Right Words Seem Wrong: Neglected in Early Chinese Philosophical Texts Author(s): Wim Reu Source: Philosophy East and West, Vol. 56, No. 2 (Apr., 2006), pp. 281-300 Published by: University of Hawai'i Press Stable URL: http://www.jstor.org/stable/4488018 Accessed: 17/10/2010 05:18

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=uhp.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

University of Hawai'i Press is collaborating with JSTOR to digitize, preserve and extend access to Philosophy East and West.

http://www.jstor.org RIGHT WORDS SEEM WRONG: NEGLECTED PARADOXES IN EARLYCHINESE PHILOSOPHICAL TEXTS

WimDe Reu,Assistant Professor Departmentof Philosophy,National Taiwan University

Well-knownversus Neglected Paradoxes

Almost all well-known early Chinese paradoxescan be found in a mere three chap- ters of literature.They appear in short lists compiled by intellectualopponents. First, the chapter "Tianxia"~T (Under heaven) of the E- gives a description of the philosophiesof some WarringStates thinkers and ends with a discussionof the thoughtof . The authorof this chapterenumerates ten paradoxicalstate- ments ascribed to Hui Shi, followed by a list of twenty-one paradoxes used by the bianzhe * (disputers)in debate with Hui Shi. Further,the chapter "Zhengming" IE? (Rectifyingnames) of the XunziAi7F presentsa threefoldclassification of para- doxes. It includes, among others, paradoxes propounded by the LaterMohists as well as by Song Xing ~fF, Hui Shi, and Long /[~.' Finally,another chapter of the , "Bugou" (Nothing indecorous),attributes a short list of five paradoxesto Hui Shi and Deng Xi (fi.2 Many paradoxesthat are included in these lists have been frequentlydiscussed and can be found in any majoroutline of early Chinese thought.3Nevertheless, in spite of their popularity,our understandingof them is complicated by a lack of con- textual information.To begin with, in the three chaptersmentioned above, the para- doxes are presented out of context without much guide to interpretation.Further, while the context of the paradoxesmust have been generally known in the Warring Statesperiod, most of it has not come down to us and had probablydisappeared at an early stage in history.Finally, even when somethingof the context is left, as is the case with bai ma fei ma (a white horse is not a Gongsun Long's l ,%#,% horse), there is no consensus as to its interpretation.Reconstructing the reasoning behind the paradoxes is thereforeproblematic and based mainly on indirectevidence and on our own creativity.4 Given the difficultiesof interpretation,one wonders why these paradoxes have attractedso much attention.Aside from the intellectualappeal of virtuallyany para- dox, two explanations in particularneed to be mentioned. One is that some of the paradoxesconstitute an importantpart of what we know about the ideas of certain thinkers.This is the case not only with, for example, Gongsun Long, but also with Song Xingand Hui Shi, none of whose writingshave survivedindependently. Study- ing the thought of these masters largely involves interpretingthe paradoxes they have left to posterity.Another, perhaps more straightforwardexplanation is that the paradoxes appear in ready-made lists. Being conveniently placed together rather than scatteredthroughout various works, they form an invitingresearch topic: there

PhilosophyEast & West Volume 56, Number2 April2006 281-300 281 ? 2006 by Universityof Hawai'iPress Table 1

Group 1A 7 X t Y ("X" +-* "Y") Group 1B X Y t Z ("X"l-"Z")

"the greatestskill seems clumsy" "the brightestway seems dark"

"the greatestsuccess seems deficient" "the most solid virtueseems weak"

is no need for the preparatorystage of searching for paradoxes in a multitudeof texts. Though this has resulted in intense scrutinyof the well-known paradoxes, it has also prevented many scholars from going beyond them. In this sense, the ready-madelists also constitutea barrierto furtherresearch into paradoxes. There is no reason to suppose that the numberof ancient Chinese paradoxes is limited to the statementsfound in the ready-madelists. Scrutinizingthe early philo- sophical writings,one readily discovers that quite a few paradoxes appear outside these lists. In contrastto their more famous counterparts,these paradoxesare often found embedded in their originalcontexts. While this contextual informationopens the way to a better-foundedinterpretation, up to now the paradoxes have been marginalizedand treated in an unsystematicway that does not take into account the immediatecontexts in which the paradoxesare uttered.As a result,these para- doxes constitutea relativelynew field of study. The present article makes a firstat- tempt in exploringthis field of neglected paradoxes.

ThreeGroups of Paradoxes

The paradoxes introduced in this article generally consist of four characters.Al- though they frequentlyappear in series, they are unlike the paradoxes in the ready- made lists in that they are scatteredthroughout the literatureand are writtendown not by philosophical rivalsbut by people who employ them in an affirmativeway. Some of the more familiarparadoxes are -kr5jt (the greatestskill seems clumsy), (the highest virtue is not virtuous),and (the speechless teaching). OnjW7KT~, the basis of semantic criteria,it is possible to distinguish`?--_ three main groups. In addition, each group can furtherbe divided into two subgroups.5The paradoxical or counterintuitivenature of the paradoxesresults from a challenge to the semantic relationsthat hold between their centralterms. As can be seen fromtable 1, paradoxesthat belong to the firstgroup are charac- terized by the semantic relation of antonymy.6According to our conventional se- mantic associations, qiao r5 (skill)is antonymouswith zhuo tdf(clumsy), and ming HA(bright) is antonymous with mei H (dark, dull). The subgroup to which these paradoxesbelong is determinedby the position of their antonyms.The antonyms in subgroup 1A come in the second and fourthposition, while those in subgroup 1B appear in positions one and four. The paradoxicaleffect of the expressionsderives

282 PhilosophyEast & West Table 2

Group 2A 7k X T X ("X" = "X") Group 2B T X ? X ("X".= X")

"the greatestbenevolence is not benevolent" "the shapeless shape"

"the highestvirtue is not virtuous" "the knowledgethat does not know"

Table 3

- - Group 3A 7k X T- Y ("X" "Y") Group 3B 7T\X Z Y ("Y" "X")

"the greatestcarving does not cut" "the speechless teaching"

