Reflexivity and metalanguage games in Buddhist causality

·期刊原文
Reflexivity and metalanguage games in Buddhist causality
Douglas D. Daye
Philosophy East and West
Vol.25 (1975)
pp.95-100
Copyright By University of Hawaii Press

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

p.95

I. INTRODUCTION

Within the context of the three papers which address the subject of Buddhist causation, I have noted (or projected) a sequence of epistemological concerns and a familiar "gap." After making brief references to the three papers, I shall make some humble methodological points and suggest an option which philosophers whose area is Buddhist thought may possibly wish to consider. The sequence is as follows: David J. Kalupahana's article gives us a clear exposition of the background, controversies, and vocabulary of the central concepts of causation in Indian Buddhist thought. He touches on the psychological aspects of causation. Luis G`omez has given us an exceedingly articulate analysis of the free-will question and dilemma as found in the Nikaayas, which clearly refer to both the psychological aspects of freely chosen actions and the bondage of karmic (neyaartha) causal relations. Additionally, Frederick Streng's brings into clear focus, the psychological and first-person epistemic considerations and values so important for the soteriological aspects of Buddhist causality. In addition to the two admirable papers, Streng's paper explicitly adds the additional "personal" first-person dimension to the concept of causation. The problem of the "gap" may be formulated as follows: given the religious and soteriological values in the Buddhist context, causation would seem ostensibly to involve only a consideration of third-person epistemological problems. However, given the soteriology of Buddhism, it seems necessary to examine the methods by which the "gap" between first-person and third-person statements about causality can be justified. It is my wish to continue the sequence I observed in these two papers; I suggest that Buddhist answers concerning the gap between the first-person and the third-person causal statements may be fruitfully examined by focusing attention on the problems of sequential metalanguages, reflexive statements, and content-restricted technical vocabulary such as santaana. I would hold that this epistemic "gap" constitutes a central problem in any epistemology; and it is of particular importance concerning Buddhist causality. I shall suggest that similar problems may be found in the Sarvaastivaada-Sautraantika controversies concerning the Sarvaastivaada dharma "praapti," and the Maadhyamika statement that "emptiness (`suunyataa) is a language-mental-construct" (praj~napti, 24:18). After reflecting upon these problems I shall make certain comments and suggestions for future research. I take it that this is one of the core concepts of what is entailed by the word "retrospect."

II. REFLEXIVITY AND FIRST- AND THIRD-PERSON LANGUAGE GAMES

I wish to call attention to certain metalanguage aspects to which Professors

p.96

Streng and G`omez allude but do not have time to consider. I would hold that these aspects are common to almost all of the major developments in Buddhist philosophy, namely (1) the methodological truism that there are multileveled metalanguages possessing a context-restricted vocabulary within the context of specific Buddhist philosophies, which are not only descriptive and/or evaluative, but are inherently reflexive ; (2) that within these context-restricted vocabularies specific first-person verbal formulations of the problem of causation may not be appropriate or "sayable." For example, within the Abhidharma-dharmavicaya vocabulary one cannot strictly use the word "self" (aatman, Jiiva, pudgala) although one can use the technical term "santaana" to talk about the problem of the dharmic and karmic relations of the "self." I am saying that the accurate and appropriate method of speaking about a problem changes with its vocabulary; in a very superficial sense, epistemology is contingent upon vocabulary. However, I am not saying that the epistemological object of an analysis is necessarily changing. An analysis of Buddhist metalanguages will enable one to clarify the accuracy of described relations between first-person and third-person reflexive causal statements.

By reflexive propositions I mean that the import and the reference of such propositions also includes the conceptual framework of the first-person reports of the speaker (or R relation is reflexive if and only if aRa holds for all a that are members of the field of R). That is, reflexive statements here strive to bridge the first-person-third-person epistemological gap. The reflexive nature of certain high level Buddhist concepts and an examination of the explicit differences of what can be appropriately said at those levels is desirable.

