Paradoxes and Parallelism in Ancient Chinese Philosophy

(Presented at the Oxford Conference: Topics in Comparative Ancient Philosophy: Greek and Chinese)

22-24 June 2006

Yiu-ming Fung Division of Humanities Hong Kong University of Science & Technology

CONTENT: (1) The Very Idea of Correlative Thinking (2) Quines Classification of Paradoxes (3) Paradoxes in Ancient Chinese Philosophy (3.1) Hui Shis Ten Theses (3.2) The Sophists Twenty-One Theses (3.3) Gongsun Longs Theses (3.4) The Theses in the Mohist Canons (3.5) The Function of the Sophists Paradoxical Expressions (4) Parallelism in the Later Mohism (5) Conclusion Appendix I: The Paradoxes of the Sophists Appendix II: The Different Cases of Parallelism (Mou Argumentation) in the Later Mohism Appendix III: The Logical Analysis of Parallelism (Mou Argumentation) in the Later Mohism
1

2 5 8 9 11 17 20 23 28 39 40

44

47

(1) The Very Idea of Correlative Thinking According to Joseph Needhams view in his Science and Civilisation in China, the thinking in ancient China, which he calls correlative thinking, is essentially different from the analytic thinking in the Western world. He describes the characteristic of the correlative thinking in the following way:1 A number of modern students -- H. Wilhelm, Eberhard, Jablonski, and above all Granet -- have named the kind of thinking with which we have here to do, coordinative thinking or associative thinking. This intuitive-associative system has its own causality and its own logic. It is not either superstition or primitive superstition, but a characteristic thought-form of its own. H. Wilhelm contrasts it with the subordinative thinking characteristic of European science, which laid such emphasis on external causation. In coordinative thinking, conceptions are not subsumed under one another, but place side by side in a pattern, and things influence one another not by acts of mechanical causation, but by a kind of inductance. In the Section on Taoism [Daoism] (pp. 55, 71, 84) I spoke of the desire of the Taoist thinkers to understand the causes in Nature, but this cannot be interpreted in quite the same sense as would suit the thought of the naturalists of ancient Greece. The key-word in Chinese thought is Order and above all Pattern (and, if I may whisper it for the first time, Organism). The symbolic correlations or correspondences all formed part of one colossal pattern. Things behaved in particular ways not necessarily because of prior actions or impulsions of other things, but because their position in the ever-moving cyclical universe was such that they were endowed with intrinsic natures which made that behaviour inevitable for them. If they did not behave in those particular ways they would lose their relational positions in the whole (which made them what they were), and turn into something other than themselves. They were thus parts in existential dependence upon the whole world-organism. And they reacted upon one another not so much by mechanical impulsion or causation as by a kind of mysterious resonance. Marcel Granet may be the first scholar who characterizes the Chinese Mind as the mind of correlative thinking. Needham follows this line to discuss correlative thinking in general and the specific symbolic correlation in Chinese cosmology in terms of a special kind of organism. In comparison to these two pioneers, A. C.
1

Joseph Needham, Science and Civilisation in China, volume 2: History of Scientific Thought (Cambridge University Press, 1956), pp. 280-81. 2

Graham addresses in more detail the nature of correlative thinking and lays much stress on its unique significance in Chinese culture. What Grahams viewpoint is divergent from Granets is that Graham does not accept Granets generalization; instead, Graham thinks that correlative thinking had not yet gained wide acceptance among social and philosophical thinkers in pre-imperial China though it was with full commitment by thinkers during the Han () dynasty. In comparison with Needhams viewpoint, Graham does not accept Needhams idea that correlative thinking is not alogical or pre-logical; instead, he thinks that correlative thinking is pre-logical and pre-linguistic in the sense that the correlative layer is submerged under the thinking which builds on it and, specifically speaking, the analytic upper layer which is thicker and denser in the West is grounded on the correlative stratum of thinking which is more fully exposed in China.2 Although David Hall and Roger Ames take side with Granet in identifying correlative thinking with a fundamental commitment of the Chinese sensibility, they think that the aesthetic ordering based on spontaneous correlation is beyond logical analysis and rational communication,3 which seems not far away from Grahams pre-logical thesis. As a matter of fact, there is no exactly the same view among these scholars when they use the term correlative thinking to describe the Chinese mode of thinking; but they all recognize the terms implication as non-logical or pre-logical, non-rational or irrational, intuitive-associative or beyond analytic thinking. Based on this presumption, some of them think that there is irreducibility from the root level of (correlative) thinking to the upper level of (analytic) thinking or that there is incommensurability between rational and correlative thinking. It seems to me that people who assert there is an essential difference between analytic and correlative thinking have never given any evidence to prove it is the case except to take it as a primary fact. I think, according to Davidsons Principle of Charity, if there is any difference between the so-called Western analytic thinking and the so-called Chinese correlative thinking, we cannot explain the difference without a common ground between them. I have argued for this point elsewhere and would not repeat here.4 I think that thinking in correlation or association is not thinking other than that of the analytic kind; it is still thinking in the rational space. Thinking analytically, discursively or rationally does not mean that someone always
A. C. Graham, Yin-Yang and the Nature of Correlative Thinking (Singapore: The Institute of East Asian Philosophies, 1986), p. 23. 3 David Hall and Roger Ames, Anticipating China: Thinking through the Narratives of Chinese and Western Culture (State University of New York Press, 1995), p. 257. 4 The detailed arguments for this point can be found in Yiu-ming Fung, Davidsons Charity in the Context of Chinese Philosophy, In Davidson Philosophy and Chinese Philosophy: Constructive Engagement, edited by Bo Mou (Leiden: Brill Academic Publishers, 2006), pp. 117-162. 3
2

thinks on basis of empirical evidence or with causal explanation; it means only that s/he thinks in the rational space and thus cannot think beyond logic. If someone thinks in the rational space, s/he cannot make an assertion by expressing a sentence of contradiction though, as mentioned by Davidson, s/he could simultaneously make two sentences mutually contradictory. If we think by making a particular sentence, say, a metaphor, which functions beyond its literal meaning, the sentence can be understood as a specific speech act other than that which functions directly on its literal meaning; but this does not mean that it goes outside the rational space. Similarly, even though thinking in correlation or association, like the Chinese thinking in the pattern with the pair of Yin-Yang () and the model of Five-Phases (wuxing), may not fully grounded on the empirical evidence of the physical world, it should not be understood as illogical (or irrational) or non-logical (or non-rational). In other words, thinking from a non-factual or non-physical perspective, such as a moral or aesthetic, metaphysical or religious perspective, or thinking not fully based on factual evidence, should not be understood as thinking other than analytic thinking or beyond the rational space. As a matter of fact, the thinkers or scientists in the ancient period of China had accumulated a huge of empirical observations from the physical and human world and they used the scheme of Yin-Yang and Five-Phases to characterize some kinds of feature of the events in the world, i.e., contrast in some aspect and complimentary in other aspect. It seems that they put too much emphasis on some kind of characteristic (and simultaneously ignored some other kind of characteristic) and thus over-generalized some kind of pattern to all the events in the world; so, sometimes their application of the patterning thinking is workable, but sometimes arbitrary. In general, thinking in correlation or association is not other thinking than the analytic kind; it is just the rational thinking in correlation or association. Graham thinks that the tendency to parallelism, including those appeared in the chapter of Minor Illustrations (Xiao-qu) of the Later Mohism (), is characteristic of correlative thinking.5 It seems to me that it is definitely a misunderstanding of the Mohist parallelism. On the contrary, I shall demonstrate the logical characteristic of parallelism in the fourth section of this paper and illustrate the semantic rather than syntactic sensibility performed in ancient Chinese philosophy. In the third section of this paper, I shall talk about some paradoxes or seeming paradoxes in the pre-Qin () texts and argue that, in comparison with the paradoxes mentioned by some Greek thinkers, the Chinese paradoxes were not mainly issued from theoretical concern but based on some kind of practical interest. This is
5

A. C. Graham, Yin-Yang and the Nature of Correlative Thinking, p. 25. 4

one of the indicators reflecting the fact that the Chinese thinkers are less syntactic sensibility and not interested too much in what Rudolf Carnap called the formal mode of speech. Furthermore, in the section on parallelism, I shall argue that the Chinese thinkers tend to think in the material mode of speech which may help them build up their semantic sensibility in rational thinking. (2) Quines Classification of Paradoxes As we know, W. V. Quine divides paradoxes into three kinds: (1) veridical or truth-telling paradoxes, (2) falsidical paradoxes, and (3) antinomies. Since to assume that there is a barber who shaves all and only those men in the village who do not shave themselves would deduce the absurdity that he shaves himself if and only if he does not, Bertrand Russells barber paradox can be understood as a reductio ad absurdum to disprove the barber.6 In this sense, according to Quine, we can identify this paradox as veridical, i.e., it indirectly tells the truth that there is no such a barber in the village. Quine thinks that the Frederic paradox and the barber paradox are alike. His reason is that: The Frederic paradox is a veridical one if we take its proposition not as something about Frederic but as the abstract truth that a man can be 4n years old on his nth birthday. Similarly, the barber paradox is a veridical one if we take its proposition as being that no village contains such a barber.7 But, I think to make a sentence with mere surprise or apparent puzzlement is not sufficient to construct a paradox, a tricky joke can produce the same effect. So, there should be a difference between the sentence about Frederics birthday and that about the barber paradox. The barber paradox as a veridical one shares with a falsidical paradox a characteristic that there is a real contradiction or seeming contradiction embedded in them which is used as a reductio ad absurdum to disprove something. In this sense, the Frederic sentence seems not qualified to be a veridical paradox because the mere surprise or apparent
The sentence There is someone who shaves all those and only those who dont shave themselves can be formulated into the following: (y)(x)[(yRx)~(xRx)] This sentence can lead to a contradiction if we assume it is true. Suppose it were true, then there would have to be someone b* who was the mysterious barber: (x)[(b*Rx)~(xRx)] But if the sentence holds for all things (x), it holds for b* himself/herself. That is: [(b*Rb*)~(b*Rb*)] Obviously this is logically impossible. However, this contradiction cannot be followed, if the sentence does not presuppose such a barber (b*). If we use Russells Theory of Definite Descriptions to deal with the sentence, no contradiction can be found. We can see that (y)(x){[(yRx)~(xRx)]&(y=x)} is (contingently) false but not logically false (i.e., self-contradictory). Hence, the barber paradox cannot be understood as an antinomy; it can only be used as an indirect proof for the non-existence of such a barber.
6

puzzlement given by the sentence does not have the same kind of function as those of veridical or falsidical paradox. It just tells a truth in a tricky way. Quine recognizes some of Zenos paradoxes as belonging under the head of falsidical paradox. For him and for most mathematicians, to assert that Achilles can never overtake the tortoise is to commit a fallacy that Zeno confuses a convergent series with a divergent series. In other words, an infinite number of segments in a convergent series can add up to a finite segment though those in a divergent series cannot. Or from a philosophical viewpoint, some may think that Zeno confuses the concept of physical divisibility with that of mathematical divisibility, so that he cannot construct a reductio ad absurdum with validity to support his teacher Parmenides monism. However, other philosophers think that Zeno may not commit such an error which is considered by modern theorists as a simple mistake. On the contrary, they believe that Zenos paradoxes provide sophisticated ideas which may challenge our pre-conception of the world. If we follow Quine to assert that Zeno commits a fallacy and to consider his paradoxes as some kind of invalid reductio ad absurdum, then these paradoxes can be identified as falsidical. If, on the other hand, we think that Zenos paradoxes are sophisticated and well-constructed as successful arguments of indirect proof, then they should be identified as belonging to the kind of veridical paradox. Nevertheless, either they are falsidical or veridical, their truth-value can be ascertained. In other words, they are either false or true. Although we do not know Zenos intention, it seems obvious that he would not like to construct a sentence the truth-value of which is not determined. These two kinds of paradox seem to expose absurdity or contradiction, but in reality it is not. Because the truth-value of these paradoxes, either veridical or falsidical, can be identified or ascertained. What cannot be identified or ascertained is the third kind that Quine calls antinomies. It is the paradoxes of this kind that bring on the crises in thought, because the truth-value of the paradoxes cannot be determined and the absurdity or contradiction arising from them cannot be eliminated. In this sense, the lair paradox (This sentence is false) and Russells set-theoretical paradox (The set of all sets which are not members of themselves is a member of itself) belong to this kind of troublemaker. Quine is quite optimistic that this kind of troublemaker can be explained away in the long run by refuting or revising the conceptual scheme in which the paradox arises. As indicated in the following paragraph, he seems to claim that an antinomy
7