"the greatesthero does not fight" "the speechless disputation" mainly from the fact that the verb ruo (to be like, to seem) suggests that the ant- - onyms are similar,though not identical, in meaning. Paradoxesthat belong to the second group are typifiedby identity(table 2). The graphic identity of the central terms leads to the assumptionof semantic identity. Differencesin structurejustify further division into two subgroups.Paradoxes of sub- group2A are independentstatements, while those of subgroup2B generallyfunction as the object of a precedingverb. In both subgroups,the identityof the centralterms is denied by a negation.7This resultsin paradoxicalsayings. Paradoxesthat belong to the thirdgroup are characterizedby implication(table 3). The termzhi $1J(carving) conventionally implies ge 9J (to cut), and jiao t (teach- ` ing) is commonly thoughtof as implyingyan (to speak). With regardto the para- doxes (the greatest carving does not cut) and (the speechless 7k•->iJ T-ZOf disputation),the relationof implicationis furtherhighlighted by the common graphic element of the central terms. The division into subgroupsis the same as with para- doxes of the second group. The paradoxicaleffect of the expressions resultsfrom negatingthe implication. Though the paradoxescan be assigned to differentgroups and subgroups,they constitutea well-definedand interrelatedset of expressions.The series in which they typically appear do not include other kinds of paradoxes,suggesting that they form a closed unit. As to their internalstructure, they are related in two ways. First,para- doxes that belong to differentgroups or subgroups sometimes appear in a single series. PT 41, for example, contains no less than twelve paradoxesthat be- long to two main groups and three differentsubgroups (1A, 1B, and 3A).8 Second, paradoxesfrom one group may be rewrittenas paradoxesfrom anothergroup. This

Wim De Reu 283 is evident from Zhuangzi 2/5/26-6/3, which first lists five paradoxes, among them (the does not and (the ~~i greatestway designate) CMT- greatestdisputation is and then these two in reverseorder as speechless),9 rephrases paradoxes T-,? (the speechless disputation)and T-,ZL (the way that does not lead the way).10 This passage indicates that paradoxesfrom groups two and three are interrelatedin a logical way. Furthermore,the paradoxesA -T-,`-and 2,? both dealing with disputation, have a counterpartin the •7 (the greatest disputation seems stuttering)from group one."1 This internalcohesion as well as the fact that the paradoxesare clearly delineated calls for a unified interpretation.

Distributionand Affiliation

In searching for paradoxes, I have examined the philosophical literaturefrom the beginning of the WarringStates period down to the end of the firstcentury of Han rule (roughly500-100 B.C.).12The search produced around 100 differentparadoxes, accounting for a total numberof approximately180 occurrences.13 This more than doubles the numberof differentparadoxes found in the ready-madelists.14 Two majorobservations can be made on the basis of the distributionof the para- doxes.15 First,the paradoxesmostly appear in writingsthat are traditionallyclassified as "Daoist."Within this group, the Zhuangzi, and even more so the Laozi, occupy prominentplaces: they constitutethe most importantsources of paradoxes.Second, outside "Daoist"works, the paradoxesmostly appear in so-called "Syncretic"writ- ings, and are occasionally also found in other categoriesof texts. Moreover,in these writings,the relevantpassages can in a substantialnumber of cases be relatedto the Laozi and the Zhuangzi.16 Though there was strictlyspeaking no Daoist "school," these findings suggest that the paradoxes constituted a typical mode of expression among thinkers we could call "Daoist."As paradoxes sometimes also appear in passages that do not show any apparent evidence of Daoist influence, it is unwarrantedto make the strongerclaim that the paradoxes are exclusively Daoist in . Nevertheless, even in these latter cases, the most plausible scenario is that the paradoxes were taken out of a Daoist context or composed in imitationof Daoist models.

Interpretingthe Paradoxes

The paradoxeschallenge the semantic relationsof antonymy, identity,and implica- tion: on the one hand, termsthat are supposed to be opposite in meaningare brought closer together,and, on the other hand, termsthat are characterizedby identityand implicationare drawn apart.This gives rise to two basic questions. First,what is the mechanism underlyingthese paradoxes?Second, what is the motive in formulating them? These questions can be approached from different angles. Some scholars have presented philosophical and religious interpretations, approaching the paradoxes from dialectics, , and mysticism.17 While valuable, these interpretations tend to ignore the immediate contexts of the paradoxes. This is unfortunate, for it is pre-

284 PhilosophyEast & West cisely this context, considered in the light of broaderobservations on language and communication,that allows for an alternativeinterpretation. This interpretationholds that the paradoxesconstitute unorthodoxredefinitions of importantterms, and that they are formulatedin order to influence the behavior and values of their intended audience. In developing this interpretation,I will first make some observationson the features and use of importantterms as a general interpretativebackground. Then I will explain the underlyingsemantic mechanism, and discuss the pragmatic and rhetoricalfunctions of the paradoxes.

The Background:Important Terms Importantterms denote things or ideas that are relevantin an aesthetic, philosophi- cal, political, or broadly culturalcontext. They often have a laudatoryconnotation and are frequentlyshared among differentgroups or subcultures.Typical examples of these importantterms include 'art,''freedom,' 'courage,' 'justice,' 'love,' 'genius,' and 'gentleman.'18 It is frequentlythe case that importantterms are interpretedin differentways. While 'art,'for example, is widely used across subculturalboundaries, its interpreta- tion varies with the intuitionsand theoreticalframeworks of differentgroups. What one group means by 'art'(the answerto the question "What is art?")differs from the views of other groups, even to the extent that an object that is labeled 'art' by one group will not necessarilybe deemed 'art' by another.Though different subcultures share a common vocabularyof importantterms, there does not appear to be a cor- respondingunity of interpretation. This differencein interpretation,when combined with the laudatoryconnotation of importantterms, frequently results in contestationand controversyabout the cor- rect use of a term.Terms such as 'art,''love,' and 'justice'are consideredto stand for valued things and ideas, acquire a positive connotation, and become themselves valued and coveted descriptions.Describing some act as 'an act of justice' gives an added value to this act, while the opposite, labelingsomething 'unjust,' condemns it. In a conflict between differentgroups, it then becomes importantto impose one's own criteriaof how a term should be used. As a result,terms become contested, an idea originallyput forwardby W. B. Gallie, who described the contested character of 'art,''religion,' 'democracy,' and 'justice'as "inevitablyinvolving endless disputes about their properuses on the partof their users"(Gallie 1968, p. 158). In disputes about the properuse of importantterms, the core membersof each group stick to their interpretationand try to persuade wavering members of other groups of their point of view. Charles Leslie Stevenson was the first to recognize that argumentsin favor of a particularview often take the form of persuasivedefini- tions. He claimed that the descriptivemeaning of terms such as 'culture,''courage,' 'justice,' and 'philosophy' is constantly being redefined "with the conscious or un- conscious purpose of changing ... the direction of people's interests" (Stevenson 1938, p. 331). In his view, such persuasive definitions should not be regarded (but frequently are) as pure intellectual analyses or as descriptions of linguistic habits, but as normative, attitude-shaping valuations.

