Naagaarjuna and analytic philosophy (I)

·期刊原文
Naagaarjuna and analytic philosophy (I)
Ives Waldo
Philosophy East and West
Vol.25 (1975)
pp.281-290
Copyright By University of Hawaii Press

--------------------------------------------------------------------------------

p.281

I

Some clarifications of the Maadhyamika and some fruitful consequences for modern philosophy arise if we view key terms of the Muulamaadhyamakakaarikaas (MK) as crossing the analytic/synthetic line. This involves a basic insight of a degree of logical sophistication comparable to that of analytic philosophy whereby ontology and epistemology collapse into a single form. The characteristic negativity of the Maadhyamika, and of the Praasa^ngika school especially, arises largely from a dilemma inherent in the attempt to state such a point of view in the essentially realistic categories of the abhidharma. Elements of a solution involving multiple reality criteria on different logical type levels are found in the Sa^mdhinirmocana Suutra and in works of the svaatantrikas. These are not fully coherent or explicit, but are clear enough to reveal doubtful presuppositions that have caused much contemporary philosophy to degenerate from a vitally significant human enterprise into methodologically sterile logicism.

II

Ordinary English terms like "cause" and "entity" have for us semiconscious connotations of empiricist logical geography that are far from parallel to those worked up by Buddhist philosophy from "ordinary Sanskrit." Kenneth Inada points out that absurdities quickly arise if we try to substitute "elementary particle or stuff" and "physical cause" for dharma and pratyaya in the abhidharma.[1] A key feature of the difference is noted by Karl Potter: Indian logic in general has little interest in distinguishing between logical necessity and empirical universality in its dependency relations.[2] It is not accidental that hetu can signify either a root cause or a premise in a formal argument. Guenther makes the important distinction between the notion of dharmas as experiential realities and the misconception that they are subjective in the Kantian sense, being mere appearances opposed to some unknowable Ding an sich.[3] For the Maadhyamika, subjective and objective are interdependent aspects of experience in such a way that, what makes a difference to experience, what can be known, and what can intelligibly be said to exist coincide. J. May presents this point of view as involving the collapse of ontology and epistemology (and so the absolute distinction of subjective and objective) into a single form, a form that Naagaarjuna himself terms the middle way.[4]

The basic question for Buddhist philosophy is "what conditions are

p.282

responsible for the arising of an intelligible situation as such in our experience ?" Examples cited include relations of proof, physical regularity, and also criterial or contextual necessities. This might seem inexcusably loose unless we remember how diffcult it can be in practice to say whether a term is being used with empirical or logical force. Is it empirically or logically true, for instance, that tracks in cloud chambers are tracks of particles, or indeed that there are such things as particles?[5] Compare Wittgenstein: "Then can one say only that a certain regularity in occurences makes induction possible. The 'possible'world of course has to be 'logically possible.' "[6] For the Maadhyamika, similar problems arise with all our basic categories of discourse. Hence the virtue of dealing with these matters at a level of abstraction prior to the arising of the analytic/synthetic, ontological/epistemological dichotomies. What this amounts to in practice is exemplified in the opening chapter of the MK. Naagaarjuna here reduces rationalism and empiricism to a single form, as attempts to find a unique and indubitable description of what there is. He then shows that form to beg the question.

III

In MK I, Naagaarjuna sets at loggerheads the concepts of svabhaava (own-being) and pratyaya (relational condition). Each of these is incompatible with the other, yet seems to require it for its own comprehensibility. Svabhaava is the notion of an existence that is ontoiogically and epistemologically self-contained. It must be sufficient reason for its own existence, and also, like Russell's logical atoms, theoretically knowable for what it is quite independently of anything else.[7] The ontological/epistemological claim is that any svabhaava is sufficient reason for its own arising in experience. Naagaarjuna says this is absurd.

asata.h pratyaya.h kasya sata`s ca pratyayena ki.m (MK I-6)
If it (already) exists, what use is a pratyaya?
If not, what sort of pratyaya is that?

