Dao Companion to Chinese Philosophy of Logic [1st ed.] 9783030290313, 9783030290337

Citation preview

Dao Companions to Chinese Philosophy 12

Yiu-ming Fung Editor

Dao Companion to Chinese Philosophy of Logic

Dao Companions to Chinese Philosophy Volume 12

Series Editor Yong Huang Department of Philosophy The Chinese University of Hong Kong Shatin, New Territories, Hong Kong E-mail: [email protected]

While “philosophy” is a Western term, philosophy is not something exclusively Western. In this increasingly globalized world, the importance of non-Western philosophy is becoming more and more obvious. Among all the non-Western traditions, Chinese philosophy is certainly one of the richest. In a history of more than 2500 years, many extremely important classics, philosophers, and schools have emerged. As China is becoming an economic power today, it is only natural that more and more people are interested in learning about the cultural traditions, including the philosophical tradition, of China. The Dao Companions to Chinese Philosophy series aims to provide the most comprehensive and most up-to-date introduction to various aspects of Chinese philosophy as well as philosophical traditions heavily influenced by it. Each volume in this series focuses on an individual school, text, or person. More information about this series at http://www.springer.com/series/8596

Yiu-ming Fung Editor

Dao Companion to Chinese Philosophy of Logic

Editor Yiu-ming Fung Department of Philosophy Tunghai University Taichung, Taiwan

ISSN 2211-0275     ISSN 2542-8780 (electronic) Dao Companions to Chinese Philosophy ISBN 978-3-030-29031-3    ISBN 978-3-030-29033-7 (eBook) https://doi.org/10.1007/978-3-030-29033-7 © Springer Nature Switzerland AG 2020 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG. The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Editor’s Acknowledgments

It may be agreed by most scholars in the field of Chinese philosophy that the main concern of traditional Chinese philosophy (hereafter, TCP) is about the problems of moral, social, and political philosophy. Traditional Chinese thinkers are not interested in making arguments and philosophical analysis. Instead, they usually use metaphors, analogies, fables, or anecdotes to express their ideas. Some even think that there are no logical and analytical methods in TCP, except figurative expressions. I think this kind of opinion on TCP is inaccurate or, at least, not totally correct. Although it is right to say that the trend of philosophical investigation in TCP is not about problems of philosophy of language and logic, we still find some interesting and significant ideas in terms of logico-philosophy in ancient China. In this volume, we are going to provide an introduction to philosophy of logic in TCP which covers topics including the thought in the School of Names and Later Mohism. I think this volume may be helpful for students and scholars who are interested in knowing some basic ideas of philosophical logic in Chinese tradition which are not similar to those in the Western tradition, though, I think, they are not incomparable to each other. To promote the idea of this book project and to make the project into reality, I have benefited from many people, especially the contributors of this volume. Here, I would like to express my gratitude to Professor Yong Huang, the editor of Dao: A Journal of Comparative Philosophy and of the present book series. Without his support, the project cannot be materialized. In addition, I am grateful to the editorial team of Springer for their outstanding and professional work for the volume. Finally, I would like to thank my wife, Shun Chee 順志, and my children, Ka Yan 家茵 and Ka King 家敬, for their constant help and inspiration.

v

Contents

1 Introduction: Chinese Philosophy of Logic ������������������������������������������    1 Yiu-ming Fung Part I Concepts in Chinese Philosophy of Logic 2 What Is Ming 名? “Name” Not “Word”������������������������������������������������   15 Jane Geaney 3 Naming, Reference and Truth����������������������������������������������������������������   33 Bo Mou 4 Sentences (Ci 辭, Ju 句) and Propositional Attitudes����������������������������   71 Yiu-ming Fung 5 Counterfactual Conditionals������������������������������������������������������������������   97 Yiu-ming Fung 6 Truth in Pre-Han Thought����������������������������������������������������������������������  113 Chris Fraser 7 Contradiction��������������������������������������������������������������������������������������������  129 Xinyan Jiang 8 Analogy ����������������������������������������������������������������������������������������������������  143 Alexei K. Volkov 9 Reasoning (Pi 譬, Mou 侔, Yuan 援, Tui 推)������������������������������������������  161 Wujin Yang 10 Argumentation (Bian 辯)������������������������������������������������������������������������  171 Lisa Indraccolo 11 Reason (Gu 故) and Principle (Li 理) ����������������������������������������������������  181 Yiu-ming Fung

vii

viii

Contents

12 Classes (Lei 類) and Individuals ������������������������������������������������������������  203 Wujin Yang and Wanqiang Zhang 13 Sameness (Tong 同) and Difference (Yi 異)��������������������������������������������  213 Yiu-ming Fung 14 Definitions in Pre-Qin Texts��������������������������������������������������������������������  233 Thierry Lucas Part II Issues and Theories in Chinese Philosophy of Logic 15 Logical Thought in Mohism and Later Mohism ����������������������������������  253 Thierry Lucas 16 Paradoxes in the School of Names����������������������������������������������������������  285 Chris Fraser 17 Logical Thinking in the Gongsun Longzi ����������������������������������������������  309 Yiu-ming Fung 18 Correcting Names in Early Confucianism��������������������������������������������  329 Hui Chieh Loy 19 Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi����������������������������������������������������������������������������������������������������  351 Kim-chong Chong 20 Problems of Language and Logic in Daoism ����������������������������������������  369 Eske J. Møllgaard 21 Paradoxical Language in Chan Buddhism��������������������������������������������  389 Chien-hsing Ho Part III Logical Thought Transplanted from India and the West 22 Yin Ming 因明 in Chinese Buddhism������������������������������������������������������  407 Mingjun Tang 23 Proper Relations of Association (Zheng 正) vs. Logical Validity of Syllogism: A Case Study of Shared Practices of Matteo Ricci, S. J and Chinese Mathematicians in Seventeenth Century��������������������������������  437 Jinmei Yuan 24 Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic in China������������������������������������������������������������������������������������������  465 Rafael Suter Part IV Logic Studies in Chinese Communities 25 Logic Studies in Mainland China ����������������������������������������������������������  511 Guoping Du and Hongguang Wang

Contents

ix

26 Logic Studies in Taiwan��������������������������������������������������������������������������  525 Zeqiang Wu and Wen-fang Wang Chinese-English Glossary��������������������������������������������������������������������������������  541 Author Index ���������������������������������������������������������������������������������������������������  549 Subject Index����������������������������������������������������������������������������������������������������  553

Contributors

Kim-chong Chong  Division of Humanities (Philosophy), Hong Kong University of Science and Technology, Kowloon, Hong Kong Guoping  Du  Institute of Philosophy, Chinese Academy of Social Sciences, Beijing, China Chris  Fraser  Department of Philosophy, University of Hong Kong, Pokfulam, Hong Kong Yiu-ming Fung  Department of Philosophy, Tunghai University, Taichung, Taiwan Jane  Geaney  Department of Religious Studies, University of Richmond, Richmond, VA, USA Chien-hsing Ho  Institute of Chinese Literature and Philosophy, Academia Sinica, Taipei, Taiwan Lisa  Indraccolo  Institute of Asian and Oriental Studies, University of Zurich, Zürich, Switzerland Xinyan Jiang  Department of Philosophy, University of Redlands, CA, USA Hui Chieh Loy  Department of Philosophy, Faculty of Arts and Social Sciences, National University of Singapore, Singapore, Singapore Thierry Lucas  Institut supérieur de philosophie, Université Catholique de Louvain (Louvain-la-Neuve), Louvain-la-Neuve, Belgium Eske  Møllgaard  Department of Philosophy, University of Rhode Island, Kingston, RI, USA Bo Mou  Department of Philosophy, San Jose State University, San Jose, CA, USA Rafael Suter  Chinese Studies, Institute of Asian and Oriental Studies, University of Zurich, Zurich, Switzerland Mingjun Tang  School of Philosophy, Fudan University, Shanghai, China xi

xii

Contributors

Alexeï K. Volkov  Center for General Education, National Tsing Hua University, Hsinchu, Taiwan Hongguang Wang  Department of Philosophy, Peking University, Beijing, China Wen-Fang Wang  Institute of Philosophy of Mind and Cognition, National Yang Ming University, Taipei, Taiwan Zeqiang Wu  Institute of Philosophy, Zhengzhou Chenggong University of Finance and Economics, Zhengzhou, China Wujin Yang  School of Philosophy, Renmin University of China, Beijing, China Jinmei Yuan  Department of Philosophy, Creighton University, Omaha, NE, USA Wanqiang Zhang  Department of Philosophy, Xidian University, Xi’an, China

Chapter 1

Introduction: Chinese Philosophy of Logic Yiu-ming Fung

1  What Is Philosophy of Logic? Following the developments of symbolic logic in the late nineteenth century and mathematical logic in the twentieth century, philosophers in the trend of analytic philosophy have been interested in using logical and other analytical methods to deal with philosophical problems or issues, on the one hand, and in explaining logical concepts to disclose their philosophical implication or significance, on the other. The most salient example of the former work is Bertrand Russell’s theory of definite descriptions (Russell 1905: 479–493) and that of the latter work is W. V. Quine’s philosophical explanation of a theory’s ontological commitment in terms of the dictum “to be is to be the value of a bound variable.” (Quine 1948: 21–38) These two kinds of work are, of course, both related to logic and philosophy and most of their topics have been identified as a special field named “philosophical logic” or “philosophy of logic.” However, philosophers do not have a consensus about the name and content of the field, as explained by Dale Jacquette in the following passage (Jacquette 2007: 1): It is standard in works dedicated to topics in philosophy and logic to distinguish between philosophy of logic and philosophical logic. There is no universally agreed upon way of dividing up these categories, and some authors prefer to conflate the two and treat the two phrases as more or less synonymous. I do not have particularly strong reasons for distinguishing or running these concepts together. If pressed, I favor a terminology in which philosophical logic involves applications of any recognized methods of logic to philosophical problems or for purposes of advancing philosophical discussions, whereas philosophy of logic is the consideration of philosophical problems that arise in relation to work in any recognized branch of logic.

Y.-m. Fung (*) Department of Philosophy, Tunghai University, Taichung, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_1

1

2

Y.-m. Fung

Here, I also prefer to conflate the two and treat the two phrases as more or less synonymous. Because, although the subjects or topics recognized as the main course of the field by some philosophers who favor one of the names are more or less different from those chosen by other philosophers who favor the other name, both still share a significant common part of content with each other. According to my personal observation, the content of the field, either named as “philosophical logic” or “philosophy of logic,” is well-understood by most philosophers as not separable from other areas of philosophy such as philosophy of language, metaphysics, epistemology, etc., let alone logic per se. In terms of content, philosophical logic or philosophy of logic partly overlaps with these areas. For example, topics like naming and reference, individuals and classes, language and meaning, truth and reality, paradoxes and argumentations, analogies and metaphors, and explicit logical theories and implicit logical thinking, etc., are shared topics of the field with the related areas mentioned above. In this sense, I think, this is a field of fusion which permeates into other related philosophical studies.

2  Studies in Chinese Philosophy of Logic It has been thought that Chinese philosophy, especially in the ancient time, is a kind of philosophical study which does not use logic and other analytical tools to express and analyze its ideas and theories. Although it is recognized by some scholars that ancient Chinese philosophers do have concepts to elaborate their ideas and reasons to argue for their views, they still think that the logic or the mode of thinking behind these acts of conceptualization and argumentation is essentially different from that in the Western tradition. Comparatists like David Hall and Roger Ames (1995: 230) and Chad Hansen (1985: 492, 1992: 238), are skeptical of the idea that there is a similar kind of logic used or thought by ancient Chinese thinkers. Some others, including Marcel Granet (1934), Joseph Needham (1956: 280), A. C. Graham (in his study of the thought in Han 漢 dynasty) (1992: 61–2), and also Roger Ames and David Hall (1995: 141), even adopt a thesis of “special mode of thinking” to identify Chinese philosophical thinking. They sometimes use the term “correlative thinking” or “associative thinking” to describe the Chinese mode of thinking. They all recognize, more or less, the term’s implication as “non-logical” or “pre-logical,” “non-rational” or “irrational,” “intuitive-associative” or “beyond analytical thinking.” Based on this presumption, some of them think that the root level of (correlative) thinking is “incomparable” to the upper level of (analytical) thinking or that the Western rational or analytical thinking and the Chinese correlative or analogical thinking are “incommensurable.”1 In contrast, some sinologists and philosophers in the field Chinese philosophy, including Shen Youding 沈有鼎 (1954, reprinted in 1980), Janusz Chmielewski

 In regard to the problem of “correlative thinking,” please see Fung 2010.

1

1  Introduction: Chinese Philosophy of Logic

3

(1962–69, reprinted in 2009), Cheng Chung-ying 成中英 (1965), A. C. Graham (in his study of Later Mohism 墨家) (1978), and Christoph Harbsmeier (1998), have demonstrated that the thinking in most philosophical views and arguments in ancient Chinese philosophy, especially in Later Mohism and the School of Names, is comparable to that in the Western tradition. This view of comparability is counter to the thesis of incommensurability. For example, based on his solid philological studies in classical Chinese and good training in Western Logic, Janusz Chmielewski’s pioneering work (1962–69, reprinted in 2009) has elaborated a clear picture of Chinese logic in the ancient time. Both Shen Youding (a famous Chinese logician in mathematical logic who was also interested in Chinese logical thought) (1954, reprinted in 1980) and A. C. Graham (1978) have given a sophisticated interpretation of the logical thought in Later Mohism respectively. In the last forty years, Cheng Chung-­ ying has written many analytical essays on topics, including the logical thought in the Gongsun Longzi 公孫龍子 and the Mohist Canons (Mojing 墨經), analogical thinking in ancient China, and semantics and paradoxes expressed in classical Chinese, which are helpful for further studies. Moreover, Christoph Harbsmeier’s contribution in Joseph Needham’s seventh volume of Science and Civilisation in China: Language and Logic (1998), in addition to providing comprehensive interpretations with solid textual evidence for ancient Chinese logical thought, has made clear that the linguistic characteristics of classical Chinese is not a hurdle of logical reasoning and the articulatory power of classical Chinese can be used in scientific investigation. As explained by Harbsmeier (1998: xviii, 261–2), the logical thought of ancient Chinese thinkers can be divided into two kinds: explicit theory of logical reasoning and implicit theory in logical practice. The only clear example of making an explicit theory of logic with philosophical implication is the theory of Later Mohism; the most sophisticated practice of logical arguments is the discourse in the Gongsun Longzi. These two kinds of logical thinking in ancient China can be understood as partial evidence to refute the thesis that the thinking in ancient China is non-­ analytical and essentially different from the analytical thinking in the Western tradition. In using logical arguments, making linguistic-conceptual analysis and elaborating logical ideas, ancient Chinese thinkers do not aim at making logical construction, meta-linguistic clarification and logical theorization consciously. Their using analytical language with logical argumentations is mainly for expressing and explaining philosophical problems. For example, in ancient China, the logical argumentations expressed in the Gongsun Longzi are used to support Gongsun Long’s ideas such as “white-horse is not horse” and “hard and white [of a stone-object] are separated.” The logical analysis of these topics is closely related to his special idea of reference which, in turn, based on the ontological commitment of his language.2 In comparison, the logical thought articulated in the Mohist Canons has intimate relation with Later Mohists’ thought of language, including reference and meaning,

 Detailed arguments for this view can be found in Fung 2020.

2

4

Y.-m. Fung

truth and reality. In regard to the problem of language rather than logic, Confucian idea of rectification of names or correcting names (zhengming 正名) is not purely a topic in linguistic philosophy or philosophy of language, it is a subject related to ethics and socio-political philosophy, especially related to the performative function of naming or using speech acts. Besides, the idea of dao in Daoism, especially in the Laozi 老子 and Zhuangzi 莊子, as the Daoists’ ultimate concern, involves the problem of language and reality and that of expressibility. All these are topics branching out of logic and entering into other areas of philosophy. In the medieval period of China, Buddhism from India brought the logical method called “yin-ming” 因明 (hetuvidyā in Sanskirt) (demonstration with reasons or science of reasons) into China. In the pre-modern time of China, especially in the late Ming 明 and early Qing 清 dynasties, the Western missionaries came to China and introduced Western Logic and mathematics to China. The former provided an analytical tool for Chinese scholars to analyze problems in Buddhism while the latter opened the door of China to welcome the Western analytical methods for scientific investigation and this has played an important role in the modernization of modern China. Since the publication of Hu Shih’s 胡適 PhD thesis (1931) at Columbia University, the Chinese name of the field “Zhongguo luoji sixiang” 中國邏輯思想 (Chinese logical thought) has been used with more or less similar meaning to the English terms like “Chinese philosophy of logic,” “Chinese philosophical logic” or “Chinese logico-philosophical thought.” The Chinese name has been used by scholars in Chinese communities to represent “logical thought” in Chinese philosophy and intellectual history (or history of ideas) in a broad sense which includes philosophical thought and history about language and logic. Some topics are about thinking on concepts of logic and modes of reasoning, some others are concerning ideas in philosophy of language, including naming and reference, truth and meaning, etc. All these are related to both logic and philosophy.

3  The Aims and Approaches of the Companion The major aim of this Dao Companion to Chinese Philosophy of Logic is to produce a source book for people in the English world to understand the logical thought and logical thinking in China. In this volume, we will comprehensively and analytically introduce the basic logical ideas and theories in Chinese thought for students and scholars who are interested in the field. We also hope it could be helpful for further studies of Chinese philosophy of logic. Specifically speaking, the Companion aims at helping senior undergraduate and postgraduate students to understand this lost tradition in Chinese philosophical studies. It also aims at attracting scholars who are interested in comparative studies of logical thought and logical thinking between China and the West. It is an introductory work to Chinese philosophy of logic with comparative and interdisciplinary perspectives. The contributors of the Companion come from

1  Introduction: Chinese Philosophy of Logic

5

different academic backgrounds. Some of their perspectives are grounded on Western logical theories and philosophy of language; some others’ are based on sinology and intellectual history. But most of them are equipped with comparative approach. In terms of historical period, it covers thoughts in ancient, pre-modern and modern China. In terms of area of study, it deals with topics in logical (including logico-philosophical) concepts and theories rooted in China and those transplanted from Indian and the West. Besides, it also provides information about the development of logic studies in contemporary Chinese communities.

4  The Skeleton and Content of the Companion In response to the aims and approaches mentioned above, the skeleton of the Companion is designed to include the following four parts: (I) Logico-philosophical concepts in ancient China (II) Logical thought and logical thinking in ancient China (III) Logical thought transplanted from India and the West (IV) Logic studies in contemporary Chinese communities Part (I) introduces some basic concepts in Chinese logical thought which include those used by ancient Chinese thinkers implicitly without theorizing and some others discussed by them at a theoretical level. In comparison with the logical thought in Western philosophical tradition, I think it is not a right thinking to overstate the sameness of conceptual theorizing between Chinese and Western philosophy, on the one hand; but it is also not an accurate way to understate the fact that both traditions have some common concerns in philosophical problems, such as those about naming and reference, truth and meaning, and reasons for making argumentations. In this part, most of the contributors introduce their topics with a comparative manner: some of them elaborate the common interest and similar ideas between both traditions, while others specify the differences. As proposed by Donald Davidson, to make clear what is the difference of concepts and thought between two people or two cultures we are compelled to use the principle of charity (Davidson 1974/1984: Ch.13). Without a common co-ordinate, it is no way to claim any sameness or difference between two sides which are maintained by two groups of rational agent. So, even though there are significant differences between Chinese and Western thought, they do share a lot of true beliefs and use similar concepts to form their beliefs. I think there is no essential difference or incommensurable factors between Chinese and Western traditions which make their language and thought untranslatable. One kind of important concepts used and theorized in traditional Chinese philosophy is about names and naming. In ancient China, one of the main uses and theorization of names focuses on the performative or pragmatic function of naming while another on referring or semantic function. In this part, Jane Geaney’s essay provides a detailed explanation for the performative function of using names which

6

Y.-m. Fung

is different from using names to express the relation between language and reality. She argues that names (ming 名) are used to name items in the cosmos that are visible and even audible. She further elaborates her view that, as indicated in some examples, the correlates (shi 實, shi 事, xing 形, xing 行, and shen 身) of names are closely related to actions and their good performance. It is not about the relation between language and reality, because the relation between names and their correlates is baptized with the evaluative or deontic function of naming. According to my understanding, it is about the assignment of status-function via naming and thus branches out from philosophy of language into the realm of social ontology. (Searle 2010) Different from Geaney’s perspective from the history of ideas, Bo Mou adopts an analytical approach to analyze the problems of naming, reference and truth in ancient Chinese philosophical texts, including the Gongsun Longzi, Mohist Canons and Laozi. In general, he proposes a view of “double-reference” semantic-­ syntactic structure to explain some of the ideas in the Gongsun Longzi such as “white-horse is not horse” and some of the theses in the Mohist Canons such as reasoning in parallelism (a kind of reasoning with the form: “A is B; so, CA is CB.”). It seems that he also uses a similar concept of “relative identity” as elaborated by P. T. Geach (1967: 3–12) to explain the problem of identity in these texts. Some scholars (including Chad Hansen 1985) reject the view that there is a concept of semantic truth in ancient Chinese philosophy. One of the reasons is that, in classical Chinese, there is no concept of a sentence and no linguistic constructions such as that-clause and counterfactual conditional which are necessary for forming true beliefs and other propositional attitudes. Against this view, in part (I), Yiu-ming Fung argues in his essay on propositional attitudes that, in classical Chinese, there is a kind of linguistic construction with paratactic form which is functionally equivalent to that-clause and, in his essay on counterfactuals, that there are also various linguistic devices to express counterfactual conditionals. So, ancient Chinese thinkers are able to use these constructions and devices to form their propositional attitudes and true beliefs. In regard to the concept of truth, Chris Fraser offers a concise interpretation and analysis for several concepts in pre-Han texts which are of conceptual role overlapping that of semantic truth. He argues that, in ancient Chinese philosophy, the absence of an explicit, thematized interest in truth does not rule out the possibility that early texts on language and logic might offer implicit or indirect explanations of what it is for an assertion to be true. He also demonstrates that, to count as an appraisal of truth, an evaluation must refer to semantic norms—it must deem an assertion correct specifically on the grounds that its content is correctly applied to the things it is asserted of. If we find terms of evaluation for assertions that are best explained as referring to semantic grounds and that perform functions associated with “… is (not) true,” then we have persuasive reasons to conclude that pre-Han philosophers indeed “have a concept of truth.” Xinyan Jiang uses a different perspective to deal with another logical concept: contradiction. In the Western philosophical tradition, a contradiction or contradictory sentence is (analytically) false. In ancient Chinese philosophy, the law of non-contradiction is acknowledged in some way by some Chinese philosophers such as Han Fei 韓非 and the Mohists, but the majority of Chinese thinkers seem to even have never paid attention to it, not

1  Introduction: Chinese Philosophy of Logic

7

alone treated it as the most basic principle of logic. In contrast, she argues, ancient Chinese philosophers stress that S can be P and not-P at the different times and/or in different aspects, even with different measurements in different contexts. Furthermore, contextualized “S is P and not-P” in Chinese philosophy reveals something very significant—dialectical relations are real and things always exist as the unity of opposites. It seems that this informal thesis is comparable to the modern formal view of dialetheism or paraconsistency (Priest et al. 2004). In part (I), in addition to addressing the topics in philosophy of logic at the level of name and that of proposition, some contributors also discuss topics in argumentation or reasoning and reasons for argumentation or reasoning. In ancient Chinese philosophy, most of the thinkers incline to use analogies to illustrate their views or to form their argumentations. Alexeï K. Volkov in his essay on analogy in pre-­ modern China demonstrates that the lack of ancient Chinese works on “explicit logic” historically was the main difficulty for studies of Chinese logical reasoning, though a number of attempts to find Western equivalents for Chinese logical methods were made at an early stage of their studies. It can be suggested that these attempts were often unproductive from the outset, because their authors looked for interpretation of analogical inferences in terms of modern formal propositional or first-order predicate logic. Volkov provides a brief summary of these attempts, offers an analysis of one representative case, and argues that the attempts of using Western formal approach cannot overcome the difficulty of interpretation. The reason is that, in his case study, the analogical inference in the form of dialogue involves sequences of analogical transfers from one domain to another, some of which are rejected, and some, further developed in the course of interaction between the interlocutors. In contrast, Yang Wujin adopts the Western logical perspective to analyze and explain four kinds of reasoning (pi 譬, mou 侔, yuan 援, tui 推) in the Mozi 墨 子. It is commonly thought that these patterns involve reasoning by analogy to an absurd conclusion as a form of reductio ad adsurdum. According to his view, although these patterns depend on inference, they are primarily concerned with various forms of proof and refutation. Among them, “illustrating” (pi) and “parallelizing” (mou) stress on proof, but “adducing” (yuan) and “inferring” (tui) stress on refutation. In comparison, based on a sinologist perspective, Lisa Indraccolo gives a brief and clear picture of the rhetorical tradition in ancient China with focus on bian 辯 and shuo 說 (argumentation) [or shui 說 (persuasion)]. She argues that argumentation and persuasion are two sides of the same coin. Through the use of a more or less fixed and predictable set of argumentative methods and rhetorical stratagems (such as chain reasoning, reductio ad absurdum, paradoxes and language jokes and so on), a case of argumentation might at times entail a deeper meaning and make a fundamental point or convey a moral lesson disguised as a highly sophisticated language play, as for instance happens in the “white horse” argument of the Gongsun Longzi. In regard to the topic of reasons for argumentation or reasoning, as discussed in the Mozi and Xunzi 荀子, three closely related concepts were mentioned as crucial for making correct reasoning or valid argumentation. They are reason/cause (gu 故), principle/reasonableness (li 理) and class/kind (lei 類, shu 屬). In discussing the

8

Y.-m. Fung

former two concepts, Yiu-ming Fung articulates their logical role in ancient philosophical texts. One of the requirements of providing good reasons and principles for making acceptable argumentation or reasoning is that the key term appeared in relevant propositions (premises and conclusion) must be of the same class. To rebut the view that there are no idea of class-membership and no concept of an individual in classical Chinese, Wujin Yang and Wanqiang Zhang in their essay on classes (lei and shu) and individuals (si 私) argue that these concepts play an important role in argumentation or reasoning respectively. They think that the emphasis on the requirement of “knowing classes” (zhi lei 知類), “understanding classes” (ming lei 明類) or “examining classes” (cha lei 察類) in ancient texts for making acceptable argumentation is an indication that the identification of the sameness and difference in classes is necessary for preventing fallacious argumentation and making acceptable one. In addition, Yiu-ming Fung elaborates from ancient texts that the sameness and difference in classes are crucial for making argumentation or reasoning and some of the examples indicate that the interpretation of lei as class is better than that as kind or type though lei can be used in other contexts with different meanings, including kind or type. In the last but not the least portion of part (I), Thierry Lucas points out that definitions are everywhere in the pre-Qin texts. Some of them are extremely accurate, concise, and almost formal: such are Hui Shi's 惠施 definitions. Confucian definitions are trying to find the fundamental characteristics of moral concepts such as ren. 仁 (humanity/benevolence) and yi 義 (righteousness/appropriateness). He also maintains that Mohist definitions constitute a remarkable, almost axiomatic attempt at constructing a philosophical system encompassing logic, ethics, epistemology and cosmology. In part (II), we will discuss some issues and theories in Chinese philosophy of logic. Later Mohism and the School of Names (Ming-jia 名家) are recognized by most scholars in the field as two main philosophical groups which adopt an analytical approach to address their philosophical problems. Thierry Lucas’s essay is a formal approach on the logical thought in Early Mohism and Later Mohism. He argues that among the different schools of philosophers of the pre-Qin period, the Mohists were the best promotors of logical research. As other schools of the time, they were interested in ethics and in socio-political problems, but their reflection evolved into a more and more precise attention to language. In Early Mohism, promotion of their theses was served by an elaborate argumentation, which remains however an instrument and is not yet the object of their reflections. In Later Mohism, from the Mohist Canons to the Daqu 大取 (Choosing the Greater) and Xiaoqu 小取 (Choosing the Lesser), one may observe a growing attention to language and to what we now call logic, epistemology and science. In dealing with the concept of paradox, Chris Fraser is very cautious to explain the phenomenon of paradoxes appeared in ancient texts, especially in the School of Names. He alerts that any discussion of early Chinese paradoxes must acknowledge a pair of important caveats. Since the extant sources provide little or no context for most of the paradoxes, interpretation is often partly conjectural and in some cases highly speculative. So, his attempt is to make explicit which aspects of the interpretations seem well

1  Introduction: Chinese Philosophy of Logic

9

grounded and which are open to doubt. In the essay, the paradoxical expressions mentioned by the School of Names and appeared in the Xunzi, Zhuangzi and Gongsun Longzi are the foci of his discussion. However, some may not treat the statement “White-horse not horse” as a paradox. According to Yiu-ming Fung’s view, the debate between Gongsun Long and his opponent in the Discourse on White-Horse of the Gongsun Longzi is embedded with logical arguments. He also tries to evaluate different interpretations of the Discourse and to make a rational reconstruction of the logical thinking in the text. Based on an analysis in terms of first-order predicate logic, it is suggested that the implicit difference between the two sides of the debate is related to the difference between using terms as predicates and as individuals and this difference in using terms indicates the difference in ontological commitment. In comparison with the analytical approach of Later Mohism and the School of Names mentioned above, the issues and thought expressed in Confucianism, Daoism and Buddhism either focus on the social dimension of speech acts or on metaphysical ideas which considered as transcending linguistic expression. In this part, some of the main issues at the methodological level are about analogical and metaphorical thinking in the Mencius, Xunzi and Zhuangzi and about the theoretical character of using paradoxical language in Chan (Zen 禪) Buddhism. The other important issues at the theoretical level are about the rectification of names in Confucianism and the ineffability of language in Daoism. Hui Chieh Loy in his essay on correcting names in early Confucianism argues that Analects (Lunyu 論語) 13.3 contains a concern with the correct use of words rather than directly with stations or roles. In addition, three different aspects of the doctrine—to do with a diagnosis of disorder, the preservation of order, and the remedy against disorder—can be distinguished. He also shows how Xunzi’s discussion in this topic both connects with the doctrine of zhengming in the Analects but also diverges from it. Kim-chong Chong’s essay on analogical and metaphorical thinking in the Mencius, Xunzi and Zhuangzi begins with a review on Irving Copi’s view of analogies and analogical argument as being inductive in nature. He shows that analogies are not inductive comparisons of similar things. The examples of analogical arguments from the texts mentioned above serve critical functions in the course of philosophical debate. These analogical arguments are not deductive in nature and therefore cannot conclusively prove a point of view. But, he thinks, this is a misleading comparison, making it seem as if philosophical argument must be conclusive. By pointing to both similarities and differences between things, analogies serve to make distinctions, force the clarification of terms, question assumptions, and reveal entrenched models of thought. He concludes that these are also analytical, argumentative functions of using language. Eske Møllgaard’s essay on problems of language and logic in Daoism shows how the Daoist experience of language has been misunderstood from the point of view of the Confucian logic of correctness. He also tries to explain the Daoist experience of language on the basis of the second chapter in the Zhuangzi. He shows that Daoists do not simply negate the logic we impose on language. Daoists accept the law of non-contradiction, but do not immediately brush aside language that falls into contradiction. They tarry with the contradiction

10

Y.-m. Fung

and show its strangeness in order to go beyond but not negate the logic we impose on language and experience the logic of language itself. So, his conclusion is that this logic of language itself, a logic beyond logic, is exhibited in Daoist stories. Chien-hsing Ho in his essay on paradoxical language in Chan (Zen) Buddhism sketches some key ideas of Chan Buddhism and examines the use of paradoxes in the sermons associated with certain masters of the southern Chan in Tang 唐 dynasty. Based on this historical understanding, the further step is to analyze the presence of paradoxical language in post-Tang encounter dialogues. He concludes that we must not be misled by the surface grammar of the expression into supposing the presence of contradiction therein. As we dig deeper and consider language’s abundant wealth of meaning and the various ways in which words can be used, we would refrain from making the supposition and even learn to appreciate Chan’s creative and skillful use of words. Although the use of logical argumentation and the theorization of logical thinking are not absent in ancient Chinese texts, there is no systematic theory originally developed by Chinese philosophers. Until seventh century when some Buddhist texts about yin-ming 因明, the science of reasons, were translated into Chinese. That is a kind of logical thought transplanted from India through the acculturation of Buddhism in China. Later in the Ming dynasty, the Christian missioners from the West went to China. Their main purpose for a long journey to a remote country is to do missionary work, but they also brought Western Science, including Aristotelian logic and mathematics to China. In part (III) we have three essays on these great phenomena of acculturation. Minjun Tang in his essay introduces some basic facts and ideas of yin-ming and gives a formal analysis for some argument forms in this tradition while, on the other hand, the other two essays written by Jinmei Yuan and Rafael Suter, respectively, are about the intellectual engagement of logical and mathematical thought between the West and China. Through the discourse between Jesuits and Chinese scholars in seventeenth century China, Yuan discovers some significant progress in building a mutual understanding between the West and the East at a level of logical practices. Through an overview on the development of the academic discipline from digestion to institutionalization, Suter’s essay explains how the arrival of foreign thought leads to the (re-)discovery of logic in traditional China. In the last part (IV) of this Companion, we have two essays which aim at giving a historical survey on logic studies in mainland China and Taiwan, the two main Chinese communities, and introducing the current state of researches and achievements in these two regions. The first essay on logic studies in contemporary mainland China is a co-contribution of Guoping Du and Hongguang Wang and the second one on logic studies in Taiwan is written by Wen-Fang Wang and Zeqiang Wu. These two reports give us a general picture of the present state of research in the field in Chinese communities.

1  Introduction: Chinese Philosophy of Logic

11

References Cheng, Chung-ying. 1965. “Inquiries into Classical Chinese Logic.” Philosophy East and West 15(3/4): 195–216. Chmielewski, Janusz. 2009. Language and Logic in Ancient China. Warszawa: Komitet Nauk Orientalistycznych PAN. (Reprinted from Rocznik Orientaistyczny (Warsaw 1962–69)) Davidson, Donald. 1974/1984. “Ch. 13: On the Very Idea of a Conceptual Scheme.” In Inquiries into Truth and Interpretation. Oxford: Clarendon Press. Fung, Yiu-ming. 2010. “On the Very Idea of Correlative Thinking.” In Philosophy Compass (Blackwell) 5(4): 296–306. Fung, Yiu-ming. 2020 (forthcoming). “Reference and Ontology in the Gongsun Longzi,” Worlds of East Asia, Swiss Asia Society. Geach, P.T. 1967. “Identity.” Review of Metaphysics 21: 3–12. (Reprinted in Geach 1972. Logic Matters. Oxford: Blackwell: 238–247.) Graham, A.C. 1978. Later Mohist Logic Ethics and Science Hong Kong: The Chinese University Press. Graham, A. C. 1992. Unreason Within Reason: Essays on the Outskirts of Rationality La Salle: Open Court. Granet, Marcel. 1934. La Pensée chinoise. Plusieurs Rééditions Hall, David and Rogers Ames. 1995. Anticipating China: Thinking Through the Narratives of Chinese and Western Culture. Albany: State University of New York Press. Hansen, Chad. 1985. “Chinese Language, Chinese Philosophy, and ‘Truth’.” The Journal of Asian Studies, 44(3): 491–519. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. New York: Oxford University Press. Harbsmeier, Christoph. 1998. Science and Civilisation in China, vol. 7, part 1, Language and Logic in China, edited by Joseph Needham. Cambridge: Cambridge University Press. Hu, Shih. 1931. The Development of the Logical Method in Ancient China. Shanghai: Commercial Press. Jacquette, Dale. 2007. “Introduction: Philosophy of Logic Today.” In Dale Jacquette ed., Philosophy of Logic (1–12). North Holland: Elsevier. Needham, Joseph. 1956. Science and Civilisation in China, vol. 2: History of Scientific Thought. Cambridge: Cambridge University Press. Priest, Graham, J. C. Beall, and Bradley Armour-Garb (eds.). 2004. The Law of Non-Contradiction: New Philosophical Essays. Oxford: Clarendon Press. Quine, W.V. 1948. “On What There Is.” Review of Metaphysics: 21–38. (Reprinted in Quine1953. From a Logical Point of View. Cambridge, Massachusetts: Harvard University Press.) Russell, Bertrand. 1905. “On Denoting.” Mind 14: 479–493. Searle, John. 2010. Making the Social World: The Structure of Human Civilization. Oxford: Oxford University Press. Shen, Youding. 1980. Logical Study in the Mohist Canons (Mojing di luoji-xue 墨經的邏輯學). Beijing: Chinese Social Sciences Press. (Reprinted from Shen’s essay of the same title in Guang Ming Daily 光明日報 1954.) Yiu-ming Fung  is Emeritus Professor at the Hong Kong University of Science and Technology and currently Visiting Professor in the Department of Philosophy at the Tunghai University, Taiwan. He is an analytic philosopher with special interest in Chinese philosophy and comparative philosophy. Fung is the author of several books, including The Methodological Problems of Chinese Philosophy (1989); Chinese Philosophy in the Ancient Period, 4 volumes (1992); The Gongsun Longzi: A Perspective of Analytic Philosophy (1999); and The Myth of Transcendent Immanence: A Perspective of Analytic Philosophy on Contemporary Neo-­Confucianism (2003). He has also published more than 100 research papers both in Chinese and English.

Part I

Concepts in Chinese Philosophy of Logic

Chapter 2

What Is Ming 名? “Name” Not “Word” Jane Geaney

It is commonplace to characterize the philosophy of Early China (ca. 500 b.c.e. to 200 c.e.) as attuned to interconnections and informed by the implications of situations and contexts. This portrait of early Chinese philosophy might encourage us to speculate about whether the same was true of concepts of “language” in Early China. Here I put “language” in quotation marks because, as I argue below, it is doubtful that modern ideas of the abstraction “language” are appropriate for understanding those of Early China. At the same time, it is crucial to rethink the habit of treating the foundational term of Chinese logic, ming 名, as meaning “word.”1 Because “word” is a linguistic abstraction, it is not a promising starting point for seeking to comprehend conceptions of ming in early Chinese texts.2 No single translation equivalent suffices, but “name” is a better option. In order to understand ming, much depends on our grasp of technical uses of translation equivalents. In scholarship on linguistics and philosophy of language, “words” and “names” are importantly different. The term “word” typically implies the abstract model of “word-tokens” and “word-type,” the latter being something static and uniform, regardless of its definitional criteria. A “name” is generally a means by which people refer to something. A “word” is the kind of thing that can be found in a dictionary, and words encompass functional terms like “in.” Again, by 1  Because certain models of “language” treat “word” as an important minimal unit of language, my comments about “language” extend to “word.” 2  The colloquial use of “word” to mean a bit of speech is not a problem in this regard, but below I focus on two other important uses of “word” that make it inappropriate as a translation equivalent of ming: “words” as having semantic meanings, and “word” as a unit of a language system. For a more in-depth analysis of contrasts between “word” and “name,” as well as an examination of the implications of using these terms to translate ming 名, see the introduction to Geaney 2018 and Geaney forthcoming.

J. Geaney (*) Department of Religious Studies, University of Richmond, Richmond, VA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_2

15

16

J. Geaney

contrast, while dictionaries include functional terms, they generally omit most names. Dictionaries comprise the words, not names, of a given language. In effect, names do not belong in dictionaries because names merely refer to entities and do not have “meanings.” (I put “meanings” in quotation marks because meaning itself is a notoriously slippery concept.)3 If we mistakenly conceive of ming through the retrospective framework of a zi 字 (graph), a ming might look more like a word than a name. It is anachronistic, however, to think that early Chinese texts used the term ming to mean what is later meant by zi (Geaney 2010).4 Terms like “word,” “language,” and “reality,” have specific histories that might prove alien to the linguistic and epistemological world of Early China. In texts from Early China, human knowledge largely derives from hearing and seeing (Geaney 2002: 50–83). The ears hear things like names, fame, speech, and music, whereas the eyes see things like walking, action, forms, shapes, colors, and patterns. Hence speech and writing—i.e. what modern readers might think of as “language”— inhabit opposing sides of an aural/visual polarity without resulting in a concept of a unified field that is “language.” Nevertheless, particularly when discussing the Later Mohists and the Xunzi’s “Zhengming” chapter, scholars typically treat ming as “words” (units of “language”) with “reality” as their primary pair term.5 If, however, in these contexts we treat ming as “name,” with its function being reference, much historical accuracy is gained and nothing of the texts’ coherence is lost. Thus, I propose that whenever passages about ming and shi seem to concern ideas about “language and reality,” we should reconsider our translations. 3  I am grateful to comments from the reviewers asking me aware to clarify my position on ming and semantics. As Jeffrey Stout argues, the twentieth-century trend in analyzing “meaning”—starting with European figures like Gottlob Frege, Max Weber, and Sigmund Freud—has produced unnecessary confusion and debate (Stout, 1982). The “meaning” of words is no exception to this lack of clarity. If we nevertheless assume the obviousness of word-meanings as a feature of ming in early Chinese texts, we would reasonably expect the texts to employ a term to mean “word-meaning.” But no such term is immediately evident. In retrospect, we might hypothesize that yi 意 and yi 義 would be likely candidates. My investigation of the emergence of the idea of word-meaning in Early China—a monograph in which I analyze the usage of yi 意, yi 義, ming 名, shi 實, yan 言, and zi 字, as well as other terms—recovers a different and more complex picture. In short, although the term yi 意 often occurs in relevant contexts, yi 意 are merely expressed by speech (yan 言)— and not at all by ming. In addition to raising questions about the precise relation of ming to yan, the account of yi 意 that I reconstruct intersects interestingly with the later development of the use of yi 義 in semantic contexts, as distinct from strictly ethical contexts. See Geaney forthcoming. 4  Recognizing that a ming is not a zi 字 has consequences for philosophy of “language” in early Chinese texts: at the very least, it undermines the likelihood that those texts attribute word-meanings to ming. See Geaney 2010. See also Geaney 2018, chapters 3–5 and 7–8, where I discuss this idea in terms of Chad Hansen’s Daoist Theory of Chinese Thought, which has been the standard text in Anglophone scholarship on early Chinese philosophy of language since its publication in 1992. 5  In Anglophone scholarship, “reality” still seems to be the most common translation equivalent for uses of shi 實 in early Chinese texts, but the related term, “actuality,” also appears frequently. See chapter 2 of Geaney 2018 for an extended argument for why “reality” is a misleading translation of shi: my argument there is largely, but not solely, based on the unlikely consequence that all preBuddhist Chinese texts take emptiness to mean unreality.