Wim De Reu 285 Stevenson also noticed that persuasive definition frequently takes place by putting modifiers such as 'real' and 'true' in front of importantterms. For exam- ple, we are told that "real culture means imaginativesensitivity" or that "true love is the communion between minds alone" (Stevenson 1938, pp. 331-332, 334). These modifiers have two major interrelatedfunctions. First,they add to the per- suasive force of the definition. The modifier attractsattention and purportsto in- troduce something of the utmost importance.Second, they signal a change in de- scriptive content. Soren Hallden has claimed that the use of modifiers such as 'true,' 'real,' and also 'genuine' is almost invariablyaccompanied by a change in the field of applicationof a term. 'Real culture' is set against some 'superficial'con- ception of culture, and 'true love' implies and is distinguishedfrom 'ordinary'or 'shallow love.' Compared to the ordinary meaning of a term, this change in de- scriptive meaning can take differentdirections: the scope of a modified term can be broader, narrower,or lie outside that of its ordinaryusage (Hallden 1960, pp. 55-58). In sum, when importantterms such as 'humor,' 'freedom,' 'philosophy,' and 'courage' are shared among differentsubcultures, they are likely to be used with mutuallycompeting interpretationscarrying different scopes and applications.In an attemptto influence the discourse, differentgroups frequentlyredefine these terms, and often do so by makinguse of persuasivemodifiers.

The UnderlyingMechanism The observationsabove on importantterms allow us to interpretthe paradoxes.War- ringStates philosophical literature abounds with importantterminology that is shared among differentsubcultures: masters of differentlineage argue about the properdao L (Way)to ordersociety, about what constituteszhi ?a (knowledge,wisdom), about who can trulybe considered de {t (virtuous),about the meaning of I (righteous- ness), and so on. Differentgroups of people place differentand competing interpre- tations on these terms.The centralterms of the paradoxesare also usually important, either in (e.g., 2f [benevolence]), in (e.g., dao de jt), in politics (e.g., $i1 [to carve, to govern]),or in a general culturalway L_, (e.g., yong A [bravery], cheng hiZ[success]). We noted that importantterms are typically used in differentways by different groups of people. An importantterm is, as it were, the center of a web aroundwhich differentinterpretations are positioned. As a result, attempts at defining important terms invariablyinvolve at least two levels of interpretation:first, the target inter- pretation(s)against which some definition is directed and, second, the new or al- ternativeinterpretation introduced in the definition.Though both levels of interpre- tation are important,they are not equally prominent.The targetinterpretation usually lies in the background. From the definition of 'culture' as "imaginative sensitivity," for example, one does not readily identify the target use "a man widely read and acquainted with the arts."'9 While this target might be mentioned in the immediate context, the definition as such does not provide any hint of this. In other cases, the target interpretation may be vaguely implied, especially when seemingly insignifi-

286 PhilosophyEast & West cant termssuch as 'also,' 'only,' or 'alone' are used. In the definitionof 'love' as "the communion between minds alone," the term 'alone' suggeststhat the targetis a view on love that includes mental as well as physical aspects. Nevertheless, even in such cases, the target interpretationlies in the backgroundand can only be known indirectly. The relativeinvisibility of the targetinterpretation is usually compensated for by a focus on the new or alternativeinterpretation. In the two examples above, we ob- tain more or less straightforwardinformation on the alternativeuse of 'culture'and 'love.' We know what someone means when he says that true love is to be found in mutualrespect and understanding.This focus on the alternativeinterpretation corre- sponds to what we implicitlyexpect of a definition.We expect it to informus about the meaning of a thing or term. This transferof informationis best accomplished by focusing on the new interpretation. Unlike ordinaryredefinition, however, the neglected paradoxes do not present direct informationon the new of terms. The interpretations important paradox_L T-, (the highest virtue is not virtuous),for example, does not make very clear what is meant by de virtue).The new of de is introducedin shang _L (highest meaning an indirect way, not by giving a straightforwarddescription, but by saying that it is differentfrom the ordinaryconception of de. Similarly,the paradox krJ1Tf (the greatestskill seems clumsy) only indirectlyconveys the new interpretationof qiao (skill). It tells us that instances of the new interpretationwould be judged 'clumsy' or 'not skillful'from the standpointof the target interpretation.In other words, the paradoxes suggest new and alternativeinterpretations of importantterms, but they do so by referringback to the targetuses of these terms.When comparedto ordinary redefinition,this unorthodoxtechnique results in a lack of information,which we can only hope to fill in by contextualanalysis. Accordingto the presentationabove, the authorsof the paradoxescombine both the targetand alternativeinterpretations of importantterms into a single expression. While it is likely that some masterscopied paradoxessolely for their literaryvalue, the use of paradoxesseems at least in a numberof cases to have been a conscious play on the multiple interpretationsof importantterms. Throughout the extant litera- ture we find passages where mastersexplicitly contrasttheir own views with those of others and subsequently present the paradoxes as a result of this opposition. A passage on le V (happiness)appearing in the Zhuangzi illustratesthis very well. After having observed that the common people find happiness in such things as wealth, old age, comfort, rich flavors, and beautifulclothing, the author continues as follows:

What today's ordinarypeople do and where they find happiness, once more I do not know whethertheir happinessis reallyhappiness or not.... Is there afterall really happinessor not? I take non-actionas true happiness,and yet ordinarypeople consider it greatlydistasteful. Therefore I say: "Ultimatehappiness is without happiness, ultimate praise is without praise."(Zhuangzi 18/47/29-48/2)

Wim De Reu 287 This passage contains two mutually exclusive interpretationsof 'happiness.' They correspond to the views of differentsubcultures. On the one hand, there is the vulgar conception of happiness (such as enjoying wealth and comfort)that is held by ordinarypeople. This is the target interpretation.On the other hand, the author considers himselfthe representativeof a higher ideal that rejectsthis vulgarinterpre- tation. The alternativehe proposes aims at attaininghappiness through an attitude of non-action. The second-to-last last sentence provides insight into the formation of paradoxes.Two interpretations(target and alternative)are taken togetherto form a paradoxicalstatement: what constitutestrue or ultimatehappiness for the authorof this passage lacks happinesswhen judgedfrom the point of view of the common peo- ple. This paradoxon happiness,and by extension also the paradoxon praise,rests on a play with the various interpretationsthat subculturesplace on importantterms. The combination of interpretationsand points of view can also be observed in Laozi 41, which contains a long series of twelve paradoxes.Immediately preceding the series, the authordescribes three types of scholars who adopt differentattitudes toward the Way:

Whena superiorscholar hears about the Way, he is diligentin practicingit; when an averagescholar hears about it, he wavers;when an inferiorscholar hears about it, he sneers.If he did notsneer, it wouldbe insufficientto be regardedas the Way.Therefore, the establishedsayings contain the followingwords: "The brightest way seemsdark,... the greatestwhite seems stained, ... the greatestsquare does not have corners."(Laozi 41/14/17-15/1)

The paradoxesare presentedas a resultof the shiftingbetween two differentpoints of view, that is, that of the superiorscholar and that of the inferiorscholar. What for the superiorscholar is brightand white looks dark(not bright)and stained (not white) from the perspectiveof the inferiorscholar. These two groups of scholars adopt op- posite interpretationsof the terms ming Ra(bright) and bai n (white),though, at least from this passage, it is unclear what these interpretationsare.20 In a similar way, what the superiorscholar considers to be the greatestsquare, that is, the most effec- tive way of governing,would be laughedat by the inferiorscholar for having no cor- ners, that is, for having no retributivesystem of punishment.21 The characterization wu yu ER (not having corners) refersback to the commonsense interpretationof fang 1Y (square),according to which governmentshave strict laws and correspond- ing sanctions. Here again, the targetand alternativeinterpretations are combined to form a paradoxicalstatement. As a last example illustratingthe underlyingmechanism, consider the following passage from the Lishi chunqiu:

Conduct has to be well considered. If it is ill considered, it is like runningtoward a deep gorge: you may regretit afterwards,but it will be of no avail. The gentleman,when con-

288 PhilosophyEast & West sideringhis conduct,thinks about his moralduty; the pettyman, when consideringhis conduct,is focusedon profit,which is not profitable.One can only talkabout the right patternwith someone who knowsthe unprofitableprofit. (LOshi chunqiu 22/143/29-30)

Justas in the two previouspassages, we see here again the contrastbetween different groups of people. The petty man thinks in terms of profit,probably conceived of as short-termmaterial benefit. After having pointed out that this, in the end, does not lead to any real profit,the authorextends the term fJ (profit)to include the moral conduct of the gentleman.At the same time, however, this new conception of 'profit' is distinguishedfrom the vulgar interpretation,and is then describedas 'unprofitable profit.'This paradoxicalsaying, too, plays on the differentinterpretations of an im- portantterm. The paradoxes introduce new interpretationsof importantterms, but they do so in an unorthodoxmanner by referringback to the target uses of those terms. As pointed out, importantterms are typically interpretedin differentand competing ways. In combining differentinterpretations into a single expression, the authorsof the paradoxesindicate their awareness of the semantic instabilityof importantterms. They play on the potential for multiple interpretations.This play on interpretation enables them to formulateparadoxes.22 It also explains the challenge to the standard semantic relations.When considered from the perspective of differentsubcultures, antonyms can be similar in meaning, identical terms can have differentuses, and implicationsbetween terms depend on particularinterpretations that are not univer- sally accepted.

Pragmaticand RhetoricalFunctions The paradoxes play on multiple interpretationsof importantterms. Insightinto the semantically unstable nature of these terms undoubtedly at some point informed the formulationof paradoxes.It would thus seem possible to construethe paradoxes as targetingphilosophical tendencies that presupposea fixed relationbetween termi- nology and interpretation.Explained in such a way, the paradoxeswould performa purely theoretical function. They would serve as an instrumentfor unveiling mis- taken semantic assumptionsabout language and terminology.I do not find it plausi- ble that the paradoxeshad such a theoreticalassumption as their main target.In any case, such a targetis not obvious from the immediatecontext. Initialobservation of the paradoxes rathersuggests that they were formulatedon account of pragmatic and rhetoricalconsiderations. Just like the kind of ordinaryredefinition described above, the centralterms of the paradoxesare not neutralbut are generally important and often have positive connotations.As to these terms,effective redefinitioncauses people to behave and make evaluations in alternativeways. This pragmaticelement is furtherdemonstrated by the presence of persuasivemodifiers such as da t (great) and zhi M (ultimate) in the majority of paradoxes. Giving full recognition to these features of persuasion opens the way for under- standing the challenge to the standard semantic relations as a kind of rhetorical de- vice. From this perspective, the paradoxes are formulated in order to produce an

Wim De Reu 289 effect of surprisewith the ultimatepurpose of transformingand redirectingthe in- tended audience toward a purportedlybetter form of knowledge, behavior, and valuation. This view is in agreement with the general objective of the sayings and writingsof the masters.Most of these sayingsand writingswere composed to instruct disciples or to advise rulerson how to ordertheir state. Ratherthan being theoretical in nature,they were in most cases designed to guide and direct their audience. In such contexts, one expects the use of rhetoricaldevices as a means of influencing people. The paradoxesattract more attentionthan ordinaryredefinition. They catch the reader'seye. However, as indicated,this is only possible at the cost of losing infor- mation. Consideredin isolation,the point of the paradoxes is difficultto grasp.This has consequences for their actual rhetoricalimpact, because in order to realize a change in attitude,the audience should at least have some idea of the message to be conveyed. This implies that the paradoxes should either be directed to people whom one can reasonablyexpect to understandthe message, or, when this is not the case, the context should make the point of the paradoxes sufficientlyclear. When neitherof these two conditions is met, the paradoxesare likely to be ineffective. There are a numberof passages that clarify the pragmaticand rhetoricalfunc- tions of the paradoxes.The most vivid illustrationsare found in dialogical passages, two of which are quoted below. The firstis an example of the effective use of para- doxes. It consists of a conversationbetween Laozi and Yang Ziju R-WJF.The open- ing lines of the story describe Laozi as being disappointed in Yang. It is initially unclear why Laozi is disappointed.A subsequent exchange, which takes place at an inn, clarifiesthe matter:

"Earlieryour disciple wanted to ask you, master,for advice, but you walkedwithout takinga rest,so I did notdare to ask.You are taking a restnow. MayI askabout my mis- take?"Laozi said: "You have a haughtyand disdainful look. Who would hold you com- pany?The greatest white seems stained, the mostabundant virtue seems lacking."Yang Ziju,embarrassed, changed countenance and said: "Respectfully I hear your command." WhenYang Ziju first arrived, the peopleat the inn came out to welcomehim. The inn- keepertook out a mat,and his wife tookout towel and comb.The guests gave up their matsand those who were preparingfood left theirplace at the stove.When he came backfrom his interviewwith Laozi, the guestscompeted with him for a mat.(Zhuangzi 27/80/26-81/4)23

