Topics on Being and Logical Reasoning

·期刊原文
Topics on Being and Logical Reasoning
By Richard S. Y. Chi
Philosophy East & West
V. 24 (1974)
pp. 293-300
Copyright 1974 by University of Hawaii Press

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

p. 293

Having carefully read Bochenski's paper, I have only admiration for his most scholarly treatment of the subject of logic and ontology and have greatly benefited from it. On one hand, I do not have any point to make in disagreement with his views and, on the other, I am unable to offer a similar treatment on the Chinese aspect of this subject to supplement his article due to a situation which might be called "cultural imbalance" between East and West. In China, Buddhistic philosophy is dominated by epistemology, and non-Buddhistic philosophy, by ethics. As Bochenski has rightly pointed out, the Chinese logic gives the appearance of being mere dialectics which never rose to the level of formal logic. Consequently, the Chinese counterpart of the subject of logic and ontology might not be significant enough to warrant a comparison. It might be possible to organize the ideas of the two disciplines existing implicitly in Chinese thought, but this would require lengthy research before one could offer a comprehensive survey comparable to Bochenski's contribution on Western thought. I hope, therefore, that I may be allowed to express my random thoughts on several topics related to the subject of being and logical reasoning. I shall confine my comments to Buddhism.

I
Buddhism in general is well known to be negativistic; the doctrine of the void (`suunya) is accepted by all schools, even including the realistic schools such as Sarvaastivaada. It is, however, by no means nihilistic; it rejects nihilism even more than realism, which is regarded as a "lesser evil." What the doctrine of the void rejects, is the existence of independent being, or being-by-itself (svabhaava), but not Being (bhaava). The rejection of the independence of anything whatsoever, whether of things or of ideas, has been the central teaching of Buddhism.

The concept of independence is characterized:by a multitude of technical terms coined throughout the ages. The meanings of these terms may not necessarily be equivalent in extension. Some of these terms are unrelated in non-Buddhistic contexts. These numerous terms can be roughly categorized into three major groups represented by the terms svabhaava (own state of being), aatman (individual self) and svatantra (own system). The three signify respectively the self-preserving aspect, the self-referent aspect and the self-asserting aspect of the concept of independence, when it is applied respectively to entities in general, to personal identity, and to logical reasoning.

II
The first, self-preserving aspect of the concept implies the qualities of substantiality, reality, durability, discreteness, inertness, etc. These qualities are

p. 294

in direct contradiction with the universal law of "dependent origination" (pratiitya samutpaada), which indicates the interdependence of all existences. According to the reasoning of the Maadhyamika school, the empirical world which appears before us is not the result of the existence of some independent substances, but is the result of their nonexistence. A substance cannot be the cause of the existence of the empirical world, but can be only its obstacle. Suppose the world consists of independent substances. Then we would not be able to see anything or to hear anything, and the entire world would be reduced to sheer nothingness. This hypothetical result is contrary to our direct experience.

The word svabhaava is just as ambiguous as the English word reality. It has at least three meanings at three different levels of truth: the apparent, the empirical and the absolute. Philosophers give various interpretations. The following one is by Vasubandhu:

1. The apparent: The existence of independent objects in our perception appears to be real. Such a reality is fictitious, because it is merely a feeling of reality created by mental fabrication. This "fabricated reality" is said to be unreal in the sense of "being nonexistent by its very nature."

2. The empirical: All processes in the empirical world, including the process of mental fabrication, are certainly not fictitious, but are devoid of independent existence because they are subject to certain conditions. This "relative reality" is said to be unreal in the sense of "being devoid of a being-in-itself."

3. The absolute: The ultimate reality is just the elimination of the fabricated reality from the relative reality. This "ultimate reality" is said to be unreal in the sense that its nature is "ultimate absence of a reality-in-its-own-right."

III
The second, self-referent aspect of the concept of independence, represented by the term aatman, is the application of the concept to personal identity (pudgala). It has a range of connotations: an agent, a possessor of life, a possessor of free will, an individual-in-itself, etc. The reflexiveness of the concept of personal identity renders the action of "identification" meaningless. Since Buddhism was motivated by the desire to solve the problem of suffering, the study of "self," which is supposed to be the basis of suffering, has been the major topic in practical Buddhism.