2  What Is Ming 名? “Name” Not “Word”

17

In the process of drawing attention to the implications of presupposed obvious concepts, we can also become more aware of the particularities of ming and the specific features associated with its correlates in early Chinese texts.6 In this chapter, I survey a broad range of early Chinese texts to undermine the apparent self-­ evidence of translating ming as a unit of “language” (or “word”) and its correlates as “reality.” While leaving ming untranslated, I show that ming often functions somewhat like the term “name.” Names point to, pick out, or indicate referents, but names do not have “meanings.” That is, whereas words have a conceptual aspect, names do not. Through emphasizing these distinctions between “word” and “name,” my overview of the usage of ming foregrounds features of early Chinese philosophical discussions that are unusual in relation to the larger history of philosophy of language and logic. Their special significance lies in the nature of both the unit of analysis (ming) and that of its correlates (shi 實, shi 事, xing 形, xing 行, and shen 身). The important aspects of ming and its correlates in early Chinese texts can be elaborated as follows. Again, like the term “name,” the texts use ming for referring. That is, as with most concepts of “names,” ming are conceived as having “referents” rather than what we might call “word-meanings.” In addition, as in some uses of “name,” early Chinese texts use ming to mean renown—the result of a name being repeatedly heard. We should not miss the importance of this overlap by which something’s name is also its reputation. While such intersecting usages have no particular significance in the dominant Western philosophical tradition, in early Chinese texts, this aspect of ming contributes to the impression of ming being audible, which makes the term different from what is often presumed about “word” and “language.”7 Other differences between ming and “words” become evident if we consider the respective entities to which they are correlated. In the context of frequent audible/ visible pairings, texts from Early China presume that ming are paradigmatically employed to name items in the cosmos that are visible, which contributes to demonstrating that ming were understood to belong to the sphere of sound.8 Not all sounds are ming, of course. A ming is a particular kind of sound that refers, typically to a visible correlate.9 Among the regular correlates of ming, we find terms used for a variety of things that are considered visible: shi 實 (actions/things), shi 事 (events/ 6  In other words, like Cao Feng 曹峰, I believe the study of traditional Chinese thought has been “thoroughly influenced by the West,” and like him, I advocate rethinking the role of ming 名 in texts from Early China through a more appropriate lens. My contribution is an embodied approach; his is that of the history of political thought (Cao 2008: 220, 224–25). 7  See Geaney 2010. We do not usually take “words” to be restricted to sound unless modified as “spoken words,” and we do not typically understand “language” to be restricted to sound unless modified as “spoken language.” 8  To be clear, my point is that, in its linguistic usage in early Chinese texts, a ming is a sound (typically made by humans) that paradigmatically refers to a visible thing. I am not saying a ming is the sound of a visible thing. 9  The idea that ming paradigmatically refer to sound does not mean it is impossible for ming to refer to other ming (or sounds of other sorts), although the notion of applying ming to ming might have seemed a bit puzzling from the perspective of early Chinese texts (Geaney 2018, 192).

18

J. Geaney

matters), xing 形 (shape), xing 行 (actions), and shen 身 (body-person).10 None of these, I argue below, quite approximate “reality.” In the following overview, I contend that the role of ming in audible/visible pairings guarantees ming some value vis-à-vis its correlates, which is one reason why “reality” is not an apt choice for translating its correlates. That is, the importance of ming depends on the worth of audible things as distinct from the value of visible things. As I explain below, texts from Early China associate audible things with the lighter, higher aspect of the cosmos, where movement occurs easily. By contrast, they associate more stable, visible things with the denser, lower sphere. Insofar as early Chinese texts affirm the value of ming, it is often due to their association with “high” things like honor (especially after death) in contrast to terms associated with vision that represent life’s creature comforts and enhancements, which are sometimes presented as inferior.11 As a result, while an expression like “language and reality” participates in the denigration of language (insofar as the phrase itself implicitly depicts language as external to reality), in pairings of ming with its correlates, the value of ming is more secure. Of course, this does not preclude contempt for ming in early Chinese texts, but it explains why that is less common than we might imagine.

1  Ming as Sound Several uses of ming in early Chinese texts indicate that ming should be understood specifically as aural/oral. Again, this constitutes an important contrast to what we might think of as “words,” consisting of types that can be tokened in writing as well as in speech.12 For instance, the interchange of ming 名  and ming 命 (command) implies a vocally articulated command to which one listens (ting 聽) as if obeying instructions. There is an embodied logic to this.13 The commanding effect of ming 名 and  For the argument that these terms are used to mean something paradigmatically visible, see the appendix to Geaney 2018, and for the case of shi 實 in particular, see also chapter 2 of that volume. The pairing of ming with wu 物 is not as common in pre-Qin texts. See n. 20. 11  Despite some potential resemblances, this is not a mind/body contrast. The terms are more like fame/fortune, or more precisely, honorable fame/fortune (Geaney 2018, chapter 2). 12  Moreover, the line between sounds that are linguistic and sound in general is less finely drawn than one might expect. I explore the implications of this for the idea of zhengming in Geaney 2011. 13  To listen is to obey, perhaps because one cannot shut one’s ears as easily as one’s eyes. Several examples should suffice to remind us that interchanges of the two characters for ming were more prevalent in Pre-Han texts than the effects of later editing would lead us to believe. The variant ming in the opening lines of the Beijing University bamboo slip manuscript of the Laozi (which has 名可命 instead of 名可名) recently stimulated new political reflections on the early Chinese concept of ming in the work of Cao Feng 曹峰 (Defoort 2013 and Cao 2011). Other examples of interchanges of ming 名 and ming 命 are common in the Mo Bian chapters of the Mozi. They also occur in discussions like this one from the Zuozhuan. 10

2  What Is Ming 名? “Name” Not “Word”

19

ming 命 suggests summoning by name. To call people by name is to require them to pay attention. The very recognition of one’s name being called entails a kind of compliance. In this sense, to name is to command. Furthermore, a personal ming also originates as a vocalization. In this passage from the Liji, it is evident that some kind of sound constitutes the naming of a child: 適子庶子見於外寢, 撫其首, 咳而名之。 A [second] son or any other son by the wife proper was presented in the outer chamber, when [the ruler] laid his hand on its head, and with gentle voice named (ming 名) it. (Legge trans. 1879, 474–75.) Liji 禮記 〈內則〉12.46

We can also see that ming is rooted in sound in the depiction of social protocol in the later text, the Baihutong (first century), wherein people show others their motivations, on the one hand, by visual gestures (like saluting and bowing) and, on the other, by offering their ming to tuqing, metaphorically coughing up what they are about. 人必有名何?所以吐情自紀, 尊事人者也。 Why must people have ming 名? To spew forth their qing 情 (feelings/motivations) for the reverential service of others. Baihutong 白虎通 姓名

Moreover, in the context of Early China, the use of ming as “fame” implies transmission through hearing, which is particularly evident when such uses occur in combination with sheng 聲, a term used to mean sound. 夫三王五帝之治天下不同, 其係聲名一也。 The Three Kings and Five Sovereigns did not rule the world in the same way, but the shengming 聲名 that accrued to them is one. Zhuangzi 莊子 天運第十四

Thus, ming involve the mouth and the ears. Furthermore, in some cases where the context would lead us to expect to see ming, sheng 聲 occurs instead. Because ming often contrasts to (or pairs with) shi 實, when sheng 聲 appears instead of ming, we can infer a similarity between sheng and ming.14 In other words, the audible aspect of ming is evident insofar as ming and

名有五, 有信, 有義, 有象, 有假, 有類。 以名生為信, 以德命 [variant名] 為義, 以類命為象, 取於物為假, 取於父為類。 Names have five [sources]: trust, yi 義 (normative modeling), xiang 象 (resemblance), borrowing, and kind. To name from birth is trust; to name (ming 命) from de 德 (power/virtue) is yi 義 (normative modeling); to name from a kind (lei 類) is xiang 象; to pick [a name] from a thing is borrowing; to pick [a name] from the father is [of a] kind (lei 類). Chunqiu Zuozhuan 春秋左傳 桓公 B 2.6《桓公六年傳》 All citations to early Chinese texts are to the CHANT 漢達文庫 database. Translations are my own unless otherwise noted. 14  Whether these are contrasts or correlations is not important for my purposes. The presence of the two sides in the pattern is what matters.

20

J. Geaney

shi pairs seem to be another version of sound and shi pairs, as in the following examples: 嚴威不能振, 惠厚不能供, 聲實有閒也。 If [the ruler’s] severity and awe are not capable of terrifying, and [his] kindness and generosity are not capable of supplying their needs, there will be a gap between the ruler’s sheng 聲 and shi 實. Guanzi 管子卷第十 君臣上第三十 是刑名異充而聲實異謂也。 This is an example of shapes (xing 刑) and ming being different in fullness, and of sheng 聲 and shi 實 referring differently. Lüshichunqiu 呂氏春秋 先識覽第四 《正名》 其於義也聲善而實惡 。 With regard to duty, this is good in sheng 聲, but bad in shi 實. Shiji 史記 世家 《趙世家第十三》 嬰聞察實者不留聲, 觀行者不譏辭。 I, Ying, have heard that those who examine the shi 實, do not heed the sheng 聲; those who observe the actions (xing 行) do not criticize the phrases (ci 辭). Lüshichunqiu 呂氏春秋 先識覽第四 《觀世》

In sum, the uses of ming in early Chinese texts that involve calling, coughing, voicing, and being heard testify to ming to be something fundamentally sound-based. The fact that early Chinese texts paradigmatically use ming to mean something that falls within the scope of sound—and furthermore something that matches things in the sphere of sight—makes it unlikely that any use of ming would be a term for a visible thing. Consequently, we can expect that the term ming was not readily employed to mean “graph” or even “word-type” (Geaney 2010). Moreover, as I argue below, contrasts of ming to something else are not likely to concern “language and reality.”

2  Ming Within Binary Pairings of Sounds and Sights In the early Chinese cosmos, earth and heaven constitutes a pair that aligns—in a yin-yang correlation—with two different modes of sensing things. Exploring this alignment sheds light on uses of ming. The subtle involvement of hearing and seeing within heaven/earth correlations is evident in a passage from the Liji that describes the following polar sides: warm versus cold, growth versus storage, and music versus ritual action.15 More specifically, the passage aligns warm weather with music, creation, growth, ren 仁 ­(kindliness), and harmony. By contrast, it aligns cold weather with ritual action (li 禮), withholding, storage, yi 義 (dutifulness), and separation.

 For a discussion of the visibility of li 禮 and my reasons for translating li as “ritual action,” rather than “ritual,” see Geaney 2018: 141–166.

15

2  What Is Ming 名? “Name” Not “Word”

21

春作夏長, 仁也; 秋斂冬藏, 義也。仁近於樂, 義近於禮。樂者敦和, 率神而從天, 禮者 別宜, 居鬼而從地。 In the spring there is creation and in the summer there is growth. This is ren 仁 . In the autumn there is holding back and in winter there is storing. This is yi 義. Ren is close to music. Yi is close to ritual action. Music is kind and harmonious. It leads spirit-souls and follows heaven. Ritual action is separating (bie 別) and appropriate. It stores ghost-souls and follows the earth. Liji 禮記〈樂記〉19.6

As these references to souls imply, in the largest frame, the contrast involves the cosmos itself: heaven and earth. The senses enter the picture in another passage in the Liji that emphasizes the importance of music. It asserts that all sound is yang. Thus we can infer that all sound must belong with heaven, because heaven is yang. 凡飲, 養陽氣也; 凡食, 養陰氣也。故春禘而秋嘗; 春饗孤子, 秋食耆老, 其義一也。而 食嘗無樂。飲, 養陽氣也, 故有樂; 食, 養陰氣也, 故無聲。凡聲, 陽也。 All drinking nourishes yang qi 氣; all eating nourishes yin qi. Therefore, there were the spring sacrifices and autumnal sacrifices. When feasting the orphaned young in spring and the feeding the aged in autumn, the model was the same. But in feeding the aged and at the autumnal sacrifice, there was no music. Drinking nourishes yang qi and therefore it occurs with music. Eating nourishes yin qi, and therefore it does not occur with sound. All sound is yang. Liji 禮記〈郊特牲〉11.3

In this formulation, sounds are fluid, yang, warm, and heavenly.16 Although the passage does not articulate it, by implication ritual action (li 禮) belongs with earth, which is yin.17 The plausible contrast to hearing sounds, which are fluid and dispersed, is seeing visible things (or their movements), which fill space in a more constrained way. Hence, because the Liji aligns ritual action to things that are yin, and because it asserts that yue 樂 (music) is yang, ritual action (as visible) is likely to be what lies at the other end of the continuum from music (as audible).18 樂由陽來者也, 禮由陰作者也, 陰陽和而萬物得。 Music is that which arrives from yang. Ritual action (li 禮) is that which makes from yin. Yin and yang harmonize and the ten-thousand things attain. Liji 禮記 〈郊特牲〉11.5

In sum, the yin-yang correlations here are as follows: Yin corresponds to condensed, cold, and earthly things that we can see. Yang corresponds to liquid, warm, heavenly things that we can hear.19 Thus, in being audible, ming aligns with things that are associated with heaven or the upper reaches of the cosmos. 16  The term used for sound, sheng 聲, is used for speech as well as other kinds of sounds (Geaney 2011). 17  The associations of yin/yang in binary constellations are still somewhat inconsistent in texts from Early China. This is the dominant pattern, but the constellations take two different forms. For a more detailed discussion of these binary constellations, see Geaney forthcoming. 18  In other words, in light of contrasts and parallels of hearing and seeing, assertions about li 禮 being visible make it likely that li is what is implied as the other pole to sound as yang. 19  Hearing and smelling (wen 聞) are often associated, perhaps because of their vacuity, as implied in this comment in the Huainanzi.

22

J. Geaney

In keeping with these yin/yang pairs of sight and sound, the world in early Chinese texts consists of things that have audible and visible aspects, which also has implications for uses of ming. Texts from Early China describe “things” (wu 物) in these terms: 物至而命之耳。發於名聲, 凝於體色, 此其可諭者也。 Things arrived, and [ancient sages] named (ming 命) them, that is all. What issued in names and sounds (mingsheng 名聲) or congealed in body and face, these were their [things] to be proclaimed. Guanzi 管子卷第十三 白心第三十八

The passage implies, on the one hand, that things (wu 物) are that which is named, and on the other, that things (wu 物) have two aspects: audible (not congealed) and visible (congealed).20 A thing (wu 物) is also sometimes described in terms of its ming and shi 實. 有名有實, 是物之居。 无名无實, 在物之虛。 It has a ming and it has a shi, this is residence of a thing. No ming and no shi, this is in the emptiness of a thing. Zhuangzi 莊子 則陽第二十五

The line specifically concerns a thing’s location or “residence,” which is constituted by the presence of both ming and shi.21 The idea that things consist of audible and visible aspects helps explain why the Guanzi would say that sound is one half of what is needed to rule the world: 故明王懼聲以感耳, 懼氣以感目, 以此二者, 有天下矣, 可毋慎乎? Therefore an astute ruler is diffident of sound as it stimulates the ear and diffident of qi as it stimulates the eyes. With these two, one possesses the whole world, so how could one not be careful? Guanzi 管子卷第十一 小稱第三十二

鼻之所以息, 耳之所以聽, 終以其無用者為用矣。  hat by which the nose breathes, that by which the ear listens: in the end it treats that which has T no use as useful. Huainanzi 淮南子 說山訓  Uses of the term wu 物 suggest an association with visual things in particular, but also a general application, so wu 物 might also be a general term. Both uses are operative in this example: 20

聲一無聽, 物一無文, 味一無果, 物一不講。 If sounds are all one, there is no listening. If things (wu 物) are all one, there is no pattern. If tastes are all one, there is no fruit. If things (wu 物) are all one, there is no thoroughness. Guoyu 國語 鄭語 《史伯為桓公論興衰》 In this sequence of sound, sight, and taste, the first occurrence of wu aligns it with vision (specifically in contrast to sound), but the second use does not link it to any particular sensory faculty. See appendix to Geaney 2018 for more examples and a longer discussion of visual associations in uses of wu 物. 21  Given this idea that the residence/location of a wu consists in having both a ming and a shi, perhaps uses of the term wu align more closely with vision than hearing because discerning something’s whereabouts poses less of a challenge to the eyes than the ears. See Geaney 2016 for an exploration of how, in early Chinese texts, location and movement determine the nature of things.

2  What Is Ming 名? “Name” Not “Word”

23

In addition, the idea that the world has these two aspects is implicit in passages that call for balancing ming and shapes: 君操其名, 臣效其形, 形名參同, 上下和調也。 The ruler upholds the ming, the ministers bring about the shape (xing 形). When the shape and the ming are comparable, above and below are in peace and harmony. Hanfeizi 韓非子 揚權第八

A version of this passage in the Huainanzi criticizes over-emphasizing the audible aspect to the detriment of the visible: 上操其名以責其實, 臣守其業以效其功, 言不得過其實, 行不得踰其法。 天下多眩於名聲, 而寡察其實, When superiors uphold their [officials’] ming to create accountability for their accomplishments (shi 實), the officials care for their affairs to make their work efficient. Speech (yan 言) is not allowed to go beyond accomplishments (shi 實); action (行 xing) is not allowed to exceed method (fa 法). But the world is mostly confused by ming and sounds (mingsheng 名聲), and rarely checks their shi 實. Huainanzi 淮南子 主術訓

In this passage, the audible aspect of the world is indicated by ming, speech (yan), and sound (sheng). The visible aspect is the accomplishments (shi) and actions (xing).22 Early Chinese cosmology aligns hearing and seeing with binaries like heaven and earth. Both the world and the things within it may be divided into audible and visible. In terms of these two modes of sensing, while the world’s things are named (ming), the things themselves also seem split into those aspects that are visible, and those that are audible (including names).

3  Comparative Value of Sounds and Sights We should not be misled, however, by comments that denigrate ming and speech like the one in the Huainanzi passage just discussed. In texts from Early China, sound is not necessarily less worthy than its visible counterpart, particularly because uses of ming often imply honor. Interestingly, this feature of ming is prominent in contrasts of farming to trading. Early Chinese texts use ming in ways that suggest farming is honorable whereas trading is disreputable. The Guanzi invokes this ming/ shi contrast in a description of the origin of trade. 地重人載, 毀敝而養不足, 事末作而民興之, 是以下名而上實也。 When the land became heavy with people, it was ruined, spoiled, and insufficient for nourishing. Trade was created and people made it flourish. This was treating ming as below and shi as above. Guanzi 管子卷第十二 侈靡第三十五

 My point is not that every term used as a pair (or contrast) with one that is associated with vision will necessarily involve an aural aspect, and I have not investigated whether such uses of fa 法 (method) with xing 行 suggest specific sensory functions.

22

24

J. Geaney

As the passage implies, because trade is dishonorable, to develop trade amounts to devaluing one’s ming. In this usage, the term ming is not simply a reputation: it is a good reputation. The unstated value of ming in its use as “good reputation” seems to arise from hard work or effort, on the presumption that effort does not feature in acts of trade. A passage in the Zhanguoce confirms that traders, who reap the rewards without the labor, have the shi without the ming, which is a condition unworthy of respect because it does not entail anything like blistering one’s back from weeding: 有其實而無其名者, 商人是也。無把銚推耨之勢, 而有積粟之實, 此有其實而無其名 者也。無其實而有其名者, 農夫是也。解凍而耕, 暴背而耨, 無積粟之實, 此無其實而 有其名者也。無其名又無其實者, 王乃是也。已立為萬乘, 無孝之名; 以千里養, 無孝 之實。 Those who have the shi but not the ming are the merchants. None of them has ever exerted himself using a hoe or pulling a rake, yet they possess the shi of accumulated grain. This is lacking the ming but owning the shi. He who has the ming but not the shi is the farmer. When the frost thaws, he turns over the earth, and blisters his back to weed. [But] he owns none of the shi of accumulated grain. This is lacking the shi but having the ming. The one who has neither ming nor shi is your majesty; you stand on [a state of] ten-­thousand chariots, but have no ming for filial piety; with a thousand li of [territory’s] nourishment, you have no shi of filial piety. Zhanguoce 戰國策 《秦四》 秦王欲見頓弱

The situation referred to here is a bit obscure, but we can see that the passage highlights the ruler’s lack of ming by contrasting it to his ample chariots, whereas it highlights his lack of shi by contrasting it to his ample means of producing nourishing goods. Setting aside the enigmatic role of the chariots, it is noteworthy that shi is a question of agricultural nourishment, and having ming is a matter of being worthy of respect. In all of these agricultural contexts, ming signifies honor. Moreover, while the Zhanguoce passage is the least clear, in all of these cases ming seems to emerge from action: specifically, hard-won efforts. The presumption that ming is rooted in action might explain why the term ming is used to mean “good reputation,” not just “reputation.”23 By contrast, while shi might sometimes involve hard work (like the shi of filial piety here), it also can be something transmitted to people like traders who do not work at all. In other words, the term shi is used to mean the fruits of hard work, as well as the hard work itself, whereas in this sort of ming/shi contrast, ming is necessarily the result of hard work. In that sense, in comparison to ming, shi is less directly linked to virtue. We can also account for the value of ming in another way: by examining examples that characterize the things to which it contrasts. A passage in the Guanzi, for instance, begins by discrediting striving for wealth and abundance in this way: 盛而不落者, 未之有也。故有道者不平其稱, 不滿其量, 不依其樂, 不致其度。爵尊即 肅士, 祿豐則務施, 功大而不伐, 業明而不矜。 Never has there been a person too concerned with wealth who did not fall. Consequently, those who possess the Way do not insist that the steelyard be exactly level, the grain mea23

 This is not to deny that early Chinese texts use e ming 惡名 to mean a bad reputation.

2  What Is Ming 名? “Name” Not “Word”

25

sure be completely full, the calculations be strictly adhered to, or metrical standards be carried out to the limit. Noble ranks are honored so that gentlemen are respected. Salaries are plentiful so that duties are assiduously performed. Though the prince’s achievements be great, he does not make a display of them. Though his cultivation be brilliant, he never boasts of it. Guanzi 管子卷第四 宙合第十一 (Rickett trans. 1985: 210).

The passage then notes that a long-standing tension between ming and shi accounts for their separation, and that a wise person chooses one over the other. 夫名實之相怨久矣, 是故絕而無交。惠者知其不可兩守, 乃取一焉, 故安而無憂。 Now antagonism between ming and shi has been around for a long time. For this reason, they are severed and have no interaction. Wise people know that they cannot preserve both. Thus they choose one of them. Therefore they are comforted and lack worries. Guanzi 管子卷第四 宙合第十一

With its unspoken choice, the passage tacitly affirms the wisdom of preferring ming over shi, but to discern why the implied correct choice is ming we must examine more contrasts of ming and shi. Claims about ming and shi do not always depict them as antagonistic, but they are always separated, in the sense that a ming is not a shi and vice versa. Methodologically, it is possible to read the Guanzi’s comment about their antagonism as unconnected to the prior lines in the passage—in which case, we would have no grounds for guessing whether the preferred option is ming or shi, nor would we have any way to justify how to interpret its uses of ming and shi .24 If, on the other hand, we choose to approach the passage as a coherent whole, we would have grounds for claiming that its use of ming does not resemble a “word.” The occurrence of ming would signify a reputation, that is, a “name.” The context would also suggest that the shi to which the ming is juxtaposed is not functioning to mean “reality.” As in the use of shi to mean “fruit,” in this passage, it would refer to the various forms of “filling out” implied by the earlier elaborations of sheng 盛: that is, greedily seeking one’s full measure and being reluctant to pay others what they are due. The point would be that such acts are at odds with ming: lust for fortune does not promote a good name.25 A comparable antagonism occurs in the Shangjunshu. The fact that, in this similar case, the correlate of ming is not shi supports my “filling” reading of the uses of sheng 盛 and shi in the Guanzi passage. The argument, which goes on to advocate farming over trade, presents ming and its correlate as goals pertaining to death and life respectively. 故民生則計利, 死則慮名。名利之所出, 不可不審也。利出於地, 則民盡力; 名出於戰, 則民致死。 Thus the masses, with regard to living, scheme for profit, and with regard to dying, think of ming. We cannot fail to examine the source of profit and ming. If profit comes from the soil,

 As others have argued, the coherence of all texts from Early China is in question because of the nature of the textual composition. 25  The passage also concludes that boasting does not help one’s reputation either, but that part does not appear to have any link to the shi to which the passage refers. 24

26

J. Geaney then people will exhaust their strength [on it]. If ming comes from battles, then the people will push to the death [in them]. Shangjunshu 商君書 筭地

In this situation, ming derives from effort in battlefields, rather than effort in agricultural fields, but it remains a matter of hard work producing honor. Here the antagonism of ming and its correlate can be understood in terms of the contrast between life and death. When it comes to dying, people value ming. When it comes to living, people value profit. These two values are somewhat at odds because, whereas ming is measured by the honor of dying bravely, profit is measured by comforts that enhance life. The same chapter contrasts robbers, who are motivated by profit, with scholars, who are motivated by ming. In its caricature of the scholarly pursuit of ming, it notes that scholars sacrifice the comforts of life in their efforts to seek ming: 衣不煖膚, 食不滿腸, 苦其志意, 勞其四肢, 傷其五臟, 而益裕廣耳, 非生之常也, 而為之 者, 名也。 Their clothes do not warm their skin, their food does not fill their stomach, they embitter their aims and intentions, work their four limbs, and harm their five internal organs and yet they increase in abundance and extent—not for the constancy of life, but for the sake of ming. Shangjunshu 商君書 筭地

Thus the antagonism between ming and its correlate can be a tension between honor—implicitly earned by strenuous efforts—versus the physical comforts of life. These are not mutually exclusive insofar as strenuous efforts can also result in physical comforts, but my point here is that when a person gains a ming, it entails unpleasant strain and effort. The enjoyments of profit, wealth, and the general idea of “fullness” are somewhat inimical to the weariness or emaciation that occurs in earning a ming. Reading the Guanzi passage’s ming/shi antagonism in light of this, we are presented with a broad contrast between on the one hand, rarified honor, and on the other, the fulsome physical enhancements of life. Thus, although the Guanzi passage does not explicitly articulate which thing the wise person would choose, the context suggests an answer (again, assuming we read the passage as a coherent piece). Ming is the preferred choice, because it is a good name. By contrast, the shi in question implies preoccupation with wealth or abundance. Thus, ming and shi can be antagonistic, not because words or language are unreliable or false, but because honor and abundance are not always entirely compatible. Early Chinese texts also separate ming from its counterpart by using honor as the criterion to assess ming, while using something different to assess its counterpart (in its various forms). For example, a passage in the Guanzi describes the different effects of virtue on one’s ming and on one’s body-person (shen 身). 故(之)身者, 使之愛惡; 名者, 使之榮辱。 Thus, with regard to shen, it causes one to be loved or hated; with regard to ming, it causes one to be illustrious or disgraced. Guanzi 管子卷第十一 小稱第三十二

Here virtue has a dual effect: being loved, in terms of one’s body-person, and being illustrious, in terms of one’s ming. The correlate to an honored ming is a loved body-­

2  What Is Ming 名? “Name” Not “Word”

27

person. In the following Hanfeizi example, someone’s military advice frames ming as a reputation for dutifulness correlated with the fruits of revenge: 此義於名而利於實, 故必有〔為〕天子誅之名, 而有報讎之實。 This will be dutiful in ming and profitable in shi. In consequence, Your Highness would have the ming of punishing the disobedient on behalf of the Son of Heaven and the shi of revenge. Hanfeizi 韓非子 外儲說左上第三十二

The twofold nature of this outcome—a dutiful ming and some kind of pleasurable profit in shi 實—is a common refrain. It also occurs in the Guanzi, which says that kingly rulers do not pursue profit and virtuous people do not pursue ming. The implication is that kingly rulers and virtuous people possess these things, but they would not have them if they had sought them. 釣名之人, 無賢士焉; 釣利之君, 無王主焉。賢人之行其身也, 忘其有名也; 王主之(☐) 〔行〕其道也, 忘其成功也。 Among people who fish for ming, there are no virtuous officials. Among lords who fish for profit, there are no kingly rulers. In the actions (xing 行) of their body-persons (shen 身), virtuous people forget about having ming. In the walking of their dao, kingly rulers’ forget about the success of their efforts. Guanzi 管子卷第六 法法第十六

While discouraging the pursuit of ming, this passage still affirms its association with honor (i.e. it implicitly asserts that ming comes to virtuous people who do not seek it). It also separates ming from profit by aligning ming with virtuous officials and profit with rulers. Thus, even if virtuous officials and rulers do not seek them, ming is the province of officials, whereas profit is the province of rulers. In a similar bifurcation, the Shangjunshu notes that whereas profit-motivated actions put one’s body-person (shen 身) in danger, their impact on ming concerns honor. It says this of robbers and thieves: 故名辱而身危; 猶不止者, 利也。 Thus, although their ming (reputations) are dishonored and their body-persons are endangered, for the sake of the profit, they do not desist. Shangjunshu 商君書 筭地

Here, one’s ming is a matter of honor/dishonor, separate from, and yet comparable to, one’s body-person, which is a matter of safety/danger. This Lüshichunqiu passage’s contrast uses analogous terms. 嘗試觀於上志, 三王之佐, 其名無不榮者, 其實無不安者, 功大故也。俗主之佐, 其欲 名實也與三王之佐同, 〔而〕其名無不辱者, 其實無不危者, 無功故也。 Considering the records of high antiquity—regarding the assistants of the Three Kings—of their ming, none were not glorious, and of their shi, none were not peaceful (qi shi wu bu an 其實無不安 ). This was because of their great efforts. Assistants to common kings have the same desire for ming and shi as the assistants of the Three Kings. But of their ming, none were not disgraced, and of their shi, none were not endangered. This is because of their lack of effort. Lüshichunqiu 呂氏春秋 士容論第六 《務大》

If we recall the two-sided pattern of honorable reputation with visible correlate, it is evident that “qi shi wu bu an” does not say they were safe “in reality.” As the

28

J. Geaney

Shangjunshu passage above indicates, just as ming is measured in terms of honor and dishonor, its correlates can be measured in terms of danger and safety. That the Lüshichunqiu passage goes on to say common kings desire both ming and shi also makes it implausible to interpret this use of shi as “reality.” Furthermore, this passage does not propose a contrast between ming and shi, but merely presents them as separate. Translating shi as “reality” might suggest that ming is in some way less real, but here the glory of the Three Kings’ assistants is not in tension with the peace that they bring, and the disgrace of the common kings’ assistants is not at odds with the danger they provoke. On the contrary, in this case ming and shi are separate but in accord. Even granted the validity of my interpretation, however, one might still question what exactly is implied by a shi that is in danger. The answer is unfortunately murky: because we do not think of the world in aural/visual polarities, there is no single term to function as a translational equivalent, but we can infer that a shi in danger is somewhat like the Shangjunshu passage’s body-person in danger ­(shenwei 身危). That is, in correlation with a disgraceful reputation that can be heard, a shi in danger involves threats to the visible aspects of the world: lifeenhancing profits, wealth, comforts, or health. Thus, even when it seems like ming is being juxtaposed to “reality,” a more plausible interpretation arises if we consider typical ways of measuring and assessing ming and shi. The idea that ming might have value that is potentially equivalent or even superior to its counterparts challenges the tendency to interpret early Chinese texts as denigrating language. Consider, for example, a common interpretation of a well-­ known passage from the Zhuangzi. The section explains the response of Xu You, the virtuous recluse, to the legendary ruler, Yao’s, offer of sovereignty. Politely rejecting the offer, Xu You responds with a metaphor that depicts ming in the role of guest. The implication, however, is not that guests are inferior to hosts. 許由曰: 「子治天下, 天下既已治也。而我猶代子, 吾將為名乎?名者, 實之賓也。吾 將為賓乎?鷦鷯巢於深林, 不過一枝; 偃鼠飲河, 不過滿腹。歸休乎君, 予無所用天下 為!庖人雖不治庖, 尸祝不越樽俎而代之矣。」 Xu You said, “You govern all under heaven, so all under heaven is already governed. Were I still to substitute for you, would I be acting for the ming? The ming is the guest (bin 賓) of the shi. Would I be acting for the guest? The wren nests deep in the forest, but does not exceed one branch. The mole drinks from the river, but does not exceed filling its belly. Return and rest in ruling. The world is of no use to me. Even if the cook were not governing cooking, the impersonator of the dead and the priest would not cross wine jars and meat stands to substitute for him.” Zhuangzi 莊子 逍遙遊第一

Xu You’s decision not to play the role of the ming (and the guest) amounts to rejecting something of equal or greater value to the shi (deed) of being a ruler. He begins by asserting that the world is already governed. By casting this as a situation where the shi of governing is already fulfilled, he forecloses the possibility that he might be replacing Yao in order to carry out acts of governing. In subsequently addressing the offer of the ming, Xu You’s assertion that the ming is the guest of the shi depicts ming/shi as a guest/host relation. The important feature of a guest/host relationship is that location determines identity. What then, is the status of the visiting guests

2  What Is Ming 名? “Name” Not “Word”

29

who the ruler typically invites? We might look to other examples. In the bin 賓 ritual of the Shang, the ancient king, who sought blessings from his royal ancestors, hosted his ancestral spirits as his guests (Keightley 1999: 260). Similarly, in the Liji’s description of guest/ruler relations in a drinking festival, the guest represents heaven, while the presiding ruler represents earth. 鄉飲酒之義, 立賓以象天, 立主以象地 。 In the model of the district drinking festival, the established guest signifies heaven and the established host signifies earth. Liji 禮記〈鄉飲酒義〉46.13

If these passages illustrate what it means for Xu You to be a guest of the ruler, then Xu You rejects the chance to function as heaven in relation to Yao’s earth. As this scenario implies, the guest is not inferior, but simply in a different location. Whether the context is the Shang ritual, the drinking ritual, or the Zhuangzi passage, the ruler, as presider (zhu主), is necessarily associated with the earth, because the ruler remains in place. Moreover, it is fitting that guest is heaven because the visiting guest moves about. The Zhuangzi passage’s emphasis on movement—which arises in reference to the wren, the cook, and the impersonator—indicates that location is decisive for ruler and guest. Recall that heaven is airy and amenable to movement. Similarly, ming disperses more loosely than visible things because it is sound (Geaney 2002: 131–132). Just as heaven does not replace earth, however, the facility of ming’s movement does not result in replacing shi. The two remain separate. That is, Xu You encourages Yao’s return to rest in the place of governing that he already fulfills. He treats this invitation as an offer to have the reputation for providing good ­governance that is already achieved. His disinterest in having such a reputation constitutes a rejection of something potentially valuable, not of something worthless. Indeed, Yao embodies the shi of ruling with its corresponding shi of life-enhancing wealth and comforts. By contrast Xu You embodies the honorable ming that is separate from, if not antagonistic to, that kind of shi. Xu You’s reason for not accepting the offer of a reputation for virtuous governing is not that reputation lacks value. Instead, like the birds, moles, and ritual participants, he already has what he needs. Thus Yao and Xu You each have their separate place and value. A ming is a guest of a shi in the same way that yang is the guest of yin. The Cheng 稱 “Weighing” section of the Mawangdui silk manuscripts notes: 制人者陽, 制人者制於人者陰。客陽主人陰。師陽役陰。言陽黑(默)陰。 One who controls someone is yang, the controlled person is yin. The guest is yang, the host is yin. The officer is yang, the servant is yin. Speaking is yang, silence is yin. Mawangdui Laozi B “Cheng” 馬王堆漢墓帛書‧老子乙本卷前古佚書-稱

The guest is in control. Both the guest and speaker are yang.26 Thus, to interpret the Zhuangzi passage as treating “language” as trivial would be to overlook the specificity of uses of terms in early Chinese texts.

 Robin Yates observes that in military terminology a “guest” implies an attacker and a “host” is a defender (Yates 1997, 279). If we press the analogy here, then the mobile, active sounds of speech and names threaten the boundaries of staid, visible things.

26

30

J. Geaney

4  Conclusion “Name” is the best translation equivalent for uses of ming in texts from Early China. Although “name” cannot fully capture the specificity of ming, it introduces fewer anachronisms than would “word” or “language.” Occurrences of ming in early Chinese texts are unlikely to have been understood by their implied readers as a unit of language. More plausibly, ming would have evoked in readers a particular kind of sound best captured by “name,” often related to people’s fame and honor, associated with upper regions, and typically employed to pick out visible features of a binary cosmos. Sound, honorable reputation, and the tenuousness of upper regions have no particular bearing on the modern concept, “language.” Moreover, that which ming names is not as implicitly superior to ming as what the translation “reality” would imply. While sound and sight are competing sides of polarities in early Chinese texts, sound often holds its own against visible aspects of the world, which means ming is no less “real” than what it names (or to put it another way, sound is no less real than sight). Does it matter that philosophy of ming, in Early China, involved “names” rather than “words”? It does because, while the principles of reasoning might be the same, in the case of Early China, the tool for philosophical thinking was a particular kind of sound referring to visible aspects of the world—an interesting alternative worthy of replacing “language and reality.” More provocatively, since ming referred directly to things, philosophizing with ming unwittingly bypassed the slippery and unnecessarily confusing concept, “meaning.”

References Chinese Classical Works27 Baihutong 白虎通 Chunqiu Zuozhuan 春秋左傳 Guanzi 管子 Guoyu 國語 Hanfeizi 韓非子 Huainanzi 淮南子 Liji 禮記 Lüshichunqiu 呂氏春秋 Mawangdui Laozi B “Cheng” 馬王堆漢墓帛書‧老子乙本卷前古佚書-稱 Shiji 史記 Shangjunshu 商君書 Zhanguoce 戰國策 Zhuangzi 莊子

27

 All references are to CHinese ANcient Texts (CHANT) 漢達文庫 database. http://www.chant.org

2  What Is Ming 名? “Name” Not “Word”

31

Secondary Sources Cao, Feng 曹峰. 2008. “A Return to Intellectual History: A New Approach to Pre-Qin Discourse on Name.” Trans. Joseph E.  Harroff. Frontiers of Philosophy in China 3.2: 213–228. (The article argues against the current logical-linguistic emphasis in interpretations of ming in pre-­ Qin texts, proposing instead a focus on ethical and political uses of ming.) Cao, Feng 曹峰. 2011. “Re-examining the Questions Related to “Ming”. In the First Chapter of Laozi: Taking the Coming-out of Laozi in the Han Dynasty’s Bamboo Slips Preserved in Peking University as an Occasion.”《老子》首章与“名”相关问题的重新审视——以北大 汉简《老子》的问世为契机《哲学研究》年第4期: 58–67. Defoort, Carine. 2013. “Excavated Manuscripts and Political Thought: Cao Feng on Early Chinese Texts.” Contemporary Chinese Thought 44: 3–9. Geaney, Jane. 2002. On the Epistemology of the Senses in Early Chinese Thought. Honolulu: University of Hawai’i Press. Geaney, Jane. 2010. “Grounding ‘Language’ in the Senses: What the Eyes and Ears Reveal about Ming 名 (Names) in Early Chinese Texts.” Philosophy East and West 60.2: 251–293. https:// doi.org/10.1353/pew.0.0097. (The article contends that in early Chinese texts, a ming is a name—understood to be something that is spoken—not a unit of writing or language.) Geaney, Jane. 2011. “The Sounds of Zhèngmíng (正名): Setting Names Straight in Early Chinese Texts.” In Chris Fraser, Dan Robins, and Timothy O’Leary, eds. Ethics in Early China: An Anthology, 107–118. Hong Kong: Hong Kong University Press. (As a step in the process of rethinking early Chinese ideas of language, the article explores the rationale for zhengming, from an approach that understands ming to be sounds.) Geaney, Jane. 2016. “Binaries in Early Chinese Texts: Locating Entities on Continuums.” International Communication of Chinese Culture 3(2): 275–292. Geaney, Jane. 2018. Language as Bodily Practice in Early China: A Chinese Grammatology. Albany: State University of New York Press. (With an emphasis on the physical experience of speaking and naming, this book reconstructs how early Chinese texts present phenomena related to what modern readers think of as “language.”) Geaney, Jane. forthcoming. The Emergence of Word-Meaning in Early China: Normative Models for Words. Albany: State University of New York Press. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. New York: Oxford University Press. Keightley, David. 1999. “The Shang: China’s First Historical Dynasty.” In Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C, 232–291. New York, Cambridge University Press. Legge, James, trans. 1879. The Lî Kî. Vol. 3 of The Sacred Books of China: The Texts of Confucianism. Oxford: The Clarendon Press. Rickett, W. Allyn trans. 1985–1998. Guanzi: Political, Economic, and Philosophical Essays from Early China, 2 vols. Princeton: Princeton University Press. Stout, Jeffrey. 1982. “What is the Meaning of a Text?.” New Literary History 14: 1–12. Yates, Robin. 1997. Five Lost Classics: Tao, Huanglao, and Yin-yang in Han China. New York: Ballantine. Jane Geaney  received her PhD at the University of Chicago. She is Professor in the Department of Religious Studies at the University of Richmond. Her publications include: The Emergence of Word-Meaning in Early China: A Normative Model for Words (forthcoming with  SUNY);

32

J. Geaney

Language as Bodily Practice in Early China: A Chinese Grammatology (SUNY, 2018); On the Epistemology of the Senses in Early Chinese Thought (Honolulu: University of Hawai’i, 2002); “Grounding ‘Language’ in the Senses: What the Eyes and Ears Reveal about Míng 名 (Names) in Early Chinese Texts,” Philosophy East and West, Vol. 60 (2010); and “Critique of A. C. Graham’s Reconstruction of the Neo-Mohist Canons,” Journal of the American Oriental Society, Vol. 119 (1999).