Ontologicaily and logically: a nonexistent being cannot bring itself into existence. Neither can an existent one, as it is there already.[8] So no being in our experience will be the sort that could be proved to have to exist. This eliminates the hope for a rationalist guarantee for our knowledge of things in experience.

Epistemologically and empirically: Nothing can serve as its own criterion of intelligibility. If the thing is not already identified as intelligibie, it will not be available to serve as a criterion. If it is, it stands in no need of a criterion.

The Sarvaastivaadins and modern empiricists both attempted to find a self-guaranteeing form of language that could describe the world with total accuracy and certainty. Both saw words ultimately as names for situations given immediately and totally in experience, and so incorrigible.

p.283

Naagaarjuna points out that this involves the same sort of question begging that invalidates rationalism. In so doing he directly parallels Wittgenstein's argument that a private language (an empiricist language) is impossible.[9] Having no logical links (criteria) to anything outside their defining situation, its words must be empty of significance or use. Wittgenstein considers a purported sensation word "S" (svabhaava!) that is allegedly identifiable without external criteria.[10] First, this model gives us no way of reusing "S". Another purporting instance would be entirely outside the certainty of the defining situation. Second, because identification of "S" is per hypothesem independent of anything else in the world, "S" carries no information about that either.

This may help explain why Naagaarjuna feels no need to argue that all intelligible situations have pratyayas, while it is a matter for subtle Kantian arguments, if it can be shown at all, that all events must be considered under the category of causation.[11] We have just seen an argument logically linking significance with an event's being conditioned by a system of external criteria. The necessary existence of such relational conditions (pratyaya) refutes both the theory of svabhaava and the possibility that significant events might arise with no relational conditions at all. Significance lies not in the substantial, experientially given, and certain; but in that which is relational, metaperceptual, and hypothetical. A man born blind who later gains his sight finds no significance in his visual held.[12] He has not developed a system of habitual expectations that will allow him to map what he experiences onto a system relating his other senses and behavior. Vision makes sense to ordinary people because what is seen is systematically related to all kinds of familiar possibilities. These relationships furnish criteria to check our visual identification, but more important they make such identification of use.

If things cannot arise in experience from themselves, it seems then that significance and existence must be understood in terms of pratyaya, which entails the arising of being and significance from another (parabhaava) (MK I-3). Significant existence has emerged as a system property, an entirely meta-perceptual affair. But how can there be other-being without there being eventually some other which has own-being (MK I-3). How can we speak of a system when we cannot specify the elements out of which such a system is built. Ontologically this is the new unfashionable problem of a first cause (by no means outworn in the second century A.D. !). Epistemologically, the problem follows a line of reasoning we find in P. F. Strawson's Individuals.[13] Strawson finds that the various sorts of particulars in the world, and such basic categories as space, time, and cause are very closely interdefined and criterially related, each serving to help explain and identify the other. The question now arises: How does the whole affair get started, unless there is some sort of basic particular that can be identified prior to identifying anything else? B and C are criteria for A; A and C criteria for B; and so forth. So without basic particulars, we will not be able to identify or characterize anything. When we see that anything

p.284

that could qualify as a basic particular would also meet the standards for being a word in a private language in Wittgenstein's sense, we can see trouble approaching. J. Moravcsik points out that Strawson's candidates, material objects, come nowhere near meeting the requirements for basic particulars, since identifying them involves questions of space, time, causation of experiences, etc.[l4] If we take this dilemma at face value, we will be forced to conclude that language is radically arbitrary and incoherent. No statement about existence, quality, causation, or anything else deserves to be taken seriously. This dilemma in the form of the svabhaava/parabhaava opposition is the central problem of the Maadhyamika. The two main interpretive traditions are differentiated by respectively negative and positive responses to it. But it is also central to modern analytic thought. Versions of it are involved in empiricist criteria of meaning, the issue of whether reference must guarantee the certainty of some propositions if language is to be nonarbitrary, the question of the absoluteness of the analytic/synthetic distinction.