W. V. Quine, The Ways of Paradox and Other Essays (New York: Random House, Inc., 1966), p. 5. 6

can be reduced to a veridical or falsidical paradox:8 A veridical paradox packs a surprise, but the surprise quickly dissipates itself as we ponder the proof. A falsidical paradox packs a surprise, but it is seen as a false alarm when we solve the underlying fallacy. An antinomy, however, packs a surprise that can be accommodated by nothing less than a repudiation of part of our conceptual heritage. I wonder whether or not we can solve the problem in the liar paradox or in the set-theoretical paradox in such a natural way as we solve the problem in some of Zenos paradoxes by exposing a fallacy in it. If it is the case that, sooner or later, we can explain away all the antinomies, it means that all the paradoxes of the third kind can be reduced to either those of the first or second kind and thus their truth-value can be identified or ascertained. I am not sure it could be done. On the contrary, if it is the case that not all antinomies are reducible, it may reflect that the limitation of our formal thinking is just right there. To recognize a veridical paradox as including a reductio ad absurdum is to identify it as a self-refuting argument to disprove something; it can be formalized into the following formula: (F1) [p(q&~q)] ~p Based on the absurdity that Achilles can never overtake the tortoise, Zeno seems to be able to prove that motion is impossible. If we accept Quines view on the paradox, the formulation of this kind of falsidical paradox could be like this: (F2) It seems that [p(contradiction or absurdity)] and thus ~p; but after examination, it is not and thus ~p cannot be proved. While a veridical paradox can be used to prove the truth of a relevant sentence, a falsidical paradox seems to do the same thing but actually it is proved to be a fallacy after examination. Although these two kinds of paradox are different in their truth-values, both paradoxes truth-value can be determined. On the other hand, an antinomys truth-value cannot be determined. For example, if we assume that the very sentence This sentence is false (L) of the liar paradox is true, This sentence is false would be true and thus L is false; on the other hand, if we assume that it is false,
8

W. V. Quine, The Ways of Paradox and Other Essays, p. 11. 7

This sentence is false would be false and thus L is true. In other words, the paradox can be formulated as: (F3) (L is true)(L is false) Since Ls truth-value cannot be obtained from (F3), the paradox cannot be explained away, i.e., it cannot be reduced to either a veridical or falsidical paradox. (3) Paradoxes in Ancient Chinese Philosophy In the pre-Qin texts, there are a few sentences identified by Sinologists and scholars in the field of Chinese philosophy as paradoxes. For example, Needham follows Fung Yu-lan () to regard Hui Shis () ten theses and the other sophists twenty-one theses reported in the chapter of Under the Heaven (Tian-xia ) in Zhuangzi () as paradoxes. Although Fung recognizes some of the twenty-one theses as paradoxes belonging to Gongsun Long (), he never considers any sentence in the main text of Gongsun Longzi as paradox. On the contrary, Needham, Graham and Chad Hansen all believe that there are paradoxes in the main text. Furthermore, Hansen and Christoph Harbsmeier identify some of the self-referring sentences in the Mohist Canons () as semantic paradoxes. Almost all these so-called paradoxes do look like ridiculous, absurd, puzzling, surprising or counter-intuitive expressions. But, in what sense they can be called paradox? Can they be put into some place of Quines classification? It seems to me that this question has not yet been addressed by the scholars mentioned above. In regard to this question, if we look carefully into the relevant sentences in the Appendix I below, it is obvious that we cannot identify any one to be in the third category of Quines classification. It means that the truth-value of all the so-called paradoxes is not logically indeterminate just like that of the liar paradox or the set-theoretical paradox. Since Hui Shi and the other sophists only make their puzzling sentences without any explanation, we cannot know the intention with which they make their sentences. In this sense, we may say that the truth-value of their sentences is not determinate. However, these sentences cannot be classified as paradoxes of the third kind (i.e., antinomy) though their truth-value has not yet been determined. Because, if we knew the sophists intentions and explanations, we could be sure whether they use the sentences to make arguments or not and we could judge whether their arguments are valid or not. When we make clear these points, I think their truth-value can be identified. In other words, the truth-value of these sentences is not
8

indeterminate in principle. Their truth-value is only factually indeterminate but can be determined if we can know more of the background knowledge in future. In this sense, we can say that these sentences can be recognized as either veridical or falsidical paradoxes if they are not just surprising or puzzling but embedded with a contradiction or seeming contradiction for making an indirect proof. (3.1) Hui Shis Ten Theses Among Hui Shis ten theses, (H1) and (H5) (see Appendix I) are his philosophical definitions of the Great Unit and the Little Unit, the Great Similarity-and-Difference and the Little Similarity-and-Difference. (H10) is about Hui Shis ideas of the Unity of the World and Universal Love, i.e., his view of the world and his attitude towards the world. All these three theses give us some kind of philosophical idea which is, of course, far away from our common sense of the world. But they are not paradoxes; just like many Western philosophical theses which are also far away from our common sense and cannot be regarded as paradoxes. The other seven theses look like paradoxes, because each of them is not only surprising but also seems to have contradiction. But, unfortunately, Hui Shi does not provide any reasons to explain why he expresses these sentences in one but not another way. Many scholars have tried to do Hui Shis own job; but none of them is satisfactory. However, if we can find out that some of Hui Shis contemporaries have some relevant ideas to his theses, it may be helpful for us to understand the meaning of his expressions. As we know, the Later Mohists also have a concept of no thickness which seems relevant to Hui Shis idea in (H2). The Later Mohists define their concept as follows:9 Canon A55: Thickness is having extension. [] Explanation A55: Thickness: Only something without thickness has no extension. [] Canon A61: [Starting-]Point is the [geometric or physical] unit [or part] without thickness and in the most front position [of a compound or whole]. [ ] Canon A69: Adjacent [or the next] is without interval but not coinciding. [ ] Explanation A69: Adjacent [or the next]: it is possible only when there is no thickness in between. []
The translation of these paragraphs is based on Graham with my modification. Please refer to A. C. Graham, Later Mohist Logic, Ethics and Science (Hong Kong: The Chinese University Press, 1978), pp. 305-306, 310, 315-316. 9
9

For Shen Youding (), the Later Mohists concept of no thickness is used to describe a point which can be considered as a spatial unit or material unit and which cannot be further divided, but Hui Shis concept of no thickness refers not only to a point but also to a line and a plane both of which are infinitely divisible.10 In some sense, it seems that the Later Mohists hold a position similar to the Greeks atomism while Hui Shi shares with Zeno the same idea of infinite divisibility. In this regard, Hui Shi can be understood as saying that a point, a line or a plane is without thickness and cannot be increased in thickness, yet in extent it may cover a thousand miles; but for the Later Mohists, no thickness means no dimension and only points are without thickness, so they would agree with Hui Shi on the first but not the second sentence of (H2). Based on this comparison, if we accept the above interpretation, there would be no real contradiction in (H2). However, If Hui Shis notion of no thickness refers only to a point but not to a line or plane, as claimed by Graham that there is no difference between the Later Mohists and Hui Shi in the concept of no thickness,11 it would be false to say that which has no thickness [i.e., a point] cannot be increased in thickness, yet in extent it may cover a thousand miles. Based on this interpretation, the apparent paradox might be regarded as a falsidical one which seems to include an unsuccessful reductio ad absurdum resulted from the confusion of the two concepts of no thickness. On the other hand, if we believe that Hui Shi does have his own idea of no thickness which allows the accumulation of length and breadth but excludes the accumulation in the third dimension, then (H2) should not be considered as having a real contradiction and there would be no paradox in it. In this regard, (H2) performs the same kind of behavior as the sentence about Frederics birthday, i.e., telling a truth in a tricky way. (H3), (H4), (H6), (H7), (H8) and (H9) appear to have contradiction, but none has been interpreted and elaborated by the Chinese and Western scholars as an antinomy. One of the reasons why they cannot be understood in this way may be that most of the antinomies are about self-referring sentence, but they all seem lacking in the characteristic of self-reference. Another reason may be that, if Hui Shi does share the view of relativity or perspectivism with his friend Zhuangzi, it seems not plausible for him to claim a thesis without truth-value. Although he believes that people can have different views of the world and cannot prove their theses absolutely true, he does not reject all the different views as nonsense and their theses as indeterminate in truth-value. He may consider these theses relatively true, his own metaphysical view more sophisticated in micro cognition and more comprehensive in macro
10

Shen Youding, Collected Essays of Shen Youding (Shen Youding Weiji) (Beijing: Peoples Publisher, 1992), p. 363. 11 A. C. Graham, Later Mohist Logic, Ethics and Science, pp. 305-306. 10

understanding, and his view higher than the view of common sense in terms of hierarchy of truth, but he does not give up truth seeking. I think he would agree that each thesis has its truth-value based on its respective perspective though the truth claim is not accepted as ultimate. If I am right at this point, the truth-value of all the six theses should be understood by Hui Shi as determinate. His intention for making the expressions may be that he wants to give some counter-intuitive examples to demonstrate that we can have other truth-claims which are different from those based on common sense and coming from physical perspective. To challenge the privileged status of the traditional viewpoint of the world may be one of the reasons why he makes his theses in such a counter-intuitive way. His metaphysical speculation gives us an impression that his theses are absurd and not in accordance with peoples ordinary views, but he does not choose a non-rational stand as Zhuangzi does. So, he does not intend to diminish or melt away all the distinctions made by human beings. He accepts that, in addition to his idea of the Great Similarity-and-Difference, there are distinctions based on the ordinary idea of the Little Similarity-and-Difference. In the chapter of Autumn Floods (Qiu-shui ), Hui Shi challenges Zhuangzis aesthetic cum mystical view by questioning how he knows the fish in the Hao () River are enjoying themselves. It is obvious that he does not give up the rational approach as Zhuangzi does though his theses also go beyond the physical perspective. (3.2) The Sophists Twenty-One Theses Fung Yu-lan divides the twenty-one theses into two groups one of which he attributes to Hui Shi and the other of which he attributes to Gongsun Long. I think his classification is arbitrary and the attribution is without evidence. Among them, I think the meaning of (S13) and (S19) (see Appendix I) is unclear and difficult to understand. So, it is no reason for us to identify them as any kind of paradox as defined by Quine. (S8) and (S18) expose contradiction literally; but we do not know what the sophists use these self-contradictory sentences for. Here I would not discuss these two groups of sentences because it seems not significant to do more work on them if we do not have any new input. The third group includes fourteen sentences, (S1)-(S7), (S9)-(S12), (S14)-(S15) and (S20), which are expressed with clear sense but look like counter-intuitive or contradictory to our common sense. The final group includes (S16), (S17) and (S21) which appear similar to some of Zenos paradoxes and theoretically interesting. I believe all the twenty-one theses cannot be interpreted as demonstrating antinomy. Because they are not self-referring sentences and there is no evidence to
11

prove that their truth-value is indeterminate. However, they could be interpreted as veridical paradoxes if we reconstruct them as successful indirect proofs, or they could be understood as falsidical ones if we demonstrate that they are unsuccessful in making self-refuting arguments, or they could be identified as tricky sentences which are used by the sophists to criticize our views from common sense or the conceptual scheme behind these views. These are the only options we can provide for understanding them. But we cannot be sure which is right without more information from the pre-Qin texts. Of the fourteen counter-intuitive theses, I think we have some source from the Later Mohism which may be helpful for interpreting (S10) and (S12). In the Mohist Canons, there are two paragraphs discussing the issue of perceptual knowledge:12 Canon B46: There is knowledge which does not come through the five roads [i.e., the five senses]. The reason for explanation is given under duration. [ ] Explanation B46: The knower sees by means of the eye, and the eye sees by means of fire [i.e., light], but fire does not see. If the only means for knowing duration were the five roads, it would not fit the fact. Seeing by means of the eye is like seeing by means of fire. [ ] Canon B47: Fire is hot. The reason for explanation is given under suddenness. [] Explanation B47: It is the fire one calls hot, one does not treat the heat of the fire as belonging to oneself. It is as when one looks at the sun. [ ] According to Shen Youdings interpretation,13 Canon and Explanation B46 appear to attack (S12) which claims that Eyes do not see and Canon and Explanation B47 seem to attack (S10) which claims that Fire is not hot. He thinks that the view of (S12) is similar to Gongsun Longs idea in his thesis of the separation of hard and white. Gongsun Long asserts that, White is seen by means of the eye which is by means of fire to see, but fire does not see. So, both fire and the eye do not see but the spirit [i.e., the mind] sees. If the sophist of (S12) takes the

