墨子大取小取篇分

Journal of Chinese Philosophy 27:4 (December 2000) 375–407 © 2000 Journal of Chinese Philosophy ian johnston CHOOSING THE GREATER AND CHOOSING THE LESSER: A TRANSLATION AND ANALYSIS OF THE DAQU AND XIAOQU CHAPTERS OF THE MOZI Introduction The Daqu and Xiaoqu are two of the six “dialectical” chapters of the Mozi , accepted as being of later composition to the main work and attributed to the aptly named “Later Mohists,” although it is not known who these Later Mohists were, exactly when they were active, or whether one or several hands were involved in fashioning these remark- able chapters. Certainly they have occasioned much puzzlement among both Chinese and Western commentators, a number of the latter ini- tially dismissing them as incomprehensible, although it has become increasingly apparent how crucial they are for a proper understanding of early Chinese philosophy. What is universally agreed upon, however, is that there are substantial textual problems, especially regarding the Daqu. The majority of scholars, starting with Bi Yuan 1 whose work sig- nalled a reawakening of interest in the Mozi in general, have attempted to bring clarity to the “dialectical” chapters with detailed textual emen- dations. Some, specifically Tan Jiefu, 2 Zhang Qihuang, 3 and A. C. Gra- ham, 4 have undertaken more or less radical, but quite different, rearrangements of the material in the Daqu and Xiaoqu. Graham, whose modifications have been the most extensive and who has been greatly instrumental in bringing an awareness of all six dialectical chap- ters to Western readers, has created a hybrid text from the Xiaoqu and parts of the Daqu which he titles “Names and Objects” and consigns most of the remainder of the Daqu to what he titles “Expounding the Canons.” Although the magnitude of Graham’s achievement cannot be denied, it must be recognized that many of his textual emendations are at odds with those of Chinese commentators, that the evidence in sup- port of such a rearrangement is by no means compelling, and that, par- ticularly, it has no counterpart in Chinese studies. Quite apart from the textual issues there is the matter of the philo- sophical context of these two chapters. Scholars generally agree that Mozi presented a major challenge to Confucianism in its initial form 376 ian johnston and that the continuing Ru–Mo debate was at least in part a stimulus to Mozi’s successors to attempt to shore up their master’s original doctrines, Mencius having by this time added his not inconsiderable weight to the Confucian cause. It is less unanimously agreed but also probably the case that the Later Mohists were responding to the chal- lenge of Hui Shi and Gongsun Long who, in their individual ways, had brought into sharp focus some of the paradoxes and uncertainties of language and ontology. More doubtful still is the relationship of the Later Mohists to the Zhuangzi , which is itself considered a work of multiple authorship. If it is accepted that the anecdotes identifying the historical Zhuang Zhou as a friend and intellectual sparring part- ner of Hui Shi are reliable and that the latter preceded the Later Mohists, then clearly one must also accept the former as doing so. Chad Hansen for one, however, sees Zhuangzi as an important com- ponent of the response to the Later Mohists. 5 This, and the much less discussed relationship to Yang Zhu, must for the present remain sub judice. In terms of successors, it seems highly probable that Xunzi was posterior to, and substantially influenced by, the Later Mohists. Indeed, in his translation of the Xunzi , Knoblock writes: “The Mohist Canons contain the final development of a logical discourse of names and reality, which is the direct source of many of Xunzi’s ideas.” 6 This would accord with the dates offered, either the relatively precise fig- ures given by Hu Shi (325–300 b.c. ), 7 or the less precise but arguably more realistic figures given by Chung-ying Cheng (early fourth to third century b.c. ). 8 Turning now to philosophical content, the overall objective of the Later Mohists could be identified as an attempt to support the basic Mohist doctrines, that is, those formulated particularly in Books II through IX of the Mozi. All the main aspects of these doctrines are at least touched upon in the Daqu and Xiaoqu and, in each case, there is support for the original formulations with no evidence of any significant doctrinal change. Two methods of defense are employed approximately corresponding to the two chapters, the first a specific examination of individual points of difficulty and the second an ambitious attempt to formulate a system of rules of language and inference that the Later Mohists could deploy against the subversive Sophists and other per- ceived threats to the main Mohist theses. An interesting