Chinese Logic

views updated


Systematic argument in Chinese philosophy began with the Moist school, founded in the fifth century BCE by the first anti-Confucian thinker, Mozi (c. 470c. 391 BCE). He laid down three tests for the validity of a doctrine: ancient authority, common observation, and practical effect. At first the controversies of the various schools over moral and political principles led to increasing rigor in argument; then to an interest in dialectic for its own sake, as evidenced in Hui Shih's paradoxes of infinity and in Kung-sun Lung's sophism "A (white) horse is not a horse"; and still later to the antirationalism of the Daoist Zhuangzi (born c. 369 BCE), who rejected all dialectic on the grounds that names have only an arbitrary connection with objects and that any point of view is right for those who accept the choice of names it assumes.

Logic of Moism

In the third century BCE the Moists responded to Zhuangzi's skepticism by systematizing dialectic in the "Moist Canons" and the slightly later Ta-ch'ü and Hsiao-ch'ü.

"moist canons."

The "Canons" confined dialectic to questions of the form "Is it this or is it not?" or, since they assumed that the proposition is merely a complex name for a complex object, "Is it or is it not the case that?" (The form is distinguished in Chinese by a verbless sentence with a final particle, not by a verb "to be.") In true dialectic the alternatives are paired ("Is it an ox or not?") so that one and only one fits the object. Dialectic excludes such questions as "Is it an ox or a horse?" (it may be neither) and "Is it a puppy or a dog?" (it may be both). Its solutions are absolutely right or wrong; being or not being "this," unlike being long or short, is not a matter of degree, since nothing is more "this" than this is. The Moists further argued that it is self-contradictory to deny or to affirm all propositions: the statement "All statements are mistaken" implies that it is itself mistaken, and one cannot "reject rejection" without refusing to reject one's own rejection.

Names are of three types, distinguished by their relations to "objects," which are assumed to be particular. "Unrestricted" names (such as "thing") apply to every object. Names "of kinds" (such as "horse") apply to every object resembling the one in question. "Private" names (for example, the proper name "Tsang") apply to one object. Whether a name fits an object is decided by appeal to a "standard." There may be more than one standard for an object; for "circle" the standard may be a circle, one's mental picture of a circle, or a compass. Some standards fit without qualification: A circle has no straight lines. Some fit only partially: In deciding whether someone is a "black man" it is not enough to point out his black eyes and hair. The "Canons" began with seventy-five definitions, evidently offered as "standards," of moral, psychological, geometrical, and occasionally logical terms. An example of a definition of a logical term is "'All' is 'none not so'" (supplemented in the Hsiao-ch'ü by "'Some' is 'not all'"). The first of the series is "The 'cause' is what is required for something to happen." ("Minor cause : With this it will not necessarily be so; without this it necessarily will not be so. Major cause : With this it will necessarily be so.") The "Canons" also distinguish the senses of twelve ambiguous terms. Thus, "same" is (1) identical ("two names for one object"), (2) belonging to one body, (3) together, and (4) of a kind ("the same in some respects").

"ta-ch'Ü" and "hsiao-ch'Ü."

The Moist Ta-ch'ü further refined the classification of names. Names indicating "number and measure" cease to apply when their objects are reduced in size; when a white stone is broken up it ceases to be "big," although it is still "white." Names indicating "residence and migration" do not apply when the population moves, as in the case of names of particular states ("Ch'i") or of kinds of administrative divisions ("country"). The claim that one knows X only if one knows that an object is X applies only to names indicating "shape and appearance" ("mountain," but not "Ch'i" or "county").

The Ta-ch'ü, and still more the Hsiao-ch'ü, also showed a shift of interest from the name to the sentence and to the deduction of one sentence from another. The Chinese never analyzed deductive forms, but the Moists noticed that the formal parallelism of sentences does not necessarily entitle us to infer from one in the same way as from another, and they developed a procedure for testing parallelism by the addition or substitution of words. For example, "Asking about a man's illness is asking about the man," but "Disliking the man's illness is not disliking the man"; "The ghost of a man is not a man," but "The ghost of my brother is my brother." In order to reconcile the execution of robbers with love for all men some Moists maintained that although a robber is a man, "killing robbers is not killing men." Enemies of Moism rejected this as sophistry, on the assumption that one can argue from "A robber is a man" to "Killing robbers is killing men," just as one can argue from "A white horse is a horse" to "Riding white horses is riding horses." The Hsiao-ch'ü replied that there are second and third sentence types of the same form, which do not allow such an inferencefor example, "Her brother is a handsome man," but "Loving her brother is not loving a handsome man"; "Cockfights are not cocks," but "Having a taste for cockfights is having a taste for cocks." A four-stage procedure was used to establish that "A robber is a man" belongs to the second type:

(1) Illustrating the topic ("robber") with things ("brother," "boat") of which formally similar statements may be made.

(2) Matching parallel sentences about the illustrations and the topicfor instance, "Her brother is a handsome man, but loving her brother is not loving a handsome man"; "A boat is wood, but entering a boat is not entering wood"; "A robber is a man, but abounding in robbers is not abounding in men, nor is being without robbers being without men."

(3) Adducing supporting arguments for the last and most relevant parallels by expanding them and showing that the parallelism still holds: "Disliking the abundance of robbers is not disliking the abundance of men; wishing to be without robbers is not wishing to be without men."

(4) Inferring, defined as "using its [the topic's] similarity to what he [the person being argued with] accepts in order to propose what he does not accept": "Although a robber is a man, loving robbers is not loving men, not loving robbers is not not loving men, and killing robbers is not killing men."


