Rejoinder to Michael Levin's Comments
on the Paradoxicality of the Koans
By Chung-Ying Cheng

The Journal of Chinese Philosophy
V. 3 (1976)
pp. 291-297

Copyright 1976 by D. Reidel Publishing Company


 

 

p.291

There are several things which I think strongly suggest that Michael Levin either misunderstands my exposition of Ch'an (Zen) paradoxes or misunderstands the genuine spirit of Ch'an Buddhism and its use of koans (Kung-an) as an exercise toward enlightenment. These I believe need clarification so that misunderstandings will not arise.

 

I

In the first place Levin confuses the set-theoretical (form of) paradoxes with semantical (form) or ontological (forms of) paradoxes by merely paying attention to the formalistic structure of paradoxes. A paradox is of course of the form 'S iff~S'. It is also assumed that when we say that 'S iff S' represents the form of paradox, it can be proved that S iff~S, as in the case of Russell's paradox. But there is no reason why one cannot make a distinction between various specialized forms of paradoxes by considering various types of predicates occurring in S. It is well known that the semantic predicate 't' has its peculiarities which lead to semantical paradoxes which are not to be identified with set theoretical paradoxes thriving on the type-ambiguous notion of membership. By the same token, predicates such as 'meaningful', 'intelligible' etc. may lead to paradoxes which depend on a semantical interpretation of the truth conditions of predication of these predicates. Thus, the different presuppositions of the truth conditions of predications of these predicates make a difference to the semantical significances of these paradoxes. When I refer to alternative forms of paradox, I have in mind precisely these different semantical significances for different semantical predicates. But this is not incompatible with the fact that all paradoxes, whether set theoretical or semantical of whatever kind, whether veridical or falsidial, share the generalized structure 'S iff~S'. Thus there is no misconception of the generalized form of paradoxes and its different semantical tokens.

 

 

p.292

II

Now Levin denies that koans are paradoxical in the sense that they do not become simultaneously meaningless and meaningful in the process of getting enlightenment. He holds rather that by uttering a meaningless assertion in a koan the Zen master makes another assertion which is meaningful and relevant for revealing the ultimate truth. Levin even ventures to suggest that the first (namely the meaningless sentence) is used to assert the second (meaningful meaning). This explanation, I suspect, is as ingenuous as it is incorrect. I suspect that it has no merit other than trying to show that koans are really not paradoxical even though they appear to be paradoxical. What is clear from this explanation is that no paradoxes are really paradoxical even though they appear to be paradoxical. In a sense what Levine suggests has a particle of truth: From some theoretical point of view or within a constructed frame of reference a paradox such as Russell's paradox will be resolved and will not occur any more. Thus, just consider the Simple Type Theory or the Zermelo Set Theory: There simply could not be any mapping of Russell's paradox into such theories. But of course this is not tantamount to saying that originally there is no Russell's paradoxes which are veridical in Quine's sense. In fact, if we have a given context or a framework in which a paradox can be shown to be paradoxical, even though there is another framework in which no corresponding paradox arises, there cannot be any denial that paradoxes do exist in the first context or framework. Now consider the following Ch'an (Zen) paradox.

        Q:     When all things are reduced to the one, where does the one reduce to?

        A:     When I was in Ching, I bought a robe that weighed seven chin.

This koan (with the form QRA) is meaningless from the point of view that Q and A are unrelated, yet at the same time, it is meaningful from the point of view that the immediate succession of A to Q producing enlightenment It might appear that the antecedent does not stand in a relation R and a relation ~R to one and the same thing which a genuine paradox in relation presupposes. But if we understand the mechanism of koans correctly, the point of view that Q and A are unrelated is what generates the point of view that A and Q are related in giving or proposing

 

 

p.293

to give enlightenment. In other words, underlying any koan statement, there is an implicit principle that the unrelated has generated the relatedness (by virtue of the occurrence or the intended or expected occurrence of enlightenment) or that the meaninglessness has generated meaningfulness. Thus with regard to the intended enlightenment from a koan, the koan should be understood as asserting that QRAÉ ~QRA. In other words it is the enlightenment that transforms meaninglessness into meaningfulness and ~R into R. But from an unenlightened point of view, and yet equipped with the understanding that QRA is meaningful from the enlightenment's point of view, one may see that ~QRAÉ QRA.