issue, and one that is so far inadequately explored, is the relationship of the Daqu and Xiaoqu to the other components of the Mohist dialectic corpus, the Canons and Explanations, in terms of chronology, authorship, and con- tent. Accepting this deficiency, and the broad subdivision above, an overview of the subject matter of each chapter will now be given prior to the translation itself. daqu and xiaoqu chapters of the MOZI 377 The Daqu The bulk of this most difficult yet intriguing chapter, bedeviled as it is by textual difficulties, is devoted to ethical matters. The main focus of these ethical deliberations is the elaboration and refinement of the important terms, ai a1 and li , l2 their interrelationship and how to accom- modate the natural gradations of love (to parents, offspring, etc.), the punishment of transgressors, and the regard of the self within the con- ceptual framework of universal love. The treatment of the first two of these issues in the original text is relatively facile, and that of the third is scanty. Thus it is said that if love is universal, appropriate filial conduct will automatically follow whilst under these circumstances no robbers will arise. 9 The reality of the world is, of course, otherwise, so at an immediate and practical level—that is until the ideal Mohist conditions obtain—these problems must be addressed. To this end the author of the Daqu argues that gradations of love are permissible, that robbers may be punished and that the self may be loved, endeavoring to recon- cile these apparent compromises with the core concepts. In particular, the concept of li l2 (benefit) receives relatively extended treatment in the Daqu, significantly amplifying that given in the main work. The issue of whether or not Mozi is a utilitarian has come under scru- tiny in recent years, starting particularly with Ahern’s challenge to the generally accepted view, as adumbrated by Fung Yu-lan, for example, that Mozi was clearly and strongly utilitarian in his ethical views. 10 The argument hinges primarily around the role played by the “will of heaven” in Mozi’s ethical schema, a matter that is somewhat equivocal in the main Mozi. Unfortunately, the position is not really clarified in the Daqu. First, in accepting gradations of love and benefit, the ethical position becomes agent relative which does not accord with true utilitar- ianism. Second, however, the killing or punishment of transgressors is clearly acceptable within the conceptual framework of act utilitarian- ism, as is the whole idea of “weighing” acts on the basis of benefit and harm, of choosing the greater or choosing the lesser. Third, what little is said to clarify the position regarding the “will of heaven” (section 3), whilst inconclusive and textually contentious, does appear to downplay its role, thus favoring a utilitarian position. Another important component of the Daqu, which is even more frag- mented than the ethical considerations, has to do, broadly, with episte- mology and language. In particular, support is sought for the Mohist objectives of consistent naming in accord with reality and reliable classi- fication. For the latter, of course, objective criteria of sameness and dif- ference are required, a matter that is considered, albeit somewhat inconclusively, in sections 12 and 15. These observations should be seen in relation to those in B1–7 of the Canons and Explanations. Despite 378 ian johnston the clear enumeration of the types of sameness in Daqu 12, it is ac- knowledged that errors are made (some of which are listed), and so acknowledged, as stated in B2: “Making inferences about classes is diffi- cult. The problem lies in understanding scope.” Finally, specific consid- eration should be given to the last two sections of the Daqu. The first of these constitutes an important attempt to clarify what ci c2 (propositions) are, how they are formed and elaborated, and then how they are brought into play. It is interesting to note that this is a term not defined in the Canons and Explanations, but it is considered in the Xiaoqu (sec- tion 1) and is of critical importance to the Later Mohists. The second (section 17) is noteworthy for its unique structure, which comprises a series of statements predominantly about love and benefit, each illus- trated by a more or less appropriate and often quite striking analogy. Although the structure is of interest, the material itself adds little to the clarification of the Later Mohist philosophical position and, again, tex- tual problems are quite intrusive. The Xiaoqu The sorts of textual constraints on interpretation and analysis that hinder clear understanding of the Daqu are much less problematic in respect to the Xiaoqu. Despite its brevity, this short treatise, prior to any dissection or transplantation, stands