastu ghano ’stiti tathagatah nastiti sa vikalpayan nirvrtasyapi kalpayet svabhavataf ca Sunye ’smimi cinta naivopapadyate param nirodhad bhavati buddho na bhavatiti vd 45 MMK. xxii. 15-16 Prapahcayantiye buddham prapahcdtitam avyayam teprapahcahatah sarve na pasyanti tathagatam tathagatoyat svabhavas tat svabhavam idam jagat tathdgato nilisvabhavo nihsvabhavam idam jagat

controversy. Bhattacharya46 is of the opinion that Nagarjuna’s opponent is a Naiyayika realist and in this he has his supporters, such as Tucci.47 Lindtner48 feels that this is incorrect. He gives five reasons to support his contention that the opponent is actually a Buddhist Abhidharmika.

Unfortunately at the present state of Buddhist studies the problem seems likely to be unsolved for sometime. However, if we do accept the opponent of MMK to be an Abhidharmika, and that MMK and W comprise a corpus with one specific end in view, then one has some reason to come down in favour of holding the

opponent in W to be from an Abhidharmic school. Now commenting on the function of the W in the Madhyamaka scheme of things Ruegg tells us that in this text ... a Mddhyamika restricts himself to a kind of philosophical destruction—and therapeutic dehabituation—with respect to dichotomising conceptualisation while refraining from propounding any propositional thesis (pratijna) of his own, but any argument adduced to combat and refute the theory of Sunyata is

devoid of cogency, and falls into line with and reinforces the Madhyamaka theory, since all things can be shown to be equally non-substantial.49 This is simply not true for Nagarjuna never successfully answers his opponent’s first objection. However, even if it is admitted that there is some substance to his replies it can hardly be held, as Ruegg would have us believe, that he is using theprasahga method. Let us examine the argument in detail. The opponent has spotted a weakness in Nagarjuna’s thought since if all is empty, then on what conceivable grounds can Nagarjuna propound, in a meaningful way, the emptiness of all views. Thus the W opens:

If own being (svabhava) does not exist anywhere in any existing thing, your statement (itself) being without own being is not capable of refuting own-being. But if that statement has (its own) own-being, then your initial proposition is refuted. There is a (logical) inconsistency here and you should explain the grounds of the difference.50

To what seems a justifiable complaint, Nagarjuna replies that either his opponent accepts that negation must always have something real as its negandum, in which case he miist accept emptiness (Sunyata), or he must

46 Bhattacharya, K. The Dialectical Method of Nagarjuna (Vigrahavyavartani) Dehli (1978) p. 38 n 2 47 Tucci, G. Pre-Dihnäga Buddhist Texts on Logic from Chinese Sources Baroda (1929) p. xxvii 48 Nägärjuniana (1982) p. 71 n. 110 49 Ruegg The Literature of the Madhyamaka School p. 22 50 W . 1-2 sarvesäm bhävänäm sarvatra na vidyate svabhavaS cet tvadvacanam asvabhävam na nivartayitum svabhävam alam atha sasvabhävam etad väkyam pürvä hatäpratijna te vaisamikatvam tasmin vifesahetuS ca vaktavyah

give up his thesis. This is confusing but, as far as Nagarjuna’s position is concerned there is no negating anything, otherwise one would be forced to accept the neganda. All he claims to be attempting to do is to suggest or indicate {jnapayate) the absence of own being.51 In his reply, Nagarjuna makes a

distinction between indicating an absence of own being and negating the existence of own being. These two activities are claimed to be completely different, and Nagarjuna claims to be doing the former and not the latter. In the accompanying auto-commentary ([svopajnavrtti) verse 64 is glossed. In the same manner, the sentence, there is no svabhava of the bhavas, does not make the svabhava without essence, but conveys the absence of svabhava in the bhavas,52

Mehta53 uses an analogy to elucidate this point in his interpretation of the argument. He says that when one makes a statement such as “Devadatta is not in the house”, the statement itself merely informs us of Devadatta’s absence in the house and does not possess the power to bring about the existence or non-existence of Devadatta as such. However the statement about Devadatta is really in no way analogous to the argument in W . The statement concerning

Devadatta is easily verifiable by sense perception and may be proved or disproved by a state of affairs beyond the structure of the sentence. Nagarjuna's contention that the statement “All things are without own-being” simply serves to make such a fact known may not be verified in the same manner. It seems that it is Nagarjuna who misses the point here. Since none of his contemporaries held a view that statements themselves have the power to bring about a state of affairs, i.e. emptiness unyata), Nagarjuna abandons any claim to be a Prasahgika. It seems that

51 W . 61-67 sata eva pratisedho yadi ¿unyatvam nanu pratisiddham idam pratisedhayate hi bhavan bhavanam nihsvabhdvatvam pratisedhayase ’tha tvam ¿unyatvam tac ca nasti iunyatvam pratisedhah sata iti te nanvesa vihiyate vadah pratisedhayami naham kimcitpratisedhyam asti na ca kimcit tasmat pratisedhayasity adhilaya esa tvaya kriyate yac cdharte vacanad asatah. pratisedhavacanasiddhir iti atra jnapayate vdg asad iti tan na pratinihanti mrgatnrsadrstante yah

punar uktas tvaya maharrrf carcah tatrapi nimayam ¿mu y ath a sa drstanta upapannah sayadi svabhavatah syad graho na syatpratitya sambhutah yaf ca pratitya bhavati graho nanu iunyata saiva yadi ca svabhavatah, syad grahah kas tam nivartayed graham fesesv apy esa vidhis tasmdd eso ’nupalambhah 52 tadva nasti svabhava bhavanam ity etad vacanam na bhavanam nihsvabhdvatvam karoti kimtu sarvabhavesu svabhdvasydbhavam jhapayati 53 Mehta, M. ‘¿unyata and Dharmata: The Madhyamika View of Inner Reality’ in Amore (1979) p. 30 n. 18.