IV
The third, self-asserting aspect of the concept of independence is its application to doctrine and logic. The discussion of this topic started as a minor controversy but ended with the split of the Maadhyamika into two subschools. One

p. 295

is called "Svaatantrika," headed by Bhaavaviveka; the other is called "Praasangika," headed by Candrakiirti. The former asserted the necessity of presenting one's own thesis, supported by a valid syllogism, in refuting the opponent's position. The latter considered that the presentation of one's own thesis was not only unnecessary but also unacceptable.

The maximum function that a syllogism can provide is its validity, and the maximum value that a major premise can offer is the acceptance by two disputing parties, or at most, that of the uncommitted audience. In actuality, however, people generally either expect more or take more for granted. What they expect from a syllogism is proof rather than mere validity, and what they expect from a major premise is self-evident truth rather than mere acceptance. The overevaluation of logical inference gives rise to dogmatism.

Buddhism, Maadhyamika in particular, is a critical philosophy which rejects all dogmatism as mere indulgence in speculation (d.r.s.ti). If Buddhism were to establish its own position by presenting in its doctrine of the void its own "first principle," as Bhaavaviveka did, then Buddhism would become an anti-dogmatic dogmatism. A close analogy would be to use shouting to stop the noise of a quarrel. If a critical school were particularly privileged to keep its own dogma when it had rejected all others, its fault would be even more irremissible than that committed by a noncritical school. Candrakiirti's solution to this problem is not proposing any independent thesis at all. His only logical instrument is reductio ad absurdum, which never stands alone but always depends on the thesis proposed by the opponent.

V
At this point I cannot help thinking about the old problem of the "four-cornered refutation" (catu.sko.ti), which has been a puzzling topic for centuries. It rejects the assertion of certain propositions, their denial, both assertion and denial, and neither assertion nor denial. It is puzzling because first, the scheme itself is both redundant and self-contradictory, According to the law of double negation, the third and the fourth alternatives are materially equivalent. The third is self-contradictory; so is the fourth. Secondly, the rejection of all four alternatives is also self-contradictory. The denial of the first is equivalent to the assertion of the second, and vice versa. Denying both the first and the second would be committing the same mistake as that committed by the fourth. These faults should be apparent even to someone without a background in logic; it would be incredible that Buddhist logicians could possibly fail to see this point.

This problem has been studied by many scholars, including T. R. V. Murti, [1] R. H. Robinson, [2] and K. N. Jayatilleke, [3] and myself. [4] Scholars agree with one another on some points, and disagree on some other points. It seems that

p. 296

on this subject, few people could reach full agreement. In the Appendix to this article, I have listed all the possible formulae on four-cornered arguments, monadic and dyadic, quantified and unquantified. These formulae can cover most examples given by Jayatilleke, [5] with the exception of the "inexpressibles" (avyaak.rta), that is, "Whether the world is finite?" etc., in which none of these formulae can fit.

These formulae include the "tetralemma" proposed by Robinson. It is, in fact, the Indian version of the so-called "categorical propositions":

"-vyaapaka" inclusion (Form A with existential import)
"-av.rtti" exclusion (Form E with existential import)
"-ekade`sav.rtti" intersecting (Conjunction of Form I and Form O)
"avidyamaana-" nonexistent (Conjunction of Form A and Form E without existential import)

The first three were introduced by Dignaaga, and the last one, by his critic Uddyotakara. [6] This scheme belongs to the quantified form of conventional two-valued logic, only slightly different from Aristotle's categorical propositions. The first two are familiar to everybody; the last two can be illustrated by the following examples respectively:

"Some philosophers are participants of this conference, and some are not."
"No unicorn is participant, and no unicorn is not."

From the above we can see that, although the four forms are indeed distinct and exhaustive, they are hardly relevant to catu.sko.ti at all. The presence of scanty examples of quantified propositions in the Muula Madhyamaka Kaarikaa is not sufficient to prove the relevance.

In order to solve the puzzle, we have to trace what motivated the Buddhists to talk about this topic. The idea of catu.sko.ti is just an extended version of the rejection of dogmatism (d.r.s.ti) in doctrine, and the rejection of establishing an independent thesis (svatantra) in logical reasoning. The Buddhists had gone further than the Greek Skeptics, who called every assertion of a positive proposition "dogmatism." According to Buddhism, one cannot escape from dogmatism even by denial, because denial of a proposition is equivalent to the assertion of its counterproposition.