as a coherent and structured argu- ment of singular significance in pre-Qin philosophy. Whether one chooses to apply the term logic to its subject matter as Hu Shi has enthusiastically and unequivocally advocated, 11 or not, it is clearly about the principles of valid inference and their application in practice, which would unquestionably fall under the broad definition of logic. Although some of the subject matter, particularly definitions of terms, is dealt with in both the Canons and Explanations and the Daqu, nowhere else does it receive such a structured and continuous presentation. The initial section offers a statement on the nature and purposes of bian , b1 a term that has occasioned much debate, not least in the endeavor to find a suitable English equivalent. It is, however, clearly and comprehensively defined within the section itself in which the whole program and apparatus of bian b1 is set out in outline. Based on the confidence that objective truth is knowable, three theoretical aims are identified, followed by three practical applications. It would seem that the Later Mohist is not interested in bian b1 for its own sake—that is, the development of a system of logic—but rather for its practical worth in resolving real issues in the quotidian world and especially in supporting fundamental Mohist doctrines as they relate to these issues. If one is seeking a suitable English rendering, perhaps the most appro- daqu and xiaoqu chapters of the MOZI 379 priate term for bian b1 is dialectic in the Aristotelian sense, in which it is contrasted with demonstration. 12 The second and third sections are devoted to terminology regarding the “instruments” of bian. b1 As with bian b1 itself, some of these terms have excited considerable debate, although again it could be argued that sufficient clarity is provided by the material itself, particularly when coupled with the additional analysis and definition in the Canons and Explanations in some instances. In the final five sections (5 to 9) the methods of bian b1 are pressed into service to provide reasoned support for basic Mohist doctrines. Indeed, in examining these examples one obtains a precise statement of a significant part of the whole raison d’être of the Later Mohists. They must deal with linguistic difficulties, particularly those raised by the Sophists, so that their thesis of names “picking out” real entities and reflecting real similarities and differences is not undermined. They must accommodate the concept of universal love within their fundamentally utilitarian ethical position and also fur- ther strengthen the arguments for the feasibility of the latter. They must defend their anti-fatalist position and they must refine their delibera- tions on naming, on the scope of terms, on the part–whole relationship, and on classification. Again, specific aspects pertaining to these argu- ments will be considered in the later sections. Translation In view of the increasing interest in the work of the Later Mohists and the concomitant increasing recognition of the significance of their brief but extraordinarily rich writings among the materials of pre-Qin philos- ophy, it seemed worthwhile to provide a full translation of these two chapters following the “standard” text, something not hitherto available in English. To this end, Wu Yujiang’s edition of the Mozi 13 has been used as the primary source, a text that is closely followed in two other, mod- ern editions of the Mozi. 14 Almost all available texts have been con- sulted, at least those included in the Mo Tzu Chi-ch’eng , 15 but particular attention has been given to those of Sun Yirang, Tan Jiefu, and Wu Feibai for both chapters, Zhang Qihuang for the Daqu, and Hu Shi and Zhou Yunzhi for the Xiaoqu. 16 The only complete translation of both chapters following the “standard” text that I could find is that of Forke into German. 17 A. C. Graham includes a full translation of the Xiaoqu, rearranged and interwoven with some sections of the Daqu in his Later Mohist Logic, Ethics and Science , as well as translations of some of the other sections of the Daqu in the various works referred to in note 4. Other, partial translations, predominantly of all or parts of the Xiaoqu 380 ian johnston that were consulted include those of Hu Shi, Fung Yu-lan, Maspero, Hughes, and Lau. 18 Choosing the Greater (Daqu) 1. Heaven’s love of man is more extensive than the sage’s love of man and its benefitting man is more substantial than the sage’s benefit- ting man. The great man’s love of the small man is more extensive than the small man’s love of the great man and the great man’s benefitting the small man is more substantial than the small man’s benefitting the great man. To consider an elaborate funeral as a manifestation of love for one’s parents is to love one’s parents. To consider an elaborate funeral as being of benefit to one’s parents is not to benefit one’s par- ents. To consider music to be of benefit to one’s son and to desire it for one’s son is to love one’s son. To consider music to be of benefit to one’s son and so seek it for one’s son is not to benefit one’s son. 2. In the matter of bodies the estimation of light and heavy is called “weighing.” [This kind of] “weighing” is not about right and wrong; it is about being accurate. 