Laozi regardsthe self-importantattitude of Yang Ziju as a source of unpopularity. The point of the story is that while haughtyconduct might result in respect and rev- erence, people will often keep a distance. In order to be treated as an equal, one should instead act in an inconspicuous way. Laozi's instruction occupies the pivotal position. He exhorts Yang Ziju to change his conduct and conveys his message by way of the paradoxes -ES (the greatest white seems stained) and $,••,

290 PhilosophyEast & West (the most abundant virtue seems lacking). These paradoxes do not have a theoretical objective. On the contrary, their goal is to produce a change in attitude. It is interesting to note that Laozi could have made his point in a more direct and informative way. Instead of saying "the greatest white seems stained," he could, for example, have claimed that "the most sociable persons adopt a low-profile attitude." While this characterization would arguably have had the same outcome on Yang Ziju's conduct, it probably would have missed the compactness and startling effect that is typical of paradoxes. The choice for a paradoxical formulation can thus be explained on stylistic and rhetorical grounds. Moreover, the loss of information is, at least in this passage, compensated by the context. The interpretation of the central terms bai and de shifts from "self-important attitude" (target interpretation) to "low- profile conduct" (alternative interpretation). The conversation between Laozi and Yang Ziju focuses more on the social than on the political aspects of behavior. By contrast, the next passage is more political in nature. It deals with conspiracy and sharing secrets. The Duke of Bai ?i (d. 479) considers usurping the throne of his native state of Chu V. According to tradition, he consults for advice. The conversation centers on the question whether it is possible to share secret plans without one's opponents finding out. The issue is phrased in terms of wei yan -- (speaking in subtle ways):

nZ-#A? AH ?p Wf_Tl?j YL#H: JfI•p-L q-_-1M -*L-L? r•ju,•7J'

The Duke of Bai asked Confucius:"Is it possible to speak with people in a subtle way?" Confuciusdid not respond, and the Duke of Bai asked: "What about throwinga stone into water?"Confucius replied: "Good divers from Wu and Yue would be capable of retrievingit." The Duke of Bai asked: "What about throwingwater into water?"Confu- cius said: "Wouldthe riversZi and Sheng meet, Yi Ya would know the differenceby tast- ing the water."The Duke of Bai asked:"If so, is it then impossibleto speak with people in a subtle way?"Confucius said: "Why should it be impossible!You only have to know the importof words! Those who know the importof words do not speak with words. Those who catch fish get wet, and those who chase animals have to run, but it is not that they enjoy doing this. Therefore,the ultimatespeech does away with speech and the ultimate action is without action. In general, those of shallow knowledge pursue inconsequential things!" The Duke of Bai did not understandthis, and as a result died in the bathing rooms. (Huainanzi12/105/20-26)

Confucius disapproves of the use of language in scheming. He suggests that there will always be someone who is able to sense the real intention behind the words, no matter how subtle and indirect the phrasing. Language, as a result, is not a safe and reliable method in conspiracy. He then changes the interpretation of wei yan in favor of a kind of communication on the level of the import of language, and gives additional impetus to his view by formulating the paradoxes M-••- (the ultimate speech does away with speech) and (the ultimate action is without action). _ .%

Wim De Reu 291 These paradoxesdo not constituteabsolute rejectionsof languageor action. The for- mer paradox,for example, does not mean that we should unconditionallyabandon all language in all circumstances,which is the kind of interpretationthat would arise from consideringthe paradoxesout of context. It has to be viewed in the context of plotting,where it serves the double goal of discouragingthe Duke of Bai from using language and instead urginghim to use another method of communication. While the paradoxesmake this passage more attractiveand provocative,there is also a substantialloss of information.Moreover, in this particularcase, the context is not very revealing.As a result,the paradoxesmiss their rhetoricalobjective, which is indicated by the Duke's death. The advice of Confucius may neverthelessbe un- derstood by looking at the chapter "Jingyu" Mrpii(Profound communication) of the Lfishichunqiu. In addition to the conversation between the Duke of Bai and Confucius,this chapter also contains another story on conspiracy. It argues that the founders of the Zhou house of rulers were able to overthrow the Shang because they shared similaraims and did not need to put their plans into words. As a result, they could not be accused of plotting.The authorof the storyuses paradoxicalstate- to describe an intuitive of ments, including--T 2 (the speechless scheming), way understandingthat is typical of like-mindedconspiratorial people. Viewed from the perspective of this story, Confucius' advice to the Duke of Bai is to conspire only with people whom he knows already have the same intention. This would allow him to usurpthe throne of Chu without actually having to speak about his plans. These two dialogical passages demonstratethe pragmaticand rhetoricalfunc- tions of the paradoxes.Whether effective or not, they are used to startletheir audi- ence and to produce a change in attitude.As non-dialogicalpassages also implicitly address rulersand disciples, it is reasonableto suppose that paradoxesappearing in these passages also performsimilar functions. This can be illustratedby the para- doxes A (the greatesttone makes little sound) and EE-~1 (the greatestves- sel takes a long time to complete).24Passages containing these paradoxes not only show the pragmaticand rhetoricalfunctions clearly, they also reveal that the same paradoxesare sometimes used in supportof differentends. In a firstexample, which is taken from LOshichunqiu 16.5, the paradoxesare placed at the very beginningof the chapterand serve as a headline to introducethe point of subsequentanecdotes:

Thegreatest knowledge does nottake shape, the greatest vessel takes a longtime to com- plete,the greatesttone makeslittle sound. While Yu was diggingchannels for rivers and streams,the peoplewere still pilingup rubbleto makebanks. After his taskwas com- pletedand his meritestablished, a myriadof generationsbenefited. Yu's way of looking at things was far-reaching,but none of the people realized this. (Lishi chunqiu 16.5/94/ 15-18) The paradoxes are placed in a political context to further the view that ministers of great potential are initially unnoticed or, in less fortunate cases, even meet with an- imosity. The first anecdote argues that Yu's A efforts were of the utmost importance, despite the fact that the common people were unaware of this. The other anecdotes