12

The translation of these paragraphs is based on the following books with my modification. Please refer to Fung Yu-lan, A History of Chinese Philosophy, volume I, translated by Derk Bodde (Princeton University Press, 1952), pp. 252, 270, and A. C. Graham, Later Mohist Logic, Ethics and Science, pp. 415-417. 13 Shen Youding, Collected Essays of Shen Youding, pp. 305-306. 12

same position as Gongsun Long, it seems that they both think the eye and fire (or light) are external conditions for seeing; the real seeing ability is from within, i.e., the perceptual mind. On the other hand, in Canon and Explanation B46, the Later Mohists seem to reject this mental conception of seeing and to maintain that the seeing sense, one of the five senses, is the seeing subject which plays a role different from fire in the sense that fire only functions as an external condition for seeing. Although they agree that the eye like fire cannot be considered as a subject of knowing duration because duration is not a perceptual object, they do not accept the sophists idea that both the eye and fire are external conditions for seeing and play the same role in seeing a perceptual object. If this confrontation is a historical fact and the sophists and Gongsun Long do have the above mental conception of seeing, then (S12) can be understood as a metaphysical view which is expressed in a tricky way. It is not a paradox in terms of Quines definitions. Similarly, Shen thinks that Canon and Explanation B47 provide an objective conception of heat to challenge the sophists subjective one. The Later Mohists assert that as a perceptual subject we do not possess the heat of fire; instead, heat is an objective property of fire. Just like looking at the sun, the heat of the sun is from the sun, not in our perceptual sense. If it were in our perceptual subject, it would not give us a strong feeling of suddenness when we approach or touch fire. If it is right to say that the sophists are subjectivism in this regard, then, again, (S10) should not be recognized as a paradox. Besides (S10) and (S12), the similar contents of (S9) and (S20) are also discussed by the sophists contemporaries. (S9) mentions that, A chicken has three legs. It looks quite strange because it is not in accord with our common sense. But it is not strange an expression in the chapter of Understanding Change (Tongbian Lun ) in Gongsun Longzi. The reason is that, when Gongsun Long says that, To name chickens leg, there is one; to count chickens leg, there are two. Two plus one is thus three. To name ox or sheeps leg, there is one; to count ox or sheeps leg, there are four. Four plus one is thus five [ ], he presupposes a view of two levels of the world which is similar to Platos idea of noumena and phenomena. This provides an example for demonstrating the thesis of Two does not have one (left or right) and Two is not one (left or right) which precedes this paragraph. It is just like the case that a concrete white-horse, a particular object in the phenomenal world, is emerged from the combination of a universal white and a universal horse, but when the combination process is finished, the combined particular white-horse would not
13

contain its parent white or horse. Similarly, a chicken usually has two particular legs that we can count, but its universal leg cannot be counted perceptually but only be named intellectually. If the sophists follow Gongsun Longs two-world view, when they say that, A chicken has three legs, it can be understood as a rhetoric change of Gongsun Longs sentence Two plus one is thus three. If I am right at this point, I think (S9) is nothing but another sentence of metaphysical speculation. It is not a paradox of any kind mentioned above. (S20) says that, An orphan colt has never had a mother. It seems physically impossible that an animal has no mother. According to the Mohist Canon B61, It is possible for there not to be; but once there is, it is impossible for it to be dismissed. The reason for explanation is given under it has been the case, and Explanation B61, It is possible for there not to be. If something has happened, then it has been the case. It is impossible for there not to be, it is impossible to say that An orphan colt has never had a mother if all animals must come out of their mothers body at the beginning of their life. Nevertheless, in the chapter of Zhongni ( Confucius) in Liezi (), there is a conceptual clarification for (S20). It says that, One who has mother is not an orphan calf [(x)(Fx~Gx)]. In other words, although every calf or colt must be born with mother, when a calf or colt becomes an orphan it is not one who has mother [(x)(Gx~Fx)]. If we accept this stipulation, I think there would be no contradiction between the Later Mohists definition of It has been the case (chang-ran ) and (S20). So, (S20) does not really violate our common sense. I think the most interesting theses of the twenty-one are (S16), (S17) and (S21). (S16) and (S17) talk about the issue of motion while (S21) discusses the problem of infinite divisibility. It seems that these three can be compared with some of Zenos paradoxes. (S16) says that, The shadow of a flying bird never moves, which looks like a contradiction when we interpret the sentence as a moving object never moves. But it is not, because it talks about the shadow of a moving object, not the moving object itself. Some may think that the shadow does not move because it does not have any dynamic magnitude. In other words, the shadow does not really have the dynamics of motion. Unlike Zenos flying arrow, the so-called moving shadow can neither possess inertia nor have forces acting directly upon it. Dynamically the shadow cannot engage in motion. It derives its apparent motion from the flying bird.14 According to Canon and Explanation B17, A shadow does not shift. The reason for explanation is given under change of happening. [ ] and Where
Cheng-yih Chen, A Comparative Study of Early Chinese and Greek Work on the Concept of Limit, in Science and Technology in Chinese Civilization (Singapore: World Scientific Publishing Co. Pte. Ltd., 1987), edited by Cheng-yih Chen, p. 8. 14
14

the light arrives the shadow disappears. If it exists, it would be at rest forever. [ ], the Later Mohists provide a different explanation for the same phenomenon. The reason for the shadow not moving is that it is not an independent phenomenon in terms of its existence; it is a phenomenon caused by another independent existent (i.e., the flying bird). Since somethings moving presupposes the independent existence of itself, the Later Mohists seem to conclude that the shadows moving is nothing but an illusion. If this interpretation is acceptable, (S16) should not be regarded as a paradox but a plain truth. Nevertheless, (S17) is quite different from (S16) though they both appear to claim that motion is impossible. The sentence of (S17), The rapid motion of a flying arrow consists of moments at which it is neither in motion nor at rest, seems to mean that, There are temporal regions when a flying arrow is neither in motion nor not in motion. Are there time regions when an event happens with such a contradictory state? It seems that this is also the question of Zenos arrow paradox. The only difference between them, I think, is that (S17) talks about temporal regions while the Zenos arrow paradox refers to spatial regions as described by Aristotle. The former provides a dilemma that, if there is a temporal region occupied by the flying arrow, it must not be in motion at that temporal region; on the other hand, if the arrow is flying with rapid motion, it is impossible for the arrow at rest. (S17) may be used as a skeptical criticism that the idea of continuity of motion cannot be in accord with the idea of division of time. It seems to me that the author of (S17) uses this example to challenge the ordinary accepted conceptual scheme. If it is right to interpret (S17) as a variant form of Zenos arrow paradox in the sense that they provide the same kind of argument with the minor difference that the former talks about temporal region instead of the latters spatial region, (S17) may be the only one of the Chinese sophists theses with the similar theoretical significance to that of the Greek sophists paradoxes. (S21) says that, If a rod one foot in length is cut every day by one half of its length, it will still have something left even after ten thousand generations. In comparison with Zenos dichotomy paradox, it also talks about the divisibility of the length of something (a rod or a race course), but unlike Zenos paradox, it does not talk about the motion within a traveling distance. In regard to the similar problem, the problem of infinite divisibility, the sophists seem to provide a dilemma in (S21): If each part for cutting has a definite, non-zero size, then, since the rod consists of infinite number of parts of that size for cutting, the rod itself must be infinitely long. But then since all rods, of whatever size, are infinitely divisible, all rods must be of infinite length, which is plainly false. But then if each part has no size, then the rod itself can only be of zero length, since even an infinite number of parts of zero size
15

cannot add up to something of non-zero size.15 One solution provided by some modern mathematicians is to accuse the sophists of making a fallacy that the first horn of the argument is based on an incorrect intuition that the sum of an infinite number of positive quantities is infinitely large, even if each such quantity is extremely small. Some philosophers, of course, do not agree that the sophists or Zenos paradoxes are so easy to explain away. They think that it is not exactly the same problem about the mathematical concept of infinite series and infinite divisibility; it is about the nature of the physical world. Here I do not want to enter into details. But, in addition to the problem of infinite divisibility, there is a problem of physical possibility I want to discuss. This is the problem of supertask, i.e., how to finish the work of cutting a rod infinitely, or how to reach the end of an endless work of cutting a rod. It seems that it is not a real problem if we believe that to cut a physical object eventually has to stop at the level of the fundamental constituents of matters and thus it is physically impossible to cut a rod infinitely. However, if we talk about the region of space occupied by the rod instead of the rod itself, it would seem no problem of physical possibility for dividing the spatial region infinitely. Here I must confess that I am not sure whether this is a problem of physical possibility or not. If the idea of spatial region can be represented as a geometric unit and the idea of infinite divisibility of a spatial region can be interpreted as that of a mathematical series, it would not be a problem of physical possibility but a problem of logical possibility. If, on the other hand, we do not accept that the mathematical representation reflects the nature of space in the real world, we may think that the problem of physical possibility is a real problem and is still unsolved. So, to identify the status of this kind of paradox depends on which understanding is right. If we regard it as committing a fallacy of confusing the divergent with convergent series, it can be identified as a falsidical paradox; if, on the other hand, we think it provides a self-refuting argument to reject the idea of infinite divisibility through a dilemma, it can be identified as a veridical one. Whether it is a successful argument based on a veridical paradox or an unsuccessful argument based on a falsidical paradox, it may be used as a tool to challenge a popular conceptual scheme that the holder of the paradox does not accept. But I am not sure whether the sophists of (S21) have this agenda behind their expression or not. Just like some of the sophists other theses, (S21) is expressed in the context that they want to propose an alternative view that is seemingly contradictory and different from the common view. Although it looks like contradictory to the common view, they never claim that the common view is false. I think Hui Shi and other sophists have no intention to challenge the relative truth of the
15

Robin Le Poidevin, Travels in Four Dimensions: The Enigmas of Space and Time (Oxford University Press, 2003), p. 103. 16

common view, what they want to challenge is the ONE TRUTH THEORY behind the common view. Furthermore, when they criticize the ONE TRUTH THEORY indirectly through their surprising or puzzling sentences, they are not criticizing the theory for its own sake; they just want to change peoples attitude to the world and to guide people to accept a new view of the world which is better for living. In this regard, I think the Chinese sophists are not so much interested in theory itself as the Western sophists are, it may be one of the reasons why they express their propositions in the material mode of speech, lacking thinking in a formal way. If I am right at this point, I think some of the Chinese sophists expressions may be functioned as the koan (in Japanese or kung-an in Chinese) of Zen Buddhism, a special kind of speech act which is used to make a perlocutionary effect through expressing a surprising sentence literally. Although each of the Chinese sophists sentences has its literal meaning which is different from each other, they all share the characteristic of counter-intuitive, making surprise and puzzlement. This characteristic, I think, would have the function to evoke people to change their attitude towards the world from a common sense perspective to a metaphysical one or from a static view to a dynamic one. This function of evocation from these sentences can be regarded as a perlocutionary force.16 (3.3) Gongsun Longs Theses Although Fung Yu-lan had never used the term paradox to characterize Gongsun Longs thesis White-horse is not horse (G1) (see Appendix I) in the Discourse on White-Horse (Bai-ma Lun), the thesis has been widely recognized as a paradox in one way or another. For example, Needham believes that, No doubt Kungsun Lungs [Gongsun Longs] aim in stating an apparent absurdity, that a white horse was not a horse, was to attract the interest of prospective thinkers. The school of logicians [i.e., the school of names] had a particular interest in paradox, as we shall see.17 But he provides no reason why this sentence of apparent absurdity can be classified as a paradox. Similarly, Graham, Hansen and Harbsmeier all use the term paradox to describe the thesis without any definition for their use. Hansen interprets the thesis as claiming either that the combined mass whole white-horse is not identical with one of its part (i.e., horse) or that the whole has one of its part (i.e., non-horse). Although Graham does not accept Hansens mass-noun hypothesis, he follows his idea of part-whole relationship to interpret the thesis as a plain truth. Harbsmeier provides good philological evidence to reject this kind of interpretation;
16

The detailed argument for this point, see Yiu-ming Fung, How to do Zen (Chan) with Words? An Approach of Speech Act Theory, forthcoming. 17 Joseph Needham, Science and Civilisation in China, volume 2, p. 187. 17

but it is a surprise that he still uses the term paradox to label what he considers as a trivial truth. I think the white-horse thesis is obviously not a paradox; it is a paradigmatic example used by Gongsun Long to rectify names from a metaphysical perspective. As I have argued elsewhere,18 it is no reason to exclude some kind of realism as one of the possible options for understanding Gongsun Longs philosophy. Basically, Fung Yu-lan is right in using a realistic interpretation for the white-horse thesis. His mistake is only that he should not regard the compound term, such as white-horse, as having the same ontological status as the single term, such as white or horse. As indicated in Gongsun Longzi, the compound term white-horse designates something what the world has, meaning something exists in the phenomenal world; but the single term horse refers to something which is separated from any other entity, meaning something exists beyond the phenomenal world. Based on the premises that the mere horse is something without selecting any color and the white-horse is something having color selection, Gongsun Long concludes that the non-selection is not identical with the selection. In the chapter of Hard and White (Jian-bai Lun), he claims that the non-selection is something separated from the phenomenal world and is self-hided in a non-phenomenal world. One of the possible interpretations is that the white-horse as a particular object is different from the universal horse as such. In response to the question in (G2), [Supposing there is a hard and white stone], is it possible to say hard, white, and stone[-shape] [i.e., the properties of this particular thing] are three?, why Gongsun Long gives a negative answer? The reason provided by him is that, seeing does not give us what is hard but only what is white, and there is nothing hard in this; touching does not give us what is white but only what is hard, and there is nothing white in this. In other words, what cannot be seen is hided from the phenomenal world and what cannot be touched is also separated from the phenomenal world. What is hided or what is separated is a universal. It is the reason why he claims that there are not three but only two phenomenal properties to be seen or to be touched. To use the idea of universal-particular relationship to explain (G1) and (G2), we can make sense of both. It is obvious that they are not paradoxical expressions. In comparison with the specific statements in (G1) and (G2), (G4) is a general description of the same issue. For Gongsun Long, the specific statement White-horse is not horse can be generalized as [particular] thing [in the phenomenal world] is
18

Yiu-ming Fung, A Logical Perspective on Discourse on White-Horse, Journal of Chinese Philosophy, forthcoming. 18

not zhi (). I think Gongsun Longs idea of zhi is similar to but not exactly the same as Platonic Idea. Generally speaking, zhi (or one specific example, horse) for Gongsun Long is like Platos universal (or one specific example, horseness) in the sense that they both exist not in the phenomenal world and both do not have the sense of concreteness defined by physical characteristics of phenomena; but they are different from each other in the sense that Gongsun Longs zhi can emerge into phenomenal thing through its jian ( combining or joining) with other zhi or through its ting ( specifying or fixing) into a concrete thing while Platos universal can be exemplified into each of its instances without any meaning of emergence. In Platos words, his Idea is absolutely transcendent and unchanged as an ontological ground which can be exemplified but cannot participate into the phenomenal world in a cosmological sense; on the other hand, Gongsun Longs zhi can emerge into phenomena though before emergence it is transcendent and unchanged as a separate simple. Just like the metaphysical Dao in Daoism, Gongsun Longs zhi plays a double role of ontological ground and cosmological origin; it is quite different from Platos and other Western philosophers metaphysics in terms of the separate role assigned to their key terms of ontological and cosmological entities respectively. In this sense, we may say that Gongsun Long has a realistic commitment in his language which is not necessarily in the Platonic sense. What (G4) (There are no things [in the world] that are without zhi, but this zhi is not zhi.) says is that all things are emergent from the combination of one zhi with another zhi; but the combinational thing (wu ) and its components (wu-zhi ) is not identical with the separated zhi. Since Gongsun Long asserts in Discourse on Zhi and Wu (Zhi-wu Lun ) that, What I have mentioned above [=this zhi is not zhi] is not to say that zhi is not zhi [=a contradiction], but that the zhi entering into wu [=wu-zhi] is not the separated zhi[=wu-zhi is not zhi], according to this general statement, we can assert that, a particular white-horse is not a universal horse [an example of wu is not zhi] and the particular attribute horse in a white-horse is also not a universal horse [an example of wu-zhi is not zhi]. (G3) is nothing but a formal expression of (G1). The accurate interpretation of (G3) is that the term two refers to something combined from fixing two independent and separate zhi [one] into a compound wu [two]. According to this interpretation, the expressions two has no left [one] / right [one] and two is combined from left [one] and right [one] can be interpreted as that the compound wu has no simple and unchanged zhi in it, though it is emerged from two unchanged zhi in terms of their fixing into something combined which is changeable.