IV

The Praasa^ngika school accepts the validity of the dilemma and the conclusion that language is fundamentally incoherent. It follows that the MK cannot prove its points by logical argument. It is viewed, as the early Wittgenstein viewed the Tractatus, as a supremely transparent example of the limits of logic and language. The comprehending reader will come to see through it the absurdity of all metaphysical projects.[15] In the context of the bodhisattva path this intellectual example is meant to resonate as an example of the folly of clinging to fixed patterns of thought, behavior, or emotion in any area of life. As Maadhyamika dialectic shows the untenability of intellectual dogmatism, skillful handling of life-situations by a Buddhist teacher reveals the untenability of any rigid behavioral-emotional stance. Eventually the practitioner must leap free of all preconceptions into the realm of `Suunyataa, which is beyond any question of being or not being. In the relative sphere the bodhisattva, realizing the impossibility of comprehensive logical planning, proceeds with intuitive panoramic awareness, employing as tools whatever behavioral and cognitive modes seem appropriate to his ends, according to the dictates of skillful means.[16]

The Praasa^ngika vehicle reveals excellent command of psychology; but interpretation of the statement that the wise have recourse neither to being nor not being (MK XV, 10) in terms of the incoherence of language is unacceptable to contemporary thought. The Maadhyamikas hold, as we do, that self-contra-dictory categories cannot be instantiated (are empty). Thus, their view entails that there are no true statements of existence or causal relationship since any such statement involves the self-contradictory category of svabhaava. But as analytic philosophers have pointed out, statements like Moore's "Here is one B

p.285

hand, and here is another,"[17] or "great oaks from little acorns grow" are meant as paradigms of true and coherent existential and causal statements. Moore is saying in effect, This is what we mean by to hold up a hand. If two svabhaavas could not share quality or causal dependence, this simply shows the logic of svabhaava is not the logic of ordinary language. Our notions of validity and invalidity have sense only within the context and presuppositions of our linguistic system. The attempt to use them to discredit that system as a whole is self-defeating. We must conclude that if Maadhyamika has a message that contemporary thought can take seriously, it will not arise from this interpretation.

In the work of Bhaavaviveka, the founder of Svaatantrika Maadhyamika we find an interpretation supportive of logic and ordinary language. Here "the wise hold not to being or not being" is taken as asserting that all categories including existence have sense only in some system involving specific criteria and uses. Thus, he maintains that since the higher truth (paramaartha satya), as expressed in the negativity of the Praj~naapaaramitaa literature, has different criteria than the truth of ordinary discourse (sa^mv.rti satya), they do not conflict.

We also accept the conventional reality which (everybody) in the world unanimously accepts as real. For in worldly experience, origination through causes and conditions is accepted as real. Therefore reality of the conditioned elements, i.e., of the eye organ, etc. is brought under the category of samv.rti-sat. Everyone including shepherds and cowherds know this. Since we make our experience from the standpoint of ultimate reality we never contradict the actual experience of the world.[18]

The higher truth is higher than the lower because it includes and is about it while the reverse is not true. This is the import of what Bhaavaviveka means in calling it more critically satisfying.[19] It amounts to the claim that the higher truth involves statements of a higher, logical type than the lower, being statements about any formulations of language on the level of the lower truth. Bhaavaviveka has a none too subtle example of a yogin (today it could be a scientist) who knows that there are certain conditions under which solid objects can pass through the earth like water.[20] Ordinary people falsely imagine that the solidity of earth is an absolute. Their usage fits all cases familiar to them, but they are wrong to think there are no other possibilities. The point is similar to one Guenther makes about the Svaatantrikas.

There is nothing wrong with saying "the apple is red." But this can be needlessly narrowing if one does not recall that this applies with full literalness only in certain standard perceptual situations. One must go beyond this to paint well, for example.