Chapter 3

Naming, Reference and Truth Bo Mou

Although the issue of reference (and naming) concerning the fundamental relationship between language, thought, and reality is primarily one central issue in the philosophy of language rather than in logic, it is examined in the philosophy of logic insofar as it involves the semantic interpretation of a logical system as an account of inferences and some basic logical principles such as the law of identity concerning the identities of referred objects in reasoning. With this specification and in view of the due scope of this book project (“Dao Companion to Chinese Philosophy of Logic”), my discussion of the issue of reference here is restricted to the relevant aspects of the issue of reference and naming to logical inference and semantic interpretation involved in logic as addressed in the philosophy of logic, instead of a comprehensive examination of all the connections of the issue of reference as explored in the philosophy of language.1 At this point, several notes are due concerning what I will do in the subsequent discussion and what I will not, and what I will emphasize among those that I will do. First, there is the need to first identify and clarify the meanigns of the key term Most of the materials in this writing are either shortend versions or excerpts of the relevant parts of Mou 2003, 2007, 2015, 2016, 2018. Thanks to their publishers for permission to re-use the materials here. I am grateful to Yiu-ming Fung for his helpful comments on an early version of this essay. As much of the content of this essay come from some of the previous relevant publications of this author, I would like to express my sincere thanks again to all those colleagues listed in the acknowledgements of these writings for their helpful comments and suggestions.  For example, I will not discuss the issue of naming and reference in the classical text Yi-Jing here, although I render philosophically interesting and signficant the relevant thoughts on the issue in the Yi-Jing text; for a detailed discussion of it by this author, see Mou 2014, 2019, chapter 2. I will not discuss Zhuang Zi’s relevant ideas here, which I examine in Mou 2019, chapter 6.

1

B. Mou (*) Department of Philosophy, San Jose State University, San Jose, CA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_3

33

34

B. Mou

‘reference’ and its closely related key term ‘truth’ (i.e., what they are used to indicate in this context); they are two basic semantic notions as addressed in the areas of logic and philosophy of logic in view of philosophy of language in a standard and cross-tradition way, neither “pragmatic” notions nor something else that one might prefer using the term ‘reference’ and/or ‘truth’ to indicate (thus mere verbal disagreement is not to be addressed here); their respective pre-theoretic understandings play significant or even indispensable explanatory roles in peoples’ folk and reflective lives in various cultural and linguistic traditions to talk about things in the world and capture the way things are. Reference as one (most) basic semantic notion addresses the fundamental “aboutness” relationship between linguistic expressions (as means and media by and through which to express and create thoughts) and extra-linguistic objects (as parts of the world or reality); its closely related semantic notion of truth, as conceived and pre-theoretically characterized by people’s pretheoretic “way-things-are-capturing” understanding of truth that plays its important and indispensable explanatory role in people’s folk and reflective lives, is fundamentally explained in terms of “aboutness” or “reference”, which points to one central characteristic feature of the “capturing” relationship between the truth bearer and the truth maker. Second, it is important to note that, on the one hand, people in the Chinese tradition might and actually use different terms and phrases to express their semantic notions of reference and truth, especially in view of the long development history of the Chinese language; on the other hand, some of those terms and phrases that have been used by some ancient persons or thinker to indicate the semantic notions of reference and truth might be, and actually have been, used to indicate something else. In this way and to this extent, what I am to focus on in this writing regarding reference and truth in the Chinese philosophy of logic is reference and truth as the two basic semantic notions themselves and thus their associated language expressions in the addressed specific contexts, rather than just on how people have actually used those specific characters and phrases that have ever been used by some persons to express such semantic notions (as they might be, and actually were, used in some non-semantic ways) and thus how to translate them;2 this strategy is especially in need in view that the (semantic) notion of truth is fundamentally explained in terms of “aboutness” or “reference” (as implicitly addressed and illustrated in Sects. 1, 2, 3 and 4, and as explicitly highlighted in the ending section). Third, instead of giving a mere historical description of the seemingly-related sayings in the relevant texts, I will emphasize due philosophical interpretation of relevant approaches and resources to the concern of this writing; this does not imply ignoring or dismissing a historical-description-giving approach that aims at collecting historical data and giving an “accurate” historical description of relevant historical matters of fact; both can be constructively complementary.

2  For my discussion of distinct linguistic ways in Chinese and Western languages of delivering pretheoretic understanding of truth, see this author’s discussion of it in Mou 2019, Section 1.4  in chapter 1.

3  Naming, Reference and Truth

35

Fourth, due to space limit, and also related to the foregoing emphasis, I will focus on those most reflectively interesting and arguably representative ideas on the issue in the classical Chinese philosophy relevant to reasoning and semantic interpretation, instead of covering all the seemingly-related sayings. Fifth, due to the nature of this writing, it is not merely legitimate but in need to resort to some relevant conceptual and explanatory resources in the contemporary literature of philosophy and logic to enhance our understanding and treatment of certain points in the texts under examination. In so doing, I am in no way claiming that the ancient thinkers of the texts actually ran these resources in their heads or anticipated the theoretical elaborations of the implications of the points of their texts. At this point, it is important to note that some of these resources might be historically put forward by some thinkers in the West; however, this fact does not necessarily imply that the addressed resources must be conceptually and intrinsically related to the West only, instead of human common spiritical wealth; I thus reflectively resist using such highly misleadning phrases as ‘Western modern logic’ that are often deems to have the foregoing implication. Sixth, also due to space and the nature of this writing, I will focus on the original texts3 and present my analysis, rather than also directly giving an evaluative examination of some other interpretations in the literature on the involved texts; nevertheless, I will give the reference information of these distinct interpretations (to my knowledge) in the footnotes part. Seventh, related to the first note above, in view of the scope and need of the philosophy of logic as well as in accordance with the standard treatment in the philosophy of language, the issue of reference is considered to be one semantic issue4 that treats reference as a two-place relation between words/phrases as the first terms and some objects as the second that the first terms are respectively about, instead of a

3  In this writing, unless otherwise indicated, the translations of the cited passages from the relevant Chinese classical texts are mine. 4  By ‘semantic’ I mean the following strict or narrow sense of the term. It is known that the term ‘semantic’ is used in its strict or narrow sense and in its broad (loose) sense. In its strict or narrow sense, ‘semantics’ means the study of the non-(purely)-linguistic, cross-categorical two-pace relations between linguistic expressions and the extra-linguistic objects for which they stand. Such a non-linguistic relation is usually called a ‘semantic relation’; the notion that captures such a semantic relation is called a ‘semantic notion’. The principal semantic relations are referring and being true; the principal semantic notions are thus the notions of reference and truth. However, in its broad sense, ‘semantics’ means the study of (a variety of) meanings of words and sentences, although often especially pointing to their relatively stable ones (such as “objective” referents, “inter-subjective” senses, etc.). If the extra-linguistic objects in question are taken to be the meanings of the linguistic expressions, semantics in the narrow sense of the term can be viewed as an investigation of the “meaning” (the referential meaning of the linguistic expressions). The two senses of ‘semantic’ are thus related. However, especially in logic and the philosophy of logic and thus in this writing, the strict or narrow sense of the term ‘semantic’ is used, unless indicated otherwise. In contrast, pragmatics is the study of situated uses of language which involve particular utterance contexts and the language users’ (particular) intentions (or their situated ‘epistemic’ attitudes), or the study of what speakers do with language in its situated uses (speech acts).

36

B. Mou

pragmatic issue or a theoretically-semantic/pragmatic-hybrid business;5 in so doing, that implies neither that the pragmatic elements are ignored nor that the semantic and the pragmatic are taken to be absolutely separate in linguistic practice and in practical reasoning; rather, on the issue of reference, the contributing pragmatic elements (if any) are legitimately presupposed for the sake of effective semantic treatment with due abstraction and in view of general theoretic need. Eighth, treating the issue of reference (as the aforementioned two-place semantic relation) per se and in view of its relation to logical reasoning that is more or less ontologically neutral, I will not inquiry into its (externally) related ontological issue concerning ontological status of referents; however, when in need, I may address some of its related metaphysical issues that are characterized as being ontologically neutral and somehow internally related to reference as a semantic relation, such as the issue of the (ontologically neutral) structure of referents (independently of the ontological debate between platonic realism, radical nominalism, and conceptualism) and of the “due-place” actuality (independently of what counts as actuality). With the foregoing notes, my strategy is this. First, in Sect. 1, I will explain several basic points as revealed by the rich resources in the classical text Gong-SunLong-Zi, focusing on the relevant points in the “Bai-Ma-Lun”, “Zhi-Wu-Lun” and “Ming-Shi-Lun”. Second, in Sect. 2, I will analyze how, essentially in line with Gongsun Long’s relevant point in the “Bai-Ma-Lun”, the Later Mohists significantly show their semantic sensitivity to the due identities of the referents in reasoning, as primarily suggested in the Later Mohist text Xiao-Qu), Third, in Sect. 3, I will look at how, essentially in line with Gongsun Long’s general point on the due-place actuality in the “Ming-Shi-Lun”, Confucius’ account of name rectification (as suggested in the Analects) together with Xun Zi’s and Wang Chong’s further developments addresses the issue of reference in moral reality. Fourth, in Sect. 4, I will explain how Lao Zi makes interesting and engaging points concerning the relationship between language engagement and the ultimate concern and between the “speakable” and the “unspeakable” as suggested in the Dao-De-Jing. Fifth, in Sect. 5, I will briefly examine a relatively recent debate on the relation of the structure of Chinese nouns to Chinese thought from the referential point of view: I will focus on how the mass-noun-semantics part of Hansen’s mass-noun hypothesis are challenged by some competing accounts in view of the referential relation between Chinese nouns and referents. Sixth, in Sect 6, I will address the intrinsic connection between the two basic semantic notions, reference and truth and briefly explain to what extent the content of this essay on the issue of reference is intrinsically related to an account of how (semantic) truth is possible in any intellectual inquiries into “how things are”, generally speaking, and in logic, specifically speaking.  In our linguistic practice and logical inference, people use words and concepts to refer to things and that the two-place semantic relation comes into existence through speakers’ referential activities. Nevertheless, Strawson renders the pragmatic dimension of people’s linguistic practice decisive indiscriminately in any theoretical treatment of meaning (Strawson 1950) and thus criticizes Russell’s semantic treatment in his Theory of Descriptions (Russell 1905); Russell responds to Strawson explaining why, in this connection, Strawson misses the point of the theoretical treatment of the semantic relation (Russell 1957). 5

3  Naming, Reference and Truth

37

1  Naming and Reference in Gongsun Long’s Account In this section I explain several basic points as revealed by the rich resources in the classical text Gong-Sun-Long-Zi (《公孫龍子》), focusing on the reflectively interesting points concerning the issue of reference in the “Bai-Ma-Lun”, “Zhi-WuLun” and “Ming-Shi-Lun” which are examined respectively in Sects. 1.1, 1.2, and 1.3.

1.1  The Relevant Thought in the “Bai-Ma-Lun” It is philosophically interesting and significant that the thought of Gongsun Long (320-250 BCE) who calls our attention to the distinct identities of referents of referring names through philosophical argumentation; his idea of reference as primarily suggested in the classical text “Bai-Ma-Lun” (白馬論 “On the White Horse”) contributes to his contemporary Later Mohists’ understanding of the semantic-syntactic structure of the parallel inference, to be explained in the next section, and can also engage with some representative approaches (in the contemporary philosophy of language) to the issue of reference in a philosophically interesting way.6 It is known that there is the debate among the modern scholars concerning how to understand Gongsun Long’s “White-Horse-Not-Horse” (白馬非馬) thesis and its significance; there are distinct dimensions of Gongsun Long’s thesis, and thus there are distinct focuses of the modern scholars’ interpretations on this thesis. One focus is on the ontological status of the referents of such names as ‘ma’ (‘horse’) and ‘bai-ma’ (‘white horse’) in the contexts under Gongsun Long’s examination.7 As indicated at 6  Due to space and the goal of this writing, I will not discuss how Gongsun Long’s thought on reference can engage with two representative approaches in the contemporary philosophy of language, i.e., the Lockean-Fregean approach (Locke 1690; Frege 1892) and the Milian-Kripke’s approaches (Mill 1881; Kripke 1980), which I explain in Mou 2018. 7  For example, Chung-ying Cheng and Yiu-ming Fung give their distinct elaborations of a representative and influential Platonic-realistic interpretation which was originally suggested by Fung Yu-lan. Fung Yu-lan renders Gongsun Long Platonic-realistic: Gongsun Long’s arguments are intended to argue that ‘white horse’ and ‘horse’ represent two distinct Platonic universals (or universal/abstract qualities) and thus the universal of white-horseness is not (identical to) the universal of horseness (Yu-lan Fung 1952–1953). Chung-ying Cheng makes a defense of a Platonic realistic interpretation in terms of his version of Platonic abstract ontology (Cheng 1983). In contrast, Yiuming Fung gives a defense of the realistic account in terms of his realistic understanding of universals, which takes it that Gongsun Long’s universals exist in their separate, transcendent and unchanged world but they can emerge, and participate, in the phenomenal world (Fung 2000). Chad Hansen proposes a radical shift of interpretation based on mereology (part-whole logic) and his mass-noun hypothesis: ‘white horse’ is a mass noun and refers to a mass-sum whole of horsemass and white stuff; the whole of white-part and horse-part is not its horse-part. Hansen’s mereological interpretation is to by-pass class-member relation but resort to whole-part division alone (Hansen 1983). In the metaphysical-interpretation part of my account, I suggest a modest mereological interpretation: the denotational semantics and deep structure of Chinese common nouns are

38

B. Mou

the outset, due to the nature of the issue under examination, this writing does not focus on the ontological implications of Gongsun Long’s thesis but on its referential-relation dimension. In the following, I briefly explain this dimension of Gongsun Long’s thesis: I first present one crucial passage from the “Bai-Ma-Lun” that gives the fundamental rationale underlying all of his specific arguments for the thesis, and then I explain why, on the issue of reference, Gongsun Long’s point in this passage is philosophcially interesting and significant.8 The crucial passage is this: 曰:有白馬, 不可謂無馬也。不可謂無馬者, 非馬也?有白馬為有馬, 白之, 非馬何也? 曰:求馬, 黃、黑馬皆可致; 求白馬, 黃、黑馬不可致。使白馬乃馬也, 是所求一 也。所求一者, 白馬1不異馬也; 所求不異, 如黃、黑馬有可有不可, 何也?可與不可, 其 相非明。故黃、黑馬一也, 而可以應有馬, 而不可以應有白馬。是白馬之非馬, 審矣! A: “When there is a horse, we cannot say that there is no horse. If one cannot say that there is no horse, then isn’t it a horse? There being a white horse means that there is a horse, why is a white horse not a horse?” B [Gongsun Long]: “When a horse is sought, either a yellow horse or a black horse may meet what is sought. When a white horse is sought, neither a yellow horse nor a black horse may meet what is sought. What makes a white horse a horse (shi-bai-ma-nai-ma) is the same [common] aspect [of a white horse and a horse] given that it is what is sought (suoqiu-yi-ye). If what is sought is the common aspect, the white horse would not differ from (bu-yi) the horse [regarding the common aspect]. If what is sought in one case [the common aspect of all the horses that is met by the yellow horse and the black horse alike] does not differ from what is sought in the other case [the distinct aspect between the white horse and the horse that is met by neither the yellow horse nor the black horse], then why is it that the yellow or black horse does meet what is sought in one case but not in the other? It is evident that the two cases are distinct. For this reason the yellow horse and the black horse, based on what are the same in them [they are horses but not white horses], respond to what ‘horse’ denotes [fall into the extension of ‘horse’] but not to what “white horse” denotes [not into the extension of ‘white horse’]. Thus [in the way in which the yellow horse and the black horse are not the white horse in regard to their distinct aspect] indeed the white horse is not the horse.”

On the one hand, what Gongsun Long emphasizes here is this: when what is sought is the common or shared aspect between the white horse and the horse [in stating how the former is related to the latter], what is expected or needs to be stated is that the whise horse is identical to or is not different from (不異 “bu-yi”), the horse with regard to the addressed common or shared aspect, because what makes a white horse a horse consists in their same [shared or common] aspect (使白馬乃 馬也, 是所求一也 “shi-bai-ma-nai-ma-ye, shi- sou-qiu-yi-ye”). On the other hand,

like those of collective nouns (in a semantic-syntatic, instead of grammatical sense), and its implicit ontology is a mereological one of collection-of-individuals with both part-whole structure and member-class structure; the denotation of ‘white horse’ is neither a realistic universal nor a mereological sum of white and horse nor an empty set but a collection of white horses with member-class structure (Mou 2007). As far as their metaphysical interpretations are concerned, Cheng’s and Fung’s accounts respectively demonstrate two distinct versions of realism concerning universals, Hansen’s account presents a strong or radical version of nominalism, while the metaphysicalinterpretation part of my account involves a kind of moderate nominalism (conceptualism). Also see Chmielewski 1962 for a set-theoretic account. 8  For my detailed discussion of the referential dimension of Gongsun Long’s thesis, see Mou 2007.

3  Naming, Reference and Truth

39

Gongsun Long stresses that what is sought in the above case is different from what is sought when the white horse is claimed to be different from the horse: in the latter case, what is sought or focused on is some distinct aspect (that of being white) between the white horse and the horse, which is thus met by neither the yellow horse nor the black horse. Exactly here does Gongsun Long make a crucial general point that underlies all his specific arguments in the Bai-Ma-Lun: while the subject expression ‘the white horse’ designates a (any) white horse (i.e., any member of the collection of white horses) as a whole, what is sought (in focus on some specific aspect of a white horse as a whole) makes a signficant difference; it is the distinct aspect or specific attribute part of being white (something that is not shared by all horses) that is focused on when an (any) individual white horse is designated; that is, a double reference, so to speak, is made in such sentential contexts. This gives the fundamental rationale behind all of Gongsun Long’s distinct arguments for the “White-Horse-Not-Horse” thesis, each of which focuses on a certain distinct aspect between the white horse and the horse (or a certain distinct “meaning” aspect between their respective names): for example, the distinctive conceptual contents of their referring names, their distinctive necessary-identity contributors, and the distinctive extensions of their referring names. It is important to note that the foregoing referential structure of the subject expression in a sentential context is neither something odd nor something that is coined for the sake of an intellectual game, because, semantically, every subject-predicate sentence or statement, as a token of the type of one basic employment of language to the effect that something is said about something, does have such a double-reference structure and does make such a double reference. Now Gongsun Long actually addresses the deep “double-reference” semantic-syntactic structure of the referential function of the subject expression in a sentential context (in terms of his key terms “what is sought” here) for the sake of calling our attention to distinct aspects of things that are often ignored when they are categorized into their kinds in which they are supposed to belong. In this way, generally speaking, Gongsun Long alerts us to the danger of overassimilating distinctions, especially when the distinctive aspects need to be emphatically focused on; specifically speaking (concerning the issue of reference), he pays due attention to more carefully distinguishing different but related referential functions of the referring terms and being sensitive to which specific aspect of an object is pointed to in focus while the object as a whole is referred to in reasoning (either deductive reasoning or non-deductive reasoning).

1.2  The Relevant Thought in the “Zhi-Wu-Lun” The ending point of the previous sub-section on Gongsun Long’s “White-HorseNot-Horse” thesis is kindred in spirit with one central point that Gongsun Long makes in his essay “Zhi-Wu-Lun” (指物論 “On Referring to Things”): those relevant contributing elements involved in the referring subject’s act of referring via a referring name (such as her purpose or her perspective focus), which can be explic-

40

B. Mou

itly or inter-subjectively presented in sentential contexts, make their intrinsic contributions to the identities of things that are referred to. Although it is highly controversial how to interpret this essay, and although this text per se is more or less “corrupted” to some extent (clearly some key words missing), in view of the principle of charity in reflective interpretation of ancient texts, I take it that this text is part of the Gong-Sun-Long-Zi and that its thought is compatible with Gongsun Long’s thought in the other essays in the Gong-Sun-Long-Zi. Although it is controversial how to interpret the text, its basic points, as I see it, are these. First, referring is not the same as what is referred to (the referents). Second, the point of emphasizing the distinction between referring process and its result (the referent) is not to simply repeat a trivial fact but addresses something significant: a certain perspective taken in referring process results in a certain identity of the referent through which the referent can be identified (thus can be called ‘a so-and-so thing’) and focused on in the world of referents. Third, the point is not (necessarily) a “subjective” view to the effect that the existence of a thing depends ontologically on the subject’s referring activity; rather, it highlights a significant semantic point: in case some things is said about an object, the object is designated as a semantic whole, a certain specific part of the object is (simultaneously) referred to as a certain identity of the thing through which something is said of the object. Fourth, the above point actually captures one basic linguistic fact concerning people’s most basic employment of language in saying something of a thing. Fifth, the ancient thinker did not have sufficient conceptual and explanatory resources to deliver the foregoing point in a refined, explicit or even consistent way; it is not something odd; we needn’t to demand too much for the ancient thinker in this connection. We can make progress in this connection, though the basic point is shared by the ancient thinker and us in the contemporary times because the following constitutes one basic employment of language in both ancient and contemporary linguistic practice: when we deal with, talk about and make inference concerning the world around us, we saying something about things in the world.

1.3  The Relevant Thought in the “Ming-Shi-Lun” The issue of the relationship between language and reality has been one of the fundamental concerns in the history of classical Chinese philosophy via the issue of the relationship between name (ming) and actuality (shi). Gongsun Long’s short essay “Ming-Shi-Lun” (名實論 “On Name and Actuality”) is the first in classical Chinese philosophy that explicitly addresses the issue in a philosophically interesting way. To my knowledge, perhaps Gongsun Long is also the first classical Chinese thinker who gave his reflective elaboration of the normative elements of people’s pre-theoretic understanding of truth in general terms through his conception of shi ( 實)/actuality, One crucial element of people’s pre-theoretic “way-things-are-capturing” understanding of truth is what is delivered by the phrase “the way things are”, which is open to various ontological or metaphysical explanations in two

3  Naming, Reference and Truth

41

major connections. First, the phrase does not commit itself to any ad hoc ontological account of what counts as reality, whether it is physical reality or social reality or some other types of reality, or whether it is a reality presently there or in the past (or even in the future)9; in this sense, the phrase is ontological neutral in this connection and to this extent. Second, it can cover a variety of ways things are, including “normative” or “due” way things are; people’s pre-theoretic understanding implies this content when people state that things go on their own way, including the possibility that things might go astray or off their own due course as well as the cases in which things go on their own “due” way (thus “capturing the way things are” means capturing the “due” way things are); to this extent, “the way things are” is both pre-theoretically compatible with, and quasi-reflectively inclusive of, both “factual” ways things are and “normative” due ways things are. If this “normative” point of people’s pre-theoretic understanding of truth is only implicitly or compatibly delivered in a variety of folk expressions of such an understanding, Gongsun Long explicitly captured and reflectively elaborated this important “normative” element or implication of people’s pre-theoretic understanding of truth through his conceptions of shi/actuality and of shi-capturing in his account of “name rectification”, which I view as one significant contribution that Gongsun Long has made both to the development of classical Chinese philosophy on the jointly-concerned philosophical issue of truth, historically speaking, and to the contemporary theoretic exploration of the issue, philosophically speaking. In his essay “Ming-Shi-Lun”, Gongsun Long presents one profound vision concerning the genuine identity10 of actuality11: 天地與其所產焉, 物也。物以物其所物而不過焉, 實也。實以實其所實而不曠焉, 位 也。出其所位, 非位。位其所位焉, 正也。以其所正, 正其所不正, 疑其所正。其正者, 正其所實也。正其所實者, 正其名也。其名正, 則唯乎其彼此焉。

9  Though in English the term “are” in the phrase “the way things are” can be understood as the present tense of “be” (a local grammatical feature of many phonetic languages like English to which presentism or more generally speaking A-theories in the philosophy of time, is intended to be sensitive), it can be also treated as a neutral and unrestricted “existential” quantifier from an atemporal perspective (as taken by eternalism or more generally speaking B-theories in the philosophy of time), whose meaning is captured in the “absolute” unrestricted existential quantifier in the first-order logic; and the latter understanding more fits the case of Chinese language which does not have its explicit counterpart “tense” expressions. 10  It is important to note this: the term ‘identity’ and its associated concept of identity both in its ordinary usage (in its dictionary literal sense) and in logic and philosophy implies neither static identity nor fixed identity nor absolute identity; what it delivers is not only totally compatible with changing process and dynamic development of things but also includes and covers the latter; it is not the same as ‘sameness’ (although a traditional formulation of the law of identity via absolute identity (in terms of ‘=’) unfortunately facilitates such an narrow or ad hoc understanding. The case is similar to the term ‘being’: ‘being’ in philosophy, in its general sense, is used to express various kinds of existence, although in the Western tradition it is used to specifically expression one kind of existence, something definite, stable, constant, essential, universal or unchanging, and although some philosophers, say, in the Western tradition argue that such a kind of existence captures the essence of being (in its broad sense). 11  For my previous discussion of this, see Mou 2006.

42

B. Mou The heaven, earth and what they produce are things. When a thing goes its own way without transgressing its limit, it achieves its actuality (shi); when its actuality goes its own way without being out of its track, it achieves its due place (wei). If a thing goes beyond its due place, it is in wrong place; if a thing is in its due place, it is in right place. One is expected to rectify a thing in wrong place into right place; [one is not expected] to challenge a thing in due place [by virtue of it being in wrong place].12 The rectification of a thing is the rectification of its actuality; the rectification of its actuality is implemented through the rectification of its name. Once its name is rectified, the standards for ‘that’ and ‘this’ will be formed up and stabilized.

Gongsun Long here emphasizes that a thing needs to go its own way without transgressing its limit to achieve its actuality; he further stresses that, once a thing achieves its actuality, there remains an issue of how to keep its actuality in due place; he explicitly points out that the so-called name rectification lies in rectifying the actuality of a thing in its due place through rectifying the due content of its name which identifies such due place and thus gives due identity condition for the thing and its actuality. Gongsun Long herer distinguishes two types of actuality: the “dueplace” actuality that goes without transgressing its due place, and the “undue-place” actuality that goes beyond its due place and needs to be rectified. From the vantage point of philosophical interpretation, Gongsun Long’s insight of the due-place of actuality provides one profound unifying basis for the (semantic) truth concern in the Confucian account of name rectification,13 which eventually (via Xun Zi and Wang Chong’s approaches) strives to unify the truth pursuit in the two distinct types of actuality (i.e., the actuality as the moral world and the actuality as the inclusive natural world). Gongsun Long’s point here is significant in two connections. First, as indicated before, he is the first thinker in classical Chinese philosophy who explicitly suggested the conceptions of the due-place actuality and of due-placeactuality capturing when elaborating the normative element in people’s pre-theoretic understanding of truth. Second, clearly, his conception of the due-place actuality is not limited to the area of moral actuality alone but was suggested in a general way covering both non-moral and moral actuality: (1) actuality is not limited to what appears to be there or the “status-quo” happenings; rather, the due place of actuality gives the genuine identity of actuality; (2) capturing the way things are thus primarily means capturing the due way thing are. One important implication of Gongsun Long’s conceptions of due-place actuality and of capturing due-place actuality is this: the normative “due” place of actuality provides the fundamental rationale and the metaphysical foundation for how it is possible for shi/actuality to include both non-moral actuality and moral actuality at the base or root level. Though Gongsun Long himself did not explicitly address this significant implication of the conception of due-place actuality, it is Xun Zi who first explicitly  It has been considered that there was something missing with “疑其所正” in the original text. Here I translate it in terms of the meaning paraphrase in view of the context. Cf., Wing-tsit Chan translation of it into “[What is correct is not used to] doubt what is correct.” (Chan 1963, 243). 13  It is noted that, for the purpose of philosophical interpretation, I first discuss Gongsun Long’s relevant thought and then Confucius’ relevant thought, although the former historically appeared after the latter. 12

3  Naming, Reference and Truth

43

addressed this idea when further elaborating and expanding relevant points of Confucius’ account of name rectification, and it is Wang Chong who further developed the idea in a more explicit and comprehensive way, to be explained below.

2  N  aming and Reference in the Later Mohist Logical Account In this section, I examine how the Later Mohists’ view on reference, which is essentially in line with Gongsun Long’s point on semantic sensitivity to distinct referential identities (what is sought in referring) in the “Bai-Ma-Lun” as explained before, bears on their treatment of the issue of validity of the mou (侔) style parallel inference (arguably a type of deductive reasoning) through their semantic sensitivity to the deep semantic-syntactic structure of such a type of inference. In the following, I present a philosophical interpretation of the ancient Mohist diagnosis of the parallel inference in the Later Mohist text Xiao-Qu (“XQ” for short) concerning its semantic-syntactic structure.14 In the opening passage, the XQ gives a general characterization of disputation in view of its strategic and tactic goals: 夫辯者, 將以明是非之分, 審治亂之紀, 明同異之處, 察名實之理, 處利害, 決嫌疑。焉 摹略萬物之然, 論求群言之比。以名舉實, 以辭抒意, 以說出故, 以類取, 以類予。有 諸己不非諸人, 無諸己不求諸人。或也者, 不盡也。假者, 今不然也。 Disputation/argumentation (辯) is to clearly distinguish between what is this/so and what is not-this/so [是非right and wrong, true and false, or adequate and inadequate], examine the patterns of order and disorder, differentiate where sameness and difference are located, investigate the principles of names and objects, make judgment of what is beneficial and harmful, and resolve suspicion and doubts. Thereby one can depict and capture (摹略) the ways ten-thousand things are (萬物之然) and explore what distinguish various sayings. Names are used to designate objects; phrases are used to express ideas; explorations are used to bring out reasons. By virtue of kinds selections are made; by virtue of kinds inferences are drawn. If one maintains it in oneself, one should not criticize it in others [not addressing it in this negative way]; if one does not maintain it in oneself, one should not demand it of others [not addressing it in this demanding way]. Some-so is not exhaustiveso; supposed-so is currently not-so.

Among others, 焉摹略萬物之然(i.e., capturing the way things are) both highlights one strategic goal of reflective disputation/argumentation and provides one strategic standard as what is modeled on. The XQ then gives the underlying rationale of a variety of argumentation via modeling: 效者, 為之法也, 所效者所以為之法也。故中效, 則是也; 不中效, 則非也。此效 也。Modeling is to have something as a standard; what is modeled on is that by which a standard is set up. Therefore, if a thing matches a model, then it is so [right or true]; if it does not match the model, then it is not so [wrong or false]. This is what modeling is.

 For a further discussion on how to formally and accurately capture the later Mohist point in this connection and give a general condition for the validity of the parallel inference via a modern logical treatment, see Mou 2016.

14

44

B. Mou

and four forms of argumentation under examination: 辟也者, 舉他物而以明之也。侔也者, 比辭而俱行也。援也者, 曰「子然, 我奚獨不可 以然也?」推也者, 以其所不取之同於其所取者, 予之也。是猶謂也者, 同也。吾豈謂 也[他]者異也。 “Drawing a metaphor-analogy inference” (辟 pi) is making one thing clear by bringing up another similar thing (舉他物); “Drawing a parallel inference” (侔 mou) is comparing/ contrasting [both] phrases [via identity or distinction regarding a certain aspect] and having both proceed in a parallel way [regarding the same aspect] (比辭而俱行); “Drawing a pulling-analogy inference” (援 yuan) is [what underlies] such sayings as “As you [the opponent in disputation] are so, why should I alone not be so?” [that is, pulling from the other’s case inward to the agent’s own case]; “Drawing a pushing-analogy inference” (推 tui) is drawing inference by recognizing what has not been selected to be the same as what has been selected [that is, pushing from what has been selected outward to what has not yet to be selected]. This amounts to saying that, if the other is the same, how can I say that the other is different?

After the XQ makes some significant remarks on how to adequately employ the foregoing forms of argumentation via specific diagnosis and some general morals, in the subsequent paragraphs, the XQ focuses on the discussion of several representative applications of the parallel inference (i.e., the “侔”-type inference) and makes its diagnosis.15 Now, in the XQ, the Later Mohists explicitly examined three distinct types of situations concerning various applications of the parallel inference, with concrete examples as illustrations: (1) the “shi-er-ran是而然” situation where a thing is this [in premise] and thus is so [in conclusion]; (2) the “shi-er-bu-ran是而不然” situation where a thing is this [in premise] but is not so [in conclusion]; (3) the “bu-shier-ran不是而然” situation where a thing is not this [in premise] but is so [in conclusion]. Some applications of the parallel inference seem to be valid, while others not. They can be further classified into two kinds of situations, i.e., the first kind of the applications are considered adequate for the sake of valid parallel-type deductive reasoning, which consists of the explicitly-given Situation (1) of “shi-erran是而然” where a thing is this [in premise] and thus is so [in conclusion] and the implicitly-addressed Situation (1)* of “bu-shi-er-bu-ran不是而不然” where a thing is not this [in premise] and thus is not so [in conclusion]”. Both are considered to be adequate applications of the parallel inference. The later Mohists gave examples and illustrations of adequate applications of Situation (1) as follows:

 In the literature on the issue of the status and nature of the parallel inference, it is controversial whether the parallel inference is a type of deductive reasoning or a kind of analogical reasoning. Those scholars such as Graham 1967, Liu 2004, and Fung 2012 explicitly render it deductive: though Graham does not explain why he thinks so, while Liu and Fung give essentially the same reason by formally presenting the parallel inference in terms of the standard first-order predicate logic resources. In contrast, some other scholars such as Hansen 1983 and Fraser 2013 render it analogical. On this issue, I agree to the former’s position, while disagreeing to the latter; however, my reason for rendering the parallel inference deductive is substantially different from that as presented in Liu 2004 and Fung 2012. Without elaborating this here, the interested reader can see Mou 2016 for the details in this connection. 15

3  Naming, Reference and Truth

45

白馬, 馬也; 乘白馬, 乘馬也。驪馬, 馬也; 乘驪馬, 乘馬也。獲, 人也; 愛獲, 愛人也。 臧, 人也; 愛臧, 愛人也。此乃是而然者也。 The white horse is the horse; [therefore] riding the white horse is riding the horse. The black horse is the horse; [therefore] riding the black horse is riding the horse. Huo [the name of a female servant] is a person; [therefore] caring for Huo is caring for a person. Zang [the name of a male servant] is a person; [therefore] caring for Zang is caring for a person. There are instances in each of which a thing is this and thus is so (是而然).16

Furthermore, the later Mohists identified and addressed the other kind of inadequate applications of the parallel inference which do not appear to be valid, that is, Situation (2) of “shi-er-bu-ran是而不然” where a thing is this [in premise] but is not so [in conclusion] and Situation (3) of the “bu-shi-er-ran不是而然” situation where a thing is not this [in premise] but is so [in conclusion]. Some examples and illustrations of Situation (2) are given as follows: …車, 木也; 乘車, 非乘木也。船, 木也; 入船, 非入木也。盜人, 人也, … 殺盜人非殺人 也, … 。此與彼同類, 世有彼而不自非也, 墨者有此而非之, 無也故焉, 所謂內膠外閉 與心毋空乎?內膠而不解也, 此乃是而不然者也。 …The cart is wood; [but] riding the cart is not riding wood. The boat is wood; [but] entering the boat is not entering wood. The robber is the person; …[but] killing the robber is not killing the person [the latter/the conclusion of this type of applications]. The latter [此ci/the conclusion] and former [彼bi/the [inadequate] premise] are of the same kind [but with their distinct focuses on distinct aspects]; the ordinary people hold the former and do not consider themselves mistaken; however, they consider it mistaken for the Mohists to [also] hold the latter. This attitude is not reasonable and amounts to what is called “ossified inside and closed off outside, which results from the closed mind that is ossified inside without being susceptible to change”. These are instances in each of which a thing is this but is not so (是 而不然).

Some examples and illustrations of Situation (3) are given as follows:

16  One note is due here about the English translation of those phrases like ‘乘馬’ (and ‘乘車’ below) in such a context of parallel inference. I hold a collective-name hypothesis concerning the semantic-syntactic structure of a common noun (for its earlier version concerning common nouns in Chinese, see the brief introduction to it in Section 5; for its expanded version concerning common nouns in general not limited to Chinese, see another forthcoming article of this author); with consideration that one standard grammantical means in English to express a collection designation is via “the definite article plus the [addressed] common noun” (for instance, grammantically adequate, one can say “the wolf is not really a dangerous animal”, with this usage of the definite article to talk about any individual member of the collection of wolfs), I thus use the English phrase ‘the horse’ to designate a collection of horses consistently both in a single subject-predicate statement [like ‘the white horse is (identical to) the horse (with regard to the shared common aspect of being horse)] and in a parallel inference here (like the inference from the premise statement ‘the white horse’ is the horse’ to the conclusion statement ‘riding the white horse is riding the horse’). Different from a kind designation or a universal designation, part of the sophisticated semantic point of a collection designation is this: individualization of a collection of individual objects consists totally in (any) individual members of the collection, instead of a collection entity that is separate from its individual members; in this way, ‘riding the horse’ means riding any one individual horse in the collection of horses (instead of some members being ridden while some others not) and more accurately captures the involved collection designation. Although, grammatically and partially semantically, one can use ‘riding a horse’ here, I think that some sophisticated semantic point of what is expressed via ‘riding the horse’ would be lost.

46

B. Mou 且夫讀書, 非, 好書也。且鬥雞, 非雞也; 好鬥雞, 好雞也。且入井, 非 入井也; 止且入井, 止入井也。… 此與彼同, 世有彼而不自非也, 墨者有此而罪非 之, 無也故焉, 所謂內膠外閉與心毋空乎?內膠而不解也。此乃不是而然者也。 Reading the book is not the book; [but] favoring reading the book is favoring books. Cockfighting is not the cock; [but] favoring cockfighting is favoring the cock. Being about to fall into the well is not falling into the well; [but] stopping being about to fall into the well is stopping falling into the well….The latter [此ci/the conclusion of this type of applications] and former [彼bi/the [inadequate] premise of this type of applications] are of the same kind [but with their distinct focuses on distinct aspects]; the ordinary people hold the former and do not consider themselves mistaken; however, they consider it mistaken for the Mohists to [also] hold the latter. This attitude is not reasonable and amounts to what is called “ossified inside and closed off outside, which results from the closed mind that is ossified inside without being susceptible to change”. These are instances in each of which a thing is not this but is so (不是而然).

Though it is known that those applications of the parallel inference in Situations (1) and (1)* are rendered adequate while those in Situations (2) and (3) problematic, what are at issue are these: Why do (2) and (3) go wrong? Can we say that those application instances of the parallel inference in Situations (2) and (3) constitute genuine counter-examples to the generality of the parallel inference and thus show the failure of the parallel inference per se? What is the point of the later Mohist diagnosis here? Based on the foregoing interpretative examination of the Later Mohist diagnosis, one central point of the diagnosis as presented in the XQ is this: generally speaking, the Later Mohists maintained due semantic sensitivity to the refined semantic relation between language expressions and the way things are in the predicative context of saying something about objects as referents of names; specifically speaking, the Later Mohists implemented this semantic sensitivity in their reflective examination of the parallel inference: such a semantic-sensitivity concern motivated the Later Mohists to call our attention to meeting certain adequate conditions for validly carrying out the parallel inference. What fails in (2) and (3) is not the parallel-inference deductive reasoning per se but its inadequate applications in (2) and (3); this indeed also shows the insufficiency of the available logic resources in the ancient times to capture the deep semantic-syntactic structure of the parallel-inference deductive reasoning. So to speak, the genuine point of the Later Mohist diagnosis of (2) and (3) lies in this: the Later Mohist diagnosis aims at keeping due semantic sensitivity to which aspect of the object (the referent of the addressed subject expression) under examination is focused on when carrying out the parallel inference and thus introducing and maintaining the genuinely relevant premise (with a due perspective focus) in an adequate application of the parallel inference.17  In this fundamental connection, the line of the later Mohists in diagnosing the parallel inference and Gongsun Long’s rationale in treating the “White-Horse-Not-Horse” thesis as presented before are essentially the same: any identity expression without being sensitive to which aspect is in reference focus would be semantically incomplete. This idea, logically speaking, is also labeled ‘relative identify’, which is often attributed to Geach 1967 in which Geach criticizes the standard notion of absolute identity in the standard first-order predicate logic (also see Deutsch 2007). For the reason explained here, the basic idea of relative identity as a kind of semantic sensitivity is quite

17

3  Naming, Reference and Truth

47

Indeed, the Later Mohists then did not have available due resources to formulate the general inference rule for the validity of the parallel inference in an effective and unified way for the sake of guaranteeing its adequate applications (in the case of “ 是而然”, explicitly addressed, and of “不是而不然”, implicitly addressed) while against introducing inadequate (i.e., contextually irrelevant) premises (in the cases of “是而不然” and “不是而然”); this has partially brought about, or at least contributed to the subsequent misunderstanding and mistreatment of the status and nature of the parallel inference. Such a historical limitation of the Later Mohists in this connection is neither alone nor accidental. The standard predicate logic with the standard form of identity does not have sufficient logical resources to formally present the foregoing point of the Later Mohist critical examination of various types of application situations of the parallel-inference deductive reasoning. However, this does not mean that there is any intrinsic problem with the parallel-inference deductive reasoning per se but only shows that the standard predicate logic (with identity) has yet to have sufficiently powerful, adequate and sensitive logical resources to formally capture the deductive reasoning in this connection. It is noted that, in view of the principle of charity in philosophical interpretation, and with the due assumption of the consistency of the Later Mohist texts within the Mohist Canons, one can further identify the foregoing point of the Later Mohist semantic sensitivity from some other major texts of the Mohist Canons. Let me consider some crucial passages in the “Jing-Shuo” (經說 “Canons A”) & the “JingShuo-Shang” (經說上 “Explanations A”) [those paraphrase remarks within bracket parentheses and the point numbers are mine]: A87 [In “Canons A”] 同, 重、體、合、類 [In “Explanations A”] 同:二名一實, 重同也。不外於兼, 體同也。俱處於室, 合同也。 有以同, 類同也。 Tong (同the same) [includes the following seemingly different but partially overlapping classification of the same things]: being duplicated; being parts; being together/united; being of a kind. Tong (同 the same): What two names designate being one (identical) object is the same of being duplicated [the referential sameness]; [Both] Not being outside the whole is the same of being parts [the sameness of belonging to the one whole]; Both residing in one [single] place is the same of being together [the sameness of being united in one single thing]; Both being the same in some aspect is the same of being a kind [the sameness in regard to some aspect, i.e., the sameness of a kind]. A88 [In “Canons A”] 異, 二、不體、不合、不類。 [In “Explanations A”] 異:二必異, 二也。不連屬, 不體也。不同所, 不合也。不有同, 不 類也。

pre-theoretic; in my view, its first explicit presentation can be traced back to one fundamental point of Gongsun Long’s “White-Horse-Not-Horse” thesis to the effect that, relative to what is sought, one can say that the white horse is identical to the horse or that the white horse is not identical to the horse. As for an explicit logical presentation of the Gongsun-Long-style concept of relative identity (in predicate logic resources), see the relevant discussion in Mou 2016.