19

Based on the above interpretation and explanation, I conclude that Gongsun Long gives us a coherent picture of his metaphysical thinking which is grounded on some kind of two-world ontology. He can make good sense of his theses and there is no paradoxical thinking in them. (3.4) The Theses in the Mohist Canons Harbsmeier thinks that, One of the important things the system of definitions and sentences in the Dialectical Chapters [i.e., the Mohist Canons] was designed to achieve was the avoidance or resolution of paradox. The Later Mohists addressed a number of paradoxes. Some of these they showed to be theoretically sound and only apparently paradoxical, other they proved to be based on conceptual confusions.19 It seems to me that what Harbsmeier refers to the former can be classified under the group of Quines broad sense of veridical paradox (including the case of Frederics birthday), while the latter under the group of falsidical paradoxes. In other words, the Later Mohists regard the former as true propositions though the propositions look like self-contradictory. On the other hand, they reject the truth of the latter and want to explain away the seeming paradoxes. Most scholars believe that the so-called paradoxes of the latter group are held by the Daoists or by the sophists who have the relativist position. In this regard, Harbsmeier identifies some quoted or embedded sentences in Canon B71 and Canon B79 as self-reflexive paradoxes. The first one is the subordinate clause, all speech is perverse, in the first sentence of Canon B71, which appears to be a self-reflexive paradox: Canon B71: To hold that all speech is perverse, is perverseness. But, the above quoted or embedded sentence is actually both not a full-fledged self-reflexive sentence and not a paradox. Although what it talks about includes the sentence itself, it also refers to all the other sentences. If we paraphrase the sentence, it amounts to say that, All sentences are false, which is not equivalent to the liar paradox, This sentence is false. So, although to hold the truth of the sentence is self-refuting, it is not self-refuting for claiming the denial of the sentence (Some sentences are not false). In other words, unlike the liar paradox, the sentence is not indeterminate in truth-value, i.e., it is not an antinomy. Actually, the sentence can derive a contradiction if we assume it is true; but to assert the denial of the sentence is not self-refuting. So, it is not an antinomy. It is not a falsidical paradox either, because
Joseph Needham, Science and Civilisation in China, volume VII: 1, Part1: Language and Logic, by Christoph Harbsmeier (Cambridge University Press, 1998), p. 342. 20
19

there is no confusion in it. According to some scholars view, it could be understood as an example of the relativists theses which is used by the Later Mohists as the main body of their self-refuting argument. Through this argument, the Later Mohists can build up a truth that there is no such kind of speech as claimed by the relativists. So, I think the complex sentence of Canon B71 could be understood as a veridical paradox in a narrow sense at the most. Similarly, the second example mentioned by Harbsmeier is the implicit sentence, All arguments (or debates) can be rejected, in the first sentence of Canon B79, which also seems to be a self-reflexive paradox: Canon B79: To assert that all arguments (or debates) can be rejected is perverse. The reason of explanation is given under not to reject. But, if we look into the sentence carefully, it is not the same sentence as This argument can be rejected (If we use K for this sentence, it can be expressed as K can be rejected.). If the very sentence K is true, it makes an argument which cannot be rejected. Since the very sentence cannot be rejected, we can say that K can be rejected is false, i.e., ~K. On the other hand, if K is false, it means that K cannot be accepted. Since the very sentence cannot be accepted, we can say that K can be rejected, i.e., K. So, K is a real antinomy. However, the quoted or embedded sentence in Canon B79 does not express a full-fledged self-reflexive proposition. In addition to referring to the sentence itself, it also talks about all the other arguments. As demonstrated by Explanation B79, to assert that all arguments can be rejected is to include the rejection of ones own argument. If one does not reject it, s/he is unable to reject it. When one cannot reject it, this means not to reject all arguments. So, the sentence is self-refuting but its denial is not. The sentence used by the Later Mohists is not as a component of an antinomy, a falsidical paradox or a veridical paradox in the broad sense; it may be considered as a main body of a veridical paradox in a narrow sense. The above quoted or embedded sentences seem to be counter-intuitive and false and are rebutted by the Later Mohists as self-defeating. However, the Later Mohists own sentence Ox-horse is not ox / non-ox or Ox-horse is not horse / non-horse in Canon B66 has been criticized by Xunzi () as a confusion and fallacy: Canon B66: To say that, Ox-horse is not ox / non-ox, and to assert this statement are the same. The reason for explanation is given under collection
21

(jian). Xunzi provides no explanation for his accusation of the holder of the statement. Some may guess that Ox-horse is not ox is similar to White-horse is not horse. So, if the latter is absurd, the former is also ridiculous. But I think this is not a paradoxical or any other kind of ridiculous sentence; this is talking about the problem of how to use a collective term in a sentence with the part-whole relationship. Let us pay ample attention to the following explanation: Explanation B66: If from the fact that there is something which is not non-ox [and something which is non-ox] in ox-horse we infer it is permissible to say that Ox-horse is non-ox, then from the fact that there is also ox, in addition to non-ox, in ox-horse we can infer it is also permissible to say that Ox-horse is ox. So, some may think that the statement Ox-horse is non-ox is not permissible and the statement Ox-horse is ox is also not permissible. But if one of them is permissible, the other must be not permissible. It is also not permissible when you say, It is not permissible to say Ox-horse is ox. Besides, Ox does not include both two and horse does not include both two; only ox-horse includes both two. Therefore, Ox is not non-ox and Horse is not non-horse, only Ox-horse is both non-ox and non-horse. It is no difficult for understanding. Literally, it looks difficult for understanding. But if we consider the sentence as expression about the use of a collective term, it may not be difficult for understanding. I think Explanation B66 can be interpreted as follows: Interpretation B66: If from the premise that the collection ox-horse has both the part of non-ox and the part of ox (i.e., the part which is not non-ox) we infer that it is permissible to say Ox-horse is identical with non-ox (i.e., The collection of ox-horse is identical with its non-ox part, i.e., horse), then from the premise that ox-horse has ox, in addition to having non-ox we can infer that it is also permissible to say Ox-horse is identical with ox. So, some may think that both sentences are not permissible. However, these two must be contradictory with each other. It is also not permissible when you say that, It is not permissible to say Ox-horse is ox, because, in this sense, the statement Ox-horse is non-ox would be permissible if its negation Ox-horse is ox were not permissible. Besides, Ox does not include both two parts and horse does not include both two parts; only ox-horse includes both two (i.e., ox and non-ox). Therefore, Ox is
22

not non-ox and Horse is not non-horse, only Ox-horse is both non-ox (=horse) and non-horse (=ox). It is no difficult for understanding, because it amounts to Ox-horse is horse-ox. If we accept the above interpretation, there can be no paradox identified in Canon B66. (3.5) The Function of the Sophists Paradoxical Expressions After the examination of the so-called paradoxical expressions mentioned by the sophists, including Hui Shi and Gongsun Long, and the Later Mohists, there are two questions should be asked. The first question is about the identification of the sophists theses: To what kind of paradox do the theses belong? If some theses can be recognized as paradoxes, my second question is: In comparison with the paradoxes in ancient Greece, is there any similarity or difference between them? If they are different, what is the distinct function of the Chinese paradoxes? What are the paradoxes used for? In regard to the first question, I think it is obvious that there is definitely no antinomy, such as Russells set-theoretical paradox or the liar paradox, in the ancient Chinese philosophical texts. As I have discussed in section (2), if we choose a narrow sense of veridical paradox, i.e., only the barber paradox but not the Frederic one can be classified as veridical, then most of the Chinese sophists examples mentioned above are not veridical paradoxes (Maybe Canon B71 and B79 are the few exceptional cases). According to the above analysis, if (H2)-(H4) and (H6)-(H9) are true, they are nothing but telling a truth in a tricky way. (S9) and (S10) seem to provide some kind of metaphysical and subjectivist view respectively, while (S20) appears to express a plain truth which does not divert away from common sense. The most interesting examples comparable to Zenos paradoxes are (S16), (S17) and (S21). In comparison with Zenos paradoxes, there are both similarity and difference between both sides. (S21) could be interpreted as either a veridical or falsidical paradox, while (S16) looks like a plain truth. I think the most significant expression of theoretical interest is (S17) which seem to construct a dilemma to challenge peoples common sense view from a non-physical perspective. Just like Zenos arrow paradox which provides a philosophical view on the relation between the continuity of motion and the division of space, (S17) provides a philosophical view on the relation between the continuity of motion and the division of time. But unlike the Zenos paradox, the sophists in (S17), (S21) and other places never claim that
23

common sense and the views of the world and life other than their own are absolutely false. For Hui Shi and the other sophists who align with him and the Daoists, they do not think that the states of affairs in the world described by common sense are illusive, but consider that the empirical perspective is nothing but one of the view to see the world. To relativize all the viewpoints can be used by the sophists as a strategy to lead people to a metaphysical vision of the world which is considered as higher in terms of the hierarchy of truth. I think the strategy of the sophists using their queer theses is mainly not to refute but to free from the ordinary thinking. What they really want is to lead people to change their attitude towards the world and to promote their wisdom from a particular view to a universal vision as what may have been done by Hui Shi, or to transcend their views in the rational space to a Daoist () or an aesthetic cum mystical view of the world and life. (S21) and some of the sophists other theses seem to play a role as the Zens koan does. The former is just like the latter that the function of their expressions is basically performative; they aim to evoking people to change their attitude towards the world and life or to transcend their mentality from the rational space to an aesthetic cum mystical spiritual realm. In regard to the question of why the sophists want to have a metaphysical view or the Daoists want to have an aesthetic cum mystical stand, I think it needs an explanation from a perspective of intellectual history. A macro explanation may be that, I think, most of the intellectuals and thinkers in the period of Warring State () were bitterly disappointed with the political situation and social environment and wanted to change to an ideal level of living. They were not satisfied with the efforts made by the Confucianists and the Mohists who were the main representatives of revisionism in that period or before. The Daoists and most of the sophists, including Zhuangzis friend, Hui Shi, did not believe any revisionist programme; they wanted to liberate people from this very bad situation. As we know from human history, if people cannot change the physical and social reality, the most plausible way for them to escape from the cage is to fly to the air. This is the way that the Daoists and most of the sophists choose. They want to liberate to an ideal place of spiritual reality which they believe that is more real or much better than the physical and social reality. It may be one of the reasons why they intent to choose a non-physical stand together with a relativist strategy. In comparison with the Greek sophists paradoxes, I think the Chinese sophists theses are less theoretical and much practical in terms of knowledge seeking. Although some of the Chinese theses appear to be similar to the Greek ones, the former cannot play, or does not mainly play, a role of promoting theoretical thinking
24