Streng notes that "karma ... is itself both the product and initiator of existence ... it is a relationship which coordinates the mementary factors (dharmas) as they pulsate in and out of existence. The arising, maintenance and disillusion of existence is an orderly sequence which should not be seen as a universal substance nor simply a mental projection onto something else," and "... thought ... itself is a contributing factor in the arising of phenomena. The formation of mental units of experience is seen as a significant force in what can be identified.... as a 'causal process'." This implicit reflexivity becomes important when we consider that these dharmas come in and out of existence known (first person) as a coordinated sequence according to the metaexplanation of karman (third person.) Dharmas then are the mutually exclusive, totally exhaustive components of a finite epistemological (and only perhaps an ontological) framework within which dharmas are "known," is also a product of the epistemological presuppositions of the knower, that is, first person. Thus when talking of dharmic causal relationships, reflexivity is implicit. Causal relations between dharmas are not seen as something "out there" divorced from the knower;Abhidharma theory suggests that the framework, at least half of which is the contribution of the knower, is a causally coordinated epistemic whole; thus dharmavicaya is an epistemic framework

p.97

by which one knows both the "knower" and the dharmas of "his" first-person epistemic field. As Streng has noted, the knower is not independent of the framework within which the known is identified and within which the coordinated sequences of the known are articulated.

Those Buddhists of the Abhidarmic schools and Maadhyamika were very careful to insure that no hypothesization or reification errors of first-person reports should take place, for example, self (aatman), svabhaava. Although there are metaexplanations of causal sequences between dharmas in which the analysis of the total field of dharmic first-person experience occurs, I would suggest that there is entailed little of ontological significance outside the epistemic framework within which the object of knowledge, the relationship of knowing and the knower are actually apprehended. Minimally, such epistemological analysis is the condition of the possibility of ontology.

An alternative way to state the same problem of first- and third-person epistemic "gaps" concerning causality, is mistakenly to assume that the knower is independently projecting causal sequences upon the object of knowledge at a metalevel about what is known. To hypothesize an ontological knower, causally independent of the relationship of knowing, is to make a reverse category mistake. The Buddhist anaatman doctrine holds that, although it is misleading to use the world "self" because it tends to hypothesize an entity apart from the fluxing components of what is known, and then to postulate that the self is the knower, it is to suggest that there is a self, independent of dharmic causation--a reverse category mistake. That is, the "knower" is just as much a component of the ordinary language epistemic field as is the relation of knowing and the characteristics of object known. To know that the particular epistemic components (dharmas) involved in the metarelationships of knowing, are members of the total class of what is knowable is to avoid making such a category mistake. That is, the epistemic components (dharmas) which are said to be provisionally "inner" first person are just as much members of the karmic factors of causation as are the "outer" third-person epistemic components of knowing. "Outer" (baahya) and "inner" (aadhyaatmika) are only provisional terms, philosophical praj~naptis, epistemic upaayas, or pedagogical rafts.

I would go further than Streng and suggest that not only are there different levels of consciousness, but furthermore, a very strong case for a series of metalanguages can be made by examining the actual uses of special language as argued. I am suggesting that what is needed is an explicit, accurate map of the metalanguage levels which are clearly present in the Buddhist texts. Furthermore such contemporary linguistic concepts as language games and the analysis of "ordinary language" are not, to paraphrase one eminent East-West philosopher, "just those 'silly games' they play at Oxford." Rather some of these descriptive methods are directly applicable as heuristic devices to the analysis of the reflexivity of metalanguage levels postulated implicitly by some Buddhists. After all, language game analysis is only an alternative method of

p.98

describing what is said, what can be said and why x is said; it is not a universal warrant to the effect that all utterances are of equal philosophical worth. To illustrate these points I wish to examine briefly certain reflexive concepts in both Maadhymaika and Abhidharma.

