go back to Zen Essays: PhilosophicalZen


When a White Horse Is Not a Horse

By Kirill Ole Thompson
National Taiwan University
Philosophy East & West, Oct, 1995, Vol. 45, Iss. 4, p481

The paradox that a white horse is not a horse (pai-ma fei ma) and the "Treatise on the White Horse" (Pai-ma lun) attributed to Master Kung-sun Lung (fl. 284-259 B.C.) have alternatively astonished and perplexed readers for over two millennia. Early critics generally attempted to explain away the paradox, and traditional readers tended to see it as a
sophistic word play.[1] Modern readers trained in philosophy and logic have taken it more seriously. Still, recent attempts to interpret the "Treatise" and explain the paradox using formal logic and linguistic theory have met with limited success. Attempts to recapitulate the "Treatise" arguments in symbolic language have so far not accounted for the paradox or shown the rationale behind it to satisfaction.[2] One linguistic approach, which trades upon perceived differences between Classical Chinese and Western languages, appears to create more problems than it succeeds in solving.[3]

An intriguing fact about the paradox is that it remains a puzzle even though the arguments offered by Kung-sun and his disputant in the "Treatise" are fairly straightforward. Moreover, the interpretations proposed by modern readers have not solved the mysteries; thus, it is doubtful that the paradox and arguments in the "Treatise" just turn on Kung-sun's tacitly invoking a special theory of meaning or metaphysics. The paradox does not simply turn on Kung-sun's tacit appeal to the senses of the terms, as opposed to their references, nor does it simply turn on the sentence expressing the denial of an identity rather than the denial of an attribution of class relationship.[4]

Signs and Signification

Let us try an entirely different tack. In another work, "Treatise on. Signifying Things" (Chih-wu lun), Kung-sun presents, among other things, arguments concerning the nature of signs and the signification function.[5] The "Treatise" is too corrupt to support a definitive interpretation, but it does nonetheless involve an argument for a conventionalist understanding of signs and the signification function.[6] Kung-sun argues there, for example, that the world consists only of things, that there are no signs as such in the world. Consequently, since "signs definitely do not signify of themselves," the signs that exist in the world have been set up by human beings to signify things in communication. Otherwise, "why would they depend upon [being related by human beings to] things so as to signify?" The "Treatise" is animated by the insight that if there were inherent, that is, natural, signs, there would be the complication of whether our human signs, such as words in language, corresponded to the natural signs or not, which would introduce serious complications into the concept of signification.[7] In sum, Kung-sun mounts a prima facie reasonable case for understanding words and language as conventional in nature in this "Treatise."

That Kung-sun views language as a communication system based upon conventionally established signs indicates that he sees languages as ultimately devised, and words as ultimately defined, by human beings for various communicative functions. He realizes that words are not inherent, natural signs that just stand for the objects: they do not simply constitute a sort of transparent medium that just presents the objects of reference. Accordingly, he sees signs (words and expressions) as human artifacts that can be viewed and discussed on their own in light of their understood significative and communicative functions.[8] And, this is what provides a possible interpretative key for explaining the white horse paradox and treatise.

White Horses and Horses

In the dialogue recorded in the "Treatise on the White Horse," Kung-sun argues that "a white horse is not a horse" is an acceptable proposition, against a disputant who argues to the contrary that "having a white horse, we cannot assert that we have no horse."[9] We will want to note that in the course of the debate, Kung-sun focuses mainly on the signs (terms, expressions) "white horse" and "horse" in their significative and communicative functions, while his disputant speaks of their objects — white horses and horses. Whenever the disputant does consider the terms as such, he views them as determining their objects in judgment, not simply in consideration of their respective significative and communicative functions.

Kung-sun's first consideration for deeming "A white horse is not a horse" an acceptable proposition is based on the respective significations of the two terms: "The term 'horse' is that by which we name the form [of the natural kind], the term 'white' is that by which we name the color." Since the term "horse" is color-neutral, it differs from the expression "white horse."[10] His second consideration is based on the communicative functions of the two terms, that is how and what they discriminate: one uses the term "horse" to select or pick out horses, regardless of color; but, one uses the term "white horse" to select or pick out only horses that are white. If these two terms were the same, they would be used to select and pick out the same things.[11]

In contrast, the disputant from the outset speaks of white horses and horses and views the terms as determining them, as in judgments.[12] He first insists that, "having a white horse, one cannot be said to not have a horse." Next, mistaking Kung-sun's claim to be that what has been judged to be a "white horse" cannot be judged to be a "horse" per se, he attempts to derive a contrary-to-fact conclusion to the effect that, since every horse can be judged to be of some color, it would follow from Kung-sun's view that there are no horses per se in the world. In responding, Kung-sun turns his attention back from the terms to horses in order to agree that, of course, horses have colors: this fact constitutes one of the reasons why people can call for a "white horse" or a "yellow horse" in the first place.