in physics, mathematics or philosophy, as the latter makes contributions to these fields. Unlike some of the sophists who seem to be an alignment with the Daoists in terms of relativism or mysticism, it is obvious that Gongsun Long and the Later Mohists are much more interested in theoretical thinking. Gongsun Long constructs a realistic metaphysics which is counter to common sense; but there is no paradox in his theses (including (G1)(G4)). The Later Mohists do not like metaphysics. Although some of the relativists self-referring statements mentioned by them seem to be paradoxical (including Canon B71 and B79), the strategy for using these statements is to demonstrate their self-refuting characteristic and to prove indirectly that at least some speech is not perverse or some arguments cannot be rejected. They are at most belonging to the veridical paradox in a narrow sense. Canon B66 is also interesting in terms of theoretical thinking. However, it is not a paradox but a statement about the use of collective terms. Both Gongsun Long and the Later Mohists really provide thinking of theoretical significance; but their theses cannot be elaborated into a rigorous theory or systematic discourse. One of the main reasons is that these thinkers cannot fully formalize their theses. They cannot fully use symbols, such as those of variable, predicate and quantifier, to formulate the logical structure of their theses. They put too much emphasis on the surface structure of the language they use to construct their view and do not aware that the surface structure may be misleading in understanding an argument. Although sometimes they notice that the surface structure is misleading, what they provide is a practical solution to the problem in a particular case. They do not want to solve the problem on a more general and theoretical level. It may be the reason why they choose the material mode of speech, instead of the formal mode of speech to deal with their theoretical problem. In the following section, I shall discuss this practical approach in the paralleling thinking of the Later Mohism. In regard to the paradoxical expressions mentioned in Canon B71 and B79, I do not think there is any good evidence to identify that the holders of the expressions are Daoists or their ally sophists as claimed by many scholars. If the Later Mohists do have this idea, it may be due to their misidentification. As we know, Zhuangzi does not consider all speech as perverse and all argumentative assertions provided by human beings as false. As a relativist, he does not need to refute all the viewpoints provided by others, what he needs to assert is that all viewpoints are not absolutely true. It seems to me that Hui Shi accepts a cognitive relativism as a strategy to lead people to change their mind from the state of exclusiveness to that of inclusiveness
25

while Zhuangzi uses a relativist strategy to promote an aesthetic mysticism. Based on relativism, Hui Shi regards all viewpoints as coming from different particular perspectives and claims that the only way out of this particular mentality is seeking for universal love. In this regard, Hui Shi does not have any non-rational stand though he chooses a relativist approach. On the other hand, Zhuangzi thinks that all viewpoints are man-made, rational-oriented and subjective and thus each of them cannot be ascertained as absolute truth. For Zhuangzi, only when people give up this rational mentality and transform their mind into a state of indifferent concern (wu-dai ) or enter into a state of harmony (), they can transcend the relative predicament. The following propositions in the chapter On the Uniformity of All things (Qi-wu Lun ) indicate this aesthetic cum mystical turn: If a man sleeping in the damp is liable to get back-ache or half-paralysis, will the same thing happen to an eel? If a man sitting in tree is liable to tremble with fear, will the same thing happen to an ape? Of the man, the eel and the ape, who knows their proper place to live in? Men eat meat; deer feed on grass; centipedes are fond of snake; owls like rats. Of the man, the deer, the centipede and the owls, who knows the right taste? An ape mates with a gibbon; a buck seeks after s doe; an eel plays with a fish. At the sight of Maoqiang and Lady Li, beauties admired by men, the fish will dive into the deep water, the bird will soar high in the sky, and the deer will take to their heels. Of the man, the fish, the bird and the deer, who knows the real beauty? As I see it, the principles of righteousness and humaneness, the standards of right and wrong, all these things are so complicated that I can hardly tell which is which. [ ] The above paragraph informs us that there is no absolute standard for us to justify one viewpoint or the other. Hence, when people argue with each other, they cannot find any objective criterion to decide which side is absolutely right: Suppose that you and I argue over something. If you win and I lose, are you indeed right and am I indeed wrong? If I win and you lose, am I indeed right and are you indeed wrong? Is one of us right and the other wrong? Are both of us right or both of us wrong? If neither you nor I can know, other people will be even more in the dark. Whom shall I ask to decide for us? Shall I ask someone
26

who agrees with you to decide? If s/he already agrees with you, how can s/he decide it? Shall I ask someone who agrees with me to decide? If s/he agrees with me, how can s/he decide it? Shall I ask someone who agrees with both of us to decide? If s/he already agrees with both of us, how can s/he decide it? If neither you nor I nor others can know, who else shall we wait for? [ ] Why there is no absolute standard for justification and there are no absolute criteria to decide the right and the wrong? It is because all viewpoints are man-made. So he says in the chapter On the Uniformity of All things that, A path is formed when we walk on it; a thing is such when we say on it. [] In the chapter of The Signs of the Fullness of Power (De-chong-fu ), he says, When you look at the myriad things from their difference, then they are different as that between the liver and the gall or as that between Chu and Yue. When you look at them from the similarities, then they are all one. ( ) In other words, the world can be seen from this way or from that way and both ways could be extremely opposite. All viewpoints are based on one-sided, subjective and particular perspective, i.e., not following the natural course of things. Furthermore, all viewpoints are give rise to each other; they are not in correspondence to the so-called reality: Everything is nothing but that; everything is [also] nothing but this. From that people do not know things can also be considered as this; from this they know things are this. Therefore, it is said that that is derived from this and this is based on that which means that this and that perspective give rise to each other. Nevertheless, where there is birth there must be death; where there is death there must be birth; where there is approval there must be disapproval; where there is disapproval there must be approval. Where there is recognition of the right there must be recognition of the wrong; where there is recognition of the wrong there must be recognition of the right. Thus, instead of following this [perspective thinking], the sage seeks enlightenment from the heaven [i.e., away from the unnatural / artifact / perspective thinking], that is to say: s/he follows the natural course of things. [
27

 ] To follow the natural course of things under the heaven, it means that we have to give up the one-sided thinking from this or that perspective and to choose a non-physical overlook which is free from all distinctions and assertions made artificially. To follow the natural course of things is to do nothing with rational calculation, i.e., wu-wei (non-doing ) and thus go beyond the rational space and enjoy a spiritual harmony which is without any distinction and calculation. A mental state of distinction and calculation is a ji-xin (mechanical mind ) or cheng-xin (prejudice mind ). What Zhuangzi wants to do with non-doing is not to do anything with ji-xin, which simultaneously amounts to do something with Dao-xin (the Daoist mind ). All competitions and fighting are coming from humans making distinctions and calculation. If people give up this kind of mental construction, they would live by following the natural course of things and entertain a spiritual harmony. It is a spiritual realm which goes beyond all mental distinctions and linguistic descriptions: There is in the world nothing greater than the tips of the downs of a bird in autumn while Mount Tai is tiny. There is no one who lives longer than a dead baby while Peng Zu, who lived over 700 years, died young. The heaven and the earth and I came into existence at the same time; all things in the world and I are one in unity. Since all things are one in unity, what else is there to say? Since I have said that all things are one unity, how can there be nothing to say? The one and its word makes two; these two and another one [explanation] makes three. If we go on like this, even the cleverest mathematician cannot keep up, let alone the ordinary people. Therefore, if, by moving from non-existence to existence, we arrive at three, how much farther would we go if we move from existence to existence! Wed better not move any farther, but follow the natural course of things. [ ] Zhuangzi and the sophists, including Hui Shi, do have interaction with each other, but they are different. The target of the Later Mohists criticism of some self-refuting expressions seems to be the sophists or Daoists, but I do not think that they are really
28

the holders of these expressions. (4) Parallelism in the Later Mohism According to Marcel Granet, the Chinese language is inherently inappropriate for logical analysis and precise scientific discourse. He thinks that, A language made for poetry and composed of images rather than concepts is not only not an instrument of analysis. It also fails to constitute a rich heritage of the work of abstraction which each generation has been able to achieve. He observes that the Chinese language works through musical and picturesque symbolization, and concludes that, In this case, one must admit that when language cannot translate the operations of thought, these operations must proceed beyond language.20 Granets understanding of the Chinese language has a strong implication of language determinism and commits to a thesis that thinking can go beyond language. Harbsmeier provides solid evidence to demonstrate that the Chinese language is not lacking grammatical forms and has rich resources for logical thinking. I think Harbsmeier is right in making a turn from Granets perspective. But his understanding of the Chinese language is still unable to escape from Granet and Grahams dichotomy between logical and correlative thinking. In my article I quote in the footnote (4), I have provided detailed arguments to demonstrate their failure of making this dichotomy. Here I do not want to continue discussing this point. A point I want to explore is what Graham thinks that making correlation is the major mode of Chinese thinking and such correlation is naturally represented in sets of parallel expressions. Both Graham and Harbsmeier think that the tendency to parallelism is characteristic of correlative thinking and of philosophical criticism of correlations.21 They regard the parallelism mentioned in the chapter of Minor Illustrations (Xiao-qu) of the Later Mohism as philosophical criticism of correlations. It seems that the parallelism and its criticism belong to two levels of thinking: the former is correlative while the latter analytic. I think they provide a wrong picture for us to understand the mou () argumentation of the Later Mohism. The mou argumentation or what most scholars translate as parallelism or parallel thinking is not a sort of thinking which differs essentially and radically from rational or logical thinking. Instead, the mou is a pretty good example of logical thinking in terms of Western logic. It is not the case that based on some kind of rational thinking the Later Mohists criticize another kind of non-rational,
Marcel Granet, Quelques particularits de la langue et de la pense chinoise, in his Etudes sociologiques sur la Chine (PUF, Paris, 1920), pp. 150, 154. This quotation is translated by Christoph Harbsmeier in his volume written for Joseph Needham, Science and Civilisation in China, volume VII: 1, Part1: Language and Logic, p. 23. 21 A. C. Graham, Yin-Yang and the Nature of Correlative Thinking (Singapore: The Institute of East Asian Philosophies, 1986), p. 25; and Christoph Harbsmeier, op. cit., pp. 104, 262. 29
20

correlative thinking. As implicitly indicated in the chapter of Minor Illustrations, all the examples of the mou argumentation share the same grammatical form, which can be formulated as follows: A is B; CA is CB. It seems that the first part of the sentence (form) is a premise and the second part a conclusion. The Later Mohists are very cautious that not all the examples of the same form are permissible argumentations. So they classify various examples into different types: examples of some type are permissible while those of other type are not. They do not provide any concrete reason to explain why some but not the other are permissible. It seems that they appeal to peoples linguistic intuition or semantic sensibility for discerning the permissible and the impermissible. They do not have the consciousness that the demarcation can be disclosed by sentence-form. In other words, their approach to the problem of validity is not formal or syntactical. It may be one of the reasons why they do not express their problems in the formal mode of speech. Instead, their expressions rely heavily on the material mode of speech. Sometimes they seem to use the word mou() as a variable,22 but, as indicated by Harbsmeier, it is only a pseudo-variable in the sense that when more than one such-and-such are involved, we find that these are all referred to [different things] by the same term.23 Nevertheless, the terms ox, horse and ox-horse in the sentence Ox-horse is non-ox of Canon B66 probably function as individual symbols, just like the role played by the terms sheep, ox, horse and chicken in the sentence Sheep combined with ox is not horse () and in the sentence Ox combined with sheep is not chicken () in Gongsun Longzi, they are obviously used as symbols for individual entities. Other examples are the terms blue, white, yellow, and jade-blue in the sentence Blue combined with white is not yellow () and in the sentence White combined with blue is not jade-blue (). These terms are clearly used as individual symbols by Gongsun Long to demonstrate two types of combination: the first type of combination is impermissible while the second type of combination is permissible. Based on this distinction, Gongsun Long can assert that only the second type of combination can be
22

Joseph Needham, Science and Civilisation in China, volume 2, and volume VII: 1, Part1: Language and Logic, by Christoph Harbsmeier (Cambridge University Press, 1998), p. 333. The example can be found in the chapter of Major Illustrations (Da-qu). The other similar symbols in Canon B68 and Explanation B1 are bi () and ci (). 23 Christoph Harbsmeier, op. cit., p. 287. 30

used to reject the general statement that Two does not have one () or Two is not one (). Since these terms are most likely used as individual constants rather than individual variables, they still cannot help in the full formalization of the relevant theses. I think it is one of the main difficulties for the ancient Chinese thinkers to adopt a formal approach to express their thought. Although the Later Mohists cannot justify which mou argumentation is valid and which one is invalid through a formal proof, they can distinguish the logical status of various argumentations by some concrete examples which are very easy for people to confirm their permissibility by peoples linguistic intuition or semantic sensibility. This kind of confirmation by examples seems to be a sort of model theory of proof, i.e., a procedure of proof based on some schematic examples. However, to find appropriate examples is not an easy job. Because linguistic intuition does not always work; sometimes our intuition may be misled by prima facie evidence of a sentence. People should appeal to a second level examination on their first level intuition if they eventually find out a counter-example to their intuition. This may be one of the reasons why the Later Mohists often stress the crisis of wild demonstration (kuang-ju ) as follows: Canon B66: To use a wild demonstration cannot help people to know difference between things. The reason for explanation is given under not permissible. [ ] Explanation B66: Although ox and horse are different, if someone says that they are different because ox has teeth and horse has a tail, it is not permissible. In fact, both have [teeth and tails], these not being [attributes] belonging to one and not the other. One has to say that horse is not the same as ox because ox has horns and horse does not; it is in this they are different in class. If a demonstration is not based on the fact that ox has horns and horse does not, it is a wild demonstration, just like the case based on the fact that ox has teeth and horse has a tail. [ ] For the Later Mohists, to know lei (class ) and to understand gu (reason ) are necessary for reasoning. If someone uses an example which is based on a wild demonstration, s/he cannot provide a permissible parallel reasoning. Because, if someone misidentifies different classes as the same class, what s/he can provide as reason in regard to one class is not a relevant reason in regard to another class in a
31

parallel reasoning. So the Later Mohists stress in Canon B2 the difficulty of the generalization of class (tui-lei zhi nan ). Although they give a very sophisticated analysis of the different concepts of similarity and difference for the classification of classes, in Canon B6 they still warn us with the principle of Different classes are not comparable (yi-lei bu bi []). They think that people should be strictly in accordance with this principle and be very cautious in making any kind of parallel reasoning. For example, in the chapter of Against Attack (Fei-gong ), Mozi () distinguishes the concept of attack (gong ) with that of execute (zhu ). He thinks that the former as an aggression with military force is unjust while the latter as a punishment with military force is just. So, if we make a permissible statement with the latter concept, it does not infer that we can make a similar permissible statement with the former concept. Based on this reason, Mozi criticizes those who do not know the class that my words refer to and do not understand the reason my words concern (). In other words, the argumentation criticized by Mozi is based on a wild illustration. What is meant by parallel reasoning in the chapter of Minor Illustrations is a kind of reasoning based on comparing two propositions both of which can be developed (). The following paragraph in the beginning of the chapter of Minor Illustrations gives serious warning to people who want to use this or the similar kind of reasoning: Of things in general, if there are respects in which they are the same, it does not follow that they are altogether the same. The parallelism of propositions is valid only as it reaches. If something is so of them there are reasons why it is so; but though its being so of them is the same, the reasons why it is so are not necessarily the same. If we accept a claim we have reasons for accepting it; but though we are the same in accepting it, the reasons why we accept it are not necessarily the same. Therefore propositions which illustrate, parallelise, adduce and infer become different as they proceed, become dangerous when they change direction, fail when carried too far, become detached from their base when we let them drift, so that we must on no account be careless with them, and must not use them too rigidly. Hence saying has many methods, separate kinds, different reasons, which must not be looked at only from one side. [ () () ]