48

B. Mou Yi (異 different/distinct) [includes the following seemingly different but not mutually exclusive modes of the same things]: two [instead of being duplicated]; not being parts; not being together/united; not being of a kind. Yi (異 different/distinct): Two certainly different is two [the referential difference]; Not being jointed is not being parts [the difference of not belonging to the one whole]; Not residing in one [single] place is not being together [the difference of not being united in one single thing]; Not having the same [in some aspect] is not being of a kind [in regard to some aspect] [the difference in regard to some aspect, i.e., difference of a kind].

Analyzing the XQ text together with examining the point of the relevant passages such as these passages cited above from the Later Mohist Canons, one can identify several relevant points of the Later Mohist logical discourse to its diagnosis concerning the semantic-syntactic structure of the parallel inference. First, the last mode of the sameness as addressed in A87 is the most general one and can include the other modes as its sub-modes; in other words, the other modes can be viewed as special cases of the sameness of a kind: the referential sameness is the sameness of being of a class with regard to two names’ having the same referent; the sameness of belonging to the one whole is the sameness of being of a class with regard to both having the same part-ship of the one whole , and the same of being unified in one single thing is the sameness of being of a class with regard to both being united by the same single thing (both contributing to the identity of the single thing). Similarly, as shown in A88, the difference of a kind can include the other modes of difference. Second, what determines the identity of a kind? There seem two contributing elements: the objective “the way-things-are” foundation: there is some aspect that constitutes a “common” attribute18 among the members of a collection kind; the aspect is what one (or a group of persons) focuses on and intends to capture for a certain purpose, which results in a certain (eligible) perspective that is intended to point to and capture the aspect. So to speak, the foregoing “the way-things-are” contributing element and “perspective” element would jointly determine the (adequate) identity of a collection kind and thus the identity standard; the former provides the “objective” foundation for the adequacy of the “shi/fei” criterion, while the latter is sensitive to our diverse purposes, interests and focuses in capturing the way things in the world are. Third, distinct perspectives, as given in distinct sentential contexts, can result in distinct identities of kinds of involved things, but distinct identities of kinds are not necessarily incoherent. For example, when (one of) those common-attribute aspects between the robber and the person are focused on, it is thus stated that the robber is the same as the person [with regard to (one of) those common aspects]; it is thus true to say that they are of the same kind; to this extent (with the focus on the common aspect), it is thus said that killing the robber is the same as killing the person. On the other hand, when some distinct aspect(s) that is(are) not shared by the robber and the person (the person meaning any member of the person as a collection) is(are) focused on, it is thus stated that the robber  It is important to note that the saying “common attribute” here does not necessarily commit itself to a platonic realism regarding universals but is open to distinct ontological interpretations.

18

3  Naming, Reference and Truth

49

is not the same as the person [with regard to that distinct aspect]; it is true to say that they are not of the same kind with regard to that distinct aspect; to this extent (with the focus on the distinct aspect), it is thus true to say that killing the robber is not the same as killing the person. It is noted that there is no contradiction here; there is no violation of the principle of non-contradiction both at the ontological level and at the level of linguistic expression: one object can possess all these distinct aspects at the same time; the foregoing distinct perspectives can be consistently maintained; the foregoing distinct cases point to a perspective shift in the agent.19 In sum, essentially along with Gongsun Long’s “double-reference” line in this connection, the Later Mohists were highly sensitive to the refined semantic relation between language expressions and the way things are in the predicative context of saying something of what is designated; the Later Mohists further implemented this “double-reference” semantic sensitivity in their reflective examination of the parallel type of deductive reasoning; such a semantic concern in the Later Mohist logical discourse motivated the Later Mohists to alert us to the failure of meeting certain adequate conditions for the sake of adequately carrying out the parallel inference, although the Later Mohists then did not have sufficient or adequate logical resources to explicitly formulate the general inference rule for the validity of the parallel inference in an effective and unified way.20

 There is one substantial implication of such understanding of identity (as labeled ‘relative identity’ in contemporary logic discourse): as one might object, it is thus possible that all different things in the world, in some sense, can be regarded as the same [or similar] from a relevant perspective. I would render such possibilities (or even some related “seemingly-bizarre” but really open-minded ways of classification) very positive and constructive; this would give a thoroughly open-minded approach to look at identities/similarities among things in the world and at how to classify them: this would be sensitive to distinct eligible perspectives that point to certain aspects which are commonly or jointly possessed by things and thus meet certain reflective needs, though some of these classifications seem trivial or can be against people’s current ready-made or habitual ways of classification. 20  However, with some enhanced predicate logic resources, the deep semantic-syntactic structure of the parallel inference can be captured and presented in a formal way, although both ancient logical resources during the later Mohist times and the so-far-available logical resources (including those currently available expanded predicate logic accounts) have yet to be refined enough to take on, or are unable to formally apprehend, this. For a detailed discussion on this, see Mou 2016; for an expanded predicate logic account with enhanced identity sign that can be used to formulate the Later Mohist way in this connection, see “Appendix”. 19

50

B. Mou

3  N  aming and Reference in Confucius’ Account of Name Rectification in View of Xun Zi’s and Wang Chong’s Developments It is known that the major concern of Confucius (551–479 BCE) as presented in the Analects is with moral and social issues. His doctrine of name rectification as suggested in the Analects serves this major concern. Confucius’ teachings in terms of his account of name ratification addresses the issue of reference in moral and social areas. The passages in the Analects that are explicitly and directly related to the issue of name rectification are three: 13.3, 12.11, and 12.17. Confucius’ doctrine of name rectification (正名) might as well be another way of presenting his teachings on moral cultivation and adequate governing which can be highlighted in terms of “sageliness within while kingliness without” (內聖外王): the teaching delivered in 12.17 is to rectify yourself to fit what those terms that signify your ranks, duties, functions and moral attributes mean (12:17), which amounts to sageliness within, while the teaching delivered in 12.11 is to participate in rectifying others to fit what those terms that signify their ranks, duties, functions, and moral attributes mean (12:11), which amounts to kingliness without. However, what really interests us here is some (explicitly or implicitly) suggested general point concerning the relation between language and reality. Let me focus on an apparent puzzle in Confucius’ account that is relevant to the discussion here: there appears to be a tension between the suggested two kinds of rectification approaches. On the one hand, the trademark title of this doctrine is “name rectification”, and, as highlighted in 13.3, Confucius emphasizes the significance of name rectification. On the other hand, 12.17 and 12.11 indicate that what is rectified is actually the persons who bear the (social-identity-title) name. Which one is the primary goal while which serves as means? What is the due relation between the two kinds of rectification? Why didn’t Confucius directly emphasize rectifying the moral agent? The reason seems to be this. To rectify the person (self and others) for the sake of self cultivation and of social reform, there needs a standard or norm that itself needs language as means or even as medium for the sake of its being carried out, communicated and passed on. This is essentially a two-level rectification process with the goal of rectifying the agent into a certain prescriptively specified person. The first step is to take a semantic accent strategy: instead of directly talking about how to rectify the agent, it is to first rectify her (social-identity-title) name under examination through this: assigning it a certain due prescriptive content that specifies the standard or norm to be met by any eligible referents of the name and thus giving the primary identity condition of such referents. The second step is to rectify the agent based on the primary identity condition of the expected referents of the name that has been established in the preceding semantic accent strategy. An interesting point concerning the relation between name and actuality, as implicitly suggested by Confucius’ account of name rectification, is this. The due identity condition of actuality of a thing (say, a ruler or a person really being a ruler) is not simply its status-quo happening or current social-identity-title-bearing appear-

3  Naming, Reference and Truth

51

ance (say, the ruler-title-bearing person); rather, it consists in realization of its due place without transgressing its due scope (say, the person who really possesses the moral character that is expected for the ruler to possess and fulfills the social duty that is expected for the ruler to fulfill); the rectification of the name ‘ruler’ will play its important or even indispensable role through the name’s carrying out and delivering the norm that specifies the due place of the ruler, which is normatively denoted by the name ‘ruler’. For Confucius, as far as a moral agent as an eligible social-identity-title-name bearer is concerned, a social-identity-title name refers not merely to her specific “social-identity-title-name-bearing” surface part but also, more substantially, to her specific “social-identity-title-associated-duty-fulfilling” part, which is actionrelated, and, more fundamentally, to her specific “social-identity-title-associatedvirtue-holding” part (contributing to cultivating her “sageliness within” part if any) which per se is more fundamental than (and thus provides the guidance to) her specific “social-duty” part (contributing to fulfilling her “kingliness without” part) in this sense: the former underlies and explains the latter; one needs to first become a good person and thus do good action). In this way, there are three closely related dimensions or levels of what a social-identity-title name can (or is supposed to) refer to: a moral agent’s specific “social-identity-title-name-bearing” surface part, her “social-identity-title-associated-virtue” part (if any), and her “social-identity-title-associated-duty” (if any). If a moral agent bears a social-identity-title name only (i.e., merely with her specific “social-identity-title-name-bearing” surface part), she fails to be a genuine social-identity-title-bearing agent because she fails to possess the “social-identity-title-associated-virtue” and fulfill the “social-identity-title-associated-duty,” both of which (such virtue and such duty) are specified by the due conceptual content of the social-identity-title name and are supposed to be possessed and fulfilled by the agent; she then fails to have her specific “social-identity-title-associated-virtue-holding” part (“specific ‘virtue’ part” for short) and her specific “social-identity-title-associated-duty-fulfilling” part (“specific ‘duty’ part” for short) and thus has the social-identity-title name become an “empty” name (or a name with incompletely-fulfilled referential function) without its substantial specific-part referents other than its specific-surface-part referent (as the agent’s “social-identity-title-name-bearing” surface part). For Confucius, rectifying a social-identity-title name of a moral agent as a whole person is to have the moral agent herself, and others help the moral agent, cultivate herself to really become an eligible social-identity-title-name bearer through having her really possessing her specific “virtue” part and her specific “duty” part as the two closelyrelated specific-part referents of the social-identity-title name, instead of merely having her surface title-bearing part. At this point, one philosophically interesting question is whether this “dueplace” status of moral and social identities can be extended to the non-moral world, as suggested in Gongsun Long’s vision concerning the “due-place” of actuality in his conception of shi as explained before. Can the name rectification be thus extended to cover two distinct types of actuality, i.e., the actuality as the moral world and the actuality as the inclusive natural world? Though Confucius himself

52

B. Mou

did not handle this issue, it is Xun Zi (荀子 298-238 BCE) in the Confucian tradition who (to my knowledge) first explicitly addressed the issue in the Confucian tradition during the pre-Han period: …制名以指實, 上以明貴賤, 下以辨同異。 “…create names to designate actual things for the sake of distinguishing the [morally] worthy from the [morally] unworthy at the upper level [concerning actuality as the moral world] and of distinguishing the same from the different at the base level [concerning actuality as the inclusive natural world]”]21

Xun Zi here actually talked about one general “way-things-are-capturing” semantic relationship between names and actual things that can be further classified into two basic categories concerning distinct types of actuality, i.e., the actuality as the inclusive natural world at the base level (including the natural basis of moral traits) and the actuality as the moral world at the upper level. What this is concerned with is essentially a “way-things-are-capturing” semantic concern or truth pursuit: at the base level, generally speaking, the goal of creating names to designate actual things is to capture and distinguish between the same and different aspects of things, while, at the upper level where people reside in the moral and social world, the goal of creating names to designate actual things is to capture and distinguish between the (morally) worthy and the (morally) unworthy. As one prominent figure in classical Confucianism with its primary jun-zi-moral-cultivation mission in view, Xun Zi focuses on the semantic truth pursuit of capturing the way things are at the upper level in the sense as specified above, while he insightfully and explicitly addresses and emphasizes the metaphysical foundation for human morality through his account of xing (性 human nature) and tian-ming (天命 the fundamental principle of the natural world).22 If Xun Zi only briefly addressed the issue of whether the name rectification can be extended to cover two distinct types of actuality, it is Wang Chong (王充 27-97? CE) during the Han Dynasty who more or less systematically addressed the issue and thus unified the truth-making basis.23 In the classical text the Lun-Heng (《論 衡》), Wang Chong comprehensively resorted to the general notion of shi (實 actuality or reality) and treated capturing shi (實) as one general thing to unify what underlines the adequacy of any accounts, both moral ones, where shi and its contrary xu (虛) (emptiness, the untrue or falsity)24 are pointed to respectively by the terms shi (是 right) and fei (非 wrong), and non-moral ones concerning the inclusive  Xun-Zi, chapter Zheng-Ming.  For this author’s detailed examination of the issue, see Mou, section 5.2. 23  For my detailed discussion of Wang Chong’s approach related to thise, see Mou 2015. 24  One might question whether xu (虛) is the negation of shi (實) (thus they are contradictory) or they are merely contrary, given that the former is stronger than the latter in the sense that in a pair of contradictory sentences, one is true and the other is false, while in a pair of contrary sentences they could not both be true but could both be false. Although the text does not explicitly tell whether Wang Chong treated the relationship between shi and xu as contradictory or contrary, the weaker “contrary” relationship is assumed here. 21 22

3  Naming, Reference and Truth

53

natural world, where shi and its contrary are pointed to respectively by the terms ran (然 what is the case) and fou (否 what is not the case). For instance,25 in chapter Dui-Zuo, Wang Chong said: 是故《論衡》之造也, 起眾書並失實, 虛妄之言勝真美也。…況論衡細說微論, 解釋 世俗之疑, 辯照是非之理, 使後進曉見然否之分。 The reason why the Lung-Heng was written is this: many books fail to capture shi [truth]; the false and absurd speeches become prevalent over true and decent ones…. Through elaboration and detailed examination, the Lun-Heng sets to explain the doubts of common people, bring to light the principles of what are right and wrong through debate, and thus help those coming later clearly distinguish between what is the case and what is not the case.26

What is at issue is this: What underlies such a unifying account? How is it possible to unify the two types of actuality into one general “shi” category for the sake of the pursuit of language capturing the way things are? How is it related to people’s pretheoretic “way-things-are-capturing” understanding of truth? Wang’s approach consists of two points. First, what fundamentally unifies the two types of actuality at the base/root level concerning the pursuit of language capturing the way things are consists essentially in Wang’s “normative” elaboration of people’s pre-theoretic “way-things-are-capturing” understanding of truth.27 Second, Wang’s reflective elaboration of the normative element of people’s pre-theoretic understanding of truth goes essentially along with, and develops, the line of a shared reflective point made among those pre-Han classical Chinese philosophers like Gongun Long, Confucius, and Xun Zi concerning one crucial “normative” parameter or contributing element to the genuine identity of the way things are (actuality), i.e., the due way things are (their due limits and due places in the world); this crucial “normative” element constitutes one intrinsic coacervating force that fundamentally unifies two types of actuality (non-moral and moral ones) into one and two types of truth pursuit into one.28

 With consideration that it is agreed that Wang treated shi as univocal, and to save space, I will not make a wide range of citations from the Lun-Heng here to show the textual evidence for this. The interesting reader can examine such chapters in the Lun-Heng as Qi-Guai (奇怪), Shu-Xu (書虛), Tan-Tian (談天), Zi-Ran (自然), Shi-Zhi (實知), Zhi-Shi (知實), Zheng-Shuo (正說) besides the chapters Wen-Kong (問孔) and Dui-Zuo (對作). 26  Lun-Heng, Chapter Dui-Zuo. (《论衡校注》, 569 and 571). 27  For a different interesting interpretation of Wang Chong’s approach via his conception of shi, see McLeod 2011. 28  For this author’s further explanation of Wang Chong’s approach, see Mou 2015; for a more complete account of the addressed issue of this section, see Mou 2019, chapter 5. 25

54

B. Mou

4  N  aming and Reference in Lao Zi’s Daoist Account: The Speakable and the Non-speakable In this section, through an examination of the opening message of the Dao-De-Jing which is delivered by its opening statements consisting of the first and second sixcharacter statements in Chapter 1 in the context of the Dao-De-Jing as a whole, I intend to spell out one of Lao Zi’s fundamental insights in the Dao-De-Jing on naming and reference, via his reflation on the relation between the ultimate concern with the eternal dao (chang-dao) and its truth-pursuit language engagement.29 By ‘the truth-pursuit language engagement with an ultimate concern’ I mean any reflective endeavor to capture (reach or characterize) the way the ultimate reality is through language. It is often believed that, in the opening message, Lao Zi30 makes his sharply negative claim in this connection to the effect that any language engagement is doomed to fail to capture the genuine dao. I intend to show that, instead of indiscriminately giving a negative claim against any language engagement with the ultimate concern, in the first six-character statement, Lao Zi reveals a two-sided transcendental insight which, on the one hand, positively affirms the role of the language-engaged finite point of view in capturing the ultimate concern and, on the other hand, alerts us to the limitation of the finite point of view and emphasizes the transcendental dimension of the Dao as the ultimate concern; and, furthermore, in the second six-character statement, through a semantic ascent of talking about the name of the Dao , Lao Zi delivers essentially the same message, though in a certain distinct way that is philosophically interesting. In the following, I first give an analysis of one traditional paraphrase in English that clearly delivers one prevailing standard interpretation, and then I explain my paraphrase of the first six-character statement of the opening passage. Second, in the context of the Dao-De-Jing, I spell out Lao Zi’s two-sided transcendental insight revealed in the first statement. Third, I then explain how, in some philosophically interesting way, Lao Zi delivers essentially the same message in the second six-character statement via a semantic ascent of talking about the name of the Dao. One of Lao Zi’s most fascinating teachings in his Dao-De-Jing is the first pair of 12-character poetized aphorisms in Chapter 1: Dao-ke-dao-fei-chang-Dao (道可道非常道);31

 For my detailed discussion of this issue, see Mou 2000, 2003.  It is known that there has been a controversy over the identity of Lao Zi. Nevertheless, throughout this essay, Lao Zi is taken as a proxy figure who speaks for the ideas delivered by the text of the Dao-De-Jing; the appearance of the name ‘Lao Zi’ thus stays neutral to the controversy. 31  In my pin-yin transliterations in this section, unless in the case of proper names, the transliteration of a Chinese character with the first letter being capital is used to indicate that the character in the relevant context of the Chinese original is used as a noun. In contrast, the transliteration of a Chinese character without the first letter being capital is used to either indicate the non-noun-status of the character in the context or suspend claiming its syntactic status. A prominent example in this aspect is the transliteration ‘Dao’ in contrast to ‘dao’; and, by the same token, ‘Ming’ in contrast 29 30

3  Naming, Reference and Truth

55

(1∗) Ming-ke-ming-fei-chang-Ming (名可名非常名). which, according to one (type of) standard, or the most prevalent, translation, is paraphrased as follows: The way that can be spoken of is not the constant way; The name that can be named is not the constant name.32 or The Dao that can be told of [in language] is not the eternal Dao; The name that can be named [in language] is not the eternal name.33 This type of translation/paraphrase might even assume the standard paraphrase.34 It is often thought that what the opening statement reveals is a fundamental Daoist fundamental insight that is strikingly similar to that of Wittgenstein’s well-known idea about the spoken and the unspoken: language expressions or formulations, or what is captured within language, cannot really capture what those expressions or formulations aim to say; the genuine Dao has to be captured in a way that is beyond language. It is often said that in Lao Zi’s case, contemplation of the Dao in silence requires sharply distinguishing the eternal Dao from what can be formulated or captured in (or by) language, for the two are simply opposed to each other. This view clearly has a bearing on the linguistic structure of the standard English translations cited previously: (2) The Dao that can be told of or formulated (in language) is not the eternal Dao; That is, as it literally says, the so-called Dao that can be told of is not the eternal Dao or is not the genuine Dao when the eternal Dao is identified as the genuine Dao; whatever can be told of in language is not the genuine Dao. In other words, the genuine Dao, as an ultimate concern, simply cannot be captured through language; any language engagement with the genuine Dao is doomed to fail to capture the genuine Dao. Note that this kind of paraphrasing implies that the first appearance of the term ‘Dao’ in noun in the opening statement of the Dao-De-Jing does not refer to or designate the genuine Dao. The preceding standard interpretation of the opening statement of the Dao-DeJing does not capture the complete message of what is delivered there, although it does partially capture the point to the extent I will explain below. Now one question is whether or not, and to what extent, the statement (2), or its Chinese counterpart

to ‘ming’. 32  Lau 1963, 5. 33  Chan 1963, 139. 34  In Creel 1983, Herrlee Creel reviews 29 English translations of the opening passage. Among them, as Creel sees it, the standard pattern is used in 14 translations, if some rearrangement and some use of synonyms are tolerated. As I see it, all those 29 translations examined in Creel’s article seem to share the same core structure regarding the opening statement to the effect that the Dao (as a noun) that can or may be dao (as a verb) is not the chang-Dao.

56

B. Mou

(2′), is an adequate paraphrase of the original statement (1). Let us examine the paraphrase in two aspects: syntax and semantics; one case see that the paraphrase (2) fails in both aspects. As far as the syntactic aspect is concerned, (2) does not have the same linguistic or syntactic structure as that of (1). In the Chinese original, “Dao-ke-dao-fei-chang-Dao”, is a conjunction which consists of two conjunct claims rather than one: the first one is ‘Dao-ke-dao’ asserting that the Dao can be talked about in language, while the second one is that what has been characterized in language is not, or cannot exhaust, the eternal Dao, which might be paraphrased back into the following Chinese expression: ‘[suo-shu-zhi-Dao] fei-chang-Dao.’ That is, a complete presentation of the two correlative messages in the context of the first six-character statement might well be given in Chinese this way: (3) Dao-ke-dao, dan-suo-shu-zhi-Dao-fei-chang-Dao; or its paraphrase in English: (3′) The Dao can be talked about in language, but the Dao that has been talked about in language is not the eternal Dao; or, more exactly, (3″) The Dao can be reached in language, but the Dao that has been characterized in language is not identical with, or does not exhaust, the eternal Dao. One thing is clear: in (1), the Chinese original ‘Dao-ke-dao’ is a positive message delivered via a positive assertive statement, whereas in (2) or (2′), the positive assertive statement is turned into a complex noun-phrase ‘the Dao that can be told of’, which consists of a noun and its attributive clause, and then makes its negative contribution to the alleged meaning of the opening statement with the notion that the Dao that can be told of is not the eternal Dao. To this extent, the difference in the syntactic structures of (1) and (2) does make difference in their meanings. This also provides one reason why, semantically speaking, (2) fails to capture what (1) delivers. Here one can see why and how a certain difference in syntactic structure bears on some difference in semantic dimension. Now let us see why (2) fails in the semantic connection for semantic reasons that are independent of the aforementioned syntactic structure. One clear textual fact is that, throughout the Dao-De-Jing, Lao Zi himself talked about the genuine Dao— the ultimate concern pursued. His language practice in the Dao-De-Jing shows his own language engagement with the ultimate concern. Lao Zi talked even about the eternal Dao (chang-Dao)—exactly in the opening statement. One might object that, although Lao Zi talked about the eternal Dao, he emphasized that he could not clearly characterize the eternal, infinite Dao as the origin of all things in the universe.35 Fair enough. However, note that the second appearance (as a verb) of the character “dao” in the opening statement does not merely mean being characterize in descriptive language; rather, it covers a variety of language-engagement activi-

35

 Cf., e.g., Chapters 14, 21, 25 of the Dao-De-Jing.

3  Naming, Reference and Truth

57

ties, including direct reference through rigid designators. It is known that there are three kinds of naming activities in language practice: descriptively assigning a name to an (any) object or thing that meets the descriptive content associated with the name; in this case, the solely descriptive name denotes whatever does fit the description; rigidly, through direct reference (without via description), assigning a name to a certain specific object or thing that does not necessarily meet the descriptive content (if any) associated with the name; in this case, the name rigidly designates that specific referent through direct reference; and rigidly, and descriptively, assigning a name to a certain specific object or thing that necessarily meets the descriptive content associated with the name; in this case, the name rigidly designates and specifies the referent. There are thus three kinds of names in regard to their naming functions: descriptive names, such as ‘the first important Daoist philosopher’ (it could be Yang Zhu or Lao Zi or Zhuang Zi or whoever meets the descriptive content of this descriptive name); rigid designators via direct reference, such as ‘Dao’ used by Lao Zi in the context of the Dao-De-Jing; descriptive (rigid-)designators, such as ‘Da’ in Chapter 25 and ‘Yi’ in Chapter 14, which were used by Lao Zi himself and can be reasonably assigned to the Dao in the context of the Dao-De-Jing. One can easily find that the genuine Dao, or even its eternal, constant and infinite dimension, is reached in language in various ways in the Dao-DeJing: In the bottom-line case regarding naming, it is rigidly designated either through direct reference (without via description) or through description36 by such rigid designators or descriptive (rigid-)designator as ‘Dao’, ‘chang-Dao’37 or ‘Da’.38 Let us further have a look at the second statement “Ming-ke-ming-fei-changMing” in the opening message of the Dao-De-Jing that explicitly addresses the issue of names and naming. In the Dao-De-Jing, the Chinese character ‘ming’/’Ming’ has altogether twenty-four appearances in ten chapters: Chapters 1, 14, 21, 25, 32, 34, 37, 41, 44 and 47. The appearance of ‘Ming’ in Chapter 44 is used to mean ‘fame’, and the appearance of ‘ming’ in Chapter 47 is used as the pseudonym of another Chinese character that shares the same pronunciation and tone meaning ‘understand’; these two appearances thus can be disregarded for the purpose here. Examining the remaining twenty-two appearances of ‘ming’/’Ming’ in regard to naming and names, one can find that Lao Zi makes two types of different or even seemingly inconsistent claims on whether the Dao could be named or whether the Dao could have its names: On the one hand, Lao Zi, as widely highlighted and long celebrated in the literature, claims that the Dao cannot be named (cf., Chapters 14 and 25) and that the Dao does not have names (cf., Chapters 1, 32, 37 and 41); but, on the other hand, Lao Zi also claims that the Dao can be named (also see Chapters 14, 21 and 25) and that the Dao does have its names (cf., Chapters 1, 21 and 32).  For the notion of rigid designation, see Kripke 1980. It is noted that the notion per se presupposes neither Kripke-style essentialism nor any other ontological commitment beyond the minimal one—the existence of what is rigidly designated; to this extent the notion of rigid designation is metaphysically neutral. 37  Cf., Chapter 1 of the Dao-De-Jing. 38  Cf., Chapters 25 and 34 of the Dao-De-Jing. 36

58

B. Mou

The traditional or standard interpretation discussed before actually treats only the first type of claims as delivering Lao Zi’s genuine ideas in this regard while dismissing the seriousness of Lao Zi’s second type of claims.39 This treatment fails to give a complete account of Lao Zi’s line of thought in this connection, which is fundamentally guided by the ancient wisdom revealed in the Yin-Yang model of complementary interaction. Essentially going along with his insight in the first sixcharacter statement in the opening passage, Lao Zi’s line of thought in the second six-character statement (or generally speaking, on the issue of language engagement with the Dao via naming) also consists of a two-sided transcendental view. On the one hand, when claiming that “Ming-ke-ming”, Lao Zi positively affirms the role of the language-engagement with the Dao via names in capturing the Dao: (i) it is presupposed in “Ming-ke-ming” that the Dao can be named [rigidly designated] by the constant name “Ming” (such as “Dao”); and (ii) it is explicitly and positively claimed that the Dao can be further [descriptively] named [designated by certain descriptive designators]: in other words, the [constant] name of the Dao can be further named [i.e., further descriptively paraphrased by certain descriptive designators].40 On the other hand, when claiming that “fei-chang-Ming”, Lao Zi warns us that any descriptive names [descriptive designators] regarding certain characteristic finite aspects, dimensions or layers are not identical with the constant name of the Dao that names [rigidly designates] the wholeness or infinite dimension of the Dao; in this way, Lao Zi alerts us to the limitation of the finiteness of any descriptive designators and emphasizes the transcendental dimension of the Dao that transcends any finite aspects of the Dao itself and any finite stages of its infinite development. In this way, the second six-character statement of the opening passage delivers essentially the same message as that delivered in the first six-character statement, though through a semantic ascent and in a distinct way.

5  T  he Issue of the Structure of Chinese Language and Chinese Thought: From the Referential Point of View The problem of relating Chinese thought to the structure and functions of the Chinese language has for generations tantalized sinologists and those philosophers who are concerned with the problem. Nevertheless, in the last three decades, some

 For example, Wing-tsit Chan commented: “Lao Tzu…rejected names in favor of the nameless…. To Lao Tzu, Tao is nameless”, suggesting that, to Lao Zi, what is named is simply not the genuine Dao. Cf., Chan 1963, 139. 40  The term ‘ming’ as a verb in ancient Chinese can be used to (rigidly or descriptively) name a non-linguistic object or descriptively paraphrase a linguistic object (say, a remark). One example of the latter case is ‘mo-ming-qi-miao’ meaning that, referring to something (often to some linguistic remark; e.g., “A remark mo-ming-qi-miao”), one cannot descriptively paraphrase or explain its subtlety; another is ‘bu-ke-ming-zhuang’ x meaning that something can hardly be descriptively paraphrased or explained. 39

3  Naming, Reference and Truth

59

substantial and significant investigation into the issue has been carried out. Although the discussion of the issue is not directly related to logic but more an issue in philosophy of language, I briefly address it as it is indirectly related to the semantic foundation of relevant logical inference. Chad Hansen (1983) advances a novel and provocative theory about the nature of the classical Chinese language. The central thesis of Hansen’s theory is his mass-noun hypothesis. Its main ideas are these: (1) the mass-noun semantics: the (folk) semantics of Chinese nouns are like those of mass-nouns (i.e., those nouns referring to the so-called interpenetrating stuffs, like the nouns ‘water’ and ‘snow’), and naming in Chinese is not grounded on the existence of, or roles for, abstract entities (either on the ontic level or on the conceptual level) but rather on finding ‘boundaries’ between things; (2) the ontological outlook: influenced by the mass-noun semantics, the classical Chinese semantic theorists and ontological theorists view words in ways that are natural to view mass nouns rather than count nouns, and Chinese theorists tend to organize the objects in the world in a mereological stuff-whole model of reality (the term “mereology,” in its technical sense, means the (mathematical) theory of the relation of parts to whole). In this way, according to Hansen, the language theory of classical Chinese philosophers differs fundamentally from the language theory of Western philosophy. Since Hansen’s mass-noun hypothesis was raised, it has been challenged from various points of view.41 With the purpose of this writing, I focus here only on two challenges that engage with the mass-noun-semantics part of Hansen’s hypothesis, as presented in (1) above, from examining the referential relation per se. One way is to challenge Hansen’s mass-noun semantics by arguing that there is a clear grammatical distinction in classical Chinese between count nouns and other nouns (Harbsmeier 1991). Claiming that there is a clear grammatical distinction in classical Chinese between count nouns and other nouns (generic nouns and mass nouns), Harbsmeier (1991) insists that the mass-noun hypothesis is “historically implausible and grammatically quite wrong-headed”. However, as Hansen himself emphasizes (1992), his mass-noun hypothesis is not a syntactic claim that classical Chinese nouns have mass-noun grammar but a semantic interpretive hypothesis that the semantics of Chinese nouns may be like those of mass nouns, and classical Chinese theorists view words in ways that are natural to view mass nouns. So it seems to Hansen that Harbsmeier systematically confuses syntax and semantics and misinterprets his semantic hypothesis. Although I agree with Hansen at this point, Harbsmeier’s criticism is not irrelevant in the following sense. Harbsmeier insists that his alleged distinction between count nouns and other nouns is not merely grammatical but also semantic (or takes the grammatical difference in question to have semantic implications); Hansen thus needs to deal with the linguistic (semantic) evidence against his hypothesis that the semantics of classical Chinese nouns may be like those of mass-nouns.

 For the challenges from other perspectives, see, for examples among the relevant English publications, Hall and Ames 1987, Cheng 1987, Roetz 1993, Fraser 2007, and Fung 2009.

41

60

B. Mou

I have responded to Hansen’s view in a similar semantic perspective and within the same mereological-analysis track; but, disagreeing with Hansen’s mass-noun hypothesis, I suggest and argue for a collective-noun hypothesis (Mou 1999). Its partial view relevant to the issue of reference takes it that Chinese common nouns typically function, semantically and syntactically, in the way collective-nouns (that is, those nouns that denote collections of individual things, like the English nouns ‘people’ or ‘cattle’ with regard to their semantic-syntactic structure) function, and the folk semantics of Chinese nouns are like those of collective-nouns;42 the referent of a Chinese common noun is a mereological collection of individuals both with the part-whole structure and with the member-class structure, which thus does justice to the role of abstraction at the conceptual level. In this connection, the collective-noun hypothesis makes a stronger claim that Chinese nouns do not function as count nouns but typically function, both syntactically and semantically, as collective-nouns.43

6  Reference and Truth It is known that reference and truth are two closely or even intrinsically related basic semantic notions by virtue of which logical reasoning and logical-inference-related resources are to hook up to the world (or, in quasi-technical terms in modern logic, the domain of quantification as a non-empty set), which is shown and illustrated by the “semantic interpretation” part of the suggested expanded predicate logic account with enhanced identity sign and collective-generic operator, as presented in “Appendix”.44 It is also noted that the (semantic) notion of truth, as conceived and pre-theoretically characterized by people’s pre-theoretic “way-things-are-capturing” understanding of truth that plays its tremendous and indispensable explanatory role in people’s folk and reflective lives, is fundamentally explained in terms

 It is noted that what is focused on in the collective-noun hypothesis is the semantic-syntactic structure of Chinese common nouns, instead of some “grammatical” feature of collective nouns in a certain natural language (say, English); in this way, not merely these nouns ‘team’, ‘jury’, ‘committee’ in English and their counterparts in other natural languages, but also ‘people’, ‘cattle’ and ‘police’ in English and their counterparts in other natural languages, are all collective names with regard to their shared semantic-syntactic structure, whether or not they would “grammatically” count as collective nouns in a natural language (if there is such a ready-made grammatical category in that natural language). 43  Actually, I now hold a stronger expanded collective-name hypothesis concerning the semanticsyntactic structure of common nouns in natural languages (not restricted to the case of common nouns in the Chinese language). For my detailed discussion of this, see Mou 2018, Section A1.2 of Appendix 1. 44  It is noted that, in the literature, different things are talked about in the name ‘truth’. To avoid mere verbal disagreement, I explicitly specify that what is addressed here is the semantic notion of truth as conceived and delivered by people’s pre-theoretic “way-things-capturing” understanding of truth, instead of something else. 42

3  Naming, Reference and Truth

61

of“aboutness” or “reference”, which points to one central characteristic feature of the “capturing” relationship between the truth bearer and the truth maker. The “reference” resources thus constitute one central “metaphysical” and explanatory foundation for truth. In this sense, and to this extent, the content of this essay on the issue of reference is also intrinsically related to an account of how (semantic) truth is possible in any intellectual inquiries into “how things are”, generally speaking, and in logic, specifically speaking. The distinct approaches to the issue of reference in Chinese philosoiphy, as partially discussed here in view of logic, essentially constitute distinct semantic-truth approaches in Chinese philosophy, a substantial point that I do not further discuss in this essay, due to its topic and space, but have given a detailed systematic examination in a monograph book of this author (Mou 2019); an early examination of its core idea related to the debate between substantivism and deflationism in the setting of contemporary philosophy is presented in my previous monograph (Mou 2009).

 ppendix: An Expanded Predicate Logic Account A with Enhanced Identity Sign and Collective-Generic Operator45 For the sake of formally and more accurately capturing, presenting and illustrating some of the addressed points that have been made in this essay (among others, the relative identity and semantic sensitivity as addressed in Gongsun Long’s and the Later Mohist accounts, the due relationship between the two basic semantic notion (reference and truth) as addressed in the last section and as resorted to in the semantic interpretation of a logical system, and the collective-name hypothesis regarding the semantic-syntactic structure of common nouns in natural languages), I propose the syntax and semantics of an enhanced expanded and strengthened account of “relative-identity-sensitive” and “collection-sensitive” predicate logic (“RI-C PC” for short) with an enhanced identity sign (called “identity with attribute-in-focus parameter”), many-sorted variable, and “collection-generic operator” symbol. In this expanded predicate logic account, there are the following additions on the basis of the standard predicate logic account, some of which are new or partially new while the others not: (1) adding the “collective-generic operator” symbol, which is partially new;46 (2) adding the sign ‘[ ]=’ for “aspect-in-focus parameter” identities, which is new;47 (3) adding the sign ι∗ for complex noun phrases (definite  This appendix writing is from Mou 2018, which is a substantial revision of an earlier version of the suggested expanded predicate logic account as given in Mou 2016 (including the addition of the coverage of collective-generic operator). 46  This “collective-generic” operator can be viewed as a further development from Krifka, M. et al. 1995 in which a “generic” operator is suggested. 47  When saying that it is new, I mean that the logical notation to be given below, i.e., “the identity symbol with ‘perspective-attribute-in-focus’ parameter [ ]=”, is new in view of a standard predicate 45

62

B. Mou

descriptions), which is a further expansion on the standard ι operator – semantically defined in a modified way; (4) adding the sign λ for complex predicates, which is not new; (5) turning one-sorted logic into many-sorted logic in the way to be defined, which is not new either; (6) this expanded logical system includes both predicate variables and predicate constants (added in our primitive vocabulary) that symbolize attributes; their semantics is partially standard one while being enhanced with modified domain and interpretation, which is partially new.48 (1) Syntax of RI-C PC (1.1) Primitive vocabulary individual variables x, y… with or without numerical or letter subscripts individual constants (names) a, b…, with or without numerical or letter subscripts sortal variables s, p…, with or without numerical or letter subscripts for each n > 0, n-place function symbols f, g…, with or without numerical or letter subscripts “definite description” symbol ι*49 for each n > 0, n-place predicate variables X, Y…, with or without numerical or letter subscripts for each n > 0, n-place predicates (predicate constants) A, B…, with or without numerical or letter subscripts “identity” symbol with “perspective-attribute-in-focus” parameter [ ]=50 logic account, instead of the basic idea of relative identity or its other logical expressions being new. The basic idea of relative identity is not odd but quite pre-theoretically intuitive in view of its close relation to our pre-theoretic understanding of the “double-reference” phenomenon of the basic language employment, which is addressed and explained in discussing Gongsun Long’s account, though he did not use the Chinese counterpart of the very phrase ‘relative identity’, and though in the logic literature it is often attributed to Geach 1967 in which Geach criticizes the standard notion of absolute identity in the standard first-order predicate logic and suggests his “modern” version of relative identity which I think presupposes absolute identity. It is noted that, in Mou 2016, the relative identity sign is given in the form of “=[ ],” in contrast to “[ ]=” given here for some substantial consideration. 48  The presentations of those added materials that are labeled “not new” are quite standard; their basic presentation lines and fashions can be found in many textbooks or more advanced source books for classical and non-classical predicate logic, such as Gamut 1991, Priest 2008, and Sider 2010, though newly introduced resources into the system unavoidably bear on some aspects of the presentations of those previous materials. 49  See its semantic interpretation (2.2) below, which distinguishes itself from the standard semantic interpretation of the “definite-description” symbol “ι.” 50  See its semantic interpretation (2.3) below, which distinguishes itself from those of the “identity” symbol “=” and of the Geach-style “relative identity” symbol (it places the “relativeidentity” parameter beneath equation symbol in the middle, indicating its symmetric character, cf., Wiggins 2001, 24-28). It is noted that Geach (1967) presents another logical notation to express relative identity (basically: x and y are the same F but x and y are different Gs, where F and G are predicates). Generally speaking, I contend that this logical notation in treating relative identity is not merely less natural and expressive (for example, it would be hard to present Gongsun Long’s “white-horse-not-horse” case) but also somehow presupposes absolute identity; the suggested

3  Naming, Reference and Truth

63

“complex predicate” symbol λ “collective-generic operator” symbol CG51 connectives: →, 〜 (universal) quantifier: ∀ collective-generic quantifier: CG ∀ parentheses: (, ) (1.2) Definition of terms: Any individual variable, sortal variable or individual constant is a term If f is an n-place function symbol and α1… αn are terms, then f (α1… αn) is a term If φ is a formula and α is an (individual or sortal) variable, then ι*αφ is a term Only strings that can be shown to be terms by the preceding clauses are terms (1.3) Definition of formulas: If Π is an n-place predicate and α1… αn are terms, then Π α1… αn is an (atomic) formula If π is an n-place predicate variable and α1… αn are terms, then π α1… αn is a formula If α, β and γ are terms, then α [γ]=β is a formula If φ is a formula, α is a variable and β is a term, then λαφ(β) is a formula If φ is a formula, α is a term, then CGφ(α) is a formula If φ and ψ are wffs, and α is any variable, then 〜φ, (φ → ψ), and ∀αφ are formulas Only strings that can be shown to be formulas using , , , , and are formulas (1.4) Definition of derivative logical symbols: Definition of ∧: “φ∧ψ” is short for “〜(φ →〜ψ)” Definition of ∨: “φ∨ψ” is short for “〜φ →ψ” Definition of ↔: “φ↔ψ” is short for “(φ→ψ)” ∧“(ψ → φ)” Definition of ∃: “∃αφ” is short for “〜∀α〜φ” Definition of ≠: “α [γ] ≠ β” is short for “〜(α [γ]=β)” identity notation with the parameter as a primitive, together with its semantic interpretation (2.3) below, treats the traditional identity sign expressing absolute identity as one special case, and it treats as not symmetric the type of relative identity that is intrinsically related to the category-belonging predication. The issue of symmetricity involved in the conception of relative identity is beyond the coverage of this book, which I will examine in another writing. 51  In Krifka, M. et al. 1995, a generic operator is suggested concerning only the feature (4) of common nouns mentioned above or the case below, having yet to take care of the feature (5) of common nouns mentioned above or the case below.