In his third counterpoint, the disputant again views the terms "white horse" and "horse" as judgments determining their objects. He attributes to Kung-sun the view that only a horse not yet judged to be white is a horse per se, and only a white patch not yet attributed to a horse is white per se (thus implying that Kung-sun doesn't accept or understand that various different judgments may be made about the selfsame object). The disputant then proceeds to claim that Kung-sun has inappropriately combined the words "white" and "horse" — which refer to different categories of entities, that is, colors and forms (of natural kinds) — to form an illicit compound term, "white horse." This presumably would entail the existence of that mixed entity as a natural kind.[13]

Kung-sun does not respond to this point directly; but the suggestion is that, in common parlance, speakers frequently create compound terms of just this sort to form qualified subjects: they use such compound terms purposely where a general term, like horse, will not answer to their communicative needs. Such compound terms are communicatively clear and
pragmatically effective in spite of their referential complexity.[14]

Use and Mention

Stepping out of the "Treatise" for the moment, we again want to note that Kung-sun's arguments focus on the terms "white horse" and "horse" and emphasize differences between these terms in significance and communicative function. In contrast, the disputant's arguments concern the relations between white horses and horses, and focus on the terms only as expressing judgments that determine them. Accordingly, we see that Kung-sun is, in effect, staking his position on an insight into what is now referred to as the use-mention distinction, an important distinction in modern logical analysis and linguistic philosophy.[15] Briefly, the use-mention distinction marks the difference between contexts in which one uses words and expressions in their normal functions, such as to refer to and talk about "things" in the world, from contexts in which one mentions words, that is, to talk about the words themselves, for example about their grammar, how they are used, their usual entailments, and even their spelling.

Kung-sun's arguments in the "Treatise on signifying things" indicate that he was cognizant of words as signs vis-a-vis the objects they signify, and that he was self-conscious enough about words as conventionally established signs to be able to focus on the terms as opposed to the objects in argument. He was in a most favorable position to discern and grasp the use-mention distinction. In this light, a natural explanation for Kung-sun's paradox of the white horse would be that he constructed the paradox by conceiving the proposition "A white
horse is not a horse" (pai-ma fei ma) in two aspects corresponding to the contexts of use and mention: viewed in the context of use, the proposition would be an absurd denial that "a white horse is a horse," and thus would appear to be inconsistent if not contradictory; viewed in the context of mention, the proposition would be a tautology, for "the term 'white horse' differs from the term 'horse'" is patently true.

The ambiguity of these two aspects was not noticed by readers down through the ages because the use-mention distinction was not widely recognized, in the West or the East. People spontaneously tend to understand words in the context of use rather than of mention, unless explicitly trained to do otherwise. Moreover, since Classical Chinese had no specific punctuation or other means to mark this distinction, Kung-sun's ambiguity was seamless. One had to be cognizant of the aspect of mention to see through it and get the point. The preface to his works present the paradox that a white horse is not a horse as bearing Kung-sun's essential teaching and as his claim to fame. We may infer that his insight into the use-mention distinction was the crucial idea behind the white horse paradox. Kung-sun apparently preserved his insight into the use-mention distinction as a school secret to be deployed as an ace up the sleeve in debates.

"White Horse" and "Horse"

Returning to the "Treatise," Kung-sun next in a relatively light-hearted argument derives a contradiction from the disputant's claim that "having a white horse, one cannot be asserted to not have a horse": he asks the disputant whether, if having a "white horse" is the same as having a "horse," it is admissible to consider that having a "white horse" is the same as having a "yellow horse." The disputant has no recourse but to answer in the negative. Kung-sun replies, "Now you have just shown that you do not consider a 'yellow horse' to be a 'horse,' yet you insist that a 'white horse' is a 'horse.'" He calls this as absurd as the idea of a flying object in a pond or the inner and the outer coffins reversed.[16]

In a final counterpoint, the disputant again speaks of horses, colors, and judgments about them: "Having a white horse, one cannot state there is 'no horse,' because when we separate out the term 'white' the term 'horse' remains." Thus, "the reason why we take it to be a 'horse' cannot be just because we call a horse a 'horse.'" In response, Kung-sun shifts back to the context of mention and observes that we do not use the term "horse" to select horses on the basis of color — that is why black or yellow steeds would be acceptable; but we use the expression "white horse" to select horses on the basis of color — that is why black or yellow steeds would not be acceptable. Since "white horse" discriminates what "horse" does not, "a 'white horse' is not a 'horse.'" Thus ends the "Treatise on the White Horse."

Significance of the White Horse Paradox

Accepting that the white horse paradox is based on the use-mention distinction, the question remains whether it is still just a sophistic sleight of hand, or whether it is indicative of some important truths about words, language, and logic.

First, the paradox thus understood underscores some real differences in the significance and implications of terms that appear when we consider them in the context of mention rather than in the usual context of use. Moreover, the paradox shows that we can gain further insight into how words and phrases operate in the complex signification system of language by considering them in the context of mention. Indeed, various features of their meaning and use remain obscure until we consider them in this context.

Second, the paradox thus understood makes us inclined to think of the instrumental value of words in communication, as opposed to simply thinking of their assigned roles in referring and in forming judgments and inferences. It becomes apparent in the course of the white horse dialogue that, although the disputant's claim appears to be true at first sight, Kung-sun's arguments based on the significative and communicative functions of the terms yield a more perspicacious view of how they can be, and are, used in common speech. The disputant's points based on the terms as determining objects in judgment increasingly appear to depart from common usage.

Kung-sun apparently noticed that the Neo-Mohist logicians' analyses of words and expressions departed from their meanings and uses in common speech. Accordingly, he formed the paradox of the white horse as a dilemma for the logicians' artificial patterns of analysis and inference, to be solved by consideration of how the terms are selected, combined, and used in ordinary speech, as revealed in the context of mention. We perhaps see in this a parallel to Ludwig Wittgenstein's later insight that the meaning of a word is a function of its use in language and communication, not just a function of its sense and reference.[17] Kung-sun in his first argument distinguishes the senses and references of the terms "white horse" and "horse" by noting that "white" is a color word while "horse" pertains to the form (of the natural kind). In his second argument, he focuses on how these terms are used to discriminate white horses and horses, respectively. On this basis, from the perspective of "meaning is use," Kung-sun shows how the qualified compound term "white horse" would express the precise meaning that speakers require in their communication, that is, in the pragmatics of actual speech, where the unqualified term would not be adequate, or perhaps not even relevant.