32

Based on peoples linguistic intuition with the Mohists careful examination, the Later Mohists provide some schematic examples for the justification of the validity of some mou argumentations. These can be classified into five cases: Of cases in general, there is a case that this [proposition] is right if that [proposition] is affirmed, or that this is not right though that is affirmed, or that this is right though that is disaffirmed, or there are cases that one is distributed but another is not distributed, or that one is affirmed but another is disaffirmed. [ ] The first case that this [proposition CA is CB] is right if that [proposition A is B] is affirmed is the basic schema of the valid examples of the mou argumentation. The first example mentioned in the chapter of Minor Illustrations seems to attack Gongsun Longs thesis of White-horse is not horse, in addition to making parallelism. The grammatical surface structure of the argumentation example (1.1) [A white horse is a horse; to ride a white horse is to ride a horse.] (See Appendix II) is A is B; CA is CB. Some Chinese scholars in the field of the traditional Western logic regard this form as showing the Method of Adding Concept to the Judgment or Inference by Added Determinants. However, its logical deep structure is more complicate. Let us use the one-place predicates S, P and M stand for the concepts white horse, horse and man, respectively; the two-place predicate R for the concept of riding, the argumentation example can be formulated into the following form: Form I: (x)(SxPx) (x){[Mx(y)(Sy&Rxy)][Mx(y)(Py&Rxy)]}

As demonstrated in the example (1) of the Appendix III, the parallel reasoning of (1.1) is valid. However, the Later Mohists do not and cannot use this kind of formal approach; they make their right judgment by their linguistic intuition, on the one hand; and by comparing the normal examples with some anomalies, on the other. The anomalies can be found in the second and third cases, i.e., the case that this [proposition] is not right though that [proposition] is affirmed; and the case that this [proposition] is right though that [proposition] is disaffirmed. The first case shows that we can infer from an affirmed premise to a right conclusion. When comparing the first case with the second and third cases, we can demonstrate that it is not permissible to infer from an affirmed premise to a right conclusion for the second case and to infer
33

from a disaffirmed premise to a disaffirmed conclusion for the third case. Why the examples of these two cases can be considered as anomalies of the first case? The Later Mohists never give a concrete answer in the chapter of Minor Illustrations, except providing some general comments which implicitly appeal to peoples linguistic intuition. For example, the literal meaning of the key words in (2.3) [A carriage is wood; but riding a carriage is not riding wood.] and in (2.4) [A boat is wood; but entering a boat is not entering wood.] is clearly ambiguous. Based on our linguistic intuition, everyone can know without further reflection that the deadly metaphorical meaning in the first occurrence of riding and entering is obviously different from the literal meaning in the second occurrence of riding and entering respectively. Therefore, we can judge that the relevant parallelism cannot be obtained and an affirmative conclusion in (2.3) and (2.4) cannot be maintained. However, some examples are not so easy to judge merely based on linguistic intuition. (1.3) [Huo is a person; to love Huo is to love a person.] looks similar to (2.7*) [Robbers are people; loving robbers is loving people.] in grammatical structure. But, why the Later Mohists reject (2.7*) by mentioning an example of case two in the beginning of (2.7) [Robbers are people; loving robbers is not loving people.]? I think the main reason is that the concept of love mentioned by the Mohists is conditioned by their philosophical idea. According to Mozis idea of jian-xiang-ai (universal mutual love) or zhou-ai (distributive love), the consequence of love is having mutual benefit ( jiao-xiang-li). Since robbers cannot contribute to mutual benefit, they are not qualified to love and to be loved by others. If we still use the term love in loving robbers or to love robbers, it does not mean jian-xiang-ai, but bie-ai (discriminative love) or ti-ai ( partial love) which is seriously criticized by Mozi himself. So, the reason behind (2.7) is that the first token of love means bie-ai while the second token of love means jian-xiang-ai. In other words, the logical symbol for the first token is R1 while that for the second token is R2; they are not the same R. For the similar reason, in the end of (2.7), when the Later Mohists claim that, Killing robbers is not killing people, they are conscious that the two tokens of killing have different meanings. As indirectly reflected in the words of Mozis teaching, it is obvious that killing people is morally impermissible; it amounts to murdering. Nevertheless, Killing robbers means to punish bad guys by taking off their life, which may not be morally prohibited. Sometimes the Later Mohists explicitly provide explanation for the anomalies in places other than the chapter of Minor Illustrations. One of the most significant examples is in Canon B40 and Canon B54:

34

Canon B54: A dog is a puppy; it is impermissible to say that to kill a dog is not to kill a puppy. The reason for explanation is given under identity. [ ] Explanation B54: Dog: A dog is a puppy; it is permissible to say that [to kill a dog is] to kill a puppy. Just like [the type identity of] twin thigh. [ ] Since this example shares a logical form with (1.1), the parallel reasoning is also valid. However, the following example indicates a different reasoning: Canon B40: If you know this is a dog, to say of yourself that you do not know this is a puppy is a fallacy. The reason for explanation is given under identity. [] Explanation B40: If knowing a dog is identical with knowing a puppy, it is a fallacy. If they are not identical with each other, there is no fallacy. [ ] As we know that, in an intensional context, although a dog is a puppy, we cannot infer from the premise that knowing this is a dog to the conclusion that knowing this is a puppy. The author of Canon B40 seems not aware that it is an intensional context and thus treats the example as a case in an extensional context as that in Canon B54, so s/he thinks that to disaffirm the conclusion is a fallacy. It seems that the author of Explanation B40 notices that, in an intensional context, it is fallacious to draw the conclusion that knowing this is a puppy from the premise that knowing this is a dog and the premise that two types of knowing are identical. If two types of knowing are not identical, even though a dog is a puppy, it is not fallacious to say that knowing that this is a dog is not to know that this is a puppy. I think most examples in the second case are related to the problem of ambiguity, except the examples about many robbers and no robbers the problem of which can be solved in terms of their syntactical structure. On the other hand, the problem of most examples in the third case is pragmatic rather than semantic. For example, in (3.5) [Being about to die prematurely is not dying prematurely; but to stop someone being about to die prematurely is to stop her/him dying prematurely. To claim that there is fate is not a fate; but to reject to claim that there is fate is to reject a fate.], the former example is about the usage of the phrase be about to (qie), a state of action which is possible but has not yet actually happened. The premise presents a common sense view that a possible action is not an actual action. Based on this
35

premise, however, we cannot draw the conclusion that to stop a possible action is not to stop an actual action. Because, being an action in an actual state presupposes a prior state which is possible but has not yet happened. So, when we stop an action in its prior state, we also stop the action in its subsequent state. The latter example is about the distinction and relation of a state of affair and a propositional attitude of it. It is clear that the state of affair that there is a fate is not the same as a claim of the state of affair. However, to reject the claim is simultaneously to reject the state of affair. Because, to reject the claim is not only to reject the claim as a claim, but also to reject what the claim claims, i.e., the state of affair. In general, before constructing a parallelism, we have to know the pragmatic use and logical function of these key words; otherwise, we may commit some kind of fallacy. A term in a sentence is distributed if the sentence refers to every member of the extension designed by the term. If a term in a sentence does not refer to every member, the term in the sentence is not distributed. In the fourth cases, the example of distribution is counter to the example of non-distribution. The non-distributive example in (4.2) [(4.2A):S/he rides horse[s] does not require her/him to ride all horses without exception before being deemed to ride horse[s]; s/he rides some horse[s], and by this criterion is deemed to ride horse[s].] can be elaborated as This is a horse; to ride this horse implies to ride some horse[s]. Its grammatical form is the same as that in the first case (A is B; CA is CB.); its logical form can be understood as follows: Form II: Pa (x){[MxRxa][Mx(y)(Py&Rxy)]} If we talk about riding white horses [in (1.1)] instead of riding this particular horse, it can also be recognized as an example of non-distributed case. The logical form of the example is the above Form I. Both Form I and Form II show that riding this particular horse and riding white horses do not require riding all horses, riding some horse[s] is qualified to call riding horse[s]. The grammatical form of the distributive example in (4.2) [(4.2B): S/he does not ride horses does require that s/he rides no horses at all; only then is s/he deemed not to ride horses.] is A is B; ~CB is ~CA. Its logical form can be understood as follows: Form III: Pa (x){[Mx(y)(Py~Rxy)][Mx~Rxa]}
36

A similar argumentation like, A white horse is a horse; not to ride horses includes not to ride white horses.[(1.1R): a reverse expression of (1.1)], also belongs to this case. The logical form of this example can be reconstructed as follows: Form IV: (x)(SxPx) (x){[Mx(y)(Py~Rxy)][Mx(y)(Sy~Rxy)]}

Both Form III and Form IV indicate that not to ride horses is distributed; it includes not to ride this particular horse and not to ride white horses, etc. Although (1.3) [Huo is a person; to love Huo is to love a person.] has a similar logical form as that of the non-distributed example of (4.2) mentioned above (i.e., Form II), some kind of reverse expression of (1.3) [(1.3R1): Huo is a person; loving a person is (=includes) loving Huo.] is permissible while that of (4.2) [This is a horse; riding a horse is (=includes) riding this horse.] is not. This is because the logical role played by the term love is different from that played by the term ride. This special use of the term love can be reflected in the distributed example in (4.1) [S/he loves people requires her/him to love all people without exception, only then is s/he deemed to love people.]. The distributed character of the reverse expression of (1.3) can be expressed in the following form (1.3R1): Form V: Ma (x){[Mx(y)(My&Rxy)][Mx Rxa]} The non-distributed example of (4.1) [S/he does not love people does not require that s/he loves no people at all; s/he does not love all without exception, and by this criterion is deemed not to love people.] can have the following form for not loving Huo (another reverse expression of (1.3): (1.3R2)): Form VI: Ma (x){[Mx~Rxa][Mx(y)(My&~Rxy)]} There are two kinds of parallelism which look like similar in surface structure but are actually different in the sense that one kind of parallelism is valid only in an affirmative form while the other kind of parallelism is valid only in a negative form. The fifth cases of parallelism are about these two kinds of parallelism. The first part of (5.4) [The ghost-soul of a man is not a man; sacrificing to a mans ghost-soul is not sacrificing to a man.] can be formulated into the following valid argument:
37

Form VII:

(x)(Sx~Px) (x){[Px(y)(Sy&Rxy)][Px(y)(Py~Rxy)]