If a man were under all circumstances immediately conscious of the medium of vision and its effect on the image of the object he would immediately see the precise effect of shifting to any other medium. He would be like a skilled

p.286

musician who can play in one key what is written in another without transcribing the score.[21]

In general the bodhisattva is one who gradually eliminates the need to cling to received views and ways of behavior as if their scope were unlimited. His world becomes not empty, but so rich that it becomes ab`Surd to think any single view or description could exhaust all its possibilities.

There is a hint to the solution of our dilemma here. It is similar to a view proposed in the "Question of Maitreya" chapter of the Sa^mdhinlirmocana Suutra, which may have been part of the scriptural material Bhaavaviveka wanted to elucidate.[22] The question arises: How can both the higher and lower truths be true? If something exists corresponding to the lower, the higher ought to be false. If the higher is true, how can there be illusion and sa^msaara at all? It makes no sense to think of ordinary truth and the truth of the Praj~naapaaraamitaa as referring to parallel worlds or points of view in any simple sense. The bodhisattva answers that since meaningfulness arises from a reaction of body, speech, and mind with the world:

1.It is all right to say within ordinary truth that things corresponding to its forms exist.

2.It makes no sense to say that such things exist prior to the significance-producing reaction.

3.There is no describing the nonlinguistic base of all such reactions aside from some particular relative articulation.

We have here the elements for a solution to Naagaarjuna's dilemma. I wish I could say that Bhaavaviveka presented it coherently. The truth seems to be that this was next to impossible without modern logical apparatus like the notion of a type distinction. The Svaatantrikas, at various points, discuss views that suggest such a solution, but do not seem to have succeeded in coherently presenting it in logical form.[23] For example, Bhaavaviveka describes the lower truth as involving the incoherent svabhaava concept and the higher as indescribable.[24] On this basis he cannot coherently differentiate the sense in which he means these terms from the usage of the Praasangikas, and so historically his school was cast into disrepute.[25]

The answer to the dilemma is that what constitutes an element of a relational system is itself a part of that system, not something existing prior to it.[26] The system provides criteria whereby its elements are to be identified. This amounts to assuring ourselves that acting on the assumption that this is a case of X will have the hoped for consequences. Asking whether we have a "real" X, over and above this, is a question with no sense attached to it. Demanding that some elements be identified prior to the system amounts to asking that we see whether the criteria apply without consulting them. It is a mistake to think that there is anything arbitrary in the way we use language. This mistake comes

p.287

from confusing what it is proper to say about linguistic systems with what can be said about reality as the base of all such systems. The base is indescribable, of course, but only because it is defined as what exists prior to descriptive articulation.

V

Having applied this bit of analytic thought to the Svaatantrika point of view we are now in a position to make sense of Naagaarjuna's assertion in MK XXIV that `suunyataa is identical with interdependent origination.

18.

yah pratiityasamutpaada.h `Suunyataa.m taa.m pracak.smahe
sa praj~naptir upaadaaya pratipat saiva madhyamaa

We declare that whatever is a relational origination is `suunyataa. It is a provisional name (i.e., thought construction) for the mutuality (of being) and indeed, it is the middle path.

19.

apratiitya samutpanno dharmah ka`scin na vidyate
yasmaat taasmad a`suunyo hi dharma.h ka`scin na vidyate

Any factor of experience which does not participate in relational origination cannot exist. Therefore, any factor of experience not in the nature of `suunya cannot exist.

The problem is that this amounts to a statement of the parabhaava view that has previously been declared incoherent. There is no logical reason whatsoever for preferring it to the svabhaava view. Naagaarjuna says that the reason for accepting it has some relationship to the distinction between the absolute and relative truths, implying that this distinction allows us to escape the nihilism of a purely logical approach.

9

ye 'nayor na vijaananti vibhaagam satyayor dvayoh
te tattvam na vijaananti gambhiira.m buddha`saasane

Those who do not know the distinction between the two kinds of truth cannot understand the profound nature of the Buddha's teaching.