    The real paradoxicality of such a paradox is to see its meaninglessness and meaningfulness of R at the same time and to see this is to face the power of the koan as a means for enlightenment. For what appears to be paradoxical will become indeed non-paradoxical when enlightenment is fully reached. I have distinguished between pre-enlightenment attitude (perception) and post-enlightenment attitude (perception). What I have said about the paradoxicality of 'QRA' is that before enlightenment A is not seen to be related to Q, and after enlightenment A is seen to be related to Q, and that to see that A is not related to Q is a way for seeing that A is after all related to Q in this koan. Thus this meta-linguistic principle of enlightenment verifies the implication from Q~RA to QRA and similarly verifies the implication from QRA to Q~RA. That the latter is the case is due to the fact that in genuine enlightenment no affirmation nor attachment of any form should be maintained so that the positive assertion of QRA must be negated into Q~RA in virtue of the enlightenment. Thus we have a perfect paradox as a result of the very defining nature of enlightenment. The very paradoxicality of the koan potentially captures the essence of enlightenment in the Ch'an (Zen) Buddhistic experiences.

    I did maintain that in a koan any sentence can be used to assert another sentence which can be substituted for it. This principle I call the principle of ontic substitution (which Levin calls ontic intersubstitution on the basis of my discussions in my original article.) Levin thinks that this cannot be proved (shown). In fact this is precisely what enlightenment requires. To be enlightened is to be able to assert 'QRA' by producing an A which does not relate (~R) to a given Q. This is to say, 'A~RQ' can be used to assert that ARQ and vice versa even if ARQ are related in na

 

 

p.294

ordinary usage. Levin misses the point of enlightenment by suggesting that one cannot use any sentence to assert any other sentence including the negation of the given sentence. His appeal to an ordinary example in ordinary language is irrelevant here. In ordinary language for an ordinary referential assertion we may use one sentence to assert another where the former has the predicate in common with the latter and we may not use any sentence to assert any other without changing the rules of language game, but to be able to change the rules of language game and to be allowed to do so is precisely what constitutes the language game of koans. Even in ordinary language, we may change the language game rules to generate paradoxes (still understood in a standard way), just like in other cases it might be that if we do not change the rules of game, we end up in paradoxes. Thus when I make a secret decision that when I say 'The man with the whiskey is not a wife-beater', and hence proceed to state the former sentence and start to treat the man with the whiskey with the attitude as if he is a wife-beater, then my behavior and my statement are still paradoxical and require explanation. Only an internal knowledge of my secret decision will enable one to see why the apparent paradox is not paradoxical even though it remains to appear to be paradoxical. The paradoxes of Ch'an (Zen) are just like this. They cease to appear paradoxical only when you come to become enlightened, but before you do so, and even after you do so, the surface form of language and its deep intent remain contradictory to each other and thus paradoxical. Levin simply fails to see how koans and its intent are to be conceived in a meta-linguistic framework.

 

III

While asserting that koans do not give us paradoxes. Levin acknowledges and is still puzzled about the 'strangeness' of the koans. For he recognizes that in koans one can use an absurd sentences to assert a non-absurd sentence S'. He seems to fail to see that S' can be the negation of S and that in the Buddhistic tradition beginning with Madhymika truth is revealed in negation, negation of negation, negation of negation of negation, etc. When I set out to explain how a koan works to reveal truth of enlightenment, I intend to provide a rational analysis in terms of which the logic of koans as instruments and expressions of enlighten-

 

 

p.295

ment can be analytically or rationally exhibited. I don't think that I misinterpret the consequence of using koans. My procedure is rather that keeping the very consequence of koan exercise in mind, namely the enlightenment, I wish to present an interpretation on a second order language level what enlightenment could mean in using a koan statement. The principle of ontological non-commitment (including the principle of ontological substitution) and the principle of contextual reconstitution are precisely proposed by me to show how the paradoxes are formed and how they can be resolved as well as how they are intended to be understood. What I have done is precisely to put the logic of koan use and logic of enlightenment in modern idioms of modern analytical philosophy. Furthermore, I see no reason for regarding my principle of ontological non-commitment as 'insufficiently general' to cover all koans. In so far koans are intended to bring about enlightenment in the Buddhistic sense, the principle of ontological non-commitment will remain the bone and blood of the koans. On the contrary, the principle may be too general and broad to explain the relevance of specific images and allusions in specific koans. Furthermore, Levin appears to miss the point of my construction and explanation. What I have done is to provide one model, namely the Fregean and Quinean model, for explaining the Ch'an (Zen) koans and their logic. There could be perhaps other consistent models of explanation in which the emphasis will not be placed on referring terms. Yet my stress on referring force in koans is not totally arbitrary.