64

B. Mou

(1.5) Definition of free and bound variables: An occurrence of a variable x in a formula φ is bound in φ if and only if that occurrence is in the context of the form ∃αφ or ∀αφ within φ. If it is not bound, it is free. A formula with no free variables is a closed formula or sentence; otherwise it is an open formula (2) Semantics of RI-C PC Definition of value-assigning interpretation, model, for RI-C PC: A RI-C PC model is an ordered pair〈D, v〉such that: (2.1) D is a non-empty set (“the domain of quantification”); D is divided into two types of primary subsets as sorts: individual-object subsets O of individual objects, d1, d2, …, which are divided into various secondary subsets (sorts or sortal collections), O1, O2, … whose “nominal” identities (or whose memberships) are given by the distinct term ι*αφ (where φ is a formula and α is a sortal variable) subset A of all specific grounded parts (specific aspects, particular attributes,…) that are grounded in, and depend on, individual objects (though the defining identities of some of them, such as relational attributes, are grounded in more than one individual objects), which can be further divided into three kinds of subsets whose members can be overlapped: universal-attribute subsets, A1, A2,…, which are various subsets of particular attributes whose memberships constitute (or are given respectively by) various universal attributes individual-object-association subsets, Ad1, Ad2, …, which are different subsets of specific grounded parts (specific aspects: such as collection-nominal-identity-determining attributes, collectively-holding attributes, generically-holding attributes, …) whose memberships are given by their respective associations with different individual objects, d1, d2, …; each of individual objects d1, d2, …, as a whole can be thus labeled d1-Ad1, d2-Ad2, …52 sortal-collection-associated subsets of attributes concerning sortal collections which might be further divided into three types (for a sortal collection of individual object Oi, it includes one, or more than one, of the following three types of sortal-collection-associated subsets of attributes):53 the subset Oi-collection-identity-attributes of collection-identity-determining attributes, which can be abstractly presented by the related universal attributes (such as the “horseness” attribute), and which are  Such identities of individual objects with its associated attributes as a whole capture our pretheoretic understanding of individual objects as “thick” objects, which constitute the semanticwhole referents of the names of such thick objects if they do have names. 53  Two notes are due. First, it is not necessary for any sortal collection to include all these types of sortal-collection-associated subsets. Second, these three subsets are not exhaustive (because some “accidental” attributes possessed by some of these objects in Oi do not fall into any of these three subsets), though they are exclusive. 52

3  Naming, Reference and Truth

65

possessed by any individual objects of Oi (for example, the “horseness” attribute or the “white-color” attribute is possessed by any individual object in the white horse as a sortal collection of individual white horses); the subset Oi-collection-generic-attributes of collection-generic attributes, which are not necessarily possessed by each individual member of Oi, (for example, the attribute of having four legs is not necessarily possessed by each of the individual members of the collection of horses but generically (typically or normally) possessed by the members of the collections); the subset Oi-collection-created-attributes of collection-created attributes, which are possessed not by any individual member of Oi, but only by Oi, as a whole (for example, the attribute of winning a team game cannot be possessed by any individual member of a team of the team members but by the team as a whole).54 (2.2) v is an (interpretation) function such that: v for terms if α is a (non-predicate) constant, then v (α) ∈D: if α is an individual object-identifying constant: v (α) is an individual object in D [which is designated by α] if α is a particular-attribute-identifying constant: v (α) is a particular attribute aij, which is a member of a “grounded” subset (sort) Ai in A (notes: the subset or the membership of the subset is identified by a “nominal” universal attribute Ai that is symbolized by a predicate Pi, the member of the subset is possessed by an individual object in D, which is a member of a subset (sort) of D, whose membership is specified and named by a “nominalized” predicate Pi, i.e., the term ι*xPi) if α is a universal-attribute-identifying constant: v (a) is a universal attribute which (as the membership) gives the identity of a universal-attribute subset Ai in A if α is an individual or sortal variable: if α is an individual variable: v (α) is an individual object in D if α is a sortal variable: v (α) is either an individual object in a subset of O, whose membership is specified and named by a  Indeed, a purely-mathematically-oriented logician would render this characterization of the domain of quantification too much metaphysically-loaded. It is noted that the primary purpose of presenting this enhanced and expanded predicate logic account is to capture the Gongsun-Longstyle relative identity and the collective-name hypothesis regarding the semantic-syntactic structure of common nouns in natural languages in a more accurate and effective way, instead of merely formal consideration.

54

66

B. Mou

“nominal” universal attribute, i.e., the term ι*αφ,55 or an (universal) attribute among various universal-attribute subsets, A1, A2,…, of A (A as a sort in D), or a particular attribute among various members of a universal-attribute subset, Ai, of A (Ai as a sort in A), or a particular attribute among various members of an individualobject-association subset, Adi, of A (Adi as a sort in A) if α is an ι* term: if φ is a formula and α is an individual variable, then v(ι*αφ) is a unique object in D if φ is a formula and α is a sortal variable, then v(ι*αφ) is a (unique) subset (sort) of D, which is named, and whose membership is specified, by ι*αφ if α is an n-place function term of the form f (α1… αn), where α1… αn are terms, and v(α1),… v(αn) are well defined, then v(α) = v(f )[v(α1)… v(αn)], which is an n-place (total) function on D v for formulas if Pi is an n-place predicate (predicate constant) and αi1… αin are terms, then vo (Pi) concerning individual objects in D, vo (Pi αi1… αin) = 1 iff〈v(αi1)… v(αin) 〉∈ vo (Pi), an n-place relation over D, i.e., a subset of n-tuples, individual objects, from D, and va (Pi) concerning attributes that are shared by individual objects in D, va (Pi), is the subset (the universal attribute Ai that identifies the membership of the subset) of particular attributes Ai1, Ai2 … which are possessed respectively by these individual objects. [In the following, by default, v (Pi) means vo (Pi) unless indicated otherwise.] if p is a predicate variable if p is an n-place predicate variable and α1… αn are terms, then v(p α1… αn) = 1 iff〈v(α1)… v(αn) 〉∈vo (p) if p is an n-place predicate variable and α1… αn are terms, then va (CG p) is concerned with the subset of collection-generic attributes in the subset Oi-collection-generic-attributes or the subset of collection-created attributes in the subset Oi-collection-created-attributes, and vo 55  Notice that the semantics for the symbol “ι*” is different from the standard one for “ι” (signifying the uniqueness of the single one object as the referent of a “definite” description) but an enhanced expansion of the latter so as to have it (in the predicate logic) more adequately capture how “definite” descriptions (descriptive noun phrases with unique referents) are used in our linguistic practice (in natural languages): a definite description as a noun phrase denotes either an unique object or a unique set of objects that meet(s) the description of the noun phrase; formally speaking, as indicated in the clause (2.2), the semantics for ‘ι*’ is presented as follows: if α is an ι* term: the case is presented this way: if φ is a formula and α is an individual variable, then v(ι*αφ) is a unique object in the domain D, while the case is presented this way: if φ is a formula and α is a sortal variable, then v(ι*αφ) is a (unique) subset (sort) of D, which is named, and whose membership is specified, by ι*αφ.

3  Naming, Reference and Truth

67

(CGp) is this: vo (CG p α1… αn) = 1 iff v(αi1)… v(αin)〉∈ vo(CGp), an n-place relation over sortal collections, i.e., a subset of n-tuples, sortal collections in D; if p is a predicate variable and φ is a formula, then vo(∀pφ) = 1 iff for every set of n-tuples from D, vo (φ) = 1. if α is a sortal variable and CGφ is a formula, then vo(CG∀(α CGφ)) = 1 iff for every subset of n-tuples, sortal collections in D, vo (CGφ) = 1. if p is an n-place predicate variable and α1… αn are terms, then v(p α1… αn) = 1 iff〈v(α1)… v(αn) 〉∈vo (p); if p is an n-place predicate variable and α1… αn are terms, then va (CG p) is concerned with the subset of collectiongeneric attributes in the subset Oi-collection-generic-attributes or the subset of collection-created attributes in the subset Oi-collection-created-attributes, and vo (CGp) is this: vo (CG p α1… αn) = 1 iff v(αi1)… v(αin)〉∈ vo(CGp), an n-place relation over sortal collections, i.e., a subset of n-tuples, sortal collections in D; if p is a predicate variable and φ is a formula, then vo(∀pφ) = 1 iff for every set of n-tuples from D, vo (φ) = 1. if α is a sortal variable and CGφ is a formula, then vo(CG∀(α CGφ)) = 1 iff for every subset of n-tuples, sortal collections in D, vo (CGφ) = 1. for any terms α, β and for a term γ which is either a universal-attribute-identifying constant or a sortal variable ranging over the universalattribute subsets in A, v(α [γ]=β) = 1 iff v(α) and v(β ) share v(γ) [v(α [γ]≠ β) = 1 iff v(〜(α [γ]=β))= 1]; that is, the formula α [γ]=β is true iff what the term α denotes is identical to what the term β denotes regarding the shared attribute which the term γ denotes. Especially, when what the term α denotes is identical to what the term β denotes regarding ALL (universal) attributes in A, the term α and β refer to the same object, which case is symbolically labeled ‘α = β’, as ∀γ α [γ]=β [that would restrict the domain of the second-order quantifier to a particular collection of “universal-attribute subsets, A1, A2,…, which are various subsets of particular attributes whose memberships constitute (or are given respectively by) various universal attributes” (see (2) (2.1) above)]. In this way, the traditional identity sign together with its semantic interpretation is one special case of the identity sign with the “aspect-in-focus-parameter”. for any formula φ, (individual or sortal) variable α and term β, v(λαφ(β)) = 1 iff v(β)∈v(φ) for any formula φ, α is a term, then v(CGφ(α)) = 1 iff v(α)∈v(CGφ). for any formulas φ, ψ, and any variable α [‘1’ is to be treated as being true while ‘0’ as being false]:

68

B. Mou

v(φ→ψ) = 1 iff either v(φ) = 0 [being false] or v(ψ) = 1 v(〜φ) = 1 iff v(φ) = 0 v(∀αφ) = 1 iff for every v(α)∈D, v(φ) = 1 It is noted that, in view of the purpose here, the inference rules of this expanded account of predicate logic are not given; to this extent, the foregoing formal system has yet to be complete. As indicated before, the purpose to present the foregoing expanded predicate logic account is to capture, present and illustrate some of the addressed points made in this essay in a more accurate and effective way: the relative identity and semantic sensitivity as addressed in Gongsun Long’s and the Later Mohist accounts, the due relationship between the two basic semantic conceptions (reference and truth) as addressed in the last section and as resorted to in the semantic interpretation of a logical system, and the collective-name hypothesis regarding the semantic-syntactic structure of common nouns in natural languages. In so doing, the current standard predicate logic is thus refined and enhanced in some relevant connections. It is rendered especially relevant to the logic-concerned subject of a volume like this one.

References Ames, Roger and David Hall. 1987. Thinking Through Confucius. Albany: SUNY Press. Confucius (Kong Zi) 孔子. Analects (Lun-Yü) 《論語》 [the Chinese original text]. Chan, Wing-tsit 陳榮捷. 1963. A Source Book in Chinese Philosophy. Princeton: Princeton University Press. Cheng, Chung-ying 成中英. 1983. “Kung-sun Lung: White Horse and Other Issues.” Philosophy East and West 33.1: 341–354. Cheng, Chung-ying. 1987. “Logic and Language in Chinese Philosophy.” Journal of Chinese Philosophy 14: 285–307. Chmielewski, Janusz. 1962. “Notes on Early Chinese Logic (I).” Rocznik Orientalistyczny 26(1) (1962): 7–21. Creel, Herrlee. 1983. “On the Opening Words of the Lao-Tzu.” In Journal of Chinese Philosophy 10: 229-329. Deutsch, H. 2007. “Relative Identity.” Stanford Encyclopedia of Philosophy. http://plato.stanford. edu/entries/identity-relative/. Fraser, C. 2007. “Language and Ontology in Early Chinese Thought.” Philosophy East and West 57.4: 420–456. Fraser, C. 2013. “Mohist Canons.” Stanford Encyclopedia of Philosophy. http://plato.stanford.edu/ entries/mohist-canons/. Frege, Gottlob. 1892. “On Sense and Reference” (English translation of “Über Sinn and Bedeutung”). In P.  Geach and M.  Black, eds., 1960. Translations from the Philosophical Writings of Gottlob Frege (2nd edition): 56–78. Oxford: Blackwell. Reprint in Martinich, A. P. and Sosa, David, eds., 2013: The Philosophy of Language (sixth edition). New York and Oxford: Oxford University Press: 35-47. Fung, Yiu-ming 馮耀明. 2000. Gong-Sun-Long-Zi 公孫籠子 Taipei: Tung-tai Book Company. Fung, Yiu-ming. 2009. “The School of Names.” In Bo Mou ed., History of Chinese Philosophy ( 164–188). London/New York: Routledge. Fung, Yiu-ming. 2012. “A Logical Perspective on the Parallelism in Later Moism.” In Journal of Chinese Philosophy 39: 333-350.

3  Naming, Reference and Truth

69

Fung, Yu-lan. 1952–1953. A History of Chinese Philosophy, 2 vols. tr. Derk Bodde. Princeton: Princeton University Press. Gamut, L.T.F. 1991. Logic, Language, and Meaning (2 vols). Chicago/London: The University of Chicago Press. Geach, P. T. 1967. “Identity.” In Review of Metaphysics 21: 2–12. Gongsun, Long 公孫龍. Gong-Sun-Long-Zi 《公孫龍子》 [the Chinese original text]. Graham, A.C. 1967. “History of Logic.” In P.  Edwards, ed., The Encyclopedia of Philosophy, Volume IV. New York: Macmillian. Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought. New York: Oxford University Press. Harbsmeier, Christoph. 1991. “The Mass Noun Hypothesis and the Part-Whole Analysis of the White Horse Dialogue.” In Henry Rosemont, Jr., ed., Chinese Texts and Philosophical Contexts (49–66). La Salle: Open Court. Krifka, M. et al. 1995. Genericity: An Introduction. In G. Carlson, and F.J. Pelletier, eds. (1995). The Generic Book (1–124). Chicago: The University of Chicago Press. Kripke, Saul. 1980. Naming and Necessity. Oxford: Basil Blackwell. Lao Zi 老子Dao-De-Jing 《道德經》 [the Chinese original text]. The English translations of this text addressed in Section 4 include those in Chan 1963 and Lau 1963. Lau, D.C. (trans.). 1963. Lao Tzu: Tao Te Ching. Penguin Books. Locke, John. 1690. An Essay Concerning Human Understanding, Book III, Chs. i, ii and iii. Reprint in Martinich, A. P. and Sosa, David, eds., 2013: The Philosophy of Language (sixth edition). New York and Oxford: Oxford University Press: 656-660. Liu, Peiyu 劉培育. 2004. “Chinese Studies of Names and Argumentation” (中國名辯學). In Part I of Zhang, J. 張家龍 ed., An Intellectual History of Logic 逻辑学思想史 (1–158). Changsha: Hunan Education Press. McLeod, Alexus. 2011. “Pluralism about Truth in Early Chinese Philosophy: A Reflection on Wang Chong’s Approach.” In Comparative Philosophy 2.1: 38–60. http://www.comparativephilosophy.org. Mill, John Stuart. 1881. “Of Names.” In his System of Logic, Book 1, Ch.2, Section 5. Reprint in Martinich, A. P. and Sosa, David, eds., 2013: The Philosophy of Language (sixth edition). New York and Oxford: Oxford University Press: 29-34. Mo-Zi 《墨子》 (the Chinese original text). [The original text on which the relevant part of this essay is based: Wu, Y. 吳毓江, ed., 2006. The Mo-Zi Text with Annotations 墨子校注. Beijing: Zhong-Hua-Shu-Ju 中華書局.] Mou, Bo 牟博. 1999. “The Structure of Chinese Language and Ontological Insights: A CollectiveNoun Hypothesis.” Philosophy East and West 49.1: 45–62. Mou, Bo. 2000. “Ultimate Concerns and Language Engagement: A Re-Examination of the Opening Message of the Daodejing,” Journal of Chinese Philosophy 27.4: 429–439. Mou, Bo. 2003. “Eternal Dao, Constant Names, and Language Engagement.” (A substantial expansion and revision of Mou 2000). In Comparative Approaches to Chinese Philosophy (245–262). Aldershot: Ashgate Publishing Ltd. Mou, Bo. 2006. “Chinese Philosophy: Language and Logic.” In Donald M.  Borchert ed., Encyclopedia of Philosophy (second edition) (202–215). Chicago: Thomson-Gale/Macmillan Reference USA. Mou, Bo. 2007. “A Double-Reference Account: Gongsun Long’s ‘White-Horse-Not-Horse’ Thesis.” The Journal of Chinese Philosophy 34.4: 493–513. Mou, Bo. 2009. Substantive Perspectivism: An Essay on Philosophical Concern with Truth (“Synthese Library: Studies in Epistemology, Logic, Methodology, and Philosophy of Science” monograph series, vol. 344). Dordrecht: Springer. Mou, Bo. 2014. “On the Double-Reference Character of ‘Hexagram’ Names in the Yijing: Engaging Fregean & Kripkean Approaches to the Issue of How Reference is Possible.” Frontiers of Philosophy in China 9.4: 523–537.

70

B. Mou

Mou, Bo. 2015. “Rooted and Rootless Pluralist Approaches to Truth: Two Distinct Interpretations of Wang Chong’s Account.” Comparative Philosophy, 6.1: 149–168. http://www.comparativephilosophy.org. Mou, Bo. 2016. “How the Validity of the Parallel Inference is Possible: From the Ancient Mohist Diagnose to a Modern Logical Treatment of Its Semantic-Syntactic Structure.” History and Philosophy of Logic: 301–324. Mou, Bo. 2018. “A Double-Reference Account of Names in Early China.” In Bo Mou, ed., Philosophy of Language, Chinese Language, Chinese Philosophy: Constructive Engagement (69–169). Leiden: Brill Academic Publishers. Mou, Bo. 2019. Semantic-Truth Approaches in Chinese Philosophy: A Unifying Pluralist Account. Lanham: Lexington Books. Priest, Graham. 2008. An Introduction to Non-Classical Logic. Cambridge: Cambridge University Press. Roetz, Heiner. 1993, “Validity in Zhou Thought, On Chad Hansen and the Pragmatic Turn in Sinology,” In H.  Lenk and G.  Paul, eds., Epistemological Issues in Classical Chinese Philosophy (69–113). Albany: State University of New York Press. Russell, Betrand. 1905. “On Denoting.” Mind 14: 479–493. Reprint in Martinich and Sosa, eds., 2013: 105-113’ to ‘Martinich, A. P. and Sosa, David, eds., 2013: The Philosophy of Language (sixth edition). New York and Oxford: Oxford University Press: 105-113. Russell, Bertrand. 1957. “Mr. Strawson on Referring.” Mind 66: 385–389. Reprint in Martinich, A. P. and Sosa, David, eds., 2013: The Philosophy of Language (sixth edition). New York and Oxford: Oxford University Press 136–139. Sider, Theodore. 2010. Logic for Philosophy. Oxford: Oxford University Press. Strawson, P.F. 1950. “On Referring.” Mind 59: 320–344. Reprint in Martinich, A. P. and Sosa, David, eds., 2013: The Philosophy of Language (sixth edition). New York and Oxford: Oxford University Press: 121–135. Wang Chong王充. Lun-Heng 《論衡》 (the Chinese original text) [The original text on which the relevant part of this essay is based: Zhang Zongxiang 張宗祥, ed., 2013. Annotated Lun-Heng 論衡校注. Shanghai: Shanghai Gu-Ji Publishing House上海古籍出版社.] Wiggins, David. 2001. Sameness and Substance Renewed. Cambridge: Cambridge University Press. Xun, Zi 荀子 . Xun-Zi 《荀子》. Bo Mou (牟博),  is Professor of Philosophy at San Jose State University, USA. After receiving B.S. in math, he obtained M.A. in philosophy from Graduate School, Chinese Academy of Social Sciences, and Ph.D. in philosophy from University of Rochester, USA. He has published in philosophy of language and logic, metaphysics, Chinese and comparative philosophy, and ethics in such journals as Synthese, Metaphilosophy, History and Philosophy of Logic, Philosophy East and West, Asian Philosophy, Journal of Chinese Philosophy. He is the author of several books, including Substantive Perspectivism (“Synthese Library” vol. 344, 2009), Chinese Philosophy A-Z (2010), and Semantic-Truth Approaches in Chinese Philosophy: A Unifying Pluralist Account (2019). He is contributing editor of a number of volumes, including: Two Roads to Wisdom (2001), Davidson’s Philosophy and Chinese Philosophy (2006), Searle’s Philosophy and Chinese Philosophy (2008), History of Chinese Philosophy (2009), Constructive Engagement of Analytic and Continental Approaches in Philosophy (co-edited with Richard Tieszen, 2013), Philosophy of Language, Chinese Language, Chinese Philosophy (2018), and Chinese Philosophy in Routledge’s “Critical Concepts in Philosophy” reference book series (2018).

Chapter 4

Sentences (Ci 辭, Ju 句) and Propositional Attitudes Yiu-ming Fung

1  Grammatical Devices for Forming Meaningful Expressions It seems that no grammatical distinctions about parts of speech were explicitly made in classical Chinese. It seems also that it is impossible or difficult to form a sentence without grammatical distinctions about parts of speech. Nevertheless, in ancient Western languages, there were also no grammatical distinctions about parts of speech in a full-fledged sense. But, were these languages unable to make meaningful sentences? I think the grammatical theorization of parts of speech is not necessary for making sentences though, for example, in making a sentence about the relation between A and B, a language user must know in advance that which word or symbol is used to represent the relation and which words or symbols are used to represent A and B respectively. Although there was no systematic theory of grammar or syntax in ancient China, some philologists or interpreters of ancient classics in the Han 漢 dynasty or earlier did use grammatical devices to recognize or construct meaningful expressions, including full-fledged or condensed sentences. Based on these devices, they were able to identify or make simple sentences or complex sentences for expressing propositional attitudes, including expressing beliefs, desires, emotions, judgments and evaluation, etc., in ancient texts. Without using these devices, these scholars cannot understand any meaningful expression in ancient texts and also cannot communicate with each other. As demonstrated by some Chinese historical linguists, some Chinese grammatical distinctions, including the distinction between shici 實詞 (notional or substantive words, including nouns, verbs, adjectives and pronouns) and xuci 虛

Y.-m. Fung (*) Department of Philosophy, Tunghai University, Taichung, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_4

71

72

Y.-m. Fung

詞 (function or empty words, including prepositions, conjunctions and particles), were made in classical commentaries and philological studies in the time of Han dynasty or earlier, including the Mao Shi 毛詩 (Mao’s Commentary of the Book of Poetry), Chunqiu Guliang Zhuan 春秋穀梁傳 (Guliang’s Commentary of the Spring-­Autumn Annals), Chunqiu Gongyang Zhuan 春秋公 羊傳 (Gongyang’s Commentary of the Spring-Autumn Annals) and Shuo Wen Jie Zi 說文解字 (Explaining Characters and Analyzing Words).1 These two kinds of words or terms, i.e., shici and xuci, are usually used together to constitute sentences or to form judgments. In the Shuo Wen Jie Zi and Chunqiu Guliang Zhuan, some clear examples of function words, including those used as sentence markers, were mentioned in explaining their usage in ancient texts:2 矣:語已詞也。(說文解字矢部) (Refer to https://ctext.org/shuo-wen-jie-zi) Yi: a word [as a sentence marker] for speech ending. (Shuo Wen Jie Zi, Shi Section) 者:別事詞也。(說文解字白部) Zhe: a word [as a nominalization marker] for distinguishing an event [from others]. (Shuo Wen Jie Zi, Bai Section) 遂, 繼事之辭也。(春秋穀梁傳桓公八年) (Refer to https://ctext.org/guliang-zhuan) Sui: a term [as a conjunction marker] for connecting a following event. (Chunqiu Guliang Zhuan, Duke of Huan Year 8) 又, 有繼之辭也。其, 緩辭也。(春秋穀梁傳成公七年) You: a term [as a conjunction marker] for having what follows. Qi: a term of slow-tone [as a mood marker]. (Chunqiu Guliang Zhuan, Duke of Cheng Year 7) 而, 緩辭也。(春秋穀梁傳宣公八年) Er: a term of slow-tone [as a mood marker]. (Chunqiu Guliang Zhuan, Duke of Xuan Year 8) 乃, 急辭也。(春秋穀梁傳定公十五年) Nai: a term of urgent-tone [as a mood marker]. (Chunqiu Guliang Zhuan, Duke of Ding Year 15)

1  In this chapter, most of the quotations from ancient texts can be found in http://ctext.org of Chinese Text Project, edited by Dr. Donald Sturgeon, and the translations by James Legge can also be found in http://ctext.org. Besides, all other translations without acknowledgement are made by myself. 2  Guo Xiliang 郭錫良 has mentioned a case in the Huainanzi 淮南子 about the difference between two kinds of function words, the assertive use and the informative use of function words. That is: “To support with hand and to take with hand, to thank and to give way, to be old and advanced, to promise and to finish something, ye and yi, are separated with a thousand miles.” (Shuolin Xun 說 林訓) (Discourse on Forests) (扶之與提, 謝之與讓, 故之與先, 諾之與已, 也之與矣, 相去千 里。) Guo’s explanation is that: Although both “ye” 也 and “yi” 矣 belong to the mood of statement, the former is used to make assertive statement while the latter is used to make informative statement. See Guo 1989: 74.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

73

Zheng Xuan 鄭玄 in his Li Ji Zhu 禮記注 (Commentary of the Book of Rites) (Refer to https://ctext.org/wiki.pl?if=en&res=368394) even uses “yuzhu” 語助 (words of assistance/particles) to name function words or structural words:3 爾, 語助。 Er: a word of assistance [used as a final particle]. 毋從從爾, 爾毋扈扈爾。(禮記檀弓上) (Refer to https://ctext.org/liji) Example: Do not make it very high, nor very broad. (Li Ji, Tangong I: 21) (Translation: Legge 1885) 居讀為, 語之助也。 Ju read as ji: it is a word of assistance [to express a question or exclamation]. 孔子曰:「三日齋, 一日用之, 猶恐不敬, 二日伐鼓, 何居?」(禮記郊特牲) Example: Confucius said, “There are three days’ fasting on hand. If one fast for the first day, he should still be afraid of not being (sufficiently) reverent. What are we to think of it, if on the second day he beat his drums?” (Li Ji, Jiaotesheng 20) (Translation: Legge 1885) 思, 皆聲之助。 Si: all these are used as sound assistance [used as a final particle]. 詩曰:「神之格思, 不可度思, 矧可射思。」(禮記中庸) Example: It is said in the Book of Poetry, “The approaches of the spirits, you cannot surmise; and can you treat them with indifference?” (Li Ji, Zhongyong 16) (Translation: Legge 1885)

In addition to function words, both texts quoted below also mention various grammatical usages or functions of words which are similar to some different parts of speech in the modern Western linguistic classification. For example: 皆:俱詞也。(說文解字白部) Jie: a word of all [the objects or events mentioned]. (Shuo Wen Jie Zi, Bai Section) 各:異辭也。(說文解字口部) Ge: a term of each [object or event mentioned]. (Shuo Wen Jie Zi, Kou Section) 人者, 眾辭也。(春秋穀梁傳宣公十五年) Ren is a term of numerousness [as a classifier for counting numerous people]. (Chunqiu Guliang Zhuan, Duke of Xuan Year 15) 曰:詞也。(說文解字曰部) Yue: a word of speaking. (Shuo Wen Jie Zi, Yue Section) 既者盡也, 有 [又] 繼之辭也。(春秋穀梁傳桓公三年) Ji is a term for a finished state of an event; it is also used for connecting what follows. (Chunqiu Guliang Zhuan, Duke of Xuan Year 3)

In the Chunqiu Guliang Zhuan, there is a question about the usage of numeral term. That is: why the numeral “five” in “meteoric stones in Song is five” (隕石於 宋五) appears at the end of the sentence while the numeral “six” in “six water birds

 The following quotations translated by me are added with my explications in the square brackets.

3

74

Y.-m. Fung

fly across the capital of Song” (六鶂退飛過宋都) appears at the beginning of the sentence? The Explanation is: 後數, 散辭也。耳治也。 What is counting later is a term about scattered things. The counting is managed by ears. 先數, 聚辭也, 目治也。(春秋穀梁傳僖公十六年) What is counting in advance is a term about assembled things. The counting is managed by eyes. (Chunqiu Guliang Zhuan, Duke of Xi Year 16)

In other words, the numeral for the former sentence is used as a complement while that for the latter sentence is used as a modifier (Hu Qiguang 2005: 41–2). So, their syntactical functions are different. The most significant explanation of the distinction between adjectives or descriptions used as general terms and nouns or names used as referring terms is provided in the Yinwenzi 尹文子 in the following two passages. (Refer to https://ctext.org/ yin-wen-zi) The first one is: 語曰:「好牛」, 又曰:不可不察也。「好」則物之通稱, 「牛」則物之定形。以通稱 隨定形, 不可窮極者也。設復言「好馬」, 則復連於「馬」矣, 則「好」所通無方 也。設復言「好人」, 則彼屬於「人」矣, 則「好」非「人」, 「人」非「好」也。 則「好牛」、「好馬」、「好人」之名自離矣, 故曰:名分不可相亂也。(大道上) Suppose one says with the words “good ox” and also says that it cannot be used without the examination of the words. Here, “good” is a general description of things, “ox” is [a name] about a thing with definite form. To use general descriptions to accompany [names of] definite form, there are no limitations. Again, if one says “good horse,” then the word “good” is also used to join with “horse” and the use of “good” leads to somewhere without any constrain. Again, if one says “good man,” then that [name] is assigned to “man.” Thus, “good” is not “man,” “man” is not “good” and the names of “good ox,” “good horse” and “good man” are naturally separated [without confusion]. Therefore, we declare that the statuses of names cannot be confused. (Great Dao I)

The second passage is more interesting. It says that: 名稱者, 何 [別] 彼此而檢虛實者也。… 今親賢而疏不肖, 賞善而罰惡, 賢、不肖、 善、惡之名, 宜在彼, 親、疏、賞、罰之稱宜屬我。我之與彼, 各得一名, 名之察者 也。名賢不肖為親疏, 名善惡為賞罰。合彼我之一稱, 而不別之, 名之混者也。故曰 名稱者, 不可不察也。(大道上) Names and descriptions are used to distinguish what is that and this and to calibrate what is unreal/abstract and real/substantive.… Now suppose that we should keep close to the virtuous and keep distance from the unworthy people and that we should bestow the good man and punish the bad man. The names of “the virtuous” and “the unworthy people,” and “the good man” and “the bad man,” should be assigned based on those persons/things outside [i.e., objective facts] while the descriptions of “keeping close” and “keeping distance,” “bestowing” and “punishing,” should be assigned based on me [i.e., my judgments]. What is related to me and what is related to that thing/person outside have each name respectively. This is the examination of [the two kinds of] names/descriptions. So, to name/describe the virtuous and the unworthy people we use “keeping close” and “keeping distance” respectively, to name/ describe the good and the bad man we use “bestowing” or “punishing” respectively. To combine the name of a thing/event outside and the name imposed by me to constitute a description, but not to make distinction between them, is a confusion of name and description. Therefore, we declare that names and descriptions cannot be used without examination. (Great Dao I)4  According to this passage, the term “ming” 名 is used with both narrow and broad senses. In the narrow sense, it means the same as “name;” in the broad sense, it also includes the meaning of “description” or “being stated” In Chinese, the latter is called “cheng” 稱.

4

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

75

In these two passages, two things deserve our attention: First, tongcheng 通稱 (general description) is distinct from the name of dingxing 定形 (definite form). Second, in general, referring terms are related to things/persons in the external world while evaluative terms are given by the speaker. Based on these distinctions, we know that there are different kinds of words whose functions in compound terms and sentences are different. In other words, it is not the case that all words in classical Chinese are used for naming whose function is only for denoting or referring, as claimed by Chad Hansen (Hansen 1985: 516).5 Some words can be used as descriptions which are not necessarily used as referring terms and, most importantly, words can be joined together to form phrases and sentences.

2  Ci 辭 or Ju 句as Sentence or Utterance In the philological tradition from the Han dynasty, the study of judou 句讀 (pausing at the end of a phrase or sentence which is used as sentence, clause and phrase markers) is very popular. Besides, a series of distinctions among zi 字 (word/character), ju 句 (sentence/utterance), zhang 章 (passage), and pian 篇 (chapter) were clearly made by Wang Chong 王充 and other scholars in and after the Han dynasty. In his Lunheng 論衡 (Balanced Discourses or Discourse of Balance), Wang Chong writes: (Refer to https://ctext.org/lunheng/zh) 夫經之有篇也, 猶有章句也, 有章句猶有文字也。文字有意以立句, 句有數以連章, 章 有體以成篇。(正說) In addition to chapters in [Confucian] Canons, there are still passages and sentences. In addition to passages and sentences, there are still words. Words with meaning are used to make a sentence; sentences accumulated in some way are used to connect in a passage; and passages incorporated in a body are used to produce a chapter. (Cheng Shuo)

Later, Liu Xie 刘勰 in his Wenxin Diaolong 文心雕龍 (The Literary Mind and the Carving of Dragons) also describes this hierarchy of constitutive relation from words, sentences, passages to chapters as follows: (Refer to https://ctext.org/ wenxin-diaolong/zh) 夫人之立言, 因字而生句, 積句而為章, 積章而稱篇。(章句) People base on words/characters to make sentence/utterance; accumulate sentences/ utterances to construct a passage, and accumulate passages to state a chapter. This is how people establish their speech. (Zhang Ju)

Based on this philological background, it is not without evidence for Xu Fang 徐 防 to say that “the unfolding/discovery of passage and sentence starts from Zi Zia

5  It is well-known that Hansen’s “mass-noun hypothesis” together with his theses of “no concept of sentence-belief” and “no concept of semantic truth” in classical Chinese (Hansen 1983, 1985, 1992, 1993) are the most influential and controversial views in the academia and it seems that he never gives up these theses.

76

Y.-m. Fung

[a disciple of Confucius]” (發明章句始於子夏).6 It seems to me that the concept of ju mentioned by the thinkers in this tradition is similar to that of sentence or utterance. Moreover, I think the concept of a sentence may be first mentioned in the Zhuangzi 莊子. In the chapter of Pian Wu 駢拇 (Webbed Toes), the sophists are criticized as cunningly contriving their sentences (cuan ju 竄句) when they enjoy in the debate of jian bai 堅白 (hard and white) and tong yi 同異 (sameness and difference). (Zhuangzi II 1: 1) (Legge 1891. Refer to https://ctext.org/zhuangzi) Before the Han dynasty, ancient thinkers usually use the term “ci” rather than “ju” to name an utterance, either in terms of a sentence or clause. For example, “ci” is frequently used in the pre-Qin 秦 texts such as the Mozi 墨子, Xunzi 荀子, Yi Zhuan 易傳 (Commentary on the Book of Change), and Lushi Chunqiu 呂氏春秋 (Spring and Autumn Annals of Mr. Lu) as a meaningful linguistic unit: 夫辭以故生, 以理長, 以類行也者。(墨子大取) (refer to https://ctext.org/mohism) Ci 辭 (sentence/utterance/expression) is made on the basis of a relevant gu 故 (condition/reason), developed by virtue of an approprite li 理 (principle/idea) and held [true] in accordance with a related lei 類 (class/kind/category). (Mozi,  Daqu, i.e., Choosing the Greater) 以名舉實, 以辭抒意, 以說出故, 以類取, 以類予。(墨子小取) To use a ming 名 (name) to identify a shi 實 (object/event), to use a ci to express an yi意 (idea/meaning), to use a shuo 說 (explanation) to bring out a reason, to use a class/kind/ category to obtain a generalization, and to use a class/kind/category to cover specific cases. (Mozi, Xiaoqu, i.e., Choosing the Lesser) 侔也者, 比辭而俱行也。(墨子小取) Mou 侔 (parallelism) is to join two cis in parallel with each other such that both of them are held [true]. (Mozi, Xiaoqu) 辭也者。各指其所之。(易傳繫辭上) (Refer to https://ctext.org/book-of-changes) Each ci 辭 is used to signify what it aims at. (Yi Zhuan, Xi Ci I, Commentary on the Appended Words I) 名也者, 所以期累 [異] 實也。辭也者, 兼異實之名以論一意也。(荀子正名) (Refer to https://ctext.org/xunzi) Ming 名 (name) is used to discriminate objects. Ci (utterance) is used to combine names for different objects/events to explicate one sense. (Xunzi, Zhengming, Rectification of Names) 言者, 以喻意也。… 夫辭者, 意之表也。鑒其表而棄其意, 悖。(呂氏春秋審應覽) (Refer to https://ctext.org/lv-shi-chun-qiu) Yan 言 is used to express meaning. …Ci is the expression of idea/meaning. To examine the expression but to give up its idea/meaning is perversity. (Lushi Chunqiu, Shenying Lan, Examining Responses)

Here, my question related to these passages is that: is ci used as a meaningful unit a sentential one? I think it is. First, as indicated in the above quotations, ming and ci 6  See the biography of Xu Fang 徐防傳 in Biographies of Tang, Zhang, Xu, Zhang and Hu 鄧張徐 張胡列傳, recorded in Fan Ye’s 范曄, Hou Han Shu 後漢書 (the Book of Later Han). Refer to the Chinese Text Project on http://ctext.org.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

77

are assigned with different roles in the sense that the former is used to identify an object or event while the latter is used to express idea or meaning. Second, as mentioned in the Huainanzi, ci is stated for communication (陳辭通意). (Lanming Xun 覽冥訓, Explication of Peering into the Obscure) (Refer to https://ctext.org/huainanzi) So, I think, without sentential unit or its elliptical form performed as an utterance, it is impossible for any language used for communication. Third, as mentioned in the Xiaoqu, we can combine two cis to make a parallelism as an inference, such as “a white horse is a horse; [therefore] to ride a white horse is to ride a horse.” If cis can be used to make inference, they must be sentential units, either performed as utterances, judgments or propositions. In Chinese communities, most scholars in the field of Chinese philosophy have a consensus that there is the concept of a sentence or judgment in ancient China. They think that a sentence or judgment as a basic unit of language with complete meaning in terms of speech act was conceptualized in terms of “ci” by Later Mohists 墨家 and Xunzi. (Li Kuangwu 1989: 243–5, 301–3) A. C. Graham in his study of Later Mohism has a similar view. He says: (Graham 1978: 25) The distinction between word and sentence, term and proposition, is much more obvious in inflected languages than in a purely analytic language such as Chinese. In Chinese any verb (hsing [xing] 行 ‘go’, ta [da] 大 ‘be-big’) can stand by itself as a sentence (‘He goes’, ‘It is big’), and the particle yeh [ye] 也 which turns ma 馬 ‘horse’ into ma yeh ‘It is a horse’ is much less obtrusive than the conjugation verb ‘is’ with the subject ‘It’ which has to be added in English. Throughout the Canons and Explanations a sentence is assumed to be simply a name or string of names, and 謂之牛 “call it ‘ox’” is used where we should choose the wording “say it is an ox” (B 35); the difference was first appreciated by the author of Names and Objects. One might say that this distinction, which Western logic could take for granted from the beginning, was the last and most difficult of the Mohist discoveries.

In this passage, Graham seems to identify some singular words used in ancient Chinese texts as speech acts which can be called “one-word sentence” or “utterance” rather than “complete sentence” in grammatical sense. Besides, he also seems to treat “ci” as a technical term for sentence. In the following passage, he claims that Later Mohists made a distinction between name and sentence (Graham 1978: 40, 480): But its most important innovation is the introduction of the t’zu [ci] 辭 ‘sentence/proposition’, for the first time distinguished from the name. The distinction, grammatically less marked in Chinese than in Indo-European languages, seems to have attracted attention only after it was noticed that “knowing is different from having a pictorial idea” (NO 3). The name ‘pillar’ evokes a mental picture which might be interpreted either as a pillar or as a piece of wood, but knowing that it is a pillar is knowing that it is the former, not knowing that it is the latter.

So, he concludes that this is a discovery of Later Mohists (Graham 1978: 40): With this discovery the Mohist’s attention shifts to the similarities and differences, not between objects or names, but between the propositions by which we describe.