A story in the preface to the works of Master Kung-sun Lung recounts a case in which Confucius distinguishes the qualified term "Ch'u man" (Ch'u-jen) from the unqualified term "man" (jen) in showing that, in that situation, the unqualified term would be the more apt choice.[18] The example thus would reiterate that the compounding of nominative terms and descriptive terms is a function of situation and communicative need, not a priori logical requirement. In fact, speakers generally do not first think of the nominative term, and then add a qualification. Rather, in forming a compound term they tend to express their qualified meaning directly without giving special thought to the meaning or range of the unqualified nominative term, which often would be irrelevant to their sphere of concern, interest, and cognizance. For instance, a man who claims that a "'white horse' is not just any old 'horse'" might be needing to obtain "white horses" for a special purpose, such as a wedding procession. In sum, if our account is at all accurate, Kung-sun made an important discovery about the formation of compound terms as viewed from the perspective of the pragmatics of actual speech rather than pure logical theory.[19]

If our general account of Kung-sun Lung's paradox of the white horse is broadly right, the early interpretations of the paradox proffered by Chuang Tzu and Hsun Tzu were well off the mark. Chuang Tzu mistakenly thought that the conventionalist views of language held by Kung-sun and others entailed that speakers are free to stipulate special meanings for their words more or less at will; thus "a white horse is not a horse" because I can arbitrarily define "white horse" to mean, for instance, "pink elephant."[20]

Hsun Tzu held that, due to a prolonged lack of stable, unified rule in the Empire, the meanings of words and expressions had become vague and unstable to the point that not only did people misunderstand each other, but measures, standards, and norms were uncertain. Thus, he advocated that the authorities undertake a form of language engineering he called
"rectifying names" (cheng-ming) in order to standardize not only measures, standards, and norms, but the general language itself. He formulated several guidelines for recognizing confusions in the uses of terms that distort the relationships between words and objects. On this basis, he viewed the paradox of the white horse as involving the distortion "of objects by confusion in the use of terms," to be refuted by testing "them by the convention for the name and us[ing] what one accepts to show that what one rejects is fallacious."[21]

According to the present account, Hsun Tzu was right in noting that Kung-sun had focused on the terms per se, rather than on their objects. He failed, however, to see that in formulating the paradox, Kung-sun was simply registering the difference in significance and communicative function between the qualified term "white horse" and the unqualified term "horse." In effect, he was "clarifying" the respective significances and communicative functions of those terms as used in ordinary language. Therefore, he did not distort the objects by means of confusing the uses of the terms. Nor was he undertaking the sort of language engineering implied in Hsun Tzu's own idea of rectifying names.[22]

What about the apparently true counterclaim that "a white horse is a horse"? Is it a significant judgment? Although it appears to take the form of a significant proposition, like "a whale is a mammal," prima facie it does not convey a positive sense in that a horse of any color is just a horse (a horse is a horse is a horse). In ordinary communicative contexts, "a white horse is a horse" would be taken as a statement that expresses no genuine attribution or classification. Applying Wittgenstein's later method, one might try to think of circumstances under which the proposition would communicate a sense, such as for the purpose of disillusioning people who hold the superstitious belief that white horses are in some sense a marvelous breed apart, not to be considered mere horses. This again underscores how Kung-sun Lung derived absurdities from the logician's tendency to view words and language through the stencil of a narrow notion of judgment and inference, thus losing cognizance of the significative and communicative functions of words and language in human life.

Concluding Reflections

In closing, we may note that the "Treatise on the White Horse" manifests a pattern of argument that later appeared in the Taoist texts, the Chuang Tzu and the Lao Tzu: the assertion of a difference, expressed by the negation "fei" (not, is not), followed by a debate or a dialectical discussion concerning the meaning and implications of the denial. We read, in Chuang Tzu, chapter 2:

Words are not just wind.[23]

Words have something to say. But if what they have to say is not fixed, then do they really say something? People suppose the words are
different from the peeps of baby birds, but is there any difference, or isn't there? And, in Lao Tzu, chapter 1:

The Tao that can be spoken of is not the Tao itself. The name that can be named is not the name itself. The unnameable is the source of the universe. The nameable is the originator of the universe. Therefore, often times with intention I see the wonder of Tao; Often times with intention I see its manifestations. Its wonder and its manifestations are one and the same. Since their emergence, they have been called by different names. Their identity is called the mystery. From mystery to further mystery: The entry of all wonders?[24]

In each of these passages, the Taoist author asserts a "difference" in the opening line, only to go on to undermine it and argue for an underlying "identity" at a deeper level. Accordingly, words clearly are not just wind, but what they say is not fixed in any ultimate sense; hence, in the broader context of nature, words are emitted as natural expressions of people — just as chirps are the natural expressions of fledglings, and rustlings are the expressions of forest leaves.[25] And, the Tao that can be spoken of is not the Tao itself, and the name which can be spoken is not the name itself; but, at a much deeper level, a Tao that can be spoken of is a particular reflection of the Tao itself and ultimately identifies with it.