292 PhilosophyEast & West also illustrate the view that worthy ministers and good policies need some time to be recognized as such. The message to the ruler is to continue to support his ministers in the face of persistent criticism. As indicated at the very end of the chapter, "aver- age rulers stop pursuing the right course of action because of such criticism, while worthy rulers establish their merit in the midst of it" (2r0 , fiJ (LCshichunqiu 16.5/95/20). 4_,-M1• ;LjLt_)j:)Another instance of these paradoxes appears in the chapter "Yu Lao" of the . While this chapter is at first sight merely a commentary on the Laozi, it also con- tains a message for the ruler for whom it was written. More specifically, the passage containing the paradoxes urges the ruler to emulate the government of King Zhuang of Chu (r. 613-591). The function as statements: VEi?_ paradoxes concluding

/)?.%r,<••,2,,R9 o• KingZhuang of Chu had been overseeingthe governmentfor three years, but had neither issued commands nor outlined policies. The senior Ministerof Warfarestood in atten- dance and hinted to the king: "Therewas a bird perched on a hill in the south. It did not flutter,fly, or sing for three years. Being silent and not making any sound, what name should one devise for it?" The king responded: "Not flutteringfor three years allowed it to grow wings. Not flying or singing allowed it to observe the practicesof the people. Though it had not flown, it surely challenged the sky when it actually did fly. Though it had not sung, it surely startledpeople when it actually did sing. Put it out of your mind, I understandyour point."Half a year later,he startedto deal with government affairshimself. For every ten things he dismissed, he raised nine. For every five senior ministershe punished, he promotedsix unknownscholars, and the state was in greator- der. He assembled the armyto punish and defeated them at Xuzhou. He conquered Jin at Heyong, gatheredthe feudal lords at Song, and subsequentlyheld hegemony over the world. King Zhuang did not act out of minor considerations,and hence achieved huge fame. He did not commit himself prematurely,and hence achieved great merit. Thereforeit is said: "The greatestvessel takes a long time to complete, the greatesttone makes little sound." (Hanfeizi21/45/15-20) While this passage also describes an initial period of anonymity, the political setting is quite different. In contrast to the previous passage, the paradoxes here do not refer to the ministers but to the ruler. Moreover, it is argued that the ruler, unlike the min- isters in the previous passage, should not at first make any conscious efforts in creat- ing order. Only after a prolonged period of non-action and observation will he be able to act in a decisive way. The use of paradoxes is thus different from the previous passage. While the paradoxes play on multiple interpretations of important terms, close attention to the immediate contexts of the paradoxes reveals that they can best be understood as rhetorical devices. The paradoxes are then probably not designed for theoretical purposes, but for influencing the behavior of disciples and rulers. The use

Wim De Reu 293 of paradoxesthus shows the rhetoricaldimension of . Moreover, as the Daoist writings in which the paradoxes typically appear are not homoge- neous, and as some other "non-Daoist"writings also contain paradoxes,it is not sur- prisingto find that the paradoxesare often used in supportof differentends.

AppendixA: Overview of Groupsand Subgroups f Group 1A )? X Y ("X" "Y") Group 1B X Y Z ("X" - "Z")

"the greatestskill seems clumsy" "the brightestway seems dark" ZZ 10/25/16-17; LZ45/15/21; LZ41/14/18 HNZ 12/111/25 M I )? ? ??A "the most solid virtue seems weak" "the greatest success seems deficient" LZ 41/14/19 LZ 45/15/21

Group 2A t X > X ("X" = "X") Group 2B -T- X 2 X ("X" = "X")

"the greatestbenevolence is not "the shapeless shape" benevolent" LZ14/5/6; WZ 7/35/6; HNZ 12/106/11; ZZ 2/5/30; (LSCQ 17.3/102/5) LSCQ 5.2/23/21; HFZ 20/39/20

"the highest virtue is not virtuous" "the knowledge that does not know" LZ 38/13/15; HFZ 20/34/8, 11; ZZ 22/62/24 XY 12/19/20; (LSCQ 17.3/102/5) Group 3A k X T Y ("X"-- "Y") Group 3B -fT X Z Y ("Y" ? "X")

"the greatest carving does not cut" "the speechless teaching" LZ 28/10/6; (HNZ 12/108/14-15) ZZ 5/13/7, 22/60/11; LZ 2/1/11, 43/15/12; 7A WZ 2/7/3, 8/42/18; HNZ 9/67/3 "theT,-I greatest hero does not fight" HNZ 17.15/169/13; LSCQ 1.4/5/2; "the speechless disputation" (ZZ 2/5/30; HNZ 14/141/1) ZZ 2/6/2, 24/71/1; WZ 2/7/4, 9/47/25-26; HNZ 6/50/23, 8/63/5

Remarks: 1. Abbreviations are as follows: ZZ = Zhuangzi, LZ= Laozi, WZ = Wenzi, HNZ = ,LSCQ = LCshichunqiu, HFZ= Han Feizi, XY= Xinyu. 2. References in parentheses indicate slightly different paradoxes.

294 PhilosophyEast & West Appendix B: Distributionof the Paradoxeson the Basis of Ban Gu's Classification

NP/NC Frequency NP/NC Frequency Category Work (work) (%o) (category) (%o)

Daoist Zhuangzi 38/65406 0.580 81/242095 0.334 Laozi 27/5676 4.756 Liezi 7/30972 0.226 6/128750 0.046 Huangdisijing 3/11291 0.265 Syncretic Huainanzi 41/133828 0.306 66/235577 0.280 Lishi chunqiu 25/101749 0.245 Legalist Han Feizi 6/107144 0.055 8/1081 75 0.073 Shenzi 2/1031 1.939 Confucian Chunqiufanlu 2/62356 0.032 3/72672 0.041 Xinyu 1/10316 0.096 Terminologist YinWenzi 3/6076 0.493 3/6076 0.493 Remarks: 1. NP/NC:'NP' standsfor "numberof paradoxes,"while 'NC' refersto "number of characters."With the exception of the Huangdisijing, the numberof char- acters can be found in the ICS concordance series. Ryden 1997, p. 298 presentsa list for the Huangdisijing. 2. The frequency refersto the numberof paradoxesper thousandcharacters. 3. The Liezi contains four relevant passages that all have close parallels with pre-Han sources. For the date of the Liezi, see Graham 1986. The relevant passages from the Guanzi, which as a whole is perhaps better classified as "Legalist,"all appear in the chapters Xinshushang (Artof the Mind, Part One), Xinshu xia LI4tCT(Art of the Mind, Part,ji'_tlLTwo), and Neiye rF?A (InnerTraining). These chapters are generally considered to representearly forms of Huang-Lao thought. Most scholars believe that the excavated manuscriptscalled Huangdi sijing also belong to Huang-Lao,though some would question the title of the work. For a brief discussion of the different views on the title, see Ryden 1997, p. 3 n. 3. The paradoxesappear in the manuscriptCheng fj (Sayings).I have not included the Wenzi in this statisti- cal overview. Virtuallyall paradoxes in the post-Han received Wenzi are copied fromthe Huainanzi. 4. The transmissionof the Shenzi seems to have stopped shortlyafter A.D. 1616. Several scholars have tried to reconstructthe Shenzi on the basis of quotations preserved in ancient writings. The paradoxes in the received version were copied from the Lushichunqiu by Ma Guohan (1794- 1857). It is, however, doubtful whether the paradoxes shouldN,, have been included (see Creel 1974, pp. 339, 375 n. 6). 5. Forthe date of the YinWenzi, see Daor 1974.