Nevertheless, in regard to the question whether there is the concept of a sentence in ancient Chinese, Hansen does not agree with Graham. Let’s see Hansen’s view in next section.

78

Y.-m. Fung

3  T  he Concept of a Sentence: Chad Hansen Versus A. C. Graham Hansen thinks that there is no concept of a sentence in classical Chinese; so he rejects Graham’s claim that Names and Objects of Later Mohism discovers the sentence. It seems that Roger Ames and David Hall (Hall and Ames 1987: 298–9 and 1995: 230) take side with Hansen7 while Christoph Harbsmeier (Harbsmeier 1998: 173–208) stands with Graham. In response to Hansen’s criticism, Graham argues that Hansen’s view is without textual evidence (Graham 1989: 394). In reply to Ames and Hall, Graham writes (Graham 1989: 395): It is a commonplace, repeated more than once in the present book, that Chinese philosophizing centres on the Way rather than on Truth. But we must insist in the first place that this difference from our own tradition has nothing to do with everyday questions of fact. Neither Western nor Chinese philosophy is concerned primarily with factual issues; to the extent that they are, as in Mohist argument for the existence of spirits, there is no significant difference from ourselves. If your wife announced ‘Dinner is on the table’, or the corresponding words in any language, you are disappointed if the dinner turns out to be still in the oven; a language without sentences in which it is impossible to affirm a fact would lack the communicative function without which it could not serve as a language. However much or little the words used to assent (Chinese jan [ran] 然 ‘so’, yu [you] 有 ‘there is’, hsin [xin] 信 ‘trustworthy’ and so forth) resemble or differ from English ‘true’, one assents if and only if dinner is on the table. Hansen has argued that even in the Mohist discussion of spirits the question is not whether it is true that they exist but whether it is socially appropriate in distinguishing the existing from the nonexistent to put the spirits on one side or the other. But the issue being purely factual (whether spirits exist, are conscious, appear before men’s eyes, reward the good and punish the wicked) with the second of the three Mohist tests being a direct appeal to reports of seeing and hearing spirits, what would this difference amount to?

7  Ames and Hall claim that: “The tendency of classical Chinese philosophers to be concerned with the ordering of names is a consequence of the dominance of the noun function. The striking claim that classical Chinese doesn’t depend upon sentences and propositions for the expression of semantic content entails the consequence that all Chinese words are names, and that compound terms, phrases, and sentences are strings of names. This consequence, in turn, requires that one appreciate the lack of interest on the part of the early Chinese in questions of ‘truth’ and ‘falsity’ Words, as names, may be judged appropriate or inappropriate; only propositions may, in the strict sense, be true or false” (Hall and Ames 1987:298–9). They also assert that: “Classical Chinese may be uncongenial to development of univocal propositions for this reason. Without such propositions, semantic notions of truth are ultimately untenable” (Hall and Ames 1995: 230). In contrast, Graham demonstrates that: “Hall and Ames seem in the quoted passage to imply that a Chinese sequence cannot be a sentence or a propositional utterance expressing a truth unless it has subjectpredicate form. On this Aristotelian principle one would have to agree that no truths were ever spoken in China, for even a verbal sentence with a subject is not conveniently analysed in subjectpredicate form” (Graham 1989: 394). So I think Ames and Hall take side with Hansen to hold the theses of “no concept of sentence-belief” and “no concept of semantic truth.” Here, thank you for one of the readers reminding me to provide evidence to demonstrate the disagreement about the view of sentences and truth in ancient China.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

79

If sentences are bearers of semantic truth, according to Graham’s view, it is impossible for ancient Chinese thinkers to express their truth-claims without sentences. But, in contrast to Graham, Hansen’s view is that: “The Mohists did indeed discover that word order was important in ci phrases. But that does not mean they discovered the sentence.” He also claims that, “Chinese theorists did not theorize that word order distinguished sentential strings from either small or larger language structures” (Hansen 1993: 45).8 It seems that his view is modest in the sense that Chinese thinkers did not have a well-defined concept of a sentence. But, actually, his view is quite radical. He believes that: “They [Chinese thinkers] defined ts’u [ci] 辭 merely [my italic] as a string of names” (Hansen 1983: 45–6). He even stresses that: “The absence of sentence function marking, the syntactic mobility of typical graphs, the mass-like grammar of nouns, and the use of predicate-only sentences contribute to viewing all words as having only a naming function [my italic] and to the failure to distinguish the sentence as a functional composite linguistic form” (Hansen 1985: 516). But, is this view reasonable? I think the reason why there is no well-defined concept of a sentence at a theoretical level in ancient China is not as Hansen’s claim about the motivation by grammatical features, but the fact that ancient Chinese thinkers are not interested in doing grammatical work, just like Plato does not have intention to be a linguist. To use language without grammatical theory does not mean using language without knowing grammatical distinctions in the language. I think it is impossible for any person who is able to express a thought of complete meaning does not have the concept of a sentence in mind. Because, without the concept of a sentence, one is unable to identify and use an appropriate expression to express their thought and what they want to express would not be interpretable or understandable by others. Speech acts are the basic or minimal units of linguistic communication.9 Every speech act is constituted of a sentence which expresses a propositional content and utters with a force. It is impossible for people to express their thought without knowing how to identify and use this basic linguistic unit in their language. Without the concept of a sentence at an operative level, whether it is about a complete sentence, sentence of one word or elliptic sentence, we cannot express any understandable thought. Besides, Hansen even claims very boldly that, “We find no concepts of beliefs, concepts, ideas, thoughts, meanings, or truth. … They [the Mohist thinkers] still do not focus on sentences but on names or longer strings of names (usually character pairs). The deep concern continues to be guidance, not description” (Hansen 1992: 235).10 However, as a matter of fact, there is no evidence to prove that ancient 8  Here, Hansen seems to treat ci as strings of words including sentences though he translates ci as phrases. 9  As pointed out by John Searle, a basic communication unit is a speech act of the sentence form and most elementary illocutionary acts are of the form F(P): they consist of an illocutionary force F with a propositional content P. (Searle 1969: 30, 2010: 66) 10  Hansen does not only claim that there is no concept of belief in ancient Chinese philosophy or theories of language, but also claims that there is no linguistic device or construction in ancient

80

Y.-m. Fung

Chinese thinkers treat ci as merely a string of names which has only a naming or denoting function. As we know, Later Mohists explicitly distinguish ming (name) and ci (utterance of sentence/clause) in the sense that the former is used (as a referring term) to refer to or (as a predicating term) to pick out shi 實 (object/thing/event) (or to fix the reference) while the latter is used to express yi 意 (idea/meaning/intention) for communication. In ancient texts, including Later Mohists’, ancient Chinese thinkers never use or mention structural or function words including grammatical particles as ming. But most of the examples of ci they use are strings of words which are composed with function words. So, ci is definitely not conceptualized by Later Mohists as merely a string of names. Even only one word appears in some passage, we can know from the context that it may be an utterance of an elliptic form. In the Mencius 孟子, the sentence, “After confirming with facial expression and expressing with sound, the meaning would be understood.” (Mencius, 12: 35) (徵於色, 發 於聲, 而後喻。) (告子下), means that the unit of language whose semantic content can be understood is an utterance in some context, including non-verbal one. As a matter of fact, we cannot find any evidence from ancient Chinese texts and Hansen has never provided any example to show that there is any meaningful string of words which is not sentential or not understood as in the context of elliptical form of sentence. He has also never provided any evidence to show that ancient Chinese thinkers were influenced by the mass-like grammatical features to view all words as having only a naming function. According to Davidson, concepts cannot be identified without propositions. If ancient Chinese philosophers do not distinguish, at least implicitly, between sentences and non-sentential expressions, they would not know any concept except that they unconsciously use sentential propositions to identify concepts. Davidson is right to say that, for example, “There is no fixed list of things you have to know about, or associate with, felinity; but unless you have a lot of beliefs about what a cat is, you don’t have the concept of a cat” (Davidson 2002: 124). In other words, to know the concept of an object, one has to know some other things about the object, including knowing the wrong application of the concept to another object which is distinct from the object. Similarly, it is impossible for ancient Chinese thinkers to make any belief and distinction based on Hansen’s so-called “term-beliefs” (Hansen 1993). They have to put their concepts in the sentential context. Hansen has complained that, “some critics of my position have confused my theory about theory of language with my theory of the language. They construe me as saying Chinese has no sentences or no abstract terms. My argument, however, was a textual one. I argued that the Chinese theories of Chinese language did not postulate sentences or abstract objects” (Hansen 1993: 46). He agrees that, “There Chinese to express clauses of “believes that.” So he stresses that: “No single character or conventional string of ancient Chinese corresponds in a straightforward way to ‘believes that’ or ‘belief that.’ No string or structure is equivalent to the word ‘believe’ or ‘belief’ in the formal sense that it takes sentences or propositions as its object” (Hansen 1985: 501). But I think this view is not right. As I shall argue below, there are some devices in classical Chinese which are functionally equivalent to that-clauses and similar to those devices in Old English.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

81

are Chinese sentences,” but, on the other hand, he asserts that, “ancient Chinese philosophical writers did not theoretically distinguish between sentences and other meaningful strings of characters” (Hansen 1985: 496). However, if there are sentences in classical Chinese but ancient Chinese philosophers did not theorize the concept of a sentence or did not theoretically make a distinction between sentences and non-sentential expressions, it doesn’t imply that they did not know how to identify a sentence and how to distinguish between sentences and non-sentential expressions. It also doesn’t imply that they treated ci only as a string of names and thus use all cis with only naming function. It only implies that they were not interested in making grammatical distinctions theoretically or they did not want to do the job of grammarians’. Metaphorically speaking, can we argue that: since John, an English school boy, doesn’t have the concept of a sentence in theory, he doesn’t know how to make or pick out a sentence to express his thought and cannot have a sentential belief? Hansen’s main point is not only that ancient Chinese philosophers did not do the theorizing work. Actually, he asserts that they did not know how to identify a sentence and thus cannot know how to distinguish sentences from non-sentential expressions. So he stresses that (Hansen 1985: 497): I am not making a claim about what cannot be thought, but about what was not written. My reference to features of language are [sic] intended to explain why certain terms and distinctions were not used in doctrines. … If the reasons Western philosophers give for making truth claims involve features of Western languages which Chinese lacks, then the counterpart features of Chinese explain why Chinese philosophers would not use the concept in their reasoning. This kind of explanation does not entail that language constrains theorizing. … If one were constructing philosophical theories in Chinese, one could more easily “get away with” a theory that makes no distinction between sentences and other strings of words and makes no distinctions among “parts of speech.”

Hansen’s argument mentioned above is based on language features; it seems like an argument as follows: “Since an underdeveloped country does not have the idea of the Western credit card, this country could more easily ‘get away with’ an economic system without non-cash transaction.” Metaphorically speaking, it seems that Hansen does not agree that there may be some functional equivalent of credit card in non-cash transaction for that country. However, Hansen seems not aware of the contributions made by some scholars of historical syntax in the field of Chinese linguistics, so he persistently claims that (Hansen 1983: 45–6): [S]ince the subject term of Chinese sentences is frequently not expressed in literary Chinese, subject-predicate sentences like ma pai (yeh) [ma bai ye 馬白也] “(it is the case that) the horse-stuff is white” and open sentences pai ma yeh [bai ma ye 白馬也] “x is white-horse” seem to differ only in the order of the graphs. Thus, as we shall see in discussing the Neo-­ Mohist semantics, Chinese semanticists had no clear concept of a sentence as a grammatical structure. They defined ts’u [ci] 辭 merely [my italic] as a string of names. That the order of graphs made a difference was clear, but their semantics included no theory of what functional elements were involved in sentences nor any use of the notion of semantic truth—the semantic correlate of sentences. Chinese semanticists simply lumped adjectives and nouns together as expressions denoting a segment of reality. [my italic] Thus, they

82

Y.-m. Fung regarded all graphs, the simple expressions, equally as ming 名 (names) and they regarded ts’u (complex expressions—phrases or sentences) as no more than [my italic] combinations of names used to express or convey speakers’ i [意 yi] (intentions). Some ming 名 (names) referred to corporeal stuffs like water, wood, man, grain, while others referred to stuffs which could “interpenetrate,” like white, hard, wisdom, and benevolence.

I think, however, this view is not grounded in evidence and is also counter to common sense. As demonstrated above, ci is not regarded by the ancient Chinese thinkers as no more than combinations of names. Since almost all cis or yans 言 (expression/speech) as units to express thought and all function words or structural words are not used for naming, referring or denoting, it is impossible for the ancient Chinese thinkers to regard each of them as a component of a string of names and they only have naming function. In addition, Hansen does not provide any clear example to show that there is no distinction between sentences and non-sentential expressions in ancient texts. The only example to illustrate this point given by him seems to me incorrect. He provides a strange example of the so-called grammatical string of words as in a topic-comment context, i.e., bai ma ye 白馬也. But we cannot find this or similar string of words in Chinese data-base, whether it is bai ma ye or ma bai ye, used independently, or used in isolation of context, as a complete meaning unit in any ancient text. We can find complete sentences in Gongsun Long’s Baima Lun白馬論 (Discourse on White Horse) such as: “But it is not acceptable to respond that there is bai ma ye” (然不可以應有白馬也) and “ma yu [and] bai ma ye” (馬與白, 馬也[耶]?). In the Hanfeizi 韓非子, we can also find an independent sentence such as: “What was it that just went out of the door? Is it a white horse?” (走出門者何?白馬也[耶]?) (Liao 2015: 864–5) Nevertheless, we cannot find any string of names which is used as a unit of complete meaning but not an utterance of sentence. Besides, I think we should not treat elliptical sentences as non-sentential expressions. One of the reasons why so many sentences in classical Chinese are elliptical and have more gapping than those in modern English is that it is too expensive or not easy to write Chinese with the book form of bamboo or silk. Even though Hui Shi 惠施 was described as a knowledgeable learner with full of five carts of books, what he had is nothing but a small capacity in comparison with what we can save in a tiny USB. To make sentence elliptical or gapping are normal use of language even now.11 It cannot be accused of inability in identifying or distinguishing the sentence from the non-sentential expressions. Chinese historical linguists have done a lot of good job to explain different constraints for compounding words, phrases and sentences respectively. It is a shame that Hansen’s example is not clear enough to demonstrate or support his peculiar view. Moreover, if, as claimed by Hansen, the so-called “Chinese semanticists” simply lumped adjectives and nouns together as expressions denoting a segment of reality but could not make sentences to do the work other than denoting or referring and thus could not make sentential beliefs, there would be no communication between them and their readers. As argued by John Searle, he writes (Searle 1967: 121):

11

 A detailed discussion on gapping phenomena can be found in Ross 1967.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

83

Now, if speaking a language is a form of behavior and if it is necessary to so regard it in order to understand communication, certain questions naturally arise; first, which of these behavioral units are the minimal complete units does this behavior naturally segment? The obvious answer to the first question is: statements, commands, promises, remarks, explanations, orders, comments, requests, and things like them. Austin baptized these with the name “illocutionary acts.”

According to Searle, it is impossible that a language which can be used for communication only has the function of referring or denoting. Referring cannot be formed as a complete speech act. So, he says (Searle 1967: 122): They need to be distinguished from acts like referring, as, for example, when I refer to John in saying “John came yesterday.” Here the reference to John is not by itself a complete speech act capable of standing on its own but is only part of the larger complete illocutionary act of making the statement that John came yesterday. The syntactical correlate of this point is that sentences and not words are units of the complete speech act. Words occur intelligibly only in sentences. (They may be one word sentences such as “look!”, but even here we need to distinguish the word from the imperative intonation contour of its utterance, the morpheme of stress, and the sentence boundary that surrounds the word. All of these are parts of the sentence, and the word without them is not the sentence.)

4  An Operative Concept of a Sentence In contrast to Hansen’s view, Harbsmeier thinks that there is an operative concept of a sentence in classical Chinese though both classical Greek and Chinese are vague in making distinction between sentences and non-sentential expressions (Harbsmeier 1998: 173–5). In addition to Harbsmeier’s examples, more examples in ancient texts demonstrate that there is an implicit or operative concept of a sentence used by ancient Chinese philosophers.12 The terms “yan” 言 (speech [act], sentence or utterance) and “shuo” 說 (explanation, discourse or theory) mentioned by Confucius 孔 子 and other ancient thinkers are all about expressions of complete meaning, either simple, compound or complex sentences, not non-sentential phrases which can only be used to express incomplete or indeterminate idea if the uses are out of context. In other words, as indicated in ancient texts, only complete (or elliptic) sentences satisfy the condition of expressing thought. For Confucius, no names cannot be used in a meaningful expression to express thought. So, he says: “Therefore a superior man considers it necessary that the names he uses may be spoken appropriately, and also that what he speaks may be carried out appropriately. What the superior man requires is just that in his words there may be nothing incorrect.” (Analects, 13: 3) (Translation: Legge 1861a) (故君子名之必可言也, 言之必可行也。君子於其言,

 As indicated by Frank Palmer, the traditional grammars do not successfully provide a kind of definition for “sentence.” Instead, “we are taught at school to recognize sentences through practice, not by a set of rules” (Palmer 1971: 71). It means that what we learn to identify a sentence is an operative concept of a sentence.

12

84

Y.-m. Fung

無所苟而已矣。) (論語子路). This passage demonstrates that he has a clear distinction between ming (name used as a referring term) and yan (speech [act] used as an utterance or meaningful expression) in the sense that what can be named can be expressed in a speech act which, in turn, can be carried out in behavior. Moreover, what Confucius regards as yi yan一言 (one expression or speech act) and liu yan 六言 (six expressions or speech acts) indicates a kind of “definition by demonstration” or “ostensive definition” of a sentence. For example: 子曰:「詩三百, 一言以蔽之, 曰『思無邪』。」(論語為政) The Master said, “In the Book of Poetry are three hundred pieces, but the design of them all may be embraced in one sentence—‘It is to think without depravity’.” (Analects, 2: 2) (Translation: Legge 1861a) 定公問:「一言而可以興邦, 有諸?」孔子對曰:「言不可以若是其幾也。人之言曰:『 為君難, 為臣不易。』如知為君之難也, 不幾乎一言而興邦乎?」曰:「一言而喪邦, 有 諸?」孔子對曰:「言不可以若是其幾也。人之言曰:『予無樂乎為君, 唯其言而莫予 違也。』如其善而莫之違也, 不亦善乎?如不善而莫之違也, 不幾乎一言而喪邦 乎?」(論語子路) The Duke Ding asked whether there was a single sentence which could make a country prosperous. Confucius replied, “Such an effect cannot be expected from one sentence. There is a saying, however, which people have—‘To be a prince is difficult; to be a minister is not easy.’ If a ruler knows this—the difficulty of being a prince—may there not be expected from this one sentence the prosperity of his country?” The duke then said, “Is there a single sentence which can ruin a country?” Confucius replied, “Such an effect as that cannot be expected from one sentence. There is, however, the saying which people have—‘I have no pleasure in being a prince, but only in that no one can offer any opposition to what I say!’ If a ruler’s words be good, is it not also good that no one oppose them? But if they are not good, and no one opposes them, may there not be expected from this one sentence the ruin of his country?” (Analects, 13: 15) (Translation: Legge 1861a) 子曰:「由也, 女聞六言六蔽矣乎?」對曰:「未也。」「居!吾語女。好仁不好學, 其 蔽也愚; 好知不好學, 其蔽也蕩; 好信不好學, 其蔽也賊; 好直不好學, 其蔽也絞; 好勇 不好學, 其蔽也亂; 好剛不好學, 其蔽也狂。」(論語陽貨) The Master said, “You, have you heard the six sentences to which are attached six becloudings?” You replied, “I have not.” “Sit down, and I will tell them to you. There is the love of being benevolent without the love of learning—the beclouding here leads to a foolish simplicity. There is the love of knowing without the love of learning—the beclouding here leads to dissipation of mind. There is the love of being sincere without the love of learning—the beclouding here leads to an injurious disregard of consequences. There is the love of straightforwardness without the love of learning—the beclouding here leads to rudeness. There is the love of boldness without the love of learning—the beclouding here leads to insubordination. There is the love of firmness without the love of learning—the beclouding here leads to extravagant conduct.” (Analects, 17: 8) (Translation: Legge 1861a)

In Maoshi Zhengyi 毛詩正義 (Explanatory Notes to Mao’s Commentary of the Book of Poetry), According to Kong Yingda 孔穎達, yi yan and liu yan in the Analects are interpreted as concrete sentences. He says: (Refer to https://ctext.org/ wiki.pl?if=en&res=871901) 句則古者謂之為言, 論語云:「詩三百, 一言以蔽之, 曰『思無邪』。」 則以「思無 邪」一句為一言。左氏曰:臣之業在「揚之水」卒章之四言。謂第四句不敢告人 也。及趙簡子稱子大叔遺我以九言, 皆以一句為一言也。

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

85

In regard to ju [sentence], it was called yan in the ancient time. In the Analects, it says that: “In the Book of Poetry are three hundred pieces, but the design of them all may be embraced in one yan—‘It is to think without depravity’.” It is to use this one ju [sentence]— “It is to think without depravity”—as one yan. Mr. Zuo [in his Chunqiu Zuo Zhuan 春秋左 傳 Zuo’s Commentary of the Spring-Autumn Annals] says that: “What I as a subject to serve is the four yan in the final passage of ‘the fretted waters’.” It means that the speaker does not dare to tell other people the fourth ju [sentence]. Moreover, Zhaojianzi mentioned that Zidoshu left after his death with nine yan. In all these examples, one ju [sentence] is treated as one yan.

I think these examples are clear enough to demonstrate that, without the concept of a sentence, it is impossible to treat yi yan 一言 (one speech act), si yan 四言 (four speech acts) and jiu yan 九言 (nine speech acts) as meaningful expressions to express thoughts. It means that yan used or identified in the Analects presupposes an operative concept of a sentence or utterance. Without the concept, Confucius cannot identify the expression that, “To be a ruler/prince is difficult” and the expression that “only in that no one can offer any opposition to what I say” as “one sentence,” respectively. “It is to think without depravity” is also used by Confucius to express his sentential belief about the real message in the Shi Jing 詩經 (Book of Poetry). What he talks about the six sentences are all of propositional content about the importance of learning and the bad consequence in behavior if without the love of learning. In comparison with the examples used by Confucius which can be regarded as definition by demonstration for the concept of a sentence as mentioned above, Xunzi is much more conscious of the distinction between ming (name) and ci 辭 (sentence, utterance or speech act) in the chapter of Zhengming 正名 (an example quoted in section two earlier). But Hansen disagrees with the view that Xunzi does make such a distinction and does regard ci as sentence. He still thinks that ci is understood by ancient thinkers as nothing but a string of names which only has the function of naming. In contrast, Heiner Roetz does not agree with Hansen that ci in the Xunzi is merely a string of names. He says: (Roetz 1993: 81) This passage poses several problems of interpretation, especially concerning the terms, which, after all, constitute the main difficulty in translating Chinese philosophical texts. Hansen only refers to it implicitly. Eager to discuss away any notion of sentences, he maintains that Hsun Tzu [Xunzi] “understood tz’u [ci] merely as strings of names” (T 496), and that for him “compounding (chien) [jian 兼] simply alters (in different ways) the conventional scope of names over a single variegated reality” (L 190 n. 26). This, however, would neither explain the clear hierarchical progression from ming via tz’u to pien [bian 辯] and shuo, nor the difference between ming and tz’u consisting in the incapability and capability respectively of expressing i (ideas) [yi 意]. Tz’u, lying between simple terms and complex disputations or explanations, can hardly be thought of as merely a “name” (with a broader scope) again. It is best understood as “clause” in the sense of a sentence. This is supported by its basic meaning “to make many words.”

I think this hierarchical progression from small linguistic unit to larger unit does indicate a clear distinction between a unit used to refer to an object and a unit used to express an idea. As explained earlier, the contextual criterion for identifying expressions as one sentence, six sentences or nine sentences can be understood as a

86

Y.-m. Fung

definition by demonstration which reflects that the speaker does have the concept of a sentence. Moreover, a final particle of a series of words plays the role of sentence marker which does not appear in a mere string of names. So, I think Later Mohists and Xunzi did use the concept of ci (speech act or utterance of sentence or clause) as a linguistic unit to express an idea of complete meaning and what they mentioned the concrete examples of ci are unexceptionally not used to express incomplete thought. In this sense, they not only used the concept of a sentence implicitly, but also mentioned the concept at a theoretical level. Another indicator of having the concept of a sentence is that, without understanding what a sentence is, the ancient Chinese thinkers, including Gongsun Long, Mozi and Later Mohists, cannot make any argument. That is, to make an argument, one has to know how to identify a sentence or a series of sentences in order to express the premises and conclusion of an argument. Without the capacity to identify a sentence or meaningful expression, one cannot form any argument; without the capacity to make an argument, one cannot dispute or debate with others.

5  That-Clauses and Propositional Attitudes As we know, one of the indicators of having sentential beliefs or propositional attitudes is the device of that-clause. I think it is obvious that ancient Chinese thinkers do express their thought with sentential beliefs or propositional attitudes. In contrast to Hansen’s view that there are no sentential beliefs or that-beliefs in ancient China, most Chinese grammarians assert that there are linguistic devices in classical Chinese which are functionally equivalent to that-clause or belief-clause in Western language. Moreover, if we agree with Davidson’s paratactic view on that-clause and indirect discourse, we can also agree that there are also that-clauses (in the sense of functional equivalent) in classical Chinese. According to Fang Wan-chuan’s 方萬全 explanation, which is based on Davidson’s paratactic analysis, there are clear cases of using expression of paratactic structure in classical Chinese which is functionally equivalent to that-clause (Fang 2006: 97–8). Davidson thinks that, for example, the sentence, “Galileo said that the earth moves,” can be paraphrased as including two parts: The earth moves. Galileo said that.

According to the Oxford English Dictionary, Davidson quotes (Davidson 2001: 106): The use of that is generally held to have arisen out of the demonstrative pronoun pointing to the clause which it introduces. Cf. (1) He once lived here: we all know that; (2) That (now this) we all know: he once lived here; (3) We all know that (or this): he once lived here; (4) We all know that he once lived here.

Based on this linguistic fact, he concludes that, “The proposal then is this: sentences in indirect discourse, as it happens, wear their logical form on their sleeves (except for one small point). They consist of an expression referring to a speaker,

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

87

the two-­place predicate ‘said’, and a demonstrative referring to an utterance, period” (Davidson 2001: 106).13 In classical Chinese, we can find a lot of examples which fit into Davidson’s analysis. One of the examples mentioned by Fang Wan-chuan is a clear case of believing-that sentence in the Mencius, that is: 如以辭而已矣, 雲漢之詩曰:「周餘黎民, 靡有孑遺。」信斯言也, 是周無遺民 也。(孟子萬章上) If we simply take single sentences, there is that in the ode called “The Milky Way”: “Of the black-haired people of the remnant of Zhou, There is not half a one left.” If we believed that utterance, then not an individual of the people of Zhou was left. (Mencius, 9: 4) (Translation: Legge 1861b, with my minor modification)

It is obvious that the demonstrative (pronoun) “si” 斯 (that/this) in the antecedent clause “we believed that utterance” of the above conditional (sentence) refers back to the quoted sentence. So, this referred sentence can be regarded as the object of the belief. In addition to Fang’s example about direct speech, we can find many examples of indirect speech, which are also embedded with paratactic structure, in ancient texts. For example: 仲尼聞是語也, 曰:「以是觀之, 人謂子產不仁, 吾不信也。」(左傳襄公三十一年) When Chung-ne [Zhong Ni, a name of Confucius] heard of these words, he said, “Looking at the matter from this, when men say that Tsze-ch’an [Zichan] was not benevolent, I do not believe it.” (Zuo Zhuan, Duke of Xiang: Year 31) (Legge 1991: 566) 子曰:「事君三違而不出竟, 則利祿也; 人雖曰不要, 吾弗信也。」(禮記表記) The Master said, “In the service of a ruler, if (an officer), after thrice leaving the court (on his advice being rejected), do not cross the borders (of the state), he is remaining for the sake of the profit and emolument. Although men say that he is not trying to force (his ruler), I will not believe that.” (Li ji, Biao Ji: 39) (Translation: Legge 1885, with my minor modification) 龐恭與太子質於邯鄲, 謂魏王曰:「今一人言市有虎, 王信之乎?」曰:「不信。」「二 人言市有虎, 王信之乎?」曰:「不信。」「三人言市有虎, 王信之乎?」王曰:「寡人 信之。」(韓非子內儲說上) When Panggong together with the Crown Prince was going to Handan as hostage, he said to the King of Wei: “Now, suppose one person says that there is a tiger in the market, will Your Majesty believe it?” [The King] said: “I will not believe it.” [Pang said:] “If two persons say that there is a tiger in the market, will Your Majesty believe it?” [The King] said: “I will not believe it.” [Pang said:] “If three persons say that there is a tiger in the market, will Your Majesty believe it?” The King said: “I [as the sovereign] will believe it.” (Hanfeizi, Inner Congeries of Sayings, Upper Series) (Liao 2015: 768–9)

All the above examples have the form: S, I believe/do not believe t.

 Davidson’s paratactic analysis is not only theoretically appropriate, but also in accordance with the development of the Present English from the Old English and Middle English. As indicated by many historical linguists of English, the complementizer (or subordinating conjunction) “that” is derived from the demonstrative (pronoun) “pæt” through a process of grammaticalization. See Williams 1975: 248. In Chinese language, there is also a similar kind of grammaticalization or xu hua 虛化 (de-substantiation).

13

88

Y.-m. Fung

or S, I believe/do not believe zhi 之 (that/this).

Here, S stands for a sentence. We use t to indicate the trace in a sentence.14 The second form uses “zhi” as a demonstrative pronoun, while the first form uses a zero form of “zhi” (that/this)15 “Zhi” in the sentences quoted above is used as an anaphora16 to refer to a relevant sentence as the object of a relevant belief.17 As found in the data-base of ancient texts, there are some interesting examples which show that a sentence of paratactic structure and a sentence with a that-clause with zero form of that are functionally equivalent. Let’s see the following examples: 不失其身而能事其親者, 吾聞之矣。(孟子离婁上) That those who do not fail to keep themselves are able to serve their parents: I have heard that.” (Mencius, 7: 19) 吾聞之也:君子不以其所以養人者害人。(孟子梁惠王下) I have heard that: a ruler does not injure his people with that wherewith he nourishes them. (Mencius, 2: 22)

Here, the word “zhi” 之 (that/this) is used as a demonstrative, either with the function of anaphora or cataphora, to refer to the sentence, either the previous nominalized sentence or the following one. I think the above sentences of paratactic construction can be regarded as functionally equivalent to the following sentences of indirect speech with the empty category (Ø) of “zhi” or “that”: 吾聞西伯善養老者。(孟子离婁上/盡心上) I have heard Ø the chief of the West knows well how to nourish the old. (Mencius, 7: 13 and 14: 22) 吾聞其以堯舜之道要湯, 未聞以割烹也。(孟子萬章上) I have heard Ø he [Yi Yin] sought an introduction to Tang by the doctrines of Yao and Shun, [but] not heard Ø he did so by his knowledge of cookery. (Mencius, 9: 7)

 In Chomskian linguistic, a trace is what is left behind in a position in which the moved constituent appeared at some deep level of syntax or some point in the derivation before surface syntax. For example, in saying “you like drinking tea,” there is a constituent appeared in the object position of the surface structure. But, in saying “what do you like eating t?” there is a moved constituent appeared in the deep structure which can be marked by t. 15  A zero subordinate conjunction (or complementizer) occurs in English in sentences like “I know Ø he likes swimming,” in which the zero conjunction plays the role of the subordinate conjunction that in its non-zero from: “I know that he likes swimming.” 16  The anaphoric (referring) term is called an anaphor. For example, in the sentence: “the music stopped and that upset everyone,” the demonstrative pronoun that is an anaphor which refers back to its antecedent sentence. 17  In addition to this kind of grammaticalization from demonstrative pronoun to subordinate complementizer (i.e., from substantive word for anaphoric or cataphoric reference to function word for subordinate conjunction), there is another kind of grammaticalization from demonstrative pronoun to conditional complementizer (i.e., from substantive word for anaphoric or cataphoric reference to function word for conditional conjunction). A detailed argument with concrete examples can be found in Hung 2010: 317–8. 14

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

89

吾聞秦楚構兵。(孟子告子下) I have heard Ø Qin and Chu are fighting together. (Mencius, 12: 24)

To compare these two groups of passages, it is obvious that there is a paratactic structure in each passage of the first group with the form: “I have known/heard zhi: S zhe.” (or “I have known/heard zhi: S [zhe].”) or “S zhe: I have known/heard zhi.” (or “S [zhe]: I have known/heard zhi.”) On the other hand, each passage of the second group has a different form from that of the first group. That is: “I have known/ heard Ø S.” or “Ø S I have known/heard.” Here, in the first group of passages, “zhi” appearing in the compound sentence of paratactic structure is used as a demonstrative to refer to a relevant sentence while, in the second group of passages, “zhi” as empty category (Ø) is implicitly used as a complementizer in an indirect speech. The word “zhi” (that/this) in the first group of sentences is used to point out the previous sentence or the following sentence. In the second group of sentences, there is an ellipsis of “zhi” just like the zero form of that in a that-clause in English. Since the sentence form of the first group is functionally equivalent to the sentence form of the second group, both of them can be transformed into each other without changing their expressive content. So, the second sentence form “I have known/heard [zero form of zhi/that] S” can be transformed into the first sentence form “I have known/heard zhi/that: S.” In other words, “I have known/heard [zhi/that] zero form of complementizer the chief of the West knows well how to nourish the old.” is functionally equivalent to “[The chief of the West knows well how to nourish the old]S zhe者nominalizer: I have known/heard zhi demonstrative.” (or “I have known/heard zhi demonstrative: [The chief of the West knows well how to nourish the old]S zhe nominalizer.”) Here, the word “zhi” (that/this) in the latter sentence is used to demonstrate the nominalized sentence or used to refer back to the antecedent sentence. In the former sentence, there is an ellipsis of “zhi.” Therefore, the sentence in the Mencius with zero form of “zhi” mentioned above can be understood as, through a process grammaticalization, transformed from the sentence which is embedded with the paratactic zhi-construction (similar to the paratactic that-construction). That is, the structure of the first sentence “I have known/heard [zero form of zhi/that] S” can be understood as transformed from the structure of the second sentence “S: I have known/heard zhi/that.” (or “I have known/heard zhi/that: S.”) Based on this analysis, “I have known/heard that: the chief of the West knows well how to nourish the old.” (吾聞 之也:西伯善養老者。) is not only grammatically acceptable, but also of the same meaning of the original sentence “I have known/heard Ø the chief of the West knows well how to nourish the old.” (吾聞Ø西伯善養老者。) in the Mencius. In comparison with English, it seems that, unlike “zhi,” the grammaticalization of “that” as complementizer cannot preserve the function of demonstrative. But it is not totally right. For example, when A says that “Does Peter know that complementizer Mary is a philosopher?” and B replies that “Peter knows that [demonstrative].”, the second token of “that” is clearly used as a demonstrative pronoun though it may also be understood as a complementizer in a skipped form of the sentence “Peter knows that [complementizer] Mary is a philosopher.” Moreover, the expression “that I can do” in the subject place of the matrix sentence “That I can do is a minimal work.” is a

90

Y.-m. Fung

subordinate clause in which “that” plays a role of a nominalizing marker to make the sentence nominalized. This is just like the work “zhe” 者 in classical Chinese. Just like the case that “I heard that demonstrative: S” is functionally equivalent to “I heard Ø [zero form of that complementizer] S,” “S: it never has been that complementizer” can also be understood as functionally equivalent to “It never has been Ø [zero form of that] S.” Let’s see the following examples: 未有仁而遺其親者也。(孟子梁惠王上) There never has been Ø one was ren but neglected his parents. (Mencius, 1: 1) 不好犯上而好作亂者, 未之有也。(論語學而) One did not like to offend against their superiors but was fond of stirring up confusion, there never has been that. (Analects, 1: 2)

After comparing these two sentence forms, I think the above first sentence with a that-clause with zero form of that can be transformed into the following sentence of paratactic structure: 仁而遺其親者, 未之有也。 One was ren but neglected his parents. There never has been that.

On the other hand, the above second sentence of paratactic structure can be transformed into the following sentence with a that-clause with zero form of that: 未有不好犯上而好作亂者也。 There never has been Ø one did not like to offend against their superiors but was fond of stirring up confusion, (Analects, 1: 2)

Besides, in classical Chinese, usually, a subordinate clause with zero form of that in a compound sentence is a nominalized sentence; but it is not necessary. For example, in the Hanfeizi, we can find a clear case that both nominalized sentence and complete sentence can be used as subordinate clause: 吾知吏之不事事也。(韓非子內儲說上) I know Ø officials’ being not have done what they should do. (Inner Congeries of Sayings, Upper Series) 吾知吏不事事也。(韓非子內儲說上) I know Ø officials have not done what they should do. (Inner Congeries of Sayings, Upper Series)

In the case of paratactic structure, it is also not necessary to use a nominalizer to nominalize a referred clause. For example: 不仁而得國者, 有之矣; 不仁而得天下[者], 未之有也。(孟子盡心下) One who was not benevolent but has got possession of a single state, there had been that; one was not benevolent but has got possession of the country under the heaven, there had never been that. (Mencius, 14: 59)

Here, zhe is used as a marker of nominalization, zhi is a demonstrative pronoun which is used as an anaphora to refer back to an event expressed by a nominalized sentence with the marker zhe or by a complete sentence without the marker zhe.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

91

Another example about the “zhi-construction” used in a sentence functionally equivalent to a compound sentence with a that-clause is much more complicated: 昔者竊聞之:子夏、子游、子張皆有聖人之一體, 冉牛、閔子、顏淵則具體而 微。(孟子公孫丑上) Formerly, I once heard this/that: Zi Xia, Zi You, and Zi Zhang had each one part of the Sage. Ran Niu, the disciple Min, and Yan Yuan had all the parts, but in small proportions. (Mencius, 3: 2) (Translation: Legge 1861b)

This example has the structure that “I heard this/that: S1 and S2.” The following example is much complicated: 子夏曰:「賢賢易色, 事父母能竭其力, 事君能致其身, 與朋友交言而有信。雖曰未學, 吾必謂之學矣。」(論語學而) Zi Xia said, “If a man withdraws his mind from the love of beauty, and applies it as sincerely to the love of the virtuous; if, in serving his parents, he can exert his utmost strength; if, in serving his prince, he can devote his life; if, in his intercourse with his friends, his words are sincere—although men say that he has not learned, I will certainly say that he has.” (Analects, 1: 7) (Translation: Legge 1861a)

This passage indicates a more complex sentence structure. Zi Xia said: “If a man can be described as having good performance in dealing with virtuous behavior, his parents, the ruler and his friends, although other people say [that] he has not learned, I will certainly say [that] he has.” It has the structure as follows: Although other people say Ø the man, who is A, B, C and D, has not learned, I will certainly say Ø he [who is A, B, C and D] has [learned].

In contrast to the above case whose clause has a zero form of “that,” some cases explicitly use “shi 是” (this/that) in the following: 德之不脩, 學之不講, 聞義不能徙, 不善不能改, 是吾憂也。(論語述而) The learning virtue without proper cultivation; the not thoroughly discussing what is learned; not being able to move towards righteousness of which a knowledge is gained; and not being able to change what is not good—that is my worry. (Analects, 7: 3) (Translation: Legge 1861a)

The structure of this sentence is: “A, B, C and D—that is my worry,” which can be regarded as the same in D-structure (deep structure) as the sentence that “My worry is that A, B, C and D.”