19 Cutting off a finger to preserve the hand is, in terms of benefit, choosing the greater; in terms of harm it is choosing the lesser. In terms of harm, choosing the lesser is not to choose harm but to choose benefit. What is chosen is determined by others. 20 Thus, in meet- ing a robber, to cut off a finger to spare the [whole] body is a benefit. It is meeting a robber that is the harm. Cutting off a finger and cutting off a hand are alike in terms of benefit to the empire. There is no choosing. Dying and living, in terms of benefit, are alike; one is not chosen. 21 Although killing one man will preserve the empire it is wrong to kill one man to benefit the empire. If killing oneself will preserve the empire it is right to kill oneself to benefit the empire. In the conduct of affairs the weighing up of light and heavy is called “seeking.” “Seeking” is about right and wrong. In situations where the lesser harm is chosen, the seek- ing may be righteous or it may not be righteous. 3. Is it right to speak of the will of heaven with respect to the tyrant? It is his nature to be a tyrant. To attribute this to the will of heaven is wrong. The tyrant says: “I am as I am through the will of heaven.” But it is through man that wrong becomes right and the nature that is not cor- rect is corrected. 22 If the various long-established beliefs have already had their effects and I act in accordance with these effects [then] it is the effects of these beliefs that cause me to act as I do. If the various long- established beliefs have not had effects yet I act in accordance with these beliefs then, in terms of the long-established beliefs, it is through me that there are effects. 23 4. In that which is beneficial there is no alternative to choosing the daqu and xiaoqu chapters of the MOZI 381 greater. In that which is harmful there is no alternative to choosing the lesser. In terms of benefit, choosing what one does not have is choosing the greater. In terms of harm, casting aside what one already has is choosing the lesser. If, according to duty, it is permissible to love [some- one] substantially then love them substantially. If, according to duty, it is permissible to love [someone] slightly then love them slightly. This is to speak of “proper sequence.” Virtuous rulers, elders, and parents all are those one should love substantially. [However] loving one’s elders sub- stantially does not entail loving those who are young slightly. If relations are close they should be loved substantially; if they are distant they should be loved slightly. One should be on close terms with one’s par- ents, whereas with respect to those other than parents, one may love slightly. It is in accord with principle to love one’s parents substantially. One must look closely at their conduct but hope only to see virtues. 24 5. To [love] Yu substantially for the sake of the empire is not because of his being Yu. To love Yu substantially for the sake of the empire is because of his love of man. To [love] Yu substantially for what he does, “adds to” the empire but to [love] Yu substantially [as an individual] does not “add to” the empire. Likewise, to abominate a robber for what he does, “adds to” the empire whereas to abominate a robber [as an individual] does not “add to” the empire. The love of mankind does not exclude the self, for the self lies within that which is loved. If the self lies within that which is loved then [necessarily] love extends to the self. Love of the self and love of mankind are without distinction. 6. The sage abhors disease and decay but does not abhor danger and difficulty. He maintains the integrity of his body and the resolve of his heart. [He] desires the people’s benefit: He does not dislike the people’s harm. The sage does not consider his own dwelling. The sage does not concern himself with the affairs of the son. The sage’s model [method] is to turn his mind from his parents on their death. With respect to the empire, to treat one’s parents substantially is divisive. One’s whole endeavor should be to bring benefit to the empire. When there is sub- stantial and slight and not “proper sequence” in bringing about benefit, this is for the self. 25 7. These are the rules of language. In seeking a white horse one may direct attention to the foal. In speaking of the negation of seeking