Wim De Reu 295 Notes

This article is based on my doctoral research.I want to thank CarineDefoort, Nico- las Standaert,Paul van Els, and KarenDesmet for their comments. A part of the article was presented at the WarringStates Working Group Conference 17, which was held on September17-18, 2003, at Leiden University.I also wish to thank the participantsin this conference for their remarks.Financial support for my research has been provided by the Fundfor Scientific Research,Flanders, Belgium (F.W.O.- Vlaanderen). 1 - One of the paradoxes included in this chapter reads ,% (a horse is not a horse). This is commonly seen as a variationon Gongsun Long's (a white horse is not a horse). 4,A,%- 2 - The paradoxesthat appear in this chapterare usuallyonly attributedto Hui Shi. Deng Xi is probablymentioned because (in Xunzi'sview) he was engaged in the same kind of pernicious intellectualactivity as Hui Shi. 3 - See, for example, Graham1989, pp. 75-95, 261-267; Hansen 1992, pp. 257- 263, 319-334; Harbsmeier1998, pp. 286-326; Needham 1956, pp. 185-197; Nivison 1999, pp. 779-783, 797-799; Schwartz1985, pp. 222-224. 4 - Reding(1985, pp. 350-385, 435-451) makes a particularlyinteresting attempt at reconstructingthe context of the paradoxes that appear in Xunzi 3 and Zhuangzi33. 5 - A general overview of all groups and subgroups is included in Appendix A. References in the Appendix as well as all other references throughoutthis article are made according to the concordances compiled by the Instituteof Chinese Studies (ICS)of the Chinese Universityof Hong Kong. 6 - In this as well as in the following tables, the term da )k (greatest)in formalrep- resentationssuch as t X Y can be replaced by other terms indicatinga de- gree, such as zhi E (ultimate) and shang -_ (highest). The signs t-, =, and symbolize antonymy, identity, and implication, respectively. Capital letters such as X and Y are variablesstanding for characters. 7 - In subgroup2A, the negation can also be a verb; for example, • (ulti- mate knowledge casts off knowledge) in LOshichunqiu J t 17.3/102/5. The same is true for subgroup3A. 8 - These paradoxes are (the brightestway seems dark), Laggg (the way that leads forwardHAl-E, seems to lead backward), aR (the smoothest way seems uneven), _Lt@ (the highest virtue is like a valley), k (the greatest white seems stained), a VA (the broadest virtue seems insuffi- cient), M (the most solid virtue seems weak), (the purest au- E, • • thenticity seems soiled), 7A 'r@ (the greatest square does not have corners), WrA (the greatest vessel takes a long time to complete), •@ (the

296 PhilosophyEast & West greatesttone makes little sound), and (the greatest image is without shape). .• 9 - The other three paradoxesare (the greatestbenevolence is not benev- olent), iKTM; (the greatesthonesty _z?TFf is not humble),and kATt-'I (the greatest hero is not cruel). 10 - Dao M also has the sense of "to speak." Hansen (1992) interpretsit as "guiding discourse."This not only clarifiesthe paradoxicalnature of the paradox f9 but also explains the replacementof cheng f4 (to designate) by dao in the>_T( paradox L-_6?j. 11 - Laozi 45/15/21-22. 12 - Forreasons of comprehensiveness,I have not a prioriexcluded writingsof sus- picious originsuch as the received versionsof the Lieziand the YinWenzi. The examined writingsare (in alphabetical order):Chunqiu fanlu EfR , Deng Xizi tJVF, Gongsun Longzi, Guanzi :F7,Han Feizi (40r, Heguanzi ff H -, Huainanzi " T, Huangdi sijing if?i , Laozi, Liezi YIJ-, LOshi chun- qiu, Lunyu~~, Mengzi tF, mT, Shangjun shu @ , Shenzi i-T, Shenzi {Iy, Shizi if, Wenzi -Zf, Xinshu Pf, Xinyu Xg, Xunzi, Yanzi chunqiu T •ti, YinWenzi JR-T, and Zhuangzi. 13 - As for groups one and two, the expressions found could readily be identified as paradoxical.Group three is more complex for there are expressionswithout the relationof implicationbut with a similargrammatical structure as the para- doxes (e.g., #~2f "innocent people"). In addition, there are some border- line cases where it is difficultto decide whether there is an implicationor not. Due to limitson space, it is impossibleto list all paradoxes. 14 - The well-known lists contain around42 differentparadoxes, with a total num- ber of 45 occurrences. Unless explicitly indicated, numbers hereafterrefer to the numberof occurrences, not to the numberof differentparadoxes. 15 - Please consult appendix B for an overview of the distributionof the paradoxes. In presentingthe data, I have listed the works according to the classification found in the bibliographicalchapter of the Hanshu•A . 16 - As for the Huainanzi, 7 out of 19 passages can be related to the Laozi (4) and the Zhuangzi (3). For the LCjshichunqiu: 2 out of 9 passages are related to the Laozi. The paradoxesin the Han Feizi all appear in the chaptersJie Lao PW?(Explaining Laozi) and YuLao ?ify (IllustratingLaozi). The single relevant passage from the Xinyu is an explicit quotationfrom the Laozi. 17 - Forthe dialectical interpretation,please referto Sun 1993, pp. 1-10, 122-128, and Zhang 1996, pp. 146-156. This interpretation, which depends on the idea of cyclical change, has indirectly been criticized in Lau 1958. Yan (1989) and Zheng (1992, pp. 151-156) have argued that the paradoxes violate the logical laws of identity and noncontradiction. Their interpretation, however, fails to