6  O  ther Grammatical Constructions for Expressing Propositional Attitudes If my analysis in the last section is right, I can conclude that classical Chinese does have some linguistic expressions similar to that-clauses or some other devices functionally equivalent to that-clauses in English. In this section, I will provide more examples to demonstrate that there are other kinds of grammatical devices which can be used to express sentential beliefs or propositional attitudes. In English, the sentence “I expect him to be a good student.” can be paraphrased as samesaying of

92

Y.-m. Fung

“I expect that he will/would be a good student.” This device is functionally equivalent to the accusative cum infinitive usage in Latin. In English, indirect statements are introduced by a subordinating conjunction that, which has no equivalent in classical Latin. Instead, a subject accusative + infinitive is used in Latin.18 We can also find some similar constructions in classical Chinese. As I know, there are at least three kinds as follows: (I) Pivotal sentence construction: In ancient Chinese grammar, there is a kind of grammatical structure which is named by most scholars of the field as “jian yu ju” 兼語句 (pivotal sentence). In a sentence of such structure, there is a noun or noun phrase, including pronoun such as zhi 之, which can be used as a jian yu, that is, can be used both as an object of the first verb and a subject of the second verb. For example, the sentence “[The old man] kept Zi Lui [PROi to stay overnight (in his house)].”(止子路宿) (Analects, 18:7) (論語微子) is embedded with an infinitival clause.19 Here, the proper name “Zi Lu” can be regarded as the accusative subject of the second verb “su” 宿 which plays a role just like an infinitive in English. Another example is the sentence “I looked at himi [PROi appears] not like a sovereign; I drew near to him and I saw nothing venerable about him.” (Mencius, 1: 6) (Translation: Legge 1861b) (望之不 似人君, 就之而不見所畏焉。) (孟子梁惠王上) In these sentences, the second sentence is a compound one while the first sentence with the zhi-construction of pivotal sentence can be understood as a functional equivalent of a to-infinitive construction or bare infinitive construction and a condensed form of the compound sentence that “I looked at him and he was/appeared not like a sovereign.” ([吾]望之 [而其/彼]不似人君。) (II) Serial verb construction: In comparison with the above pivotal sentence, there is another kind of infinitival clause in form of serial verb construction which is also very popular in classical Chinese. The sentence “Fishes see them to dive deep in the water.” (Zhuangzi, I 2: 11) (魚見之深入) (莊子齊物論) This is not a pivotal sentence. It is because there is no jian yu which is used as both the object of the main verb and the subject of the second verb. Unlike this example which has the form “S+V1+O+V2,” there are some  For example, the following English sentence uses a that-clause to express an indirect discourse: “Julia says that she is a good student.” Instead, the samesaying in Latin is expressed in accusative cum infinitive structure: “Iūlia dīcit sē bonam discipulam esse.” (literal translation: “Julia says herself to be a good student.”) Sē here is an accusative reflexive pronoun referring back to the subject of the main verb, i.e., Iūlia; esse is the infinitive “to be.” In classical Chinese, we can find that a similar kind of grammatical device is usually used in indirect discourse, and thus used to express that-beliefs. This point can be found in Parsons 2008: 216. 19  In generative linguistics, the motivation for a PRO subject comes from the grammatical of a sentence like “Peter asked Billk [PROk to sleep],” where the subject of the infinitival to-clause, though not overtly expressed, is understood to be controlled by an argument (i.e. an expression or syntactic element in a sentence that serves to complete the meaning of the verb.) of the main clause. 18

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

93

sentences of serial verb construction which have the form “S+ V1+V2.” The serial verb construction in the sentence “Zi Lu came in [to] see him” (Zhuangzi, II 10: 9) (子路入見) (莊子秋水) can be treated as a bare infinitival construction. (III) Small clause construction: In classical Chinese, another kind of device for making clauses is the construction of verbless small clause. Some Chinese grammarians call the sentence embedded with this kind of clause “xiao ju” 小句 (small sentence/clause). For example, the sentence “My lord forces the people lazy” (君使民慢) (Zuo Zhuan, Zhuang Gong 8) (左傳莊公八年) is embedded with a verbless small clause. The sentence “People with regard to ren is more [serious] than with regard to water and fire.” (民 之於仁也, 甚於水火。) (Analects, 15: 35) (論語衛靈公) is also embedded with a subordinate part. Here the operator “zhi” (之) as a self-referring marker added into the first part of the main sentence transforms the first part from an independent sentence into a subordinate clause in form of nominalization. This is just like the nominalized form of that-clause used as the subject of a matrix sentence in English. Although the word “zhi” in the zhi-clause construction is needed for making this transformation; but sometimes it can be removed. For example, although the sentence, “That you not knowing the happiness of fishes is the whole fact.” (子之不知 魚之樂全矣。), is used with a zhi-construction, the other sentence, “How do you know Ø [that] I do not know the happiness of fishes?” (Zhuangzi, II 10: 13) (安知 我不知魚之樂?) (莊子秋水), can be used for making this transformation without a zhi-construction. When the “zhi” is removed, the small clause in the original sentence will be transformed into a that-clause with a zero form of that. Last but not the least, a more complicated case deserves our attention. In the Analects, we can find an example of paratactic structure which is embedded with two pivotal sentences: 既欲其生, 又欲其死, 是惑也。(論語顏淵) [You] wish him to live and also wish him to die. That is the delusion. (Analects, 12: 10)

In the first two sentences, the two tokens of “qi” 其 (its/his/their) are jian yu, that is, used both as the object of the verb “wish” (yu 欲) and the subject of the verb “live” (sheng 生) and “die” (si 死). The demonstrative “shi” 是 (this/that) in the matrix sentence refers back to this compound sentence. So, the structure is: “(S+V1+qi+V2) and (S+V1+qi+V3). That is the delusion.” I think it can be translated into English as: “It is the delusion that (S+V1+qi+V2) and (S+V1+qi+V3).”

References Davidson, Donald. 2001. Inquiries into Truth and Interpretation, 2nd edition. Oxford: Oxford University Press. Davidson, Donald. 2002. Subjective, Objective, and Intersubjective. Oxford: Oxford University Press.

94

Y.-m. Fung

Fang, Wan-chuan. 2006. “The Concept of Truth and Pre-Qin Chinese Philosophy: On Chad Hansen’s View” (真理概念與先秦哲學). Journal of Nanjing University (Philosophy, Humanities and Social Sciences), 2.2006: 91–102. Graham, A.  C. 1978. Later Mohist Logic Ethics and Science. Hong Kong: The Chinese University Press. Graham, A. C. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle: Open Court. Guo Xiliang 郭錫良. 1989. “A New Investigation into the Terms of Mood in Pre-Qin” (先秦語氣 詞新探) (2). Research in Ancient Chinese Language 古漢語研究: 74–82. Hall, David, and Ames, Roger. 1987. Thinking Through Confucius. Albany: State University of New York Press. Hall, David and Ames, Rogers. 1995. Anticipating China: Thinking through the Narratives of Chinese and Western Culture. Albany: State University of New York Press. Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press. Hansen, Chad. 1985. “Chinese Language, Chinese Philosophy, and ‘Truth’.” The Journal of Asian Studies, 44.3: 491–519. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. New York: Oxford University Press. Hansen, Chad. 1993. “Term-Belief in Action: Sentences and Terms in Early Chinese Philosophy.” In Hans Lenk and Gregor Paul ed., Epistemological Issues in Classical Chinese Philosophy (45–68). Albany: State University of New York Press. Harbsmeier, Christoph. 1998. Science and Civilisation in China, vol. 7, part 1, Language and Logic in China, edited by Joseph Needham. Cambridge: Cambridge University Press. Hu Qiguang 胡奇光. 2005. A Chinese History of Small Learning (中國小學史). Shanghai: Shanghai People’s Publishing House. Hung, Bo 洪波. 2010. A Study of Chinese Historical Syntax (漢語歷史語法研究). Beijing: Commercial Press. Legge, James (translation). 1861a. The Analects of Confucius. In The Chinese Classics with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes: Volume 1. Hong Kong: at the author’s. London: Trübner & Co. Chinese Text Project. http://ctext.org/ analects Legge, James (translation). 1861b. The Works of Mencius. In The Analects of Confucius. In The Chinese Classics with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes: Volume 2. Hong Kong: at the author’s; London: Trübner & Co. Revised second edition (1895), Oxford: Clarendon Press. Reprinted by Dover Books in 1990 Dover Publications. Chinese Text Project, http://ctext.org/mengzi Legge, James (translation). 1885. The Book of Rites. In Sacred Books of the East, volume 28, part 4: The Li Ki. Chinese Text Project. http://ctext.org/liji. Legge, James (translation). 1891. The Writings of Chuang Tzu. Oxford: Oxford University Press. Chinese Text Project. http://ctext.org/zhuangzi. Legge, James (translation). 1991. The Chinese Classics: With a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes, vol. V. (The Chun Ts’ew, with the Tso Chuen). Hong Kong: at the author’s; London: Trübner & Co., 1861. Reprinted in 1991 by Taipei: Southern Materials Center Publishing (SMC Pub). Li Kuangwu 李匡武, et  al. 1989. A History of Chinese Logic (中國邏輯史), Pre-Qin volume. Lanzhou: Gansu People’s Publishing House. Liao, W. K. (translation). 2015. Han Fei Zi. Beijing: The Commercial Press. Palmer, Frank. 1971. Grammar. Harmondsworth: Penguin Books. Parsons, Terence. 2008. “The Development of Supposition Theory in the Later 12th through 14th Centuries.” In Dov M. Gabbay and John Woods, eds., Medieval and Renaissance Logic, Handbook of the History of Logic vol. 2: 157–280. Amsterdam: Elsevier.

4  Sentences (Ci 辭, Ju 句) and Propositional Attitudes

95

Roetz, Heiner. 1993. “Validity in Chou Thought: On Chad Hansen and the Pragmatic Turn in Sinology.” In Hans Lenk and Gregor Paul, eds., Epistemological Issues in Classical Chinese Philosophy. Albany: State University of New York Press. Ross, John. 1967. Constraints on Variables in Syntax (Cambridge: MIT dissertation). Searle, John. 1967. “Human Communication Theory and the Philosophy of Language.” In Frank E.  X. Dance ed., Human Communication Theory (116–129). New  York: Holt, Rinehart and Winston. Searle, John. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Searle, John. 2010. Making the Social World: The Structure of Human Civilization. Oxford: Oxford University Press. Williams, Joseph M. 1975. Origins of the English Language. New York: The Free Press. http://ctext.org of Chinese Text Project, edited by Dr. Donald Sturgeon. Yiu-ming Fung  is Emeritus Professor at the Hong Kong University of Science and Technology and currently Visiting Professor in the Department of Philosophy at the Tunghai University, Taiwan. He is an analytic philosopher with special interest in Chinese philosophy and comparative philosophy. Fung is the author of several books, including The Methodological Problems of Chinese Philosophy (1989); Chinese Philosophy in the Ancient Period, 4 volumes (1992); The Gongsun Longzi: A Perspective of Analytic Philosophy (1999); and The Myth of Transcendent Immanence: A Perspective of Analytic Philosophy on Contemporary Neo-­Confucianism (2003). He has also published more than 100 research papers both in Chinese and English.

Chapter 5

Counterfactual Conditionals Yiu-ming Fung

1  Alfred Bloom’s Challenge Marcel Granet may be the first scholar who describes the Chinese way of thinking as “correlative” which is regarded as contrary to the Western “analytical” thinking (Granet 1934). Joseph Needham follows this line to assign the term “correlative” to the cosmology in Han 漢 dynasty. He also stresses that “differences of linguistic structure between Chinese and the Indo-European languages had influence on the differences between Chinese and Western logical formulations” (Needham 1956: 199). Although A. C. Graham does not think that ancient Chinese thinkers did not have logical thinking and their thinking was totally incomparable to the Western logic, he still agrees that “this general indifference to logical problems is somehow connected with the structure of the Chinese language” (Graham 1971: 55). In his later writings, he further elaborates the idea of “correlative thinking” and believes that the difference between the Chinese correlative way of thinking and the Western analytical way of thinking is at the bottom level of conceptual schemes though both of their thinking at surface level are logical.1 This kind of comparative view implies the thesis of incomparability, if not global incommensurability, between the Chinese mode of thinking and the Western mode of thinking. A similar thesis of incomparability is based on the linguistic relativity hypothesis.2 That is the proposal that the particular language one speaks influences

 A detailed discussion about Graham’s view on “correlative thinking” in ancient China can be found in (Fung 2006: 117–62). 2  The linguistic relativity hypothesis holds that the structure of a language affects its speakers’ world view or cognition. This is popularly known as the Sapir-Whorf hypothesis or Whorfianism. 1

Y.-m. Fung (*) Department of Philosophy, Tunghai University, Taichung, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_5

97

98

Y.-m. Fung

the way one thinks about reality. Despite long-standing historical interest in the hypothesis, there is relatively little empirical research directly addressing it. In this regard, a more challenging view proposed by Alfred Bloom in his book, titled: The Linguistic Shaping of Thought: A Study in the Impact of Language on Thinking in China and the West (1981), is grounded in empirical evidence which is supposed to demonstrate how linguistic differences between Chinese and Western languages are paralleled by important conceptual differences. According to Bloom’s view, cognitive activities depend on highly abstracted representations of reality and only linguistically shaped schemas can effectively serve. So, it is in such activities that a speaker’s thinking might be expected to be most reflective of the constraints and conveniences inherent in her/his language. His empirical observation is that (Bloom 1979: 585): Specific differences between the Chinese and English languages suggest themselves as potentially significant sources of distinctive influence on the abstract cognitive activities of speakers of these two languages. Chinese and English both have precise means for expressing descriptive statements … but Chinese, unlike English, does not have any means for expressing counterfactual implicational statements such as “If John were to go to the library, he would see Mary” or “If John had gone to the library, he would have seen Mary” as distinct from the descriptive and straightforward implicational alternatives. Chinese, in other words, has no way to express distinctly that mood which in English and other Indo-­ European languages invites the reader or listener to shunt aside reality considerations and consider a state of affairs known to be false for the purpose of drawing implications as to what might be or might have been if that state of affairs were true.

One of the main points in Bloom’s research is about the relation between counterfactual markers and speakers’ facility with counterfactual or hypothetical reasoning. In the course of doing research on moral reasoning, Bloom’s observation was that Chinese speakers had difficulty with the counterfactual questions used in such research. In searching for a reason, Bloom came to suspect that the difficulty stemmed from the way counterfactuals were marked in the Chinese language. He then designed several counterfactual reasoning experiments where he presented various controlled stories to English and Chinese speakers and concluded that systematic marking of counterfactuals (along with other linguistic resources) aided sustaining a theoretical mode of thought. He also discussed the disadvantages of this mode of thought from a Chinese perspective. So, according to Bloom, the question why Chinese speakers do not have counterfactual thinking can be explained by the fact that the Chinese language does not have linguistic markers or other devices to express counterfactuals. Moreover, if sometimes argumentation needs to form

As mentioned in (Wolff and Holmes 2011: 253): “Linguistic relativity comprises three main ideas. First, it assumes that languages can differ significantly in the meanings of their words and syntactic constructions—an assumption that is strongly supported by linguistic, anthropological, and psychological studies of word and phrasal meaning across languages. Second, the proposal holds that the semantics of a language can affect the way in which its speakers perceive and conceptualize the world, and in the extreme, completely shape thought, a position known as linguistic determinism. Finally, given that language can affect thinking, linguistic relativity holds that speakers of different languages think differently.”

5  Counterfactual Conditionals

99

with counterfactual, Chinese speakers are incapable to make such kind of argumentation.3 Bloom’s conclusion is that, historically speaking, the fact that Chinese has not offered its speakers incentives for thinking about the world in counterfactual and entificational ways is likely to have contributed substantially to sustaining an intellectual climate in which these modes of thinking were less likely to arise. Here, I think, Bloom’s account raises two interesting questions: First, is it true that there is no way of making explicit counterfactuals in the Chinese language, including classical and modern Chinese? Second, is it true that the Chinese language unlikely uses counterfactuals to make reasoning?4

2  I s Chinese Language Handicapped to Express Counterfactuals? One of the major voices against Bloom’s view is coming from Terry Kit-fong Au’s articles (Au 1983: 155–87, 1984: 289–302). She had reservations about Bloom’s findings. In reading Bloom’s “counterfactual story,” she found the Chinese text used by Bloom rather unidiomatic. The low counterfactual response rates given to Bloom’s Chinese versions of the story might have been due to the unidiomatic text rather than to difficulty of reasoning counterfactually in Chinese per se. So she tried to replicate his findings with both his story and a new counterfactual story. The Chinese bilinguals in her studies read either Bloom’s or her story written in either Chinese or English. They showed little difficulty in understanding the counterfactual in either story in either language, insofar as the English and Chinese were idiomatic. When given her story written in Chinese, nearly monolingual Chinese also gave mostly counterfactual responses. She therefore concluded that the absence of a distinct counterfactual marker did not seem to hinder counterfactual reasoning in Chinese, and that the mastery of the English subjunctive by Chinese bilinguals might be quite tangential to their reasoning counterfactually in Chinese (Au 1984: 289). With solid evidence, Au’s response does rectify the conclusion of Bloom’s empirical study. In addition to their studies about modern Chinese, I think we can also rectify Boom’s view in the context of classical Chinese. In classical Chinese, I don’t think there is no counterfactual marker or other linguistic device to express counterfactuals which can be used to make counterfactual arguments. In this 3  In regard to this point, Bloom explains that: “the Chinese language has no distinct lexical, grammatical, or intonational device to signal entry into the counterfactual realm, to indicate explicitly that the events referred to have definitely not occurred and are being discussed for the purpose only of exploring the might-have-been or the might-be” (Bloom 1981: 16). 4  With respect to the Sapir-Whorf hypothesis, the problem of counterfactuals and their relation to Chinese language, Robert Wardy has provided a detailed discussion from a comparative perspective. Please see (Wardy 2004: § 3.1 and § 4). Here, thank you for one of the readers reminding me the contribution of Wardy in this regard.

100

Y.-m. Fung

chapter, I will argue for this point with textual evidence in the coming sections. In this section, I would like to demonstrate that the hypothesis of linguistic relativity together with its consequential thesis of incomparability or incommensurability is not sustainable. One of the reasons for my argument is that: if Chinese people can use their own language to describe or to translate the difference as argued by Bloom, there would be no real difficulty for them to use counterfactuals to express fictional reality and to make counterfactual reasoning. Or, put it another way, if Bloom were right, Chinese people would be unable to use their own language to understand Bloom’s findings. In this situation, they would be totally ignorant of Bloom’s idea in Chinese. But, this is an impossible case. Indeed, I think, this view is not in accordance with Donald Davidson’s principle of charity.5 That is: without a common coordinate, it is impossible for Chinese people to identify what has been said in English, including the counterfactual thought, by others. According to Davidson’s principle of charity, we cannot identify any thought of another person if there is no common ground between us (the interpreter) and her/ him (the speaker). Even if there is any thought with the characteristic of irrationality or unreason, we can only explain their irrationality in the seat of rationality. There is no massive errors in interpretation or communication and there is also no transcendence of rationality or logic in human thought. Davidson writes (Davidson 2001: 148–9): The principle [of charity] directs the interpreter to translate or interpret so as to read some of his own standards of truth into the pattern of sentences held true by the speaker. The point of the principle is to make the speaker intelligible, since too great deviations from consistency and correctness leave no common ground on which to judge either conformity or difference. From a formal point of view, the principle of charity helps solve the problem of the interaction of meaning and belief by restraining the degrees of freedom allowed belief while determining how to interpret words.

Based on this principle, Davidson thinks that “we could not understand someone whom we were forced to treat as departing radically and predominantly from all such [rational] norms. This would not be an example of irrationality, or of an alien set of standards: it would be an absence of rationality, something that could not be reckoned as thought” (Davidson 1993: 307). So, cognitive relativism or skepticism, including linguistic relativity, does not have a place in human thought. It is because, Davidson says, “If what we share provides a common standard of truth and objectivity, difference of opinion makes sense. But cognitive relativism about standards and schema requires what there cannot be, a position beyond all standard and schema” (Davidson 1993: 307).

5  Donald Davidson’s formulation of the principle of charity can be summarized as that: We make maximum sense of the words and thoughts of others when we interpret in a way that optimizes agreement. The principle may be invoked to make sense of a speaker’s utterances when one is unsure of their meaning. In regard to the theoretical character of the principle and its application in the Chinese context, please refer to (Fung 2006: 117–62).

5  Counterfactual Conditionals

101

3  L  inguistic Devices for Expressing Counterfactuals in Classical Chinese As indicated by Au, Bloom’s argument based on empirical findings is misled by his inadequate Chinese translation of the English version of the questionnaire. But Au does not go into details about the question whether there are any counterfactual markers or any other linguistic devices in Chinese language which are functionally equivalent to the form of counterfactual in English. Let alone the case in modern Mandarin. As demonstrated by many Chinese and Western sinologists and linguists, classical Chinese did have counterfactual markers and other functionally equivalent devices. For example, Edwin Pulleyblank points out that (Pulleyblank 1995: 149): [In classical Chinese] Subordination between clauses may be indicated by (a) a particle in the first clause, (b) a particle in the second clause, or (c) particles in both clauses. Thus an “if” clause may be marked by a particle such as ru 如 or gou 苟, or may be implied by the particle ze 則 introducing the conclusion, or may have both kinds of markers.

As with temporal and causal relationships, conditionals are often expressed in a paratactic form, simply by juxtaposing two simple predicates, such as “bu duo bu yan” 不奪不饜 (If they are not snatching, they are not satisfied. Literally: not snatch not satisfied.) (Mencius 孟子 I: 1) (Refer to http://ctext.org/mengzi) Pulleyblank’s explanation is that (Pulleyblank 1995: 149): There is nothing but the sense to tell us that we should translate in this way rather than as two coordinate clauses: They do not snatch and are not satisfied. It is a kind of conditional clause with paratactic structure.

In this case, the conditional “bu duo bu yan” as a condensed hypothetical sentence without an English-like counterfactual marker is used to express a warning to the king Hui of Liang 梁惠王 about a hypothetical situation of the consequence of invading other countries. Mencius’ warning is not used to express a real case, but a case which may happen in future.6 The conditional mentioned above is of the form of parataxis which does not have the English-like conditional marker. However, most of the conditionals in classical Chinese do have conditional marker similar to that in English. Pulleyblank shows that some particles in the subordinate clause often perform as a conditional marker. These particles include ruo 若, ru 如, er 而 (if); shi 使, ling 令 (supposing); gou 苟, cheng 誠, xin 信 (really); ji 即, fei 非, wei 微 (unless). Besides, some other particles in the main clause also perform as a conditional marker. These particles are ze 則, si 斯, and ji 即 (then). If these conditional markers are used in a hypothetical or fictional context, they perform the same logical function of counterfactual markers. Concerning the usage of counterfactual markers in classical Chinese, Christoph Harbsmeier, another influential Western sinologist with specialty in classical Chinese syntax, provides some concrete examples to illustrate the fact that classical  Most Chinese grammarians treat the double-negative construction, say, “bu … bu …” (“ 不…不…”) as a special kind of counterfactual marker. This point will be discussed later.

6

102

Y.-m. Fung

Chinese has its own patterns to express counterfactuals. He says (Harbsmeier 1998: 117): Classical Chinese writers regularly use patterns like ‘shih [shi] 使 p, (tse [ze] 則) q’ or, much more rarely, “chi [jie] 藉 p, (tse) q.” Chinese counterfactuals about a contingent event in the future cannot per definitionem be strictly counterfactual: “If he came, I would simply send him home again” speaks about a remote hypothesis, or about an inconceivable but strictly speaking possible future event. In just the same way counterfactual shih 使, when followed by a clause describing a contingent future event, will introduce a remote or “unrealistic” possibility.

Moreover, as pointed out by Harbsmeier, the counterfactual with the marker shi can be reinforced by adding the word jia 假 or ruo 若. So, jia shi or ruo shi is also used as markers in expressing counterfactuals (Harbsmeier 1981: 277). In this way, he says (Harbsmeier 1998: 117): [T]he combinations chia shih [jia shi] 假使 and jo shih [ruo shi] 若使 always mark counterfactual clauses. Hsiin Tzu’s [Xunzi 荀子] combination chia chih [jia zhi] 假之 “let us assume this as an arbitrary hypothesis” is not so much counterfactual as theoretically hypothetical. The counterfactual particle wei 微 “if it had not been for, but for,” on the other hand, like its English paraphrases, is entirely limited to counterfactual usage: the noun mentioned after wei must refer to something that is presupposed to have been non-existent.

Wei in classical Chinese is often used in a hypothetical situation to mean “unless” or “if not,” “without” or “if it had not been.” I think the most salient example to show the counterfactual situation in a sentence using negative particle wei (if it were not for, but for) is in the Lunyu 論語 (Analects XIV: 17): Wei Guan Zhong wu qi pifa zuoren yi. 微管仲吾其披髮左袵矣。 If it had been the case without Guan Zhong we would have been wearing our hair down and folding our robes to the left [just like the barbarian].

To use counterfactuals to make reasoning, there are a lot of examples in the Han thinker Wang Chong’s 王充 Lunheng 論衡 (Balanced Discourses or Discourse of Balance). One of the examples is to use counterfactual reasoning to demonstrate the unacceptability of traditional superstitions. Harbsmeier quotes the following passage as an example (Translation from Forke 1911: 320): 堯、高祖審龍之子, 子性類父, 龍能乘雲, 堯與高祖亦宜能焉。(奇怪) If Yao and Kao Tsu [Gao Zu] had really been the sons of dragons, then, since it is in the nature of offspring to be like their parents, and since a dragon can fly on the backs of clouds, then it follows that Yao and Kao Tsu should have been able to mount clouds and fly. (XV: 7)

Harbsmeier’s explanation is that: It turns out that for counterfactual conditionals of the sort: (C) if (contrary to fact) it were the case that p, then q. (Harbsmeier 1998: 117) Moreover, in regard to the counterfactual marker, Harbsmeier also provides clear examples to demonstrate that classical Chinese does have counterfactual markers which are distinct from ordinary conditional markers. His explanation is (Harbsmeier 1981: 273–4):

5  Counterfactual Conditionals

103

To my knowledge, it has not been noticed so far by grammarians of ancient Chinese that the language does make a clear distinction between ordinary conditionals and counterfactuals. Compare the following: (4a) 吾若言則死矣。 If I speak up, I shall die. (4b) 使吾言則死矣。 Supposing I had spoken up, I would have died. Not: “Supposing I speak up, I shall die.” Apparently, ruo 若- or ru 如 cannot replace shi [使] in counterfactuals like: (5) 使吾言則死久矣。 If I had spoken up, I would have died a long time ago! (Xin Xu 5.23.) (6) 使武安侯在者則族矣。 If the Lord of Wu-an was alive, I would exterminate him and his family. (Shi Ji 107.29.) Not: “Supposing he turns out to be alive, I shall exterminate him with his family!” The counterfactual nature of shi is conveniently explicit in: (7) 使寡人治信若是則民雖不治寡人弗怨, 意者未至然乎。 If I really did govern in this way, then I could not be angry even if the people were unruly. But for all I know things have not got to this point. (LSQ 16.8.) Here again, ruo cannot be used for shi.

In addition to the markers mentioned above, there are many other words which can be used to form a counterfactual in classical Chinese, such as tang 儻 (in case), dang 當 (should be), she 設 (supposing), etc. As discussed above, a pair of negations (negation terms such as bu 不, fei 非, fu 弗 (not), and wu 無 (no), etc.) in paratactic form can be used to express a necessary condition. Their forms include: “bu … bu …,” “bu … wu …,” “wu … fei …” or “fei … fu …,” etc. Some other counterfactual markers which are not so attractive to the Western sinologists include xiang 鄉 or xiang 向 (formerly, before), and xi 昔 (past); jin 今 (now), yu ci 於此 (here, this), and you 有 (have) in “you … yu ci …” These are temporal or spatial adverbials: The former three words are about cases in the past while the latter three words are about cases at present or here now. If each word is used in a context which indicates that the relevant case did not happen in the past or does not happen at present, the relevant sentence is usually used in a hypothetical conditional to express a counterfactual situation. As the examples demonstrated above, classical Chinese does have counterfactual markers to form conditionals to express fictional reality.7 If, for the sake of argument, we accept that there is no counterfactual marker in classical and modern Chinese, it does not imply that we cannot use Chinese language to express counterfactual situations and to make counterfactual arguments. The counterfactual is not the conditional in general but a special kind of the conditional.8 It is because an  More examples can be found in (Behr 2006).  The popular view in the field of counterfactuals is that the indicative conditional is different from the counterfactual conditional. However, if the context of a conditional indicates that the domain of possible worlds is narrowed down to one of the counterfactual, this kind of conditional can be treated as a special kind of ordinary conditional. Besides, I think the difference between the counterfactual with the tensed marker in Western languages and the counterfactual with the tenseless marker in Chinese language is that: the counterfactual information of the former is indicated by the background knowledge which is signified by the built-in tensed marker while that of the latter is indicated by the background knowledge reflected in the context of utterance.

7 8

104

Y.-m. Fung

ordinary conditional, say (A→B), is true in the following three types of possible world: ( 1) Both A and B are true. (2) A is false and B is true. (3) Both A and B are false. But, a counterfactual, say (ACF→B) [formulated also as (ACF→B) or (A>B)], is only true in the possible world (3). So, if we make a conditional (A→B) together with an assertion of (3) or with the background knowledge of (3) reflected in the context of utterance, it is tantamount to making a counterfactual (ACF→B). In other words, on the one hand, (ACF→B), i.e., a formulation of the counterfactual conditional: “If it were A, it would be B.” or “If it had been A, it would have been B.”) entails or presupposes both (A→B) and ~B and, on the other hand, usually, in a context of hypothetical situation, the assertion of both (A→B) and ~B is nothing more or less than the assertion of (ACF→B). If I am right, we can have the following formal definition for the counterfactual: (Definition 1) (ACF→B) =df [(A→B)&(~A&~B)] [in the context that both A and B are in the past unreal or present unreal situation.]

It can be simplified as: (Definition 2) (ACF→B) =df [(A→B)&~B] [in the context that B is in the past unreal or present unreal situation.]9

4  Counterfactual Reasoning in Classical Chinese If, for the sake of argument, we agree that there is no explicit counterfactual marker in Chinese language, it does not imply that Chinese people cannot use their language to express their thought in a counterfactual situation. As we know, (A→B) is logically equivalent to (~AVB) or ~(A&~B). For example, the sentence “People make effort to do this [easy work] but not [too much] to do that [hard work].” (人力 為此, 而寡為彼。) (Xunzi IV 10) (Refer to http://ctext.org/xunzi) is used in a hypothetical situation. Its form (A&~B) is logical equivalent to ~(A→B). So, it can be translated as “It is not the case that if people make effort to do this [easy work] then they would make effort to do that [hard work].” In other words, it can be transformed into a negation of a hypothetical situation. Let’s see another example in the Analects bout a similar kind of transformation:

9  A special case of this formula is that: the consequent B is embedded with a contradiction. In this case, we can make an argument of Reductio ad Absurdum (RAA) to reject the truth of the antecedent A.

5  Counterfactual Conditionals

105

為人謀而不忠乎?與朋友交而不信乎?傳不習乎?(論語學而) Whether, in transacting business for others, I may have been not faithful; whether, in intercourse with friends, I may have been not sincere; whether I may have not mastered and practiced the instructions of my teacher. (Analects I 4) (Translation from Legge 1861a) (Refer to http://ctext.org/analects)

All these questions have the form that: “(A and ~B)?”. But these questions are used rhetorically to reject the conjunctive sentence (A and ~B). So, their real form is; “~(A&~B).” If my interpretation is acceptable, I think these rhetorical sentences can be transformed into the conditional sentences with the logical form that: (A→B). If this kind of implicit conditional is used in the context of an unreal situation, it can be understood as a counterfactual. Moreover, if one utters a Chinese sentence with the form (AVB), which is logically equivalent to (~A→B), and asserts that B is false, it implies that ~A is also false. That is a counterfactual reasoning in terms of an implicit Modus Tollens. In the Zhuangzi 莊子 (II 4) (Refer to http://ctext.org/zhuangzi), the compound sentence: “One has no predetermined mind but has his affirmations and negations; this is [like the case] that one went to Yue today but arrived at it yesterday.” (Qiwu Lun) (未成乎心而有是非, 是今日適越而昔至也。) (齊物論) is a clear example of counterfactual. Here, the word “shi” 是 is a demonstrative pronoun which is used anaphorically to refer back to the former sentence. It has the form: (“Fa; this/that (i.e., shi) is Ga” in which “this/that” refers to “Fa.”) and can be simplified as: (Fa is Ga). So, its logical form can be elaborated as: (Fa→Ga). (Here, “a” stands for any person) It means that: “That one has no predetermined mind but has his affirmations and negations is that one went to Yue today but arrived at it yesterday.” Since the consequent (Ga) of the conditional is not true to the fact, it implies that, for Zhuangzi, this true conditional is used counterfactually. A more explicit counterfactual reasoning with a counterfactual marker can be found in the Discourse on the White-Horse (Baima Lun 白馬論) of the Gongsun Longzi 公孫龍子. In response to the objector’s challenge, Gongsun Long provides a counterfactual argument to rebut the claim that “white-horse is horse” in the following passage: (Refer to http://ctext.org/gongsunlongzi) 求馬, 黃、黑馬皆可致, 求白馬, 黃、黑馬不可致。使白馬乃馬也, 是所求一也; 所求 一者, 白者不異馬也。所求不異, 如 [而] 黃、黑馬有可有不可, 何也?可與不可, 其相 非明。故黃、黑馬一也, 而可以應有馬, 而不可以應有白馬, 白馬之非馬, 審矣。 Someone who seeks horse will be just as satisfied with yellow-horse or black-horse; someone who seeks white-horse will not be satisfied with yellow-horse or black-horse. Supposing that white-horse were after all horse, what they seek would be one and the same; that what they seek would be one and the same is because the white-thing would not be different from the horse. If what they seek is not different, why is it that yellow-horse or black-horse is admissible in the former case but not in the latter? Admissible and inadmissible are plainly contradictory. Therefore that yellow-horse and black-horse are one and the same in that they may answer to “having horse” but not to “having white-horse” is conclusive proof that white-horse is not horse. (Translation from Graham 1989: 88–9)

The main premise of Gongsun Long’s argument here is based on an instance of the Leibniz’s Law, i.e., the Principle of the Indiscernability of Identicals; and the argument is obviously of the form of Modus Tollens.

106

Y.-m. Fung

(1) 使白馬乃馬也, 則所求一也。 If white-horse is/were horse, it will/would be no difference between yellow-­ horse (or black-horse) in response to seeking white-horse and seeking horse. (2) 所求不一。 There is a difference. (3) 故白馬非馬。 Therefore, white-horse is not horse. The argument’s form can be elaborated as follows: (1F) [(a=b)→(Fa↔Fb)] (2F) ~(Fa↔Fb) (3F) ∴ ~(a=b) Here “Fa” and “Fb” are abbreviations of “Rac1c2” and “Rbc1c2” respectively, which represent “yellow-horse or black-horse is able to satisfy the request of seeking horse” and “yellow-horse or black-horse is able to satisfy the request of seeking white-horse” respectively. This argument is not only valid but also sound if both sides agree that there is a difference between “Fa” and “Fb” in seeking horse and seeking white-horse or both recognize that it is a fact. In this argument, Gongsun Long asserts that: the supposition that white-horse is/were horse leads to an unacceptable consequence that there will/would be no difference between yellow-horse (or black-horse) in response to seeking white-horse and seeking horse, which is counterfactual. So, based on the rule of Modus Tollens, he does make a beautiful counterfactual argument to reject the objector’s view.10 In conclusion, for Gongsun Long, this argument can be expressed in a counterfactual situation. That is: 以黃馬為非馬, 而以白馬為有馬, 此飛者入池而棺槨異處, 此天下之悖言亂辭也。 To take yellow horses as not horses, and yet take white horses as being horses, this is [like the case that] to have the flying in a pond and the inner and outer coffins in different places: a contradictory claim and misuse of statements as there is under heaven! (My translation)

This is a counterfactual conditional with the logical form: (A; this/that (i.e., shi) is B). Since B (to have the flying in a pond and the inner and outer coffins in different places) is counterfactual, A (to take yellow horses as not horses, and yet take white horses as being horses) can be rejected by the rule of Modus Tollens. In criticizing Confucians’ antitheism, Mozi 墨子 argues that: (Translation from Mei 1929) (Refer to http://ctext.org/mozi) 執無鬼而學祭禮, 是猶無客而學客禮也, 是猶無魚而為魚 摤 也。(公孟) To hold there are no spirits and learn sacrificial ceremonials, this is comparable to learning the ceremonials of hospitality while there is no guest or to making fishing nets while there are no fish. (Mozi XXXXVIII: 9)

10  In regard to the problem why I use individuals rather than predicates for the words such as ma 馬 and baima 白馬 to formulate the argument, the reason can be found in (Fung 2007) and (Fung 2020).

5  Counterfactual Conditionals

107

Again, this is also a counterfactual conditional with the logical form: (C; this/that (i.e., shi) is D). Since D (learning the ceremonials of hospitality while there is no guest or to making fishing nets while there are no fish) is counterfactual or unreasonable, C (to hold there are no spirits and learn sacrificial ceremonials) can be rejected by the rule of Modus Tollens In the Mencius, there is another example of counterfactual reasoning which is a passage on humans’ essential nature: (Translation from Legge 1861b) (Refer to http://ctext.org/mengzi) 所以謂人皆有不忍人之心者, 今人乍見孺子將入於井, 皆有怵惕惻隱之心。非所以內 交於孺子之父母也, 非所以要譽於鄉黨朋友也, 非惡其聲而然也。由是觀之, 無惻隱 之心, 非人也; 無羞惡之心, 非人也; 無辭讓之心, 非人也; 無是非之心, 非人 也。(公孫丑上) When I say that all men have a mind which cannot bear to see the sufferings of others, my meaning may be illustrated thus: even now-a-days, if men suddenly see a child about to fall into a well, they will without exception experience a feeling of alarm, distress and commiseration. They will feel so, not as a ground on which they may gain the favour of the child’s parents, nor as a ground on which they may seek the praise of their neighbours and friends, nor from a dislike to the reputation of having been unmoved by such a thing. From this case we may perceive that one who has no feeling of commiseration is not a human being, that one who has no feeling of shame and dislike is not a human being, that one who has no feeling of modesty and complaisance is not a human being, and that one who has no feeling of approving and disapproving is not a human being. (Mencius III 6)

This passage includes four parallel conditionals with the form “wu … fei …” ( 無 … 非 …) (no… not…). Each is paratactically formed with a pair of negations for expressing a necessary condition. It means that: without the fact described by the former part (B: “one has the feeling of …”), there would be no fact described by latter part (A: “one is a human being”) of the sentence. Its logical form is: (~B→~A). If we accept the fact that we are all human beings (i.e., A is true), to suppose that we are not (i.e., ~A) is counterfactual. Based on the fact that A (“one is a human being”) and the hypothetical conditional that (~B→~A), we can use the rule of Modus Tollens to reject the claim or the state of affairs that we do not have the moral feelings as mentioned in the passage (i.e., ~B is false). Let’s see one more example in the Mencius: (Translation from Legge 1861b) (Refer to http://ctext.org/mengzi) 楊氏為我, 是無君也; 墨氏兼愛, 是無父也。無父無君, 是禽獸也。(滕文公下) Mr. Yang’s principle is “each one for himself,” which does not acknowledge the claims of the sovereign. Mr. Mo’s principle is “to love all equally,” which does not acknowledge the peculiar affection due to a father. But to acknowledge neither king nor father is to be in the state of a beast. (VI: 14)

Here, for Mencius, the society in which “each one cares only for himself” is a society “without sovereign” (which entails “without government”) and the society in which “people love others all equally” is a society “without father” (which entails “without family”). As we all know, a society “without government” and “without family” is an “animal” society. Based on this background knowledge, the embedded argument is that: (1) The society in which “each one cares only for himself,” that (shi 是) is a society “without sovereign” (or “without government”) and the society in which “peo-

108

Y.-m. Fung

ple love others all equally,” that is a society “without father” (or “without family”). To claim or suppose this kind of view as indicated in the antecedents is counterfactual. [(∀x)(E1x→~Gx) and (∀x)(E2x→~Fx)] (2) In the actual world, our society does have government and family. [Ga and Fa] So, we can have the following conclusion (by the elimination of universal quantifier and the Modus Tollens): (3) Our society is not a society of egoism (E1) nor a society of equalitarianism [E2]. [~E1a and ~E2a]

5  Fables and Thought Experiments Used as Counterfactuals As pointed out by Jon Elster, “possible worlds are a hidden and implicit aspect of all model-building and of all theorizing . . . A theory must have implications for possible worlds by specifying the set of jointly realizable values of the relevant variables.” (Elster 1987: 7) Following this line, I think some fables, anecdotes, analogies and thought experiments can be treated as a kind of model-theoretical construction which provides reasoning in possible worlds, including the counterfactual one. In recent philosophical literature, there is an influential view advocated by Timothy Williamson and others that a philosophical thought experiment is just a piece of ordinary counterfactual reasoning. (Williamson 2004: 1–23, 2007). A story in the Zhuangzi (II: X 13) can be understood as a thought experiment. That is, a possible world in which the story players argue in terms of counterfactuality. This is a story or fiction of the debate on whether Zhuangzi knows the happiness of a fish. Zhuangzi and his friend, Hui Shi 惠施, have the following dialogue: (Refer to http://ctext.org/zhuangzi) (Translation from Watson 2013: 137–8 with my minor revision) 莊子與惠子遊於濠梁之上。 莊子曰:「儵魚出遊從容, 是魚之樂也。」[Z1] 惠子曰:「子非魚, 安知魚之樂?」[H1] 莊子曰:「子非我, 安知我不知魚之樂?」[Z2] 惠子曰:「我非子, 固不知子矣; 子固非魚矣, 子之不知魚之樂全矣。」[H2] 莊子曰:「請循其本。子曰『汝安知魚樂』云者, 既已知吾知之而問我, 我知之濠 上也。」[Z3] (秋水) Zhuangzi and Huizi were strolling along the dam of the Hao River when Zhuangzi said, “See how the minnows come out and dart around where they please! That’s what fish really enjoy!” [Z1] Huizi said, “You’re not a fish—how do you know the happiness of a fish?” [H1] Zhuangzi said, “You’re not I, so how do you know that I don’t know the happiness of a fish?” [Z2] Huizi said, “I’m not you, so I certainly don’t know what you know. On the other hand, you’re certainly not a fish—so that still proves that you don’t know the happiness of a fish” [H2]

5  Counterfactual Conditionals

109

Zhuangzi said, “Let’s go back to your original question, please. You asked me how I know the happiness of a fish—so you already knew that I knew it when you asked the question. I know it by standing here beside the Hao.” [Z3] (Zhuangzi II 10: 13)

When Hui Shi asks the question: “You’re not a fish—how do you know the happiness of a fish?” in (H1), his question is rhetoric and thus performs as a rejection of Zhuangzi’s knowing the happiness of a fish. It is because his assertion that: “On the other hand, you’re certainly not a fish—so that still proves that you don’t know the happiness of a fish!” in (H2) confirms this rejection. If (H2) for Hui Shi is a truth and (H1) is a rhetoric form of a conditional: “If you’re not a fish then you do not know the happiness of a fish.” (~A→~B) or “Since you’re not a fish, you do not know the happiness of a fish” (~A /∴ ~B), it seems that Hui Shi accepts the principle of the inaccessibility of other minds. Based on this principle, he can make the argument in (~A /∴ ~B). If (H2) is understood as (~A→~B) rather than (~A /∴ ~B), since (H2) for Hui Shi is a truth, the conditional (B→A) (If you know the happiness of a fish then you are a fish.), which is a logical equivalent of (~A→~B), must be a counterfactual conditional. Moreover, the reasoning either from the premises ~A and (~A→~B) to the conclusion ~B or from the premises ~A and (B→A) to the conclusion ~B is valid, though the former, in general, is a conditional reasoning while the latter, in specific, is a counterfactual reasoning. So, I think, there is an implicit argument in Hui Shi’s thinking which can be elaborated in modern Western logic.11 Furthermore, Hui Shi’s question and assertion imply or presuppose that if Zhuangzi’s view were right (that he knows the happiness of a fish) it would lead Zhuangzi to a commitment that he is a fish. But, this is counterfactual. The analogy used as a thought experiment in the Hanfeizi 韓非子 also includes a counterfactual argument. It says that (Refer to http://ctext.org/hanfeizi) (Liao 2015: 1490–1): 楚人有鬻楯與矛者, 譽之曰:「吾楯之堅, 莫能陷也。」又譽其矛曰:「吾矛之利, 於物 無不陷也。」或曰:「以子之矛陷子之楯, 何如?」其人弗能應也。夫不可陷之楯與無 不陷之矛, 不可同世而立。(難一) Once there was a man of Ch`u selling shields and halberds. In praising his shields he said, “My shields are so solid that nothing can penetrate them.” Again, in praising his halberds, he said, “My halberds are so sharp that they can penetrate anything.” In response to his words somebody asked, “How about using your halberds to pierce through your shields?” To this the man could not give any reply. Indeed, impenetrable shields and absolutely penetrative halberds cannot stand together at the same time. (Hanfeizi XXXVI: 4)

Since “impenetrable shields and absolutely penetrative halberds cannot stand together at the same time,” if we assert that “my halberds are so sharp that they can penetrate anything,” it would lead to the consequence that they can also penetrate my impenetrable shields. A case similar to “To move an unmovable stone, which is God’s creation, by God,” that is counterfactual. So, we can use this counterfactual consequent of the conditional to reject its antecedent. That is, on the one hand,  The elaboration of the arguments in the debate between Zhuangzi and Hui Shi can be found in Fung 2020? (forthcoming).