In the "Treatise on the White Horse," Master Kung-sun Lung opens by affirming a "difference" where most readers at first see a natural inference if not an identity between a white horse and a horse. But, unlike the authors of the Chuang Tzu and the Lao Tzu, who attempt to show the implicit identity of items first asserted to be different, Kung-sun persists in arguing for the difference between two things that look to be basically the same — a "white horse" and a "horse," while his disputant persists in arguing for their identity or for their intimate inferential connection, without any clear synthesis or resolution.

Why does Kung-sun display this difference in purpose and approach from the Taoist masters? Broadly, the Taoist masters sought by such dialectical reflections, as well as by assorted other means, to undermine the sense that discriminative judgments are in any way final and to project a sense of the experience of ultimate identity at the heart of their philosophic program.[26] In the "Treatise on the White Horse," Kung-sun set for himself the more modest goal of displaying a dilemma inherent in the way logicians analyze words, judgments, and
things and then of hinting, tantalizingly, at a solution. He was teasing us.

It is lamentable that Master Kung-sun Lung's insights and teachings were not developed and transmitted beyond his school, and that over time his works became fragmented and, in places, virtually indecipherable. Had they been understood aright in the beginning, they might have had a salutary effect on the early development of Chinese logic and philosophy. As it turned out, later Neo-Mohists and Hsun Tzu got in the last word by classifying the white horse paradox as a fallacy under their respective criteria for terms and judgments, thus reducing this suggestive and intriguing proposition and the related arguments to trivial wordplay for the subsequent tradition.

Although the present study is not the last word on these issues, hopefully it charts the course to a more adequate comprehension of Master Kung-sun Lung's essays and insights.


Pai-ma lun (Treatise on a White Horse)

A disputant inquired, "Is 'A white horse is not a horse' an admissible

Master Kung-sun replied, "It is."

"How can that be the case?"

"The term 'horse' is that by which we name the form, the term 'white' is that by which we name the color. The term by which we name the form cannot be used to name the color. Therefore, I affirm, 'A white horse is not a horse.'"

The disputant pursued, "Having a white horse, we cannot assert that we have no horse. How could 'we cannot assert that we have no horse' mean that it 'is not a horse'? To have a white horse is to have a horse; how can its being white make it not a horse?"

Master Kung-sun replied, "ln calling for a horse, either a yellow or a black horse will do, but in calling for a white horse, a yellow or a black horse will not do. If having a 'white horse' is indeed to have a 'horse,' what we would call for [in using each term] would be one and the same; if 'what we would call for would be one and the same,' 'white' would not differ from 'horse.' What we would call for [by using either term] would not differ. However, yellow and black horses are acceptable and unacceptable. How can that be the case? That acceptable and
unacceptable are contradictory is manifest. Therefore, yellow and black horses equally can answer to the term 'horse' but not to the term 'white horse.' This is my case for asserting that 'A white horse is not a horse.'"

The disputant continued, "You take a horse's having color to deny that it is a horse, but there are no colorless horses in the world. Is the assertion 'There are no horses in the world' then admissible?"

Master Kung-sun replied, 'Horses certainly have color. That is why there are white horses. If horses were colorless and there were just horses per se, how could we pick out a 'white horse'? That is why a white horse is not a horse. The expression 'white horse' is composed of 'horse' and 'white'; 'horse' thus differs from 'white horse'; hence I affirm that 'A white horse is not a horse.'"

The disputant pursued, "You regard only a horse that has not yet been judged white to be a horse, and only whiteness that has not yet been judged a horse to be white; but, to combine 'horse' and 'white' to make the composite name 'white horse' is to pair two dissimilar things in producing a name; that is inadmissible. Therefore, I maintain that the assertion 'A white horse is not a horse' is inadmissible."

Master Kung-sun replied, "If you consider having a white horse to be having a horse, is it admissible to consider 'having a white horse to be having a yellow horse'?"

"It is not admissible."

"If you consider having a horse to be different from having a yellow horse, isn't it because you distinguish a yellow horse from a horse? Therefore, a yellow horse is not considered to be a horse. To consider a yellow horse not a horse while considering a white horse to be a horse is for 'this flying object to be in the pond' or for 'the inner and outer coffins to be in different places! [i.e., contradictory,]. This is what the world considers 'perverse speech and disorderly expressions.'"

The disputant pursued, "Having a white horse, one cannot state that there is no horse, because when we separate out the term 'white' there remains 'horse.' This separating out [shows that] having a white horse, one cannot go on to say there is no horse. On your view, however, the sole reason why there is the term 'horse' is that we say 'horse' to indicate there is a horse; and it is not the case that we say 'white horse' to mean there is a horse. Therefore, the reason why we take it to be a horse cannot be just because we call a horse 'a horse.'"

Master Kung-sun replied, "The term 'white' does not determine what thing is white; that can be neglected. The expression 'white horse' designates what thing is white. That which is designated as white is not whiteness per se. We do not select horses on the basis of color; that is why yellow and black ones will suffice. But, we select white horses on the basis of color; that is why yellow and black ones are rejected. Therefore only a white horse will suffice; that which rejects [on the basis of color] is not what is rejected; therefore, I say, 'A white
horse is not a horse.'"

II. Chih-wu lun (Treatise on Signifying Things)

[Master Kung-sun asserted,] "There is no thing that is not signified; but, what signify are not [inherently] signs."

[A disputant questioned, "Yet, if] the world had no signs, things could not be called things. If what signify are not [inherently] signs, how could the world and things be said to be signified?"

[Master Kung-sun in reply stated the thesis,] "Signs are not what the world possesses; things are what the world possesses. To take what the world possesses to be what the world does not possess is inadmissible."