10.

vyavahaaram anaa`sritya paramaartho na de`syate
paramaartham anaagamya nirvaa.na.m naadhigamyate

Without relying on common practice (i.e., relative truths) the absolute truth cannot be expressed. Without approaching the absolute truth, nirvaa.na cannot be obtained.

As verse 7 suggests, it is a matter of pragmatics that one distinguishes two truths. The lower truth is necessary both for the business of ordinary life and for use as a sort of medicine that can bring people to the higher truth. The Praasa`ngika interpretation of this is that the doctrine of dependent origination is useful as a teaching device, but that ultimately it is incoherent.[27]

The Svaatantrikas, as presented, tried to define a limited realm in which

p.288

ordinary statements, interpreted via the parabhaava view, would be valid. Candrakiirti points out that to claim with Bhaavaviveka that there are valid Indian syllogisms (which require examples) is also to claim that there is such a realm.[28] Interpreting the equation of `suunyataa and dependent origination is then quite straightforward: Since all existing things are necessarily causally and criterially interdependent, they have no absolute existence in the sense of svabhaava.

This line of reasoning has important consequences for our ideas about philosophy. It breaks down the empiricist's absolute distinction between science as the study of fact and philosophy as the study of logical relations. If reality, as the significant elements of fact, arises only after systematic interpretational operations on a presignificant base, it follows that the factual and logical relations of the elements arise only after the articulation of such a system. The setting up of the system is not properly described either as an analytic or a synthetic operation. Einstein was neither (or both) scientist and philosopher when he suggested cogent reasons for changing the ways we spoke of space and time in the context of the relativity theory. The progression from Mosaic physical adultery to Jesus' adultery in the heart, to Freud's unconscious adultery involves paradoxical shifts in usage that allow us to work with facts that we had no way to articulate before.[29]

Kuhn, Fodor, and Chihara and others have presented an evolutionary picture of science and its language in the light of this point of view.[30] For the Maadhyamika no concepts are immune from this dynamic process, in the sense that they refer to an absolute stratum of reality presupposed by all other concepts.[31] Two very important consequences follow for the nature of philosophy. First, philosophy is not simply a means of analyzing away metaphysical illusions, or kinks, in ordinary usage. It becomes a creative enterprise, that without being arbitrarily speculative can conceive changes in the way we view the world that will make a real difference in the course of our lives. On this point of view philosophy is one pole of an interplay with practical activity to whose evolution we can set no absolute limits. Science is one product of such interplay and the changes of psychology and apprehension of the world accompanying progress on the Buddhist path another.

The second consequence is that because construction and articulation are now seen as prior to factual meaning, the human element in reality is more basic than any construction of "hard fact." Humanity as constructor is of a higher logical type than the articulations of nature, including human nature, that exist at any particular point. Again it is the interplay of philosophy and practice that makes this possibility more than empty verbiage. For the Maadhyamika it is of course the bodhisattva path that is the paradigm of such interplay.

p.289

NOTES

1.Naagaarjuna. A Translation of his Muulamadhyamakaakarikaa (Tokyo:Hokuseido, 1970); see Introduction to MK I.

2.Presuppositions of India's Philosophies(Englewood Cliffs,N.J.: Prentice-Hall, 1963), especially chap. 5.

3.Herbert V. Guenther, Buddhist Philosophy in Theory and Practice (Baltimore, Md.: Pelican Books, 197?), pp. 90ff.

4."La Philosophie Bouddhiquz de la vacuit`e,"Studia Philosophica,vol,XVIII,1958, specially pp.128,136-137.For a.discussion of collapsing of contexts to bring out formal analogies see G. S.Brown, Laws of Form(New York: Bantam Books, 1972), Preface, Introduction, and Appendix 2,who compares such a process to the use of imaginary numbers. Of course there is a claim that the bodhisattva path involves experiential correlates to the collapsed form, though ordinary experience does not.