11

110

Y.-m. Fung

based on the assumption of impenetrable shields, we can reject the existence of absolutely penetrative halberds; on the other hand, based on the assumption of absolutely penetrative halberds, we can reject the existence of impenetrable shields. As we know, Wang Chong is a critical thinker who rejects many fictional ideas which are without empirical evidence to confirm their objective validity. In the Lunheng 論衡 (Balanced Discourses or Discourse of Balance), for example, in regard to the legend that Yao is able to shoot nine of the ten suns, he argues in the story in terms of a hypothetical form to express a counterfactual with the marker jia shi as follows: 假使堯時天地相近, 堯射得之, … 能射火而滅之, 則當射水而除之。… 堯能射日, 使 火不為害, 不能射河, 使水不為害。夫射水不能卻水, 則知射日之語虛, 非實也。(感 虛) (Refer to http://ctext.org/lunheng) Falsely supposing that the heaven is near the earth in Yao’s time and he was successful to shoot the sun…. If he was able to shoot the fire [from the sun] and thus to exterminate the fire, then he could be also able to shoot the water [the flood from the river] and thus to exterminate the water [flood]. Here, Yao was able to shoot the sun and thus to remove the harm of the fire; but he was unable to shoot the river and thus to remove the harm of the water [flood]. (Lunheng XIX 4–5)

Suppose that Yao could shoot the sun and thus remove the fire from the sun, he could also shoot the river and thus remove the flood from the river (ACF→B). But, Yao could not remove the flood from the river (~B). So, based on the rule of Modus Tollens, Yao could not shoot the sun (~A).

References Au, Terry Kit-fong. 1983. “Chinese and English counterfactuals: The Sapir-Whorf hypothesis revisited.” Cognition 15.1–3: 155–187. Au, Terry Kit-fong. 1984. “Counterfactuals: In reply to Alfred Bloom.” Cognition 17.3: 289–302. Behr, Wolfgang. 2006. “Morphological Notes on the Old Chinese Counterfactual.” Bochumer Jahrbuch zur Ostasienforschung 30: 55–88. Bloom, Alfred H. 1979. “The Impact of Chinese Linguistic Structure on Cognitive Style.” Current Anthropology 20.3: 585–586. Bloom, Alfred H. 1981. The Linguistic Shaping of Thought: A Study in the Impact of Language on Thinking in China and the West. New Jersey: Lawrence Erlbaum Associates, Hillsdale. Davidson, Donald. 1993. “Locating Literary Language.” In Reed W.  Dasenbrock ed., Literary Theory after Davidson (295–308). University Park: Pennsylvania State University Press: . Davidson, Donald. 2001. Subjective, Intersubjective, Objective. Oxford/New York: Oxford University Press. Elster, Jon. 1987. Logic and Society: Contradictions and Possible Worlds. Chichester/New York: Wiley. Forke, Alfred (translation). 1911. Lun Heng, Philosophical Essays of Wang Chhung vol. I. Kelley & Walsh, Shanghai; Luzac, London; Harrassowitz, Leipzig. Fung, Yiu-ming. 2006. “Davidson’s Charity in the Context of Chinese Philosophy.” In Bo Mou ed., Davidson Philosophy and Chinese Philosophy: Constructive Engagement. Leiden/Boston: Brill: 117–62. Fung, Yiu-ming. 2007. “A Logical Perspective on ‘Discourse on White-Horse’.” Journal of Chinese Philosophy 34.4: 515–536.

5  Counterfactual Conditionals

111

Fung, Yiu-ming. 2020. “Reference and Ontology in the Gongsun Longzi.” In Worlds of East Asia, Swiss Asia Society (forthcoming). Fung, Yiu-ming. 2020 (?). “Zhuangzi’s Idea of Wei-yi (Being One): with Special Reference to the Story of Happy Fish.” In Chong Kim-chong ed., Dao Companion to the Zhuangzi. Dordrecht/ Heidelberg/New York/London: Springer (forthcoming). Graham, A. C. 1971. “The Place of Reason in the Chinese Philosophical Tradition,” In Raymond Dawson ed., The Legacy of China. New York: Oxford University Press. Graham, A. C. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle: Open Court. Granet, Marcel. 1934. La Pensée chinoise. Paris: La Renaissance du livre. Harbsmeier, Christoph. 1981. Aspects of Classical Chinese Syntax. London: Curzon Press. Harbsmeier, Christoph. 1998. Science and Civilisation in China (Joseph Needham), vol. 7, part I: Language and Logic. Cambridge: Cambridge University Press. Legge, James (translation). 1861a. The Analects of Confucius. In The Chinese Classics with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes: Volume 1. Hong Kong: at the author’s; London: Trübner & Co. (Refer to http://ctext.org/analects) Legge, James (translation). 1861b. The Works of Mencius.In The Analects of Confucius. In The Chinese Classics with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes: Volume 2. Hong Kong: at the author’s; London: Trübner & Co. Revised second edition (1895), Oxford: Clarendon Press. Reprinted by Dover Books in 1990 Dover Publications. (Refer to http://ctext.org/mengzi) Liao, W. K. (translation). 2015. Han Fei Zi. Beijing: The Commercial Press. Mei, Y. P. 1929. The Ethical and Political Works of Motse. London: Probsthain. (Refer to http:// ctext.org/mozi) Needham, Joseph. 1956. Science and Civilisation in China, volume 2: History of Scientific Thought. Cambridge: Cambridge University Press. Pulleyblank, Edwin G. 1995. Outline of Classical Chinese Grammar. Vancouver: UBC Press. Wardy, Robert. 2004. Aristotle in China: Language, Categories and Translation. Cambridge: Cambridge University Press. Watson, Burton (translation). 2013. The Complete Works of Zhuangzi. New York/Chichester/West Sussex: Columbia University Press. Williamson, Timothy. 2004. “Armchair Philosophy, Metaphysical Modality and Counterfactual Thinking” In Proceedings of the Aristotelian Society, CV: 1–23. Williamson, Timothy. 2007. The Philosophy of Philosophy. Oxford: Blackwell. Wolff, Philip and Holmes Kevin. 2011. “Linguistic relativity.” Wiley Interdisciplinary Reviews: Cognitive Science 2. 3: 253–265. http://ctext.org of Chinese Text Project, edited by Dr. Donald Sturgeon. Yiu-ming Fung  , is Emeritus Professor at the Hong Kong University of Science and Technology and currently Visiting Professor in the Department of Philosophy at the Tunghai University, Taiwan. He is an analytic philosopher with special interest in Chinese philosophy and comparative philosophy. Fung is the author of several books, including The Methodological Problems of Chinese Philosophy (1989); Chinese Philosophy in the Ancient Period, 4 volumes (1992); The Gongsun Longzi: A Perspective of Analytic Philosophy (1999); and The Myth of Transcendent Immanence: A Perspective of Analytic Philosophy on Contemporary ­Neo-­Confucianism (2003). He has also published more than 100 research papers both in Chinese and English.

Chapter 6

Truth in Pre-Han Thought Chris Fraser

1  Introduction The role of truth in pre-Han thought has been a focus of interest and controversy since Munro first suggested that early Chinese thinkers were concerned primarily with the consequences of a belief or proposition for action, not its truth (Munro 1969: 55). Scholars have defended a range of interpretations of the place of truth in early Chinese thought, from the view that pre-Han philosophy has no concept of semantic truth, for example, to the view that it has several concepts with roles overlapping those of truth, to the view that early Chinese thinkers actually offer theories of truth.1 One reason for intense interest in the topic is signaled by Graham’s well-known remark that for pre-Han philosophers “the crucial question…is not the Western philosopher’s ‘What is the truth?’ but ‘Where is the way?, the way to order the state and conduct personal life” (Graham 1989: 3). How issues related to truth are framed may reflect central features of the theoretical orientation of early Chinese philosophical discourse. For example, if pre-Han inquiries into logic, language, knowledge, and metaphysics focused on accurate description of the structure of reality, they might accordingly devote much attention to explaining the nature of truth and evaluating which descriptions are indeed true. On the other hand, if such inquiries centered on questions concerning the proper dao 道 (way) of conduct and how it relates to the course of the natural world, they might instead be concerned primarily

 For an example of the first of these views, see (Hansen 1985); the second, (Fraser 2012); the third, (McLeod 2015). 1

C. Fraser (*) Department of Philosophy, University of Hong Kong, Hong Kong, People’s Republic of China e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_6

113

114

C. Fraser

with questions regarding the appropriate use of language and the basis for norms of correct use—whether they are purely products of human convention, for example, or are grounded in natural patterns. A clearer grasp of how pre-Han philosophers thought about truth may help us better understand the overall framework and orientation of early Chinese philosophy. This article first reviews the long-standing debate in the scholarly literature concerning the role of truth in pre-Han philosophy. Next, it considers how we might effectively and informatively frame inquiry into the topic of truth in Chinese thought. It then offers concise interpretations of several concepts in pre-Han texts with conceptual roles overlapping that of truth.

2  Debates on Truth Discussions of truth in the international literature on Chinese philosophy over the past several decades were sparked by Munro’s generalization that in evaluating a belief or proposition, early Chinese thinkers were less interested in its truth or falsity than in its “behavioral implications”—in what consequences adherence to the claim might have for personal and social conduct (Munro 1969: 55). Munro’s aim was to underscore the emphasis in Chinese thought on the action-guiding functions of word choices and verbal pronouncements—a theme stressed in both the Confucian Analects and the Mozi, for instance—and the tight conceptual connection between knowledge and action. In many contexts in Chinese philosophy, knowledge is understood primarily as knowing how to act, not knowing the truth. Developing Munro’s observation, Hansen suggested that in pre-Han thought, language was regarded as having mainly a regulative function, producing and expressing attitudes with implications for action (Hansen 1983: 60). Texts such as the Xunzi “Correct Names” essay clearly indicate that early theories about words and statements have a predominantly practical orientation, a salient concern being correcting speakers’ use of “names” (ming 名) so that all speakers use the same names to refer to the same, conventionally designated objects. Once names are corrected, they can be used to direct action effectively, as those issuing and following orders, for example, will interpret them in a unified way.2 The Mohist doctrine of the “three standards” for evaluating statements or doctrines implies that, as Munro originally suggested, the Mohists’ primary interest lies in how statements affect people’s behavior. In evaluating statements, the Mohists assign as much or more weight to their practical utility and conformity to precedent as to whether they correspond to what people see and hear.3 Drawing on such features of pre-Han philosophical discourse, Hansen proposed that “Chinese philosophy has no concept of truth” 2  For a detailed account, along with an overview of major themes in early Chinese philosophy of language and logic, see (Fraser 2016). 3  See Mozi (1948), book 35. For an overview of the Mohist theory, see (Fraser 2015, sect. 4). For further discussion, see (Loy 2008) and (Fraser 2012).

6  Truth in Pre-Han Thought

115

(Hansen 1985: 492), insofar as, in his view, the concept of truth and theoretical concerns related to truth play effectively no role in early Chinese thought. Hansen offered three reasons for this audacious claim (Hansen 1985: 492–496, 504–509; Hansen 1992: 44, 116). First, he suggested, early Chinese theories of language had a pragmatic, not semantic, orientation, and thus assigned no role to a concept of truth. Chinese theorists were concerned with the proper use of language and the action-guiding consequences of its use, not with whether descriptive statements about the world were true. Second, pre-Han discussions of language focused on names (ming), or words, and devoted little or no attention to the role of sentences or statements, the units of language usually regarded as bearing a truth value. Because of this formal feature of their inquiries, Chinese theorists could not be expected to address questions of truth. Third, pre-Han thinkers’ main concern in evaluating utterances was to ask whether they were ke 可 (“assertible,” “permissible”), a pragmatic rather than a semantic concept. Instead of being concerned with the semantic issue of whether utterances were true, then, they were concerned with the pragmatic issue of whether they were assertible. Dovetailing with the second and third of these arguments, Hall and Ames contended that since Classical Chinese expresses semantic content through “strings of names,” not sentences or propositional utterances, early Chinese thinkers lacked an interest in questions of truth and falsity (Hall and Ames 1987: 298–299). “Words, as names, may be judged appropriate or inappropriate,” they suggested, while “only propositions may, in the strict sense, be true or false” (Hall and Ames 1987: 299). Against these interpretations, Graham maintained that a theoretical focus on dao rather than on truth has nothing to do with “everyday questions of fact,” regarding which “there is no significant difference” between Chinese thinkers and ourselves (Graham 1989: 395). Classical Chinese must have possessed the expressive resources needed for speakers to “affirm a fact,” he argued, for otherwise it would have lacked “the communicative function needed to serve as a language” (Graham 1989: 395). Graham thought it unimportant whether early thinkers had a concept of truth that maps closely onto ours, provided they had words by which to express assent to the announcement “Dinner is on the table” if and only if dinner is in fact on the table (Graham 1989: 395). His underlying point seems to have been the potentially cogent claim that a language cannot be fully adequate for communication unless it allows for semantic evaluation of assertions as true or false. Unfortunately, Graham’s discussion at times seems to conflate assenting to an utterance with affirming its truth, leaving open the possibility that, as Hansen or Hall and Ames might rejoin, a listener could assent on grounds of social appropriateness or utility rather than truth.4 In the ensuing decades, numerous writers have discussed Hansen’s and Hall and Ames’s contentions, some offering lists of examples purporting to illustrate—with varying degrees of plausibility—instances of truth-like evaluations of assertions in

4  For Graham’s full response to Hansen, see (Graham 1989: 395–396). For Hansen’s reply, see (Hansen 1992: 392, n. 86).

116

C. Fraser

pre-Han texts.5 Until relatively recently, however, little research offered detailed, thorough explanations of how evaluations of truth might fit into early Chinese philosophy of language and logic. Several interpretations along these lines have now appeared, including Robins’s account of how Later Mohist logic uses shi 是 and ran 然 as terms of semantic appraisal (Robins 2010: 46), Fraser’s treatment of how the roles of dang 當, ran 然, shi 是, and ke 可 in Later Mohist thought overlap with truth (Fraser 2012), and Leong’s study of ran 然 in pre-Han texts (Leong 2015).6 Complementing these studies, Robins and Saunders have shown that Later Mohist texts themselves appear to implicitly recognize the difference between semantics and pragmatics, insofar as they distinguish between the normal semantic content of a word—to them, its reference in typical contexts—and the content the same word may take on because of pragmatic factors in an unusual context, as when a general term such as “crane” is “borrowed” to refer to dogs or is used as a proper name (Robins 2012; Saunders 2014). The Later Mohists’ discussion of such contexts undermines claims that pre-Han thinkers were concerned purely with pragmatics as opposed to semantics. What of Hansen’s original argument, partly shared by Hall and Ames, that the theoretical orientation and structural features of pre-Han discourse directed thinkers away from issues of truth and even from an explicit concept of truth? The points Hansen and Hall and Ames cite indeed reflect signal features of pre-Han thought. Nevertheless, they fall short of supporting the radical conclusion that early Chinese thinkers do not address truth at all. Pre-Han philosophical discourse could indeed have had a fundamentally pragmatic orientation, being concerned mainly with the appropriate use of words and statements to guide conduct, while nevertheless leaving conceptual space for a concept of truth and explanations of what it is for an assertion to be true. Such explanations could be grounded in a conception of correct word use, thus cohering with the Chinese focus on dao and norms of appropriate conduct. A concern with pragmatics—with how words are used—need not exclude a concern with semantic issues—such as whether some utterances correctly describe how things stand. Similarly, a formal focus on names or words, rather than sentences, does not preclude evaluations or explanations of truth. Strictly speaking, the units of discourse that are evaluated as true or false are assertions, not sentences.7 In English, assertions are typically made by uttering grammatically complete subject-verb 5  See, for example, (Roetz 1993: 85–96), (Harbsmeier 1998: 193–209), (Van Norden 2007: 361– 380), and, more recently, (McLeod 2015). On Harbsmeier’s methodology, see (Robins 2010: 24, n. 45). 6  These studies follow in the footsteps of Chmielweski’s pioneering study in English of the evaluation of assertions in the later Mohist texts. See, e.g., (Chmielweski 1966: 36, 41–46). 7  An assertion is a speech act by means of which a speaker claims that some fact holds, whereas a sentence is not a speech act but a grammatical unit that can be used to perform a speech act. For the purposes of this discussion, I will set aside the question of whether it is sentences or propositions that should be regarded as the bearers of truth values. The point is simply that many sentence types, such as “The cat is on the mat,” are not associated with a truth value. Rather, particular tokens of such sentences take on a truth value when asserted in some context.

6  Truth in Pre-Han Thought

117

s­ entences, but an assertion can be expressed using only a term, provided contextual clues make clear what object the term is predicated of. In the context of an ancient Chinese distinction-drawing debate (bian 辯), if a Mohist dialectician were to indicate a certain animal and say “niu 牛” (ox) or “fei niu 非牛” (non-ox), he would be making an assertion functionally analogous to the assertion we might express by saying “That animal is an ox” or “That animal is not an ox.” Regarding Hansen’s third point, as the discussion in Sect. 4 will show, ke 可 (assertible, permissible) is in fact not the only term pre-Han thinkers used to evaluate utterances. Nor is ke purely a pragmatic concept, since appraisal of whether utterances are ke rests at least partly on semantic and logical grounds. Pre-Han thought thus could indeed have had a pragmatic, practical orientation, have focused on issues pertaining to dao rather than truth, and formally have been concerned with terms rather than sentences, while nonetheless addressing issues of truth and employing a concept of truth.

3  Framing the Inquiry The foregoing discussion suggests that in considering interpretations of the role of truth in early Chinese thought, it is vital to distinguish two sets of questions. One concerns the extent to which truth was a focus of theoretical attention. Did Chinese thinkers present explicit accounts of truth? Did they take giving true descriptions of the world to be a major aim of inquiry? Did they explain other aspects of their thought by appeal to truth? To answer such questions in the affirmative, we would expect to find explicit, sustained discussions in pre-Han texts addressing the role of one or more concepts best explained as counterparts to truth. The other set of questions concerns whether pre-Han thought “has a concept of truth.” Do pre-Han texts employ terms whose function overlaps “…is true”? Do they evaluate assertions as to their truth? Do they have a conception of an assertion being correct according to specifically semantic norms, as opposed to norms of social propriety or practical utility? To answer these questions in the affirmative, we need only identify passages in which assertions are endorsed or rejected using evaluative terminology best explained as invoking semantic rather than other grounds. Regarding the first set of questions, the principal discussions of language and logic in pre-Han texts8 are concerned mainly with how the appropriate use of names (ming 名) and statements (yan 言) to guide action is based on distinctions between similar and dissimilar kinds (lei 類) of “stuff” (shi 實), or objects and situations.9 8  I have in mind here the Mohist “Condemning Fate” essays (Mozi books 35–37), the Later Mohist dialectical writings (Mozi books 40–45), the Xunzi “Correct Names” and “Correct Discourse” essays (Xunzi books 21–22), and the major essays on language in The Annals of Lü Buwei (see Knoblock and Riegel 2000, sections 16/8, 17/1, 18/4, and 18/5, for example). 9  Early texts on language treat all words as various kinds of “names” (ming 名). “Statements” (yan 言) and “expressions” (ci 辭) are regarded as combinations of names used to express a single

118

C. Fraser

Questions about the correct way to draw such distinctions are answered by reference to models or standards (fa 法) of the kinds of things a name typically refers to, which exemplify practical norms for distinguishing things, using words, and making statements. To justify asserting a name of something, speakers cite models of the kind of object or situation the name applies to and propose analogical inferences to support treating the case at hand as similar to the model. Among the models proposed for evaluating statements are precedents set by wise leaders, public observation, and useful practical consequences—criteria that align at best only partly with semantic correctness.10 Much of the early literature thus emphasizes the action-­ guiding role of language, the importance of coordinating norms of word use, and appeals to precedent and utility in determining correct use. Given the prevailing focus on such themes, there are strong grounds for accepting the shared underlying insight, articulated in various ways by Munro, Hansen, Hall and Ames, and Graham, that truth is indeed not a central theme of inquiry in early Chinese thought. Instead, early texts emphasize issues related to identifying and practicing the proper dao, a concept that incorporates norms for using words and making statements. Responses to such issues were typically not assessed or explained in terms of truth. Indeed, they may not have been conceptualized as truth-bearing entities, such as assertions, but as ways, methods, styles, manners, guidelines, and related notions. Truth as an explicit topic was of at most only peripheral interest.11 This answer to the first set of questions of course leaves the second set open, as Graham recognized when he insisted that a philosophical focus on dao does not preclude thinkers from having a vocabulary by which to affirm facts (Graham 1989: 395). The absence of an explicit, thematized interest in truth also does not rule out the possibility that early texts on language and logic might offer implicit or indirect explanations of what it is for an assertion to be true. Indeed, as an inductive generalization, one might argue that any language with reasonably complete expressive resources, such as Classical Chinese, will likely include semantic terminology that permits speakers to appraise the status of assertions as true or false. Such terminology is expressively useful, most fundamentally because it allows speakers to express evaluations of others’ or their own previous assertions by referring to their content and explicitly endorsing or rejecting it as fact. To borrow Graham’s example, having been told that dinner is on the table yet finding it is not, a speaker can remark that “‘Dinner is on the table’ is not true.” Without a concept of truth, speakers cannot explicitly attribute or deny factual correctness to an assertion this way; they can only repeat or negate the assertion in question (“Dinner is not on the table”). Other expressive functions of truth include allowing quantification over endorsed thought or intention. “Explanations” (shuo 說) or “expositions” (yi 議) are longer pieces of discourse that present the reasons for something. Names refer to “stuff” or “objects” (shi 實), which include both physical objects and situations or states of affairs. 10  The generalizations in this paragraph draw on the extended discussions in (Fraser 2012, 2013, 2015 and 2016). 11  McLeod rejects this characterization, but the strained interpretations he offers as evidence of purported early Chinese theories of truth tend to reinforce it (McLeod 2015).

6  Truth in Pre-Han Thought

119

a­ ssertions (“Some of what she said is true”), endorsement or rejection of assertions referred to using nominalizations or noun phrases (“What he said is true,” “Her claim is not true”), and embedding of such constructions within other assertions, without necessarily endorsing their content (“If everything she said is true, then we should cancel our plans”). A language that provides for such functions is expressively richer than one that does not. Since the second set of questions remain open, how might we go about investigating them? We need to look for terms in pre-Han texts that are used to evaluate the status of assertions and that perform expressive functions similar to those of “…is true.” At the same time, we need to carefully examine the context of such evaluations to determine whether they are best explained as appraising assertions according to specifically semantic norms. Utterances may be evaluated positively or negatively on a variety of non-semantic grounds, such as whether they are useful, polite, or aesthetically pleasing. To count as an appraisal of truth, an evaluation must refer to semantic norms—it must deem an assertion correct specifically on the grounds that its content is correctly applied to the things it is asserted of. If we find terms of evaluation for assertions that are best explained as referring to semantic grounds and that perform functions associated with “…is (not) true,” then we have persuasive reasons to conclude that pre-Han philosophers indeed “have a concept of truth.” The concepts expressed by such terms could of course diverge in various respects from our concept of truth, but their roles would be similar enough to justify interpreting them as counterparts to truth in certain discursive contexts. The next section surveys several such concepts.

4  Concepts Related to Truth The terms in pre-Han discourse most frequently considered to have roles similar to that of “true” include dang 當 (“fit,” “coincide”), ran 然 (“so,” “like this”), and shi 是 (“this, right”).12 A further concept used to evaluate assertions is ke 可 (“permissible,” “acceptable”). A survey of the use of these terms shows that in some contexts, although not all, the first three function as terms of semantic appraisal with roles overlapping that of truth, while the fourth seems to function mainly as a term of logical appraisal. Contexts in which these terms play roles similar to that of truth also hint at pre-Han thinkers’ implicit explanation of what it is for an assertion to be true.

 See Graham (1989: 395), Harbsmeier (1998: 201), Robins (2010: 46), Fraser (2012: 362), and Leong (2015: 68). Another term used in some contexts to evaluate statements is xin 信 (“trustworthy,” “sincere,” “reliable”). However, as the Mohists explain, in many such contexts, xin refers not to the semantic status of an utterance but to whether it conforms to the speaker’s actual thoughts (Graham 2003: 275). In other contexts, xin may refer to whether an utterance is reliable or trustworthy. 12

120

C. Fraser

4.1  Dang 當 Dang can be used to express whether a term asserted of an object correctly fits it. In the Later Mohist canons, for example, asserting a term of a thing, such as by calling it an ox, is a speech act that can be dang or not. With respect to a term, such as “ox,” if one side in a distinction-drawing debate (bian 辯) calls something “ox” and the other calls it “non-ox,” exactly one of these two assertions must dang (fit, coincide with) the animal in question (Graham 2003: 318, A74). A speaker wins such a debate if the term she asserts of a thing fits (dang) it (Graham 2003: 402, B35). Dang can also be used of sentential utterances. An example in the canons is “in debates there is no winning” (bian wu sheng 辯無勝), which the text contends is not dang (Graham 2003: 402, B35). This evaluation implies that dang can be used to express endorsement of and not-dang disagreement with an assertion that is a subject-­predicate sentence. Yan 言 (statements) too can be dang or not, again indicating that dang can be used to evaluate not only term use but longer utterances expressing thoughts. Dang can also be used with noun phrases referring back to the content of an antecedent assertion, as in “qi yan zhi dang 其言之當” (“the dang of his statements” or “that his statements fit”) (Graham 2003: 275, A14). It can be used of noun phrases embedded in longer sentences: “If this person’s yan (statement) is not permissible, then to take it as dang is surely injudicious” (Graham 2003: 445, B71). Referring to an assertion in a debate as dang amounts to endorsing it as correct, the winner between two opposing claims only one of which can stand. Dang thus fulfills a central function of truth, namely to express, from the standpoint of the speaker, endorsement of the correctness of an assertion. In these contexts, do such endorsements reflect specifically semantic appraisals? Dang applies to the winning assertion in a dispute concerning which of two contrasting general terms applies to an object. The examples the Mohists give—general terms such as “ox,” “horse,” and “dog”—have no ethical or other normative significance, and asserting them of an object is dang if and only if the object is the kind of thing normally denoted by that term. Evaluation of whether an assertion is dang in such contexts thus seems a purely semantic matter. So it is likely that dang here is indeed a term of semantic appraisal tantamount to “true.”13 In broader philosophical discourse, dang plays other roles beyond semantic evaluation. Besides utterances, a person’s conduct can be dang (fitting, apt).14 Indeed,  Hansen notes this role of dang but suggests dang is nevertheless a term of pragmatic, not semantic, evaluation, which he interprets as “is appropriately predicable of” rather than “is semantically satisfied by” or “is true of” (Hansen 1985: 509, n. 22). Like Robins, I see no difference between Hansen’s suggested interpretation and taking dang to refer to a semantic status (Robins 2010, 46: n. 72). Either way, as Robins points out, a predicate F is dang with respect to a thing a if only if a is F, a relation sufficient for dang to refer to a semantic status. 14  The Xunzi refers to a gentleman’s speech, conduct, and cognition as being “largely dang” (Xunzi 8/57–58) and describes the sage as “dang right and wrong” (Xunzi 8/103)—the point presumably being that the sage acts and speaks in conformity to the norms of right and wrong. Citations to Xunzi give chapter and line numbers in the Harvard-Yenching concordance (Xunzi 1966). 13

6  Truth in Pre-Han Thought

121

even when applied to utterances, in some contexts dang probably has the connotation of endorsing them as socially appropriate, rather than true. A passage in Xunzi describes the gentleman as “correcting his [use of] names and making his [use of] expressions dang in order to clearly express his thoughts” (Xunzi 22/49). Here dang probably concerns not the truth of the gentleman’s assertions, but the suitability of the expressions he uses with respect to facilitating communication. A passage in The Annals of Lü Buwei warns that the explanations and expositions of a gentleman must “dang [coincide with] righteousness” (Knoblock and Riegel 2000: 184, sect. 7/5), suggesting that the moral status of the gentleman’s utterances may be more important than their semantic status. In some contexts, then, dang seems to express semantic appraisals and functions much as “true” does, while in others it may connote a general notion of propriety grounded in rhetorical, prudential, and ethical norms, not only semantic ones. Indeed, outside of the Later Mohist writings, contexts in which dang unambiguously expresses semantic evaluation are uncommon.

4.2  Ran 然 and Shi 是 Examples of semantic appraisal in pre-Han philosophical texts often occur in passages devoted to citing and rejecting opponents’ assertions. Such rejections may claim that the opponent’s statement (yan 言) is “bu ran 不然,” or “not so.” For example, the Xunzi cites an assertion, attributed to Mencius, that “people who study, their nature is good” and responds that “shi bu ran 是不然,” or “this is not so.”15 The ensuing rebuttal adduces distinguishing features of people’s nature (xing 性) and non-spontaneous, intentional activity (wei 偽) to argue that the Mencian claim is mistaken because it conflates the distinction between the two. What people acquire through study is not their nature, Xunzi contends, and therefore that people who study become good does not show their nature is good (Xunzi 23/10–20). The grounds on which Xunzi rejects the claim, then, are norms for distinguishing different kinds of things, namely nature (xing) and non-spontaneous, deliberate activity (wei). According to how ran (so) was commonly understood at the time, such norms are precisely the basis for semantic appraisals. For example, the Later Mohists explain that something is ran with respect to a kind or name when it is similar to a model associated with that kind or name (Graham 2003: 316, A70–71). Assessments of whether things are ran thus rest on norms for distinguishing similar from different kinds of things. By extension, assessments of whether assertions are ran rest on norms for distinguishing the kinds of things mentioned in the assertions. Xunzi’s line of argument against Mencius thus renders it likely that bu ran (not so) in this

 Xunzi 23/10. From the context, the only reasonable interpretation of the pronoun “shi” (this) is that its antecedent is Mencius’s assertion, as the next line of the text clearly uses “shi” again to refer to the assertion, complaining that “this [Mencius’s assertion] does not attain to knowledge of people’s nature.”

15

122

C. Fraser

context invokes semantic norms. Probably, “shi bu ran” here expresses the attitude that Mencius’s claim is not semantically correct—in other words, it is not true. The Mohist “Lesser Selection” uses ran and also shi 是 (this, right) to express semantic assessments of terms asserted of things. On one plausible interpretation, the text uses shi to endorse predication of ming 名 (names), or nominal predication, and ran to endorse predication of ci 辭 (expressions, phrasings), or verbal predication (Robins 2010: 35). For example, the writers refer to utterances of the form “White horses are horses, riding white horses is riding horses” as “shi er ran 是而 然” (“this and so”) (Graham 2003: 485). The implication is that, asserted of white horses, “horse” is shi (right), and, asserted of someone’s riding white horses, “riding horses” is ran (so). In such cases, shi and ran seem best explained as terms of semantic appraisal, since claiming that a predicate is shi or ran when applied to a thing is equivalent to asserting the predicate of that thing (Robins 2010: 45–46). Saying “shi” of “white horses are horses” or “ran” of “riding white horses is riding horses” is thus tantamount to deeming these assertions true. Since the Mohists give a reasonably clear account of the concept of ran, their dialectical writings furnish the basis for a plausible conjecture about how they might have explained truth, had they been interested in formulating a theory of truth. Ran in Classical Chinese is a contraction of “ru zhi 如之” (“like this,” “resemble it”) (Pulleyblank 1995: 10).16 To say something is ran is to say it is like or similar to some antecedently specified thing or has some antecedently specified feature. The Mohists seem to have explained most forms of predication as a matter of asserting that a relevant likeness or similarity obtains between the object of which a term is predicated and paradigmatic referents of that term (Fraser 2013: 13–14; Fraser 2015, sect. 6). Assertions predicating kind terms of things enable us to communicate, they thought, because they show us what the subject of the assertion is like— namely, it is like other, standard referents of the term predicated, with which we are already familiar (Fraser 2013: 11–12; Fraser 2015, sect. 4.2). By extension, to evaluate an assertion as ran—and thus true—is to affirm that what is asserted of things is indeed like those things. “Riding white horses is riding horses” is ran because riding white horses is indeed an action relevantly similar to other actions referred to by “riding horses.” To evaluate an assertion as not ran—and thus false—is to deny that the things referred to are like what the assertion states. “Riding horses is riding oxen” is not ran because riding oxen is a different kind of action from riding horses, and thus the asserted predicate is not like the action referred to. What determines whether two things should count as similar, and thus whether an assertion is true when it predicates of one of them a term normally associated with the other? This question can be regarded as a variant formulation of a pivotal issue in early Chinese philosophy of language, epistemology, and metaphysics: what determines the norms—the proper dao—for drawing distinctions between different kinds of things? In the theoretical framework of early Chinese philosophy,  For example, Xunzi cites a feature in children’s development and says “教使之然 education makes them ran [‘like this’]” (Xunzi 1/4–5). Similarly, Mengzi cites a feature and remarks that “ 物皆然 things are all ran [‘like this’] (Mengzi 1995, 1.7/5/12).

16

6  Truth in Pre-Han Thought

123

what we regard as issues concerning truth can be expected to instead be formulated as issues concerning the dao of language use, specifically how to draw distinctions properly. Suitably elaborated, then, early Chinese thinkers’ different accounts of the practical norms of distinction-drawing underlying the proper use of words could also yield explanations of semantic correctness and thus truth. Although ran (so) and shi (this, right) can indeed be used to express semantic appraisals, contexts in which these terms are unequivocally applied to assertions are relatively uncommon in pre-Han texts. Both terms are more frequently predicated of objects or situations, to describe how they are, to indicate that they are like other objects or situations, or to assert that something previously said also applies to them. Of more than three dozen occurrences of ran in the Mohist canons, for instance, none are unambiguously applied to utterances.17 Typical examples of how ran is used include the explanation of a “minor reason”—a necessary but not sufficient condition—as “having it, [things are] not necessarily so (ran); lacking it, [things are] necessarily not so (ran)” (Graham 2003: 263, A1) and the statement that “how things are so (ran), and how we know it, and how we make others know it need not be the same” (Graham 2003: 359, B9). As to shi, even in logical and semantic contexts it is more often used of objects than of assertions, as when the Mohists explain distinction-drawing debates (bian) as “one side calls it shi (this), one side calls it fei 非 (not-this), and the one that fits (dang) wins” (Graham 2003: 403, B35). The Xunzi regularly uses ran as a term of semantic assessment only in contexts concerned with rebutting opponents’ views18; outside of such contexts examples of ran used this way are unusual. Even within such contexts, ran may be applied to things or persons. Rejecting Song Xing’s doctrine that to be insulted is not disgraceful, for instance, Xunzi contends that everyone from the sage-kings on down follows the same traditional norms in distinguishing honor from disgrace; only Song Xing is not ran (not so), or not like everyone else (Xunzi 18/112). Another example occurs in a passage explaining why Mencius’s assertion that people’s nature is good is not ran. The passage contends that, counterfactually, if people’s nature were indeed good, they would spontaneously be “upright, well-organized, peaceful, and orderly,” but “now [it’s] not so” (jin bu ran 今不然)—in fact, people are not like this (Xunzi 23/39). To sum up, although ran and shi can be used to appraise the semantic correctness—and thus truth—of assertions, they are used this way only infrequently. More often, they are predicated of things, people, or situations. Moreover, the most prominent early Chinese texts that treat philosophy of language hardly address the topic of semantic evaluation using these terms.19 The paucity of attention devoted to them again supports the generalization that truth was not a focus of theoretical interest in early Chinese philosophy.  The examples of semantic appraisal using ran considered above are from the “Lesser Selection” (Mozi book 45), not the canons (books 40–43). 18  See Xunzi books 15, 18, and 23. Even in these contexts, not all instances of ran are unambiguously predicated of utterances rather than of things. 19  Examples include the Mohist canons, the Xunzi “Correct Names” essay, and the essays cited earlier on language use in The Annals of Lü Buwei. 17

124

C. Fraser

4.3  Ke 可 As we saw, one argument proposed against ascribing a concept of truth to pre-Han philosophical discourse is that, rather than assessing whether assertions are true, early Chinese thinkers supposedly evaluated whether utterances are pragmatically assertible, using the term “ke 可” (permissible, acceptable, possible). Throughout the pre-Han literature, ke is regularly used to express assessments of conduct or policies as socially or ethically acceptable or permissible. As it turns out, however, outside of the Later Mohist canons and the Gongsun Longzi 公孫龍子, pre-Han texts rarely use ke to evaluate utterances. Moreover, in the Mohist canons, assessments of whether utterances are ke seem best explained as resting on logical and semantic grounds, not pragmatic ones. Appraising whether utterances are ke is thus intertwined with, not an alternative to, appraising whether they are true. In non-semantic contexts, ke is sometimes used to express possibility, as in “If the south has a limit, then it can (ke) be reached” (Graham 2003: 448, B73). In semantic contexts, when predicated of utterances, it seems to express a related notion of logical-semantic possibility or permissibility. An utterance is ke if it is logically and semantically possible for it to be correct. To be ke, an utterance must comply with logical and semantic norms and be consistent with contextually understood premises. The utterance must be free of contradiction, inconsistency, and other logical and semantic errors, perhaps including practical contradiction with the speaker’s actions. Ke and dang 當 (fitting) are interrelated. One Mohist canon indicates that if yan 言 (statements) are not ke, then “to take them as dang would surely be injudicious” (Graham 2003: 445, B71). This remark suggests that a statement can be dang only if it is also ke—it can fit what it is asserted of only if it is a permissible utterance, one that conforms to relevant logical and semantic norms. Conversely, a statement can probably be ke without being dang in a particular context—it can be permissible (ke) as long as it could potentially fit (dang) in some conceivable context, without necessarily fitting in the actual context. Another canon indicates that when asserted of an object, two contradictory terms, such as “ox” and “non-ox,” cannot both be not-ke, or “impermissible” (Graham 2003: 317, A73). At least one must be ke. Assessment of whether an utterance is ke is thus partly grounded in logical and semantic norms governing the use of contradictory terms, specifically including some version of the principle of excluded middle. According to the Mohists’ conception of a distinction-drawing debate (bian 辯), exactly one of any pair of contradictory terms, such as “ox” and “non-ox,” asserted of an object must fit (dang) it (Graham 2003: 318, A74). Consequently, of two assertions that cannot both be impermissible (not-ke), at least one must fit (dang) its object. Another canon indicates that, having distinguished two opposing terms and their objects, it is ­impermissible (not-ke) to apply one of the terms to the referents of both, as doing so leaves us without any semantic relation between name and object that can be applied consistently (Graham 2003, 446, B72). Using a term this way would be a logical

6  Truth in Pre-Han Thought

125

error, presumably a type of self-contradiction, as well as a semantic error, in that the term would not be used consistently to refer to the same kind of object. As this quick survey of the Later Mohists’ use of ke shows, in contexts addressing the use of words and statements, ke is not primarily a term of pragmatic evaluation. Assessments of whether an utterance is ke appeal to logical norms, such as excluded middle and non-contradiction, and depend partly on semantic norms. They may contribute to semantic appraisals, as they pertain to the conditions under which assertions can fit (dang) their objects. Appraisal in terms of ke thus complements appraisal in terms of truth, and the Mohists’ concern with ke provides no reason for thinking they may not also employ concepts with roles that overlap that of truth.

5  Concluding Remarks Pre-Han philosophical texts employ at least three terms—dang 當, ran 然, and shi 是—that in some contexts are best explained as expressing evaluations of the semantic correctness of asserted words, phrases, and statements. Since these evaluations seem to appeal specifically to semantic norms, they are functionally analogous to appraisals of truth. Pre-Han thinkers can thus defensibly be said to employ terms whose function is similar to “…is true,” to evaluate assertions as true or false, and to have a conception of assertions being correct according to semantic norms, rather than only pragmatic norms. At the same time, however, the interpretive effort required to extract early Chinese concepts of and views on truth from the texts only underscores how marginal issues of truth were in pre-Han philosophy. The emphasis of Chinese philosophy of language and logic lay elsewhere, on the role of names and statements in communicating, guiding action, and preserving social order. The central semantic issue concerned how the use of terms is grounded in practical norms for drawing distinctions between what is similar or different. Logical inquiries mainly investigated norms of