    The Buddhists from the very beginning of their preaching attach great importance to problem of attachment to objects, things, principles, all referred to as dharmas in Sanskrit. They are making points about referring to nothing as a starting point for detachment. This is true of the Madhymika Theory of Nagarjuna as any student of Buddhism knows. Thus to apply the logical idioms and first order logic framework of Frege and Quine would provide the necessary contrast with Buddhist views on reference and referring terms. Our purpose of course is to use them in order to see how they are negated and denied in a genuine effort in Ch'an Buddhism to get enlightenment. Levin does recognize that my ontic interpretation of koans is one possible interpretation. It is sufficient for our purpose to recognize this for I nowhere claim uniqueness. But I do sincerely doubt that any other interpretation will bring out the logic of koans as an expression and as a means of achieving enlightenment as

 

 

p.296

elegantly and as vigorously as the ontic interpretation. Again if one remembers that ontic problem is indeed the main problem of Buddhistic thinking, in the history of Buddhism, and for that matter, in the history of Taoism which plays an important role in transforming early Mahaayaana Buddhism into Ch'an (Zen) Buddhism, there is no arbitrariness nor anomaly in my focussing on the ontic issue and the ontic interpretation of the koans. Levin has missed both the historical development of Ch'an (Zen) spirit and the purpose of my reconstruction and exposition.

 

IV

Now I come to examine how Levin proceeds to suggest another 'gloss' on koans which is alternative to mine. He suggests that our tendency to theorize produces an intermediary (presumably our ideas of things between ourselves and external things) and this is assumed to be an unhappy and undesirable state. Thus the question is how to reject these intermediary ideas and restore or establish our immediate consciousness of external objects. Levin regards the koans as a way to stop theorizing or verbalizing for the purpose of restoring or establishing the immediate consciousness of objects outside us. Now I do not wish to deny the plausibility of this interpretation of Ch'an (Zen) koans as an interpretation. But it cannot be the correct one nor the profound one. It has forgotten that the original question of Ch'an (Zen) Buddhism is not how not to think or theorize in order to experience external objects, but how not to think and theorize without reference to any objects at all! Levin's proposal has too much phenomenology and psychology and dichotomy to satisfy any Zen master's mentality. Without reducing the ontology of our discourse and statement to zero, no immediate experience of objects can be significant (necessary and sufficient) for the experience of enlightenment.

    When I introduce the principle of contextual reconstitution, I do have the immediate experience of reality in mind as an element of Ch'an (Zen) enlightenment. But I want to point out that it is only after the introduction of the principle of ontological non-commitment, including that of ontic reduction, that I speak of immediate experience of reality by relying upon
the principle of contextual demonstration. Furthermore, I intend the Ch'an (Zen) experience to be a free and creative immediate experience of reality. It is not an immediate experience of any fixed object in reality.

 

 

p.297

There simply cannot be any attachment and/or importance for attachment to external objects by immediate experience in genuine Ch'an (Zen) enlightenment. To understand this is to understand why one must use the principle of contextual reconstitution only after the principle of ontic reduction (or non-commitment) is applied in an interpretation of Ch'an (Zen) experience. When reality is voided in enlightenment, reality can be reinstituted by free experience, but without such voiding in the first place, no free experience will be possible. Levin fails to see this crucial point of enlightenment and felt puzzled by my presentation of the principle of contextual reconstitution as consequent upon ontic reduction, because he does not grasp the philosophical intent of Ch'an (Zen) Buddhism in both its historical context and in its practical uses in China and Japan. His conclusion that I have erred is unfortunately a non sequitus out of ignorance.

    I do think however that Levin has had some good point in suggesting that Ch'an (Zen) paradoxes serve the purpose of reminding the student to live his life. Perhaps in reminding the student to live his life in simplicity and without burden of mind and thought, one should learn to forget ontology and metaphysics. But then my whole central position in interpreting koans is to show that koans are means for shedding off the ontological and metaphysical language one carries in one's head. Then clearly for the metaphysically vary, one cannot succeed in doing this without facing the ontological issue and voiding it. The practical benefits follow from a rational self-revoking understanding.