It is quite natural, when people have learned that both "extremes" have been rejected as dogmatism, that they try to avoid them by remaining in a middle position, in the form either of the mean, or of a synthesis, in the Hegelian sense. It is also natural that others try to avoid dogmatism by denying everything, whether affirmation or negation, and thus leading to skepticism or agnosticism. The catu.sko.ti merely shows that, as long as one holds one's own view to be ultimate, one can never avoid dogmatism, not in assertion, or even in denial, synthesis, or skepticism.

p. 297

Objection might arise at this point. If Buddhist refutation is a kind of denial, then it would be committing the same mistake as that committed in the second and the fourth alternatives which have been refuted by Buddhists. If it is other than denial, then what is the difference between refutation and denial? The answer is that the Praasangika's method in refutation is not proving the falsity of a proposition by means of a syllogism but by reductio ad absurdum, which shows the absurdity of a proposition rather than its falsity. The difference between falsity and absurdity is that the falsity of a proposition "p" is equivalent to the truth of its negative proposition "~-p," while the absurdity of "p" is not equivalent to the truth of "~p," because "p" and "~-p" may be equally absurd when they are nonsensical, ill formed, or self-contradictory.

The next question is whether the system of catu.sko.ti is sound in the sense that the four alternatives are logically distinct and exhaustive. The answer is as follows:

First, the system is a linguistic classification, that is, single assertion and denial, double assertion and denial. Linguistically speaking, the four are both distinct and exhaustive.

Second, even if it is considered as a logical classification, since it is supposed to include many conflicting schools, conventional and unconventional, logical and illogical, the system should reserve room for the illogical. Consequently, there is no reason why one should impose any law of logic at all, whether the law of the excluded middle or anything else.

In this respect, it is farfetched to think that catu.sko.ti belongs to a particular kind of non-Aristotelian logic. In many Chinese and Tibetan texts, the third alternative is condemned as "self-contradictory" and the fourth, as "sheer play of words without content." Therefore Buddhist logic belongs to strictly conventional two-valued logic. Jayatilleke's comment, "The Buddhist four-fold logic is in this respect no more true or false than the Aristotelian...." does not seem to be applicable. [7]

At this point I must confess that I have made a similar mistake by comparing catu.sko.ti with intuitionism. It happens that the system of catu.sko.ti can be formulated by the logic of intuitionism, in which some conventional laws of logic, the law of excluded middle, etc., are rejected. [8] Despite the superficial resemblance, it is incredible that Naagaarjuna and Brouwer could possibly think in the same way. Dialectics and pure mathematics are, after all, two different disciplines. The agreement of the two systems is a sheer coincidence; they reach the same result for different reasons. Establishing a system in which certain laws are rejected is totally different from not imposing any law at all.

Third, in reality we do need four different categories for classifying thinkers; two or three are not enough. Synthesis and skepticism should not be put in one and the same category.

p. 298

Although the subject of catu.sko.ti is extensively mentioned by Buddhists, it is not at all "Buddhist logic" because Buddhists are its critics, not its sympathizers. Consequently, Buddhists are not responsible for its being sound or not; it is not necessary that a system should be sound before one can reject it.

After many years of dispute, I still think that Murti's view on this subject is the right one, if one minor point is clarified. His comment, "Four alternative views are possible on any subject," might cause objection. [9] There are two possible explanations. If the system of "four alternative views" in this context is taken in its broader sense, as in the Appendix to this paper, then there is nothing wrong in his statement. If he means the system in its narrow sense, that is, the catu.sko.ti refuted by Buddhists, then this statement might be the result of a minor slip of the pen, because some propositions are definitely true or definitely false and leave no room for alternatives. As a matter of fact, catu.sko.ti is applicable to metaphysical speculations only, and Buddhists would most unlikely apply it to propositions in general.

VI
When the rejection of independence is carried out thoroughly, it eventually reaches a point at which abstraction and theorization are abolished. In reality, there was indeed such a school of Buddhism in history; it is the Ch'an School of China, called Zen in Japan. In the vast Ch'an literature, excepting its early period, abstract language disappears completely. Instead of abstract theories, there are thousands of concrete "cases" (kung-an) which record individual "situations" (chi) in daily life. Some cases are simple. For instance, several cases can be reduced to the problem of personal identity: "Who identifies whom with what?" Some others involve very complicated paradoxes.

The existence of a paradox is generally due to the existence of a flaw somewhere in the argument. Usually it is the unnoticed existence of some implicit premise, which has been hidden because it appears to be too self-evident to be mentioned at all, though it is usually false. The paradoxes can be solved, according to the methodology of the Ch'an school, by utilizing a psychological state created by the urge to solve the paradox, called "the state of being puzzled" (i-ch'ing).