he is putting forward this view himself. The opponent is therefore not objecting to this particular thesis but to the logical form of Nagarjuna’s central theme. In other words “If all things are empty, how can you demonstrate, given the fact that your own words are empty, emptiness?” The logical structure of a sentence such as “Devadatta is not in the house” is simply an inadequate analogy to the Madhyamaka contention that: ... iunyata does not have the function of making dharmas empty since this is what they are; a sentence concerning iunyatd therefore serves to make this fact known.54

All sentences must presumably serve to make something known, otherwise one would be left with an absurd theory of language. Here then we have evidence of Nagarjuna’s technique at work. He does not attempt to answer the objection, but rather sidesteps it, proposes a theory that his opponent does not hold, which has the effect of introducing confusion, and finally introduces a conclusion which, because of the foregoing argument, seems acceptable when viewed

not too critically. It is simply not the case that by a remorseless application of logic based on reductio ad absurdum of the opponent’s thesis, Nagarjuna achieves a crushing victory, and it is certainly not the case, as Ruegg55 would have us believe that the Madhyamaka theory is immune from refutation. One cannot help but agree with Streng here when he says that Nagarjuna’s work occasionally is “an analysis which appears to be rather arid and often simply a play on words”.56 With reference to this particular argument in VV Betty has recently observed:

It is as if the objector had said to Nagarjuna, “Your’re wrong”, and Nagarjuna had answered “Of course I’m wrong, that’s precisely what makes me right”. As alluring, as stunning, as Taoistically fascinating as such an answer is, it is not really an answer; it is not cogent in an argument where the rules of logic apply, as they do here. Nagarjuna has evaded the issue; he has seen the problem, but he has not treated it seriously: he has not “accepted” it.57 Another apparent inconsistency arises in connection with W . 29 which says:

If I would make any proposition whatever, then by that I would have a logical error; but I do not make a proposition; therefore I am not in error.58 The autocommentary goes on to say:

... when all entities are empty, altogether still and devoid of a nature how could there be a proposition (presenting them as being something or other)?59 54 Ruegg op. cit. p. 22 55 ibid. p. 23 56 Streng (1967) p. 181-182. 57 Betty (1983) p. 128. 58 W . 29 yadi ká cana pratijña syan me tata esa me bhaved dosah. ndsti ca mama pratijña tasman naivasti me dosah.

The objection must be raised, that here again Nagarjuna is up to something fishy. Is it not true that the statement I do not make a proposition”, is itself a proposition (pratijna)? Since it is, how is this compatible with the autocommentary in which we are told that there are no such things as propositions. The problem from the logical point of view, here is quite analogous to our examination of statements concerning Sunyata, above. However, in this case Nagarjuna does not attempt to follow up the problems. Ruegg attempts to dispense with them by saying

... this interpretation assimilates two distinct uses of the term “proposition”, and it would hold good only if pratijna meant here any sentence or statement... But this sentence (i.e. nasti ca mama pratijna W .29) is not a pratijna in Nagarjuna’s sense; for in his way pratijna denotes an assertion and more specifically a thesis which seeks to establish something.60 So according to Ruegg the term (i.e. pratijna) may have one of two meanings. Firstly it

may mean any sentence. Secondly it may mean a thesis which seeks to establish something. If we accept Ruegg’s belief (unsupported by reference to sources) that all that Nagarjuna is saying in W . 29 is that he does not make propositions which seek to establish something, we are still back to square one and

Ruegg has done nothing to extricate himself and Nagarjuna from the problem. The objection still exists, “Is not your statement, that you do not make propositions seeking to establish theses, itself a proposition?” W is actually full of such inconsistencies and in the light of what we have said regarding

both it and MMK, we must be forced into a different interpretation of these two works than that provided by Ruegg and others. There can be little doubt that Nagarjuna does not abide by the prasahga method in argumentation. If he was a Prasahgika we could accept that he has no thesis of his own to put

forward, but this is simply not the case. Once we are able to bring in to question this putative connection with prasahga logic there will be no obstacle in our way to accepting Nagarjuna’s adherence to partial (paryudasa) as opposed to total (prasajya) negation (patisedha). This interpretation is certainly

consistent with the texts themselves. Using these conclusions as our foundation we shall be able to pursue the thesis that the Madhyamaka is not as dissimilar to the Yogacara as is generally thought. Since we now understand that Nagarjuna, particularly in his apparent use of a three valued logic, may

be implicitly able to hold “positive” positions concerning the nature of things, the idea that only the Yogacara adopted such an outlook seems onesided.61 It should now be possible to examine more parallels between the two “schools”, particularly 59 sarvabhavesu Sunyesv atyantopa Santesuprakrtiviviktesu kutah pratijna 60 Ruegg (1977-78) p. 49 61 Conze’s point of view. This will be treated in more detail in chapters 6 and 8 supra. when they are treated against the background of the early Buddhist tradition.