5.Cf. C. S. Chihara and J. A. Fodor,"Operationalism and Ordinary Language," reprinted in George Pitcher,ed.,Wittgenstein:Philosophical lnvestigations(EnglewoodCliffs, N.J.:Prentice -Hall, 1964), pp. 409-13.

6.Ludwig Wittgenstein, On Certainty, G. E. M. Anscombe and G. H. van Wright, eds. (Oxford: Basil Blackwell, 1969), no. 618. Cf., ibid., no. 318ff.

7.Bertrand Russell, "Logical Atomism" and "The Philosophy of Logical Atomism," in Logic and Knowledge, R. C. Marsh, ed. (London: George Allen and Unwin, 1964).

8.In example,MK XV-IO,"It is'is a notion of eternity." Elsewhere Naagaarjuna allows that, as far as the notion of svabhaava goes,the way is left open for an eternal,unchanging being. But such a being would be quite irrelevant to our experience, since per hypothesem it could not in any way react with it.

9.Ludwig Wittgenstein, Philosophical Investigations, G. E. M. Anscombe, trans. (Oxford: Basil Blackwell, 1958), no. 243ff.

10.Ibid., no. 258ff

11.See Jonathan Bennett in Kant's Analytic(London: Cambridge University Press, 1966), for discussions on senses in which it makes sense to see Kant's principle of universal causality as analytic or self-evident, and in which discussions are very conveniently indexed.

12.R. L. Gregory, Eye and Brain (New York: McGraw-Hill, 1966), pp. 191ff.

13.(London: Methuen, 1964), pp. 16-17.

14."Strawson and Ontological Priority,"in Analytical Philosophy,R.J.Butler,ed. (New York: Barnes and Noble, Second Series, 1965).

15.Ludwig Wittgenstein, Tractatus Logico-Philosophlicus. D. F. Pears and McGuinness, trans. (London: Routledge and Kegan Paul, 1962); see especially no. 6.53-54.

16.That this is a fair rendering of the traditional understanding of the Praasa~ngika view can be verified by comparing Mipham's account in Guenther, p. 151.

17.G. E. Moore, "A Proof of the External World," in Philosophical Papers (London: George Alien and Unwin, 1959), p. 146.

18.Iida, "An Introduction to Svaatantrika Maadhyamika," Ann Arbor, Michigan, Ph.D. Thesis, University Microfilm,University of Wisconsin,Madison,1968, p.260. This is the single clearest locus,but as this is Bhaavaviveka's central distinction many other passages are relevant. Iida's thesis is at this time the only available extended treatment of the Svaatantrikas in English.

19.Ibid., p. 94.

20.Ibid., p. 115.

21.Guenther, p. 126.

22.Iida,pp.280ff.Iida also discusses the suutra's relevance to and possible influence on Bhaavaviveka.

23.This is the impression arising from work currently done,for example,Guenther,124ff. It must be emphasized that this remains a largely unexplored area.

24.Iida, p. 94. See also the final discussion of the two truths.

25.Theodore Stcherbatsky, The Conception of Buddhist Nirvana (The Hague: Mouton, 1965), pp.105-119 especially. This is a translation of Candrakiirti s commentary on MK I.It is dubious, but good enough for our purposes if equations of `Suunyataa and monism and the like are systematically

p.290

corrected.It is primarily concerned with a refutation of Bhaavaviveka's attempt to stare the Maadhyamika point of view in positive terms.

26.Cf. Wittgenstein, Philosophical Invesrigarions, nos. 46-64.

27.Cf., notes 26 and 15 herein.

28.Cf., note 25 herein. See also Guenther's and lida's discussions.

29.J. Wisdom, "Lectures on Explanation at the University of Virginia."

30.T. S. Kuhn, The Structure of Scientific Revolutions (Chicago, 111.: University of Chicago Press, 1962). See also Chihara and Fodor.

31.In his last years Wittgenstein seems to have been working toward a similar idea: that the groundlessness of language games was also the possibility of their change or evolution. Cf. On Certainty, nos. 140-145, 166, 247, 256, 375, 559, 618.