VII
To conclude, I would like to give a brief account of "logic" in China during three periods.

1. The logic of ancient China. As Bochenski has said, it is dialectics rather than formal logic. This dialectics was short-lived and exerted little influence on Chinese thought.

p. 299

2. The importation of Buddhist logic. Catu.sko.ti is by no means "Buddhist logic." The real Buddhist logic established by Dignaaga belongs to conventional two-valued formal logic. The difference between Aristotle's four-operator system and Dignaaga's three-operator system, and that between Aristotle's three-member syllogism and Dignaaga's three-member syllogism (which was reduced from a five-member syllogism) is not substantial. [10]

The influence of Buddhist logic on Chinese thought is almost nil. According to some record, when Buddhist logic was introduced by Hsuan-tsang to China in the eighth century, his headquarters in the capital, where important doctrinal discussions were held, was "bugged" by his rival, Yuan-ts'e. As a sort of counterintelligence measure, the teaching of logic, which was regarded as the weapon for doctrinal combat, was kept as a highly guarded secret. [11] I cannot judge the authenticity of the strange story, but it: is quite certain that without verbal instruction, the huge literature on logic written by Hsuan-tsang's disciples, a literature which is full of professional jargon, is almost unintelligible to even learned Chinese scholars.

3. The logic of Ch'an. I have heard too often that "Zen is illogical and irrational." I do not know what the word "illogical" is meant to convey in this statement. If one looked into a book on elementary logic, one would find that tautological formulae in material implication, such as , , seem to be "illogical," too. Since people already have the preconceived idea that it is illogical, thereby adding some mysterious religious flavor, which is rather irrelevant, this field is unexplored, despite the recent appearance of many books about this subject on the market.

The content of Ch'an literature is unusually rich; in thousands of cases, there must be a great deal to be studied. Since Ch'an deals with extremely complicated paradoxes and offers a detection procedure for logical fallacies hidden in propositions usually considered "logical," its subject matter inevitably involves logic. Since it is unexplored, it is hard to tell how "logical" it is. What I can tell is that, if it is truly "logical," this kind of logic is implicit and situational but not formularized. It is more likely to be overlogical than illogical.

NOTES
1. T. R.V. Murti, The Central Philosophy of Buddhism (London, 1960), p. 130.

2. R. H. Robinson, "Some Logical Aspects of Naagaarjuna's System," in Philosophy East and West, 6, no. 4 (Jan., 1957): 302-303.

3. K. N. Jayatilleke, Early Buddhist Theory of Knowledge (London,1963), pp. 333-351; K. N. Jayatilleke, "The Logic of Four Alternatives," Philosophy East and West, 17, nos. 1-4 (1967): 69-83.

4. R. S. Y. Chi, Buddhist Formal Logic (London, 1969), pp. 156-163.

5. Early Buddhist Theory of Knowledge, pp. 334-347.

6. Buddhist Formal Logic, pp. 8-10, 45.

p.300

7. Early Buddhist Theory of Knowledge, p. 350.

8. Buddhist Formal Logic, pp. 162-163.

9. The Central Philosophy of Buddhism, p. 129.

10. Buddhist Formal Logic, pp. xxvi-xxvii.

11. Ibid., p. lxxvi.

APPENDIX
________________________________________________________________________
Unquantified Quantified
________________________________________________________________________
Monadic p (Ex)fx.~(Ex)~fx
~p ~(Ex)fx.(Ex)~fx
pv~p(tautology) (Ex)fx. (Ex)~fx
p˙~p(contradiction) ~(Ex)fx.~(Ex)~fx
_________________________________________________________________________
Dyadic p˙~q (Ex)(fx.gx).~(Ex)( fx.~gx)
~p˙q ~(Ex)(fx.gx).(Ex) (fx.~gx)
p˙q (Ex) (fx.gx).(Ex) (fx.~gx)
~p˙~q ~(Ex)(fx.gx) .~(Ex) (fx.~gx)
_________________________________________________________________________

a. The last group of quantified dyadic forms is equivalent to tetralemma:

b. None of the above formulae can fit in the scheme of catu.sko.ti. The following formulae with the symbols of intuitionism can formally fit in the scheme, but the interpretation of intuitionism differs from that of catu.sko.ti.

c. The terms "monadic" and "dyadic" are borrowed, because I have not found concise names for the two groups of formulae.