The second part of (5.4) [The ghost-soul of your elder brother is your elder brother; sacrificing to your elder brothers ghost-soul is sacrificing to your elder brother.] can be formulated into a valid argument of the similar form as Form I. I think the reason why the Later Mohists treat the premise of the first parallelism as positive while that of the second one as negative may be that the permissibility of a sentence is based on the normal usage of the relevant terms in a sentence. In accordance to this pragmatic criterion, however, we cannot determine which argumentation is valid. The validity of an argumentation still depends on the permissible models recognized by peoples linguistic intuition and the comparison between the normal cases and their anomalies. The Later Mohists provide permissible models of parallelism through the examples of the first case and demonstrate the validity of the examples through their comparison with the examples of the second and the third case. The permissible models are found from linguistic intuition while their validity is justified by reflective thinking through their comparison with the anomalies in the second and third case. Furthermore, they indicate the distribution and non-distribution of terms through the two kinds of examples in the fourth cases and they also disclose the affirmative and negative version of parallelism in the fifth cases. These schematic contrasts further help people know how the semantic or pragmatic role of various usages of terms is played in different kinds of parallelism. Here, the main approach to understanding or identifying the permissible or valid argumentations of parallelism is pragmatic in the sense that the criteria are based on peoples semantic sensibility instead of formal thinking. The merit of this approach is that people can identify the permissible argumentations of parallelism in the normal case easily and quickly. But the disadvantage is that people would neglect some sophisticated and profound difference and similarity between examples of the same case and different cases. For example, the examples 1-4 mentioned in the Appendix III are all valid argumentations of the first case; but they do not have the same logical form. The Later Mohists cannot discern the different structures in these examples. Example 1 [(1.1): A white horse is a horse; to ride a white horse is to ride a horse.] and Example 2 [(1.3): Huo is a person; to love Huo is to love a person.] or Example 4 [(Major Illustration): A gem is a jade; this being a gem is this being a jade.] seem to share the same form. However, the subject of (1.1) is a concept which can be formulated into a symbol of logical predicate; the subject of (1.3) is a proper name and the subject of the example in the chapter of Major Illustrations is a pronoun, both of which can be formulated into a
38

symbol of individual constant. Why the Later Mohists cannot make this distinction between predicate and individual? As I have discussed in the beginning of this section, the word mou() used by the ancient Chinese philosophers is merely a pseudo-variable and the terms sheep, ox, horse and chicken and the terms blue, white, yellow, and jade-blue used by Gongsun Long are only functioned as individual constants. There is no real individual variable used by these thinkers. Since they cannot create independent symbols for logical predicates and for quantifiers, only individual constants and pseudo-variables are not sufficient to suggest a formal approach. Example 3 [(Canon B54): A dog is not a puppy; to kill a puppy is to kill a dog.] also seems to have the same form as Example 1; but they are different. Because the logical formulation of the linking verbs of Example 1 is a conditional connective (implication) while that of Example 3 is a bi-conditional connective (equivalence). To appeal to semantic sensibility is very difficult to make this distinction. I think this is one of the disadvantages of the Later Mohists pragmatic approach. (5) Conclusion In this paper, I have one general question and two specific questions to answer. The general question is whether the thinking in the ancient China is mainly not analytic but correlative, or whether it is incommensurable in conceptual scheme between ancient China and the Western world. I think the answer is negative. Based on Davidsons Principle of Charity, I have generally given an answer to this general question elsewhere. In the previous sections, I have given a specific answer which is based on my answers to two specific questions concerning paradoxes and parallelism in ancient Chinese philosophy. The first question is about the characteristics and functions of the so-called paradoxes in ancient Chinese philosophy. I have argued that they are not antinomies and most of them do not have the same kind of theoretical ground as those in the Western philosophy. In comparison with the Greek sophists paradoxes, I think the Chinese sophists theses are less theoretical and much practical in terms of knowledge seeking. Although some of the Chinese theses appear to be similar to the Greek ones, the former cannot play, or does not mainly play, a role to promote theoretical thinking in physics, mathematics or philosophy, as the latter makes contributions to these fields. The main function of their theses is to change peoples attitude towards the world and life from a physical perspective to a non-physical one (Hui Shis wing) or to promote peoples vision of the world and life from a rational space to an aesthetic cum
39

mystical realm (Zhuangzis wing). Almost all the sophists theses look counter-intuitive and some may have a special linguistic function, i.e., performative force. Although most of the theses are not theory-oriented, they are still thinking in the rational space. The second question is about the characteristics and functions of the Later Mohists parallelism. The expressions and discussions of the validity of the parallel argumentations in the chapter of Minor Illustrations are mainly in the material mode of speech; and their analytic power is usually reflected in their semantic sensibility instead of syntactic construction. In this regard, their argumentations are delivered in a different way from the Western ones, but this does not mean that these discussions are related to the so-called correlative thinking or that they are characteristic of correlative thinking. According to my analysis of the mou argumentation, it demonstrates that the Chinese thinking in the Later Mohism is analytic and discursive, that the difference between Chinese and the Western thinking is not essential in the sense that the former is analytic while the latter is correlative. They both have analytic thinking though the former expresses its analytic thinking in a more or less formal way while the latter in a more or less pragmatic way. The conclusion of this paper is: the paradoxes and parallelism in ancient Chinese philosophy indicate that the Chinese sophists and thinkers adopt a pragmatic approach to deliver their conceptual thinking of paradoxes and choose the material mode of speech to express their logical thinking of parallelism. These all are not the so-called correlative thinking; they belong to the same kind of analytic and discursive thinking as those addressed by the Western thinkers. To make a distinction of correlative and analytic thinking is inevitably to commit a self-refuting argument. Appendix I: The Paradoxes of the Sophists24 Hui Shi is one of the famous sophists and a pioneer of the School of Names ( ) in the pre-Qin period. Some of his main theses are reported in the chapter of Under the Heaven in Zhuangzi. His ten theses are expressed in the sense quite counter-intuitive and look like the Western paradoxes. They are as follows:25

24

The translation of the paragraphs here quoted from the ancient texts is based on the following books with my modification. Please refer to Fung Yu-lan, A History of Chinese Philosophy, volume I, A. C. Graham, Later Mohist Logic, Ethics and Science, Joseph Needham, Science and Civilisation in China, volume 2, and volume VII: 1, Part1: Language and Logic, by Christoph Harbsmeier (Cambridge University Press, 1998). 25 Fung Yu-lan, A History of Chinese Philosophy, volume I, pp. 197-200. 40

(H1) The greatest has nothing beyond itself, and is called the Great Unit (da-yi ); the smallest has nothing within itself and is called the Little Unit (xiao-yi ). [] (H2) That which has no thickness cannot be increased in thickness, yet in extent it may cover a thousand miles. [] (H3) The heavens are as low as the earth; mountains are on the same level as marshes. [] (H4) The sun at noon is the sun declining; the creature born is the creature dying. [] (H5) The great similarity differs from a little similarity. This is called the Little Similarity-and-Difference (xiao-tong-yi ). All things are in one way similar, in another way all different. This is called the Great Similarity-and-Difference (da-tong-yi ). [ ] (H6) The south has no limit and has a limit. [] (H7) I go to the state of Yue to-day and arrived there yesterday. [ ] (H8) Connected rings can be separated. [] (H9) I know the center of the world; it is north of Yan [ the northernmost state] and south of Yue [ the southernmost]. [ ] (H10) Love all things equally; the universe is one. [] In addition to the ten theses attributed directly to Hui Shi, the chapter of Under the Heaven records twenty-one others made by many (other) sophists. According to Fung Yu-lan, they can be divided into two groups as follows:26 Group I: Unity of Similarity and Difference: (S1) The egg has hair. [] (S2) Ying [ the capital of Chu ] contains the whole world. [] (S3) A dog may be a sheep. [] (S4) The horse has eggs. [] (S5) The frog has a tail. [] (S6) Mountains produce mouths. [] (S7) Tortoises are longer than snakes. [ ] (S8) A white dog is black. []
26

Fung Yu-lan, A History of Chinese Philosophy, volume I, pp. 215-217. 41

Group II: Separateness of Hard and White: (S9) A chicken has three legs. [] (S10) Fire is not hot. [] (S11) Wheels do not touch the ground. [] (S12) Eyes do not see. [] (S13) Zhi () do not reach; things never come to the end. [] (S14) T-squares are not square; compasses cannot make circles. [ ] (S15) Chisels do not surround their handles. [] (S16) The shadow of a flying bird never moves. [] (S17) The rapid motion of a flying arrow consists of moments at which it is neither in motion nor at rest. [] (S18) A dog is not a puppy. [] (S19) A brown horse and dark ox make three. [] (S20) An orphan colt has never had a mother. [] (S21) If a rod one foot in length is cut every day by one half of its length, it will still have something left even after ten thousand generations. [ ] Gongsun Long, another important figure of the School of Names in the pre-Qin period, had been widely identified by his contemporaries as the most cunning sophist. He has at least four major theses which are recognized by the modern scholars as paradoxes. They are as follows: (G1) White-horse is not horse. [] (G2) [Supposing there is a hard and white stone], is it possible to say hard, white, and stone[-shape] [i.e., the properties of this particular object] are three? No. Can they be two? Yes. How? When without hard[-ness] one finds what is white, this gives two. When without white[-ness] one finds what is hard, this gives two. Seeing does not give us what is hard but only what is white, and there is nothing hard in this. Touching does not give us what is white but only what is hard, and there is nothing white in this. [ ]

42

(G3) Does two contain one? Two does not contain one. Does two contain right? Two has no right. Does two contain left? Two has no left. Can left and right together be called two? They can. [ ] (G4) There are no things [in the world] that are without zhi, but this zhi is not zhi. [] In the Mohist Canons, there are some paragraphs recognized by some scholars as criticism on some paradoxes quoted from the relativists or skeptics. They are as follows: Canon B71: To hold that all speech is perverse, is perverseness. The reason for explanation is given under his speech. [] Explanation B71: To hold that all speech is perverse is not permissible. If the speech of the person [who urges this doctrine] is permissible and is not perverse, then some speech is permissible. But if her/his speech is not permissible, to suppose that it fits the fact is necessarily ill-considered. [ ] Canon B79: To assert that all arguments can be rejected is perverse. The reason of explanation is given under not to reject. [] Explanation B79: To assert that all arguments can be rejected is to include the rejection of ones own argument. If one does not reject it, s/he is unable to reject it. When one cannot reject it, this means not to reject all arguments. [ ] Canon B66: To say that, Ox-horse is non-ox and to assert this statement are the same. The reason for explanation is given under combination. [ ] Explanation B66: If from the fact that there is something which is not non-ox [and something which is non-ox] in ox-horse we infer it is permissible to say that, Ox-horse is non-ox, then from the fact that there is also ox, in addition to non-ox, in ox-horse we can infer it is also permissible to say that, Ox-horse is ox. So, some may think that the statement Ox-horse is non-ox is not permissible and the statement Ox-horse is ox is also not permissible. But if one
43

of them is permissible, the other must be not permissible. It is also not permissible when you say, It is not permissible to say Ox-horse is ox. Besides, Ox does not include both two and horse does not include both two; only ox-horse includes both two. Therefore, Ox is not non-ox and Horse is not non-horse, only Ox-horse is both non-ox and non-horse. It is no difficult for understanding. [[] ] Appendix II: The Different Cases of Parallelism (Mou Argumentation) in the Later Mohism The schematic examples of the mou argumentation in the chapter of Minor Illustrations can be classified into the following five cases: (1) The case that this [proposition] is right if that [proposition] is affirmed ( ): (1.1) A white horse is a horse; to ride a white horse is to ride a horse. (1.2) A black horse is a horse; to ride a white horse is to ride a horse. (1.3) Huo is a person; to love Huo is to love a person. (1.4) Zang is a person; to love Zang is to love a person. These belong to the case that this is right if that is affirmed. [ ] (2) The case that this [proposition] is not right though that [proposition] is affirmed (): (2.1) Huos parents are jen (people); but Huos serving her/his parents is not serving jen. (2.2) Her/his younger [brother or sister] is a beautiful person; but loving her/his younger is not loving a beautiful person. (2.3) A carriage is wood; but riding a carriage is not riding wood. (2.4) A boat is wood; but entering a boat is not entering wood. (2.5) Robbers are people; but abounding in robbers is not abounding in people;
44

being without robbers is not being without people. How shall we make this clear? (2.6) Disliking the abundance of robbers is not disliking the abundance of people; desiring to be without robbers is not desiring to be without people. (2.7) The people of the whole world agree that these are right; but if such is the case, there is no longer any difficulty in allowing that although robbers are people, loving robbers is not loving people, not-loving robbers is not not-loving people, killing robbers is not killing people. The later claims are the same in kind as the former; the people in the world do not think themselves wrong to hold the former, yet think the Mohists wrong to hold the later. Is there any reason for it but being, as the saying goes, clogged within and closed without? (Gloss: Having no empty space in the heart; it is indissolubly clogged). These belong to the case that this is not right though that is affirmed. [ ] (3) The case that this [proposition] is right though that [proposition] is disaffirmed (): (3.1) Reading books is not books; but to like reading books is to like books. (3.2) A cockfight is not a cock; but to like a cockfight is to like a cock. (3.3) Being about to fall into a well is not falling into a well; but to stop someone being about to fall into a well is to stop him falling into the well. (3.4) Being about to go out of doors is not going out of doors; but to stop someone being about to go out of doors is to stop him going out of doors. (3.5) If such is the case, there is no difficulty in allowing that: Being about to die prematurely is not dying prematurely; but to stop someone being about to die prematurely is to stop her/him dying prematurely. To claim that there is fate is not a fate; but to reject to claim that there is fate is to reject a fate.
45

The later claims are the same in kind as the former; the people in the world do not think themselves wrong to hold the former, yet think the Mohists wrong to hold the later. Is there any reason for it but being, as the saying goes, clogged within and closed without? (Gloss: Having no empty space in the heart; it is indissolubly clogged). These belong to the case that this is right though that is disaffirmed. [ [] [] [] [] ] (4) The cases that one is distributed but another is not distributed ( ): (4.1) S/he loves people requires her/him to love all people without exception, only then is s/he deemed to love people. S/he does not love people does not require that s/he loves no people at all; s/he does not love all without exception, and by this criterion is deemed not to love people. (4.2) S/he rides horse[s] does not require her/him to ride all horses without exception before being deemed to ride horse[s]; s/he rides some horse[s], and by this criterion is deemed to ride horse[s]. On the other hand, s/he does not ride horses does require that s/he rides no horses at all; only then is s/he deemed not to ride horses. These are the cases that one is distributed but another is not distributive. [ ] (5) The cases that one is affirmed but another is disaffirmed ( ): (5.1) If you inhabit somewhere in the state you are deemed to inhabit the state; when you own one house in the state you are not deemed to own the state. (5.2) The fruit of the peach is a peach; the fruit of the bramble is not a bramble. (5.3) Asking about the mans illness is asking about the man; disliking the mans illness is not disliking the man.
46