Outside the Moist school only the Confucian Xunzi (c. 313c. 238 BCE) left a consecutive treatise on logical questions. According to his "Correct Use of Names" the purpose of names is to point out objects, thereby distinguishing the noble from the base and the similar from the different. Names are fixed by convention and are mutable, but to use them idiosyncratically when their usage is fixed is a crime akin to falsifying weights and measures. Objects are different if they differ in place although not in form; they remain the same if they change in form without dividing. Objects of the same kind are perceived by the senses as similar and are given the same name. Names may be of any degree of generality; we may assimilate objects under the name "thing" or distinguish them as "bird" and "beast." (Like the Moists, Xunzi took for granted a nominalist position.) The sentence is a series of names conveying one idea, and a name is understood when we grasp both the object to which it points and its interconnections in the sentence.

Xunzi distinguished three sorts of fallacies, which he illustrated with unexplained examples (two are explained by his refutations of them in his "Treatise of Corrections"). Fallacies that abuse names are exposed by an appeal to established usage, and fallacies that abuse objects are exposed by an appeal to the evidence of the senses. The first fallacy, "confusing names by misuse of names," Xunzi illustrated by "To be insulted is not disgraceful." This is a violation of the established use of "disgrace" in two senses, for social and for moral degradation. The second fallacy, "confusing names by misuse of objects," was exemplified by "Our genuine desires are few." Xunzi criticized this as a factual error about humankind. The third fallacy is "confusing objects by misuse of names." Kung-sun Lung (born 380 BCE) had defended the sophism "A (white) horse is not a horse" on grounds which assume that the question is one of identity, not one of class membership. Xunzi would presumably have replied simply that a white horse is commonly called a "horse."

Later Logical Thought

The classical period of Chinese philosophy ended about 200 BCE. The next important movement, the neo-Daoism of the third and fourth centuries CE, revived the study of the sophists and the Moist "Canons." Indian treatises on logic were available in translation from the seventh century on; Buddhists wrote commentaries on them during the Tang dynasty (618907), and in Japan they have continued to do so. But there is little evidence of progress by either Daoists or Buddhists. Neo-Confucianism, the main philosophical movement after the Song dynasty (9601279), entirely neglected logical inquiry.

Chinese Neglect of Logic

It is well known that almost all Chinese philosophical "systems" are practical, moral, or mystical philosophies of life, indifferent to abstract speculation. It is therefore not surprising that Chinese thinkers have cared little for the forms of reasoning, except under the pressure of the acute controversies of the third century BCE. What is surprising is the almost exclusive interest of Chinese philosophers in the problem of names and the fact that even those who advanced from the name to the sentence studied the parallelism of sentences rather than their analysis.

A reason for this interest can be found in the Chinese language, which organizes uninflected words solely according to word order and the placing of particles. Without the inflections that expose the structure of Sanskrit, Greek, or Arabic sentences and encourage the simultaneous growth of grammar and logic the Chinese sentence, until recently, almost defied analysis; the Chinese have been lexicographers but not grammarians. On the other hand, strict parallelism of clausesin which noun is matched with noun, adjective with adjective, adverb with adverb, verb with verbis part of the ordinary resources of the Chinese language and easily calls attention to the logical dangers of formal parallelism.

See also Chinese Philosophy; Gongsun Long; Hui Shi; Mozi; Proper Names and Descriptions; Xunzi; Zhuangzi.


chinese logic

Chao, Y. R. "Notes on Chinese Grammar and Logic." Philosophy East and West 20 (2) (1970): 137154.

Cheng Chung-ying. "On Implication (tse ) and Inference (ku ) in Chinese Grammar and Chinese Logic." Journal of Chinese Philosophy 2 (1975): 225244.

Chmielewski, Janusz. "Notes on Early Chinese Logic." Rocznik Orientalistyczny 26 (1) (1962): 721; 26 (2) (1963): 91105; 27 (1) (1963): 103121; 28 (2) (1965): 87111; 29 (2) (1965): 117138; 30 (1) (1966): 3152; 31 (1) (1968): 117136; 32 (2) (1969): 83103.

Graham, Angus C. Disputers of the Tao. La Salle, IL: Open Court, 1989.

Graham, Angus C. Later Mohist Logic, Ethics and Science. Hong Kong: Chinese University Press, 1978; reprinted, 2003.

Hansen, Chad. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press, 1983.

Harbsmeier, Christoph. Science and Civilisation in China, Vol. 7: The Social Background, Part 1: Language and Logic in Traditional China. Cambridge, U.K.: Cambridge University Press, 1998.

Hu Shi. The Development of the Logical Method in Ancient China. Shanghai: Oriental, 1922.

Lau, D. C. "On Mencius' Use of the Method of Analogy in Argument." Asia Major 10 (1963). Reprinted in D. C. Lau, tr. Mencius. London: Penguin, 1970, pp. 235263.

Lau, D. C. "Some Logical Problems in Ancient China." PAS, n.s. 53 (19521953): 189204.

Lenk, Hans, and Gregor Paul, eds. Epistemological Issues in Classical Chinese Philosophy. Albany: State University of New York Press, 1993.

Leslie, Donald. Argument by Contradiction in Pre-Buddhist Chinese Reasoning. Canberra: Australian National University, 1964.

Reding, Jean-Paul. Les fondements philosophiques de la rhétorique chez les sophistes grecs et chez les sophistes chinois. Bern: Peter Lang, 1985.

A. C. Graham (1967)

Bibliography updated by Huichieh Loy (2005)

About this article

Chinese Logic

Updated About content Print Article