[He pressed, "By arguing that] 'if the world did not possess signs, things could not be said to be signified [as things],' aren't you merely contending [that they] 'could not be said to be signified,' because what signify are not [inherently] signs? There aren't such signs, because there is no thing that is not signified."

[He continued,] "You contend that, 'If the world did not possess signs, things could not be called signified [as things],' but there are not any unsignified things. There could not be any unsignified things, because there is no thing that is not signified. 'There is no thing that is not signified; but, what signify are not [inherently] signs.' "

[The disputant pursued, "I then venture to say that] 'the world does not possess signs,' because each kind of thing has its own name, which was not simply made up to signify it. By calling these signs, when they were not made up to signify, is how you consider all things to be signified. To conceive what have not been made up as signs to have been made up as
signs is surely inadmissible."

[Master Kung-sun replied, "I maintain that] 'signs are what the world does not possess.' That the world has no signs is because no thing can be said to be unsignified. No thing can said to be unsignified, because no thing fails to be signified. And, no thing fails to be signified, because 'there is no thing that is not signified.' The problem is not just that what signify are not [inherently] signs, but that [the relationship between] signs and things cannot be signified."

[Master Kung-sun concluded,] "If the world possessed no signified things, who could even say that [signs do] not signify? If the world possessed no things, who could say that [something is] signified? If the world possessed signs, but things were not signified, who could say that [signs do] not signify? Who could say that 'there is no thing that is not signified'?"

"In sum, signs definitely do not of themselves signify. [Otherwise,] why must they depend upon [being related to] things in order to be made to signify?"

Further Readings


This study benefited considerably from the critical input of friends Bill McCurdy, Harry White, Karen Chung, Huang Yi-hsuan, Bob Christiansen, and Mike Bybee. The author bears sole responsibility for whatever flaws remain.

I have relied mainly on two modern annotated editions of the Kung-sun Lung Tzu in developing translations of the "Pai-ma lun" and "Chih-wu lun": Ch'en Kuei-miao ed., comp., and trans., Kung-sun Lung Tzu chin-chu chin-i, 2d ed. (Taipei: Commercial Press, 1987), and P'ang P'u, ed., comp, and trans., Kung-sun Lung Tzu yen-chiu (Peking: Chung hua, 1979).

1 - Well-known criticisms are implied or stated in the Chuang Tzu, chap. 2, "Making Things Equal" (Ch'i-wu lun) (see Burton Watson, trans., The Complete Works of Chuang Tzu [New York: Columbia University, Press, 1968], p. 40, and A. C. Graham, trans., The Inner Chapters [London: George Allen and Unwin, 1981 ], pp. 53, 284-285), in the Hsun Tzu, chap. 22, "Rectifying Names" (Cheng-ming) (see Burton Watson, trans., Hsun Tzu: Basic Works [New York: Columbia University Press, 1963], p. 146, and A. C. Graham, Later Mohist Logic, Ethics and Science [Hong Kong: The Chinese University Press, 1978], pp. 234-235), and in the Neo-Mohist Canons (Mo ching) and the Explanations (Shuo) (see Graham, Later Mohist Logic, pp. 245, 440). Graham gives representative assessments of later commentators in Later Mohist Logic, pp. 174-175.

2 - Interestingly, Wang Kuan used simple set diagrams in discussing the paradox in 1925 in Kung-sun Lung Tzu suan-chieh (reprint, Taipei: Taiwan Chung-hua, 1980), p. 44. The pioneering work in Western languages is Janusz Chmielewski's "Notes on Early Chinese Logic," pt. 1, Rocznik Orientalistyczny 26 (1) (1962): 7-22. A. C. Graham critiques Chmieiewski in "Three Studies of Kung-sun Lung," in Studies in Chinese Philosophy and Philosophical Literature (Singapore: Singapore University Press, 1986), pp. 181-187. Two new logical studies appeared in journal of Chinese Philosophy 20 (2) (June 1993): Ernst Joachim Vierheller, "Object Language and Meta-Language in the Gongsun-long-zi" (pp. 181-209), and Thierry Lucas, "Hui Shih and Kung Sun Lung: An Approach from Contemporary Logic" (pp. 211-255).

3 - Proposed by Chad Hansen in Language and Logic in Ancient China (Ann Arbor: The University of Michigan Press, 1983), and A Daoist Theory of Chinese Thought: A Philosophical Interpretation (Oxford: Oxford University Press, 1992). We cannot go into specifics here. Hansen's approach builds on an array of controversial linguistic and grammatical assumptions about Classical Chinese and yields a very counterintuitive interpretation of the white horse treatise and paradox. If Hansen's assumptions were true, we would have expected early Chinese readers to have understood the paradox and treatise without too much ado. As it was, they were as much in the dark as are modern readers.

Hansen's work has been criticized on linguistic and grammatical grounds. In a series of writings, Bao Zhi-ming has criticized Hansen's linguistic approach to and philosophical conclusions about Classical Chinese, starting from a critical review of Language and Logic in Ancient China, in Philosophy East and West 35 (2) (April 1985): 203-213. A subsequent issue of Philosophy East and West (35 [4] [October 1985]: 419-430) carried Hansen's response and Bao's reply. For a general statement of Bao's position, see his "Language and World View in Ancient Chinese Language," Philosophy East and West 40 (2) (April 1990).