Wim De Reu 297 explain the semantic play on importantterms that is typical of the paradoxes (infra).For mystical interpretations,see, for example, Qian 1980, pp. 463- 465, and Schwartz1985, p. 198. Apophaticmystical readings of the paradoxes seem to have been typical of Chan-influencedor post-BuddhistConfucian commentaries(see note 22 below). 18 - These terms have been the subject of study in a numberof articles, most nota- bly in Stevenson 1938, Hallden 1960, and Gallie 1968. The expression'impor- tant term' is my own. Gallie speaks of 'essentially contested concepts,' while Stevenson and Hallden name the linguistic structuresin which these terms appear 'persuasive definitions' and 'essence statements,'respectively. I have no intentionto give an exhaustive overview of the problems and discussions related to importantterms. I will describe these terms only insofaras relevant to the interpretationof the paradoxes. 19 - I here employ the alternativeand target interpretationsof 'culture'put forward in Stevenson 1938, pp. 331-332. 20 - The paradoxHal M does not appearanywhere else. Otherpassages contain- ing the paradox 7kC • indicate that bai and ru stand for differentkinds of behavior. I will discuss one of these passages furtheron. 21 - The paradox k 1E is very similarto the paradox kIJrlTJ (the greatestcarv- ing does not cut). Both yu rA(corner) and ge NJ(to cut) suggest harshlaws and punishment.This political interpretationis supportedby the fact that the term zhi $•1has the double meaning of 'carving'and 'governing.' 22 - Mark Csikszentmihalyi(1999) has shown that early Chinese commentators favored a semantic interpretationof Laozi paradoxes. By contrast, laterChan- influenced or post-BuddhistConfucian commentariesgenerally present apo- phatic mystical interpretations.I would furtherargue that early commentators favored a semantic interpretationprecisely because the paradoxeswere origi- nally formulatedas a semantic play on importantterms. 23 - This storyalso appearswith minorvariations in Liezi2/13/16-22. 24 - The latterparadox is not obviously paradoxical(see the remarkon borderline cases in note 13 above). Nevertheless, it always appears together with the clearly paradoxical)I , and is, in Laozi 41, part of a series of twelve paradoxes. It presumably had some paradoxicalflavor for the masters who used it.

References

Creel, HerrleeGlessner. 1974. -hai:A Chinese Political Philosopherof the FourthCentury B.C. Chicago: Universityof Chicago Press. Csikszentmihalyi,Mark. 1999. "Mysticismand Apophatic Discourse in the Laozi."

298 PhilosophyEast & West In Religiousand PhilosophicalAspects of the Laozi, edited by MarkCsikszent- mihalyiand PhilipJ. Ivanhoe.Albany: State Universityof New YorkPress. Daor, Dan. 1974. "The Yin Wenzi and the Renaissanceof Philosophy in Wei-Jin China."Ph.D. diss., Universityof London. Gallie, W. B. 1968. "EssentiallyContested Concepts." In Philosophyand the Histor- ical Understanding.2nd ed. New York:Schocken Books. (A shorterversion was published in Proceedingsof the AristotelianSociety, 1956, pp. 167-198). Graham,Angus Charles.1986. "The Date and Compositionof Lieh-tzi." In Studies in Chinese Philosophyand PhilosophicalLiterature, edited by AngusC. Graham. Singapore:National Universityof Singapore,Institute of EastAsian Philosophies. . 1989. Disputersof the :Philosophical Argument in Ancient China. La Salle, IL:Open Court. Halld n, Soren. 1960. TrueLove, TrueHumour and TrueReligion: A SemanticStudy. Libraryof Theoria,no. 6. Lund:Gleerup; Copenhagen: Munksgaard. Hansen, Chad. 1992. A Daoist Theoryof Chinese Thought:A PhilosophicalInterpre- tation. New Yorkand Oxford:Oxford UniversityPress. Harbsmeier,Christoph. 1998. Languageand Logic in Ancient China. Editedby Ken- neth Robinson. Vol. 7, pt. 1, of Science and Civilisationin China, edited by Joseph Needham. Cambridge:Cambridge University Press. Instituteof Chinese Studies (ICS),Chinese Universityof Hong Kong~~& r7 f2t~lf~ ?.1992- . Xianqinlianghan guji zhuzi suoyin congkan $MV A- WH@-••IJ (ICSancient Chinese text concordance series). Editedby D. C. Lau. Hong Kong:Commercial Press. Lau,D. C. 1958. "TheTreatment of Opposites in LaoTzui." Bulletin of the School of Orientaland AfricanStudies 21 (2): 344-360. Needham, Joseph. 1956. Historyof Scientific Thought.Vol. 2 of Science and Civili- sation in China, edited by Joseph Needham. Cambridge:Cambridge University Press.

Nivison, David Shepherd. 1999. "The Classical Philosophical Writings."In The CambridgeHistory of China:From the Originsof Civilizationto 221 B.C.,edited by Michael Loewe and EdwardL. Shaughnessy.Cambridge: Cambridge Univer- sity Press. Qian Zhongshu ,JN. 1980. Guan zhui bian f (The tube and awl chapters), vol. 2. Hong Kong:Zhonghua Shuju. Reding, Jean-Paul. 1985. Les fondements philosophiques de la rhetorique chez les sophistes grecs et chez les sophistes chinois. Berne: Peter Lang. Ryden, Edmund. 1997. A Literary Study and Edition of the Text from Mawangdui. Ricci Institute Varietis Sinologiques New Series, no. 85. Taipei: Kuangchi Press.

Wim De Reu 299 Schwartz, BenjaminI. 1985. The Worldof Thoughtin Ancient China. Cambridge: HarvardUniversity Press. Stevenson,Charles Leslie. 1938. "PersuasiveDefinitions." Mind 47: 331-350. (Also published in Stevenson 1963.) . 1963. Factsand Values:Studies in EthicalAnalysis. New Haven: Yale Uni- versity Press. Sun Zhongyuan Ttt. 1993. Zhongguo luojixue Yr• 4 (Chinese logic). Tai- pei: Shuiniu. Yan Lingfeng . 1989. "Laozi de 'zheng yan ruo fan' de luoji ji qi lishi yuanyuan" f lR (Laozi's logic of 'Rightwords E(•- FI- jn seem to be wrong' and its historicalorigins). Dongfang zazhi 21 (12): 12-16. ,r , Zhang Xiaomang $6. 1996. Xianqinbianxue faze shilun (His- '•J•A•i•J-•i torical overview of pre-Qin dialectical principles). Beijing:Zhongguo Renmin Daxue Chubanshe.

Zheng Chenghai $:fIh. 1992. Laozi?K-aozi xueshuo yanjiu (Researchinto Laozi'steachings). Taipei: Huazheng Shuju.

300 PhilosophyEast & West