III. AN OVERLAP BETWEEN TWO BUDDHIST LANGUAGE-GAMES

An example of the overlap between two Buddhist language-games, one of which is a metalanguage to the other, can be seen in the Abhidharma use of a surrogateself concept : santaana. We realize from a neyaartha point of view, certain concepts belong to "common sense," the first language-game; these are pragmatically useful at a relatively low level of religious or philosophic insight. Certain questions are postulated at this level for which the Buddhist philosophical solutions can be answered only at a metalanguage level--from the niitaartha level--the second language-game. The word "santaana" was postulated to talk about the dharmic concantinations which (metaphorically speaking at a lower level of linguistic usage) can be said to "belong" to an "individual." Explicit here is the warning that no self or ontological entity is suggested or implied by the use of the word "santaana"; however, the Buddhists maintain that it is a fruitful way to talk about wholesome or unwholesome karmic actions. These are present in the karmic causal stream (santaana) about which we postulate the word "self" in order to designate the most "personal" dharmas which metaphorically "belong" to an "individual." It is at this point in the second language-game of Abhidharma (dharmavicaya) that we find that there is no vocabulary for speaking about the ordinary language concept of an ontological self; it is also at this point that these two language-games overlap. This overlapping then generates the need for a naataartha surrogate concept for the self-concept, that is, santaana. These low level overlapping neyaartha terms are generated by a motivation to answer low level questions with high-level niitaartha answers, that is, "ordinary language" questions with, for example, Abhidharma answers. Thus the term "santaana" grows out of the need to give more accurate metalanguage answers to logically complex, inaccurate common sense questions. Theoretically, if one spoke only in Abhidharmic language, one would never feel the need for talking about "one's own" dharmas, santaana is a linguistic accommodation to ill-put questions and inaccurate (avidyaa) presuppositions which Buddhist Abhidharma is designed to rectify.

In Maadhyamika we find the same problems of reflexivity and metalanguages at work in MK 24:18 where emptiness is said to be only a language construct (praj~napti). In order to warn against the hypothesization of emptiness as an ontological entity rather than, as I take it to be, a reflexive rule about the evaluation of other metalanguage concepts, we find this warning about hypothesizing emptiness. It is a warning for those who would take emptiness as a "thing"--word having ontological overtones, rather than as a high order

p.99

metalanguage description. Here "emptiness" refers to the absence of any absolute ontological fulcrum from which one can make assertions about a reality which may be, by definition, beyond the confines and restrictions of what is actually and potentially known within the common human language neyaartha framework. The reflexivity of praj~napti needs to be made explicit. Just as the concept of self as denied as an extracategorical concept among the epistemic descriptive components of Abhidharmic language, just so is there the Maadhyamika warning against ontological postulation made by means of the word "praj~napti." This insures the nonreification of certain high level metalanguage concepts such as emptiness. Thus praj~napti is reflexive.

IV. CAUSALITY, IDENTITY AND THE SHIFT OF THE LANGUAGE GAMES

It seems also clear within the Abhidharmic language framework that we find there is no noncircular justification for postulated sequential causal relationships. As an example, let us consider a general argument concerning personal identity through time. An argument for the denial of such causal relations can be made in the Abhidharmic framework by saying that any past identity, when compared with the present identity, cannot be postulated as the same known (neyaartha) entity through time. This is because the Abhidharmist holds that the "content" of the component memory dharma (smrrti) which enables one to compare an image of a past identity with a perception of a present dharmic events, a present identity, all occur at in present time. In other words we are not identifying an object through time by comparing it with something in the past and with something in the present; on the contrary, what we are comparing is two things in the very present, one of which is the "content" of a memory dharma of the past and the other is an immediate field of coordinated dharmic synergies which occur wholly in the present. All memories of the past are merely immediate components of present moment. Thus, the causal sequences which might be expressed by " É " become circular staccato-like epistemic equivalences "=". However, epistemological solipsism may be partially "escaped" by shifting the gears of language from niitaartha metalanguage back to "ordinary" neyaartha language. The reflexive doctrines of "upaaya" and "praj~napti" seem to be reflexive metaconcepts which allude to this shift of linguistic gears from high level philosophical niitaartha concepts to the low-level neyaartha "ordinary" language concepts. Tentatively, this is also one way to approach Buddhist answers regarding the persistent confusion surroundding the propositions that there is perception but no perceiver, thoughts but no independent ontological thinker.

In conclusion, it is my suggestion that by focusing attention upon metalanguage levels and implicitly reflexive relationships, the "gap" between third-person discussions of causal relations, and first-person causal statements about "psychological" aspects involving soteriological values--religious upaayas

p.100

--may be explicitly bridged and carefully articulated. Language game analysis and reflexivity are merely conceptual means, heuristic praj~naptis, to help us describe the particular Buddhist answers to this universal epistemological problem.

To quote Mikel Dufrenne, "the universe of discourse becomes the discourse of the universe."