More importantly, Christoph Harbsmeier has challenged Hansen's grammatical assumptions and philosophical conclusions about Classical Chinese. In particular, see his study, "The Mass Noun Hypothesis and the Part-Whole Analysis of the White Horse Dialogue," in 'Henry Rosemont, ed., Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus C. Graham (La Salle: Open Court, 1991), pp. 49-66. A. C. Graham's comments are given on pp. 274-278. A. C. Graham approvingly restates Harbsmeier's view in Disputers of the Tao (La Salle: Open Court, 1989), p. 402. Harbsmeier's works have resulted in an advance of our understanding of Classical Chinese grammar and thought. Hansen replies, rather unpersuasively, that Harbsmeier's grammatical evidence is syntactic and doesn't touch his assumptions, which are semantic (see Hansen, A Taoist Theory, p. 48).

4 - See Graham, Disputers of the Tao, pp. 85-86. Fung Yu-lan presents Kung-sun's first argument as based on the "difference in the intention of the terms 'horse,' 'white,' and 'white horse,' " his second argument as based on different extensions of these terms, and his third argument as based on "the distinction between the universal, 'horseness,' and the universal, 'white-horseness'" (Fung Yu-lan, A Short History of Chinese Philosophy, ed. Derk Bodde [New York: Macmillan, 1948], pp. 87-88. See also Fung Yu-lan,A History of Chinese Philosophy, Vol. 1 , ed. Derk Bodde [Princeton: Princeton University Press, 1952], vol. 1, pp. 203-205).

With its suggestion that the arguments turn on arcane logical and conceptual reflections, Fung's work aroused renewed interest in Kung-sun Lung's works among Chinese specialists trained in modern Western logic and philosophy. The annotations in new Chinese-language editions of Kung-sun Lung's treatises tend to follow Fung's view that Kung-sun is
referring to objective concepts in the first argument and to the extensions of these concepts in the second argument, and so forth.

Other studies along these lines include: Chung-ying Cheng and Richard Swain, "Logic and Ontology in the Chih Wu Lun of the Kung-sun Lung Tzu," Philosophy East and West 5 (2) (March 1970): 137-154; Kao Kung-yi and Diane B. Obenchain, "Kung-sun Lung's Chih wu lun and Semantics of Reference and Predication," Journal of Chinese Philosophy 2 (3) (September 1978): 285-324; F. Rieman, "Kung-sun Lung, Designated Things, and Logic," Philosophy East and West 30 (4) (October 1980): 349374; and Chung-ying Cheng, "Logic and Language in Chinese Philosophy," Journal of Chinese Philosophy 14 (3) (September 1987): 285-308.

Graham's "Three Studies of Kung-sun Lung Tzu," pp. 125215, and of course Hansen's books mark significant departures from these lines of interpretation.

5 - A full translation is given at the end of this essay. Cf. A. C. Graham's annotated translation in Disputers of the Tao, pp. 90-94. I render the title "Chih-wu lun" as "On Signifying Things" so that it parallels the title of Chuang Tzu's essay "Ch'i-wu lun" (On making things equal). Kung-sun is showing how signs are used to select or pick out (kinds of) things in the world, i.e., to discriminate them from other kinds of things, while Chuang Tzu seeks to show that, beneath all the perceived discriminations among things, such as are enshrined in language, things are ultimately equal and identify with each other.

6 - On this point, I agree with Chad Hansen that Kung-sun shows that "There is no natural realistic relation between names [signs] and things. They depend on conventions." But I disagree with his further claim that Kung-sun infers from this that "we can construct the conventions according to whatever ideal principles we want. We can have a language in which white horse is not horse . . ." (A Daoist Theory, p. 260). I find nothing in Kung-sun's treatises to suggest that he advocated language reform of any sort. The present study indicates that his views were far closer to common usage than were the views of the disputant, whom Hansen characterizes as either a Neo-Mohist or an advocate of common sense.

7 - Such complications would resemble those attending the relationship between the deep structure and the surface structure of propositions and language postulated by Ludwig Wittgenstein in Tractatus logico-philosophicus. The German text of Logisch -philosophische Abhandlung, with a new translation by D. F. Pears and B. F. McGuinness, trans. D. F. Pears and B. F. McGuinness (London: Routledge and Kegan Paul, 1961). It was due to realizing a collection of such intractable complications that Wittgenstein eventually abandoned the Tractatus line of inquiry based on meaning as a function of sense and reference, and began to investigate the meanings of words and expressions as a function of their uses in language, viewed in the context of language games and understood in the flow of human life.

8 - We shall see below that some of Kung-sun's arguments in the "Treatise on the White Horse" turn on the uses the terms have in ordinary language and communication.

9 - See the full translation of this treatise at the end of this essay. Cf. A. C. Graham's annotated translation in Disputers of the Tao: Philosophical Argument in Ancient China, pp. 8590.

10 - Early commentators saw this as the main argument of the "Treatise." See Graham, Studies in Chinese Philosophy, pp. 157-158. Kung-sun uses the term hsing (form, shape) in an abbreviated sense; that the term ma (horse) refers to a kind of animal is left unstated. Given that implied understanding, hsing refers to the features by which people distinguish and identify horses. Oddly, Hansen's claim that ma refers mereologically to "horse-stuff" in this dialogue cannot be squared with Kung-sun's own words (see Language and Logic, p. 142, and Harbsmeier, "The Mass Noun Hypothesis").

11 - This argument signals Kung-sun's alertness to the communicative values of the terms in human life, i.e., the pragmatic dimension of their meaning. Surprisingly, Hansen, who capitalizes on the pragmatic orientation of Classical Chinese, in contrast to the propositional orientation of modern English, does not seize upon the pragmatic character of this argument (see, e.g., Language and Logic, p. 164).