(5.4) The ghost-soul of a man is not a man; the ghost-soul of your elder brother is your elder brother. Sacrificing to a mans ghost-soul is not sacrificing to a man; sacrificing to your elder brothers ghost-soul is sacrificing to your elder brother. (5.5) If this horses eyes are blind we say that this horse is blind; though this horses eyes are big, we do not say that this horse is big. (5.6) If these oxens hairs are yellow we say that these oxen are yellow; though these oxens hairs are many, we do not say that these oxen are many. (5.7) Both one horse and two horses are horses. Horses have four feet implies four feet to one horse, not to a pair of horses. (5.8) Both one horse and two horses are horses. One-or-other of horses is white implies one-or-other not of one but two horses being white. These are the cases that one is affirmed but another is disaffirmed. [ ] Appendix III: The Logical Analysis of Parallelism (Mou Argumentation) in the Later Mohism The grammatical structure of the mou argumentation is: A is B; CA is CB. Let us see the following examples: (1) The case that this is right if that is affirmed: Example 1: A white horse is a horse; to ride a white horse is to ride a horse. (1.1) [] Example 2: Huo is a person; to love Huo is to love a person. (1.3) [ ] Example 3: A dog is not a puppy; to kill a puppy is to kill a dog. (Canon B54) [ ]
47

Example 4: A gem is a jade; this being a gem is this being a jade. (Major Illustrations) [ ] (2) The case that this is not right though that is affirmed: Example 5: A boat is wood; but entering a boat is not entering wood. (2.4) [] (3) The case that this is right but that is disaffirmed: Example 6: Being about to fall into a well is not falling into a well; but to stop someone being about to fall into a well is to stop her/him falling into the well. (3.3) [] # Both (2) and (3) are cases that exemplify invalid mou argumentation. Some of the failure is due to the semantic imperfection, the other the pragmatic deviation. (4) The cases that one is distributed but another is not distributed: Example 7: This is a horse; to ride this horse is to ride a horse. (4.2A) [ ] Example 8: This is a horse; not to ride horse[s] implies not to ride this horse. (4.2B) [ ] Example 9: A white horse is a horse; not to ride a horse is not to ride a white horse. (1.1R) [] Example 10: Huo is a person; to love a person is to love Huo. (1.3R1) [ ] Example 11: Huo is a person; not to love Huo is not to love a person. (1.3R2) [] # Example 1, Example 2, Example 7 and Example 11 are non-distributed; while Example 8, Example 9 and Example 10 are distributed. (5) The cases that one is affirmed but another is disaffirmed: Example 12: The ghost-soul of a man is not a man; the ghost-soul of your elder
48

brother is your elder brother. Sacrificing to a mans ghost-soul is not sacrificing to a man; sacrificing to your elder brothers ghost-soul is sacrificing to your elder brother. (5.4) [ ] # The second part of the parallelism (The ghost-soul of your elder brother is your elder brother; sacrificing to your brothers ghost-soul is sacrificing to your elder brother.) is an affirmative case while the first part (The ghost-soul of a man is not a man; sacrificing to a mans ghost-soul is not sacrificing to a man.) is not. Example 1: (1.1) Argument Form: (x)(SxPx) (x){ [Mx(y)(Sy&Rxy)][Mx(y)(Py&Rxy)]}

Proof: [1] 1. (x)(SxPx) [2] 2. [Ma(y)(Sy&Ray)] [3] 3. Ma [2,3] 4. (y)(Sy&Ray) [2,3] 5. (Sb&Rab) [1] 6. [2,3] 7. [1,2,3] 8. [2,3] 9. [1,2,3] 10. (SbPb) Sb Pb Rab (Pb&Rab) Assumption Assumption Assumption 2,3, MP 4, EE 1, UE 5, Simp 6,7, MP 5, Simp 8,9 Conj 10, EI 11, RCP 12, RCP 13, UI

[1,2,3] 11. (y)(Py&Ray) [1,2] 12. [Ma(y)(Py&Ray)] [1] 13. {[Ma(y)(Sy&Ray)][Ma(y)(Py&Ray)]} [1] 14. (x){[Mx(y)(Sy&Rxy)][Mx(y)(Py&Rxy)]} Example 2: (1.3) Argument Form: Proof: Ma (x){(MxRxa)[Mx(y)(My&Rxy)]}

49

[1] 1. [2] 2. [3] 3. [2,3] 4. [1,2,3] 5.

Ma (MbRba) Mb Rba (Ma&Rba)

Assumption Assumption Assumption 2,3, MP 1,4, Conj 5, EI 6, RCP 7, RCP 8, UI

[1,2,3] 6. (y)(My&Rby) [1,2] 7. [Mb(y)(My&Rby)] [1] 8. {(MbRba)[Mb(y)(My&Rby)]} [1] 9. (x){(MxRxa)[Mx(y)(My&Rxy)]} Example 3: Canon B54 Argument Form: Proof: [1] 1. [2] 2. [3] 3. [2,3] 4. [2,3] 5. [1] 6. [1] 7. [1] 8. [2,3] 9. [1,2,3] 10. [2,3] 11. [1,2,3] 12. [1,2,3] 13. [1,2] 14. [1] 15. [16] 16. [3,16] 17. [3,16] 18. [1] 19. [3,16] 20. [1,3,16] 21. [1,3,16] 22. [1,3,16] 23. (x)(SxPx) (Ma(y)(Sy& Ray)]) Ma (y)(Sy&Ray) (Sb&Rab) (x)[(SxPx)&(PxSx)] [(SbPb)&(PbSb) ] (SbPb) Sb Pb Rab (Pb&Rab) (y)(Py&Ray) [Ma(y)(Py&Ray)] {[Ma(y)(Sy&Ray)][Ma(y)(Py&Ray)]} [Ma(y)(Py&Ray)] (y)(Py&Ray) (Pb&Rab) (PbSb) Pb Sb Rab (Sb&Rab)
50

(x)(SxPx) (x){[Mx(y)(Sy&Rxy)][Mx(y)(Py&Rxy)]} Assumption Assumption Assumption 2,3, MP 4, EE 1, Definition 6, UE 7, Simp 5, Simp 8,9, MP 5, Simp 10,11, Conj 12, EI 13, RCP 14, RCP Assumption 16,3, MP 17, EE 7, Simp 18, Simp 19,20, MP 18, Simp 21,22, Conj

[1,3,16] 24. [1,16] 25. [1] 26. [1] 27.

(y)(Sy&Ray) [Ma(y)(Sy&Ray)] {[Ma(y)(Py&Ray)][Ma(y)(Sy&Ray)]} [(15)&(26)]

23, EI 24, RCP 25, RCP 15,26, Conj

[1] 28. {[Ma(y)(Sy&Ray)][Ma(y)(Py&Ray)]} 27, Definition [1] 29. (x){[Mx(y)(Sy&Rxy)][Mx(y)(Py&Rxy)]} 28, UI Example 4: Major Illustration Argument Form: Proof: [1] 1. [2] 2. (x)(SxPx) Sa Assumption Assumption 1, UE 2,3, MP (x)(SxPx) / Pa Sa

[1] 3. (SaPa) [1,2] 4. Pa Example 7: (4.2A) Argument Form: Proof: [1] 1. Pa [2] 2. (MbRba) [3] 3. Mb [2,3] 4. Rba [1,2,3] 5. (Pa&Rba) [1,2,3] 6. (y)(Py&Rby) [1,2] 7. [Mb(y)(Py&Rby)] [1] 8. {[MbRba][Mb(y)(Py&Rby)]} [1] 9. (x){[MxRxa][Mx(y)(Py&Rxy)]} Pa (x){[MxRxa][Mx(y)(Py&Rxy)]}

Assumption Assumption Assumption 2,3, MP 1,4, Conj 5, EI 6, RCP 7, RCP 8, UI

# This is the example of to ride horse[s] which is not distributed to other than this horse. Example 8: (4.2B) Argument Form:

Pa (x){[Mx(y)(Py~Rxy)][Mx~Rxa]}
51

Proof: [1] 1. [2] 2. [3] 3. Pa [Mb(y)(Py~Rby)] Mb Assumption Assumption Assumption 2,3, MP 4, UE 1,5, MP 6, RCP 7, RCP 8, UI

[2,3] 4. (y)(Py~Rby) [2,3] 5. (Pa~Rba) [1,2,3] 6. ~Rba [1,2] 7. [1] 8. [1] 9. (Mb~Rba) {[Mb(y)(Py~Rby)][Mb~Rba]} (x){[Mx(y)(Py~Rxy)][Mx~Rxa]}

# This is the example of not to ride horse[s] which is distributed to all horses including this horse. Example 9: (1.1R) Argument Form: Proof: [1] 1. (x)(SxPx) [2] 2. [Ma(y)(Py~Ray)] [3] 3. Ma [2,3] 4. (y)(Py~Ray) [2,3] 5. (Pb~Rab) [1] 6. (SbPb) [7] 7. Sb [1,7] 8. Pb [1,2,3,7] 9. ~Rab Assumption Assumption Assumption 2,3, MP 4, UE 1, Simp Assumption 6,7, MP 5,8, MP (x)(SxPx) (x){[Mx(y)(Py~Rxy)][Mx(y)(Sy~Rxy)]}

[1,2,3] 10. (Sb~Rab) 9, RCP [1,2,3] 11. (y)(Sy~Ray) 10, UI [1,2] 12. [Ma(y)(Sy~Ray)] 11, RCP [1] 13. {[Ma(y)(Py~Ray)][Ma(y)(Sy~Ray)]} 12, RCP [1] 14. (x){[Mx(y)(Py~Rxy)][Mx(y)(Sy~Rxy)]} 13, UI # This is the example of not to ride horse[s] which is distributed to all horses including white horse[s].

52

Example 10: (1.3R1) Argument Form: Proof: [1] 1. Ma [2] 2. [Mb(y)(My&Rby)] [3] 3. Mb [2,3] 4. (y)(My&Rby)] [2,3] 5. (Ma&Rba) [2,3] 6. Rba [2] 7. (MbRba) 8. {[Mb(y)(My&Rby)][Mb Rba]} 9. (x){[Mx(y)(My&Rxy)][Mx Rxa]} Ma (x){[Mx(y)(My&Rxy)][Mx Rxa]} Assumption Assumption Assumption 2,3, MP 4, UE 5, Simp 6, RCP 7, RCP 8, UI

# This is the example of to love a person in terms of Mozis universal love which is distributed to all people including Huo and other people. Example 11: (1.3R2) Argument Form: Proof: [1] 1. Ma [2] 2. (Mb~Rba) [3] 3. Mb [2,3] 4. ~Rba [1,2,3] 5. (Ma&~Rba) [2,3] 6. (y)(My&~Rby) [2] 7. [Mb(y)(My&~Rby)] 8. {[Mb~Rba][Mb(y)(My&~Rby)]} 9. (x){[Mx~Rxa][Mx(y)(My&~Rxy)]} Ma (x){[Mx~Rxa][Mx(y)(My&~Rxy)]} Assumption Assumption Assumption 2,3, MP 1,4, Conj 5, EI 6, RCP 7, RCP 8, UI

# This is the example of not to love a person which is not distributed to other than Huo. Example 12: The first part of (5.4)

53

Argument Form: Proof: [1] 1. [2] 2. [3] 3. [2,3] 4. [2,3] 5. [1] 6. [2,3] 7. [1,2,3] 8. [9] 9. [10] 10. [9,10] 11. [9,10] 12. [1,2,3,9,10] 13. [1,2,3,10] 14. [1,2,3] 15. [1,2,3] 16. [1,2] 17. [1] 18. [1] 19.

(x)(Sx~Px) (x){[Px(y)(Sy&Rxy)][Px(y)(Py~Rxy)]}

(x)(Sx~Px) [Pa(y)(Sy&Ray)] Pa (y)(Sy&Ray) (Sb&Rab) (Sb~Pb) Sb ~Pb Rab Pb (Pb&Rab) Pb (Pb&~Pb) ~Rab

Assumption Assumption Assumption 2,3, MP 4, EE 1, UE 5, Simp 6,7 MP Assumption Assumption 9,10, Conj 11, Simp 12,8, Conj 13, RAA

(Pb~Rab) 14, RCP (y)(Py~Ray) 15, UI [Pa(y)(Py~Ray)] 16, RCP {[Pa(y)(Sy&Ray)][Pa(y)(Py~Ray)]} 17, RCP (x){[Px(y)(Sy&Rxy)][Px(y)(Py~Rxy)]} 18, UI

# The case of sacrificing to a mans ghost-soul is what can be affirmed both in the premise and the conclusion while the case of sacrificing to your elder brothers ghost-soul is what cannot be affirmed both in the premise and the conclusion. The former parallelism has the same logical form as (1.1) (i.e., Example 1); the latter parallelism has the above logical form (Example 12).

54