12 - The disputant thus reads the proposition at issue sometimes as "a white horse is not a horse" and sometimes as "that which is determined/judged a 'white horse' is not determined/judged a 'horse.' "

13 - Mainly a logic of general nouns and descriptive terms, Mohist logic focuses on rules governing the ways in which terms may be validly combined and separated, and the various scopes afforded by the various combinations of terms. See Graham, Disputers of the Tao, pp. 150-155 and 167-170. Hansen attempts to interpret Kung-sun's arguments in the light of these rules and several linguistic hypotheses (Language and Logic, pp. 150 ff.).

14 - Kung-sun's emerging view appears to be that whether "white horse" is a term or not turns on the speaker's intentions and communicative needs, not on the ideal requirements of logical theory. Thus, it is not as if he were treating the compound expression as a single term, like palomino, as suggested by Hansen (ibid., p. 166).

15 - For discussion of the meaning and importance of the use-mention distinction, see Logic , Logic, 3d ed. (Englewood Cliffs: Prentice-Hall, 1984), pp. 140-144, and K Codell Carter,A contemporary introduction to logic with applications (Beverly Hills: Glencoe Press, 1977), pp. 6-8.

16 - This argument is the closest to sophistry of the lot. Still, if the argument concerns the difference between the terms "white horse" and "horse", Kung-sun's argument is quite valid.

17 - In his later thought, Wittgenstein "displayed a movement awayfrom focusing on forms of expressions and their patterns of relationship towards concentrating on uses--away from viewing discourse as a patterned array of symbols towards seeing speech as part of the web of human life, interwoven with a multitude of acts, activities, reactions and responses" (G.P. Baker and P.M.S. Hacker, Wittgenstein: Rules, Grammar and Necessity [Oxford: Basil Blackwell, 1985], p. 39). For instance, in Philosophical Investigations , trans. G.E.M. Anscombe (Oxford: Basil Blackwell, 1958), par. 43, Wittgenstein writes: "For a large class of cases . . . in which we employ the word 'meaning' it can be defined thus: the meaning of a word is its use in the language." And Norman Malcolm records Wittgenstein as saying, "An expression only has meaning in the stream of life" (Ludwig Wittgenstein: A Memoire [London: Oxford University Press, 1958], p. 93). A vast literature exists on Wittgenstein's seminal idea that meaning is use.

18 - The King of Ch'u has said that a Ch'u man will recover his lost bow, to which Confucius replies that the term "Ch'u man" is too restrictive, for the bow could be recovered by a man from any state, thus, as Kung-sun notes, distinguishing "Ch'u man" from "man." Notice that the sentence "A Ch'u man is not a man" (Ch'u-jen fei jen) could be plugged into Kung-sun's white horse pattern of argument to yield the same pattern of results. This example shows that Kung-sun's arguments hold under different circumstances and for various speaker's intentions. Thus, K'ung Ch'uan's reply that "to widen 'man' you have to leave out the 'Ch'u,' if you wish to specify the name of the color you must not leave out the 'white'" (quoted by Graham from an independent source, in Disputers of the Dao, p. 84) does not apply to Kung-sun's position. Kung-sun is only arguing for the practical difference between the qualified and the unqualified terms in significance and communicative function: in the case of the horses, the distinction of white horses is deemed important; in Confucius' case concerning men, the qualification of Ch'u man is deemed unsuitable to the circumstances.

The preface contains three stories probably composed during the Han, all of which maintain that the white horse paradox represents Master Kung-sun's essential teaching and attempt variously to defend its assertability.

The first story attributes to Kung-sun motivations of (1) upholding the independent significance of descriptive terms, like "white" (presumably as opposed to nominal terms), and (2) rectifying the relationship between terms and objects. It recounts the first two arguments of the "Treatise" and gives the example of how a black horse in the stable will not answer to one's request for a "white horse." It concludes that "Kung-sun intended to extend this argument in order to rectify the relationship between terms and objects so as to transform the Empire."

The second story tells about the King of Ch'u and Confucius in order to give another illustration of Kung-sun's paradox. The third story illustrates Kung-sun's ideas in terms of examples of character, obligation, and law.

Although based on extant records about Kung-sun, these stories were written centuries later, probably by Han thinkers interested in casting Kung-sun as a distant Master for their sect, which combined Confucianism and Legalism. Thus, they attributed to him a variation of Confucius' and Hsun Tzu's idea of rectifying names (cheng-ming), i.e., rectifying the relationships between terms and objects, and claimed he was out to transform the Empire. They attempted to link his doctrine to a point made by Confucius, thus showing him to be closer in spirit to the Sage than was Confucius' own descendent, K'ung Ch'uan. Finally, they illustrated his idea with an example about character, obligation, and law.

Hansen's claim that these stories contain "hints about the motivation for Gongsun Long's theorizing" cannot be taken at all seriously. And, neglecting that the stories were written centuries later by people with their own agendas, Hansen unaccountably portrays Kung-sun himself as the author: "He represents himself as defending a Confucian position . . . . He is . . . associating [himself] with the divine sage . . ." (A Daoist Theory, p. 256). Hansen somehow forgets that a mark of the proponents of the School of Names (ming-chia), the "Sophists" in Graham's rendering, was their pure focus on issues in language and logic. There is no record of their recognizing any sages or favoring any particular interventionist philosophies, such as Confucianism, Mohism, or Legalism.

It is possible that Kung-sun Lung thought that, if understood, his ideas would have the beneficial effect of encouraging people to be more self-conscious and precise in their uses of terms. That is why I prefer to speak of Kung-sun as "clarifying the relationships between terms and objects" rather than as "rectifying" them. He didn't advocate making changes and reforms, just being more "perspicacious" about terms and objects.

19 - Moreover, this "meaning is use" perspective provides a way to understand the formation of figurative meanings; that is, besides having its literal sense and reference, a term or expression can be used to convey other, extended meanings and nuances. Thus, "white horse" has come to signify the noble steed of a young girl's Prince Charming, and the term "dark horse" to signify the steed of an evil knight, or unknown participants in elections or races. Sometimes the original literal meaning is forgotten, while the figurative meaning remains.

20 - According to Graham, Chuang Tzu thought Kung-sun was "wasting his time [debating whether a white horse is a horse], since all disputation starts with arbitrary acts of naming, he only has to pick something else as the meaning of the word, name something else 'horse', and then for him what the rest of us call a horse would not be a horse" (Chuang Tzu, p. 53).

21 - Graham, Neo-Mohist Logic p. 43.

22 - Thus, I disagree strongly with Hansen's view that "Gongsun Long, like ideal-language theorists in the Western tradition, is proposing a language reform. He has an antecedent theory of strict clarity and wants to rectify ordinary language. He requires language to conform to a general principle of strict clarity" (A Daoist Theory, p. 258).

23 - Watson, Chuang Tzu, p. 39.

24 - Chang Chung-yuan, trans. and comp., Tao: A New Way of Thinking — A Translation of the Tao Te Ching with an Introduction and Commentaries (New York: Harper, 1975), p. 3. See pp. 4-5 for Chang's commentary. (Translator's note: The meaning of this chapter varies a great deal with the different punctuations adopted by commentators. This translation follows the traditional punctuation, which is most natural to the authentic Chinese literary style. However, in the most important sentence of this chapter, a full stop is used after the word tung, or identity. With this punctuation, the sentence reads, "its wonder and its manifestations are one and the same." The sentence thus carries the meaning of the middle way philosophy, according to which reality and appearance are identified.)

25 - Paraphrasing Chuang Tzu, Graham asks, "Then why go to the trouble of arguing with Kung-sun Lung over the problems of the white horse . . .? From the ultimate viewpoint, where dividing has not yet begun, the cosmos is the horse and there are no divisions of it to be horses. . . . ". . . Rather than use the horse to show that 'a horse is not a horse' use what is not a horse" (Disputers of the Tao, pp. 179-180). Chuang Tzu, in chap. 2, uses the idea of the music of Heaven (nature) encompassing the music of man and the music of the earth to illustrate the idea of original identity (see Watson, Chuang Tzu, pp. 37-38). Chang Chung-yuan provides the commentary of the seventeenth-century scholar Yao-Nai: "To a man who has achieved the Self of Non-Self, all music, whether from pipes or flutes or the wind through nature's apertures, is Heavenly music. But to the man who has not achieved this Non-Self, these sounds are still heard as the Music of Man and the Music of Earth" (Creativity and Taoism: A Study of Chinese Philosophy, Art, and Poetry [Harper and Row, 1963], pp. 110-111).

26 - For discussion, see Chang, Tao: A New Way, pp. xv-xvii, and Creativity and Taoism, pp. 33-37 and 96-105; and Graham, Chuang Tzu, pp. 20-22. The respective titles of Kung-sun and Chuang Tzu's essays are suggestive in this regard: "Treatise on Signifying Things" and "Treatise on Making Things Equal."

Further Readings

Chung-ying Cheng: On Zen (Ch’an) Language and Zen Paradoxes

this essay inquires into the logical and semantical significances of the dialogic exchanges (kung-an, koan) in Zen language and discourse as well as to clarify their methodological and ontological basis…In what logically intelligible way does a puzzle or a paradox as generated in a dialogic exchange derive its extraordinary meaningfulness as a tool for reaching or revealing the ultimate truth?…How is the paradoxicality or puzzlement of such a puzzle or paradox to be rationally explained and logically dissolved?

Philip Goodchild : Speech And Silence In The Mumokan: And Examination of the Use of Language In Light of the Philosophy of Gilles Deleuze 

This essay attempts to "extract interpretative methods and concepts from Deleuze's work in order to make use of them in observing exactly how language is used in the Mumonkan" — a language-based interpretation of koans.

Jin Y. Park: Zen Language in Our Times: the case of Pojo Chinul's huatau meditation 

Park discuss the role of language in huatau meditation, focusing on Korean Zen Buddhism.

James D. Sellmann & Hans Julius Schneider: Liberating Language in Linji and Wittgenstein

The aim in this paper is to explicate some unexpected and striking similarities and equally important differences, which have not been discussed in the literature, between Wittgenstein's methodology and the approach of Chinese Chan or Japanese Zen Buddhism.

Dale S. Wright: Rethinking Transcendence: The Role of Language in Zen Experience

 Wright questions whether enlightenment "stands altogether beyond the shaping power of language and culture". He also looks at the role language played in the origins and development of the monsastic community, a community that made the Zen experience of awakening possible.

Desheng Zong: Three Language-Related Methods In Early Chinese Chan Buddhism 

The primary concern of this essay is the history and philosophical significance of three language-related methods widely used in Chan practice during the golden age of Chinese Chan Buddhism, roughly from the eighth to the twelfth centuries. Zong looks at "the Bodhidharma Method", "the naming game" and "the four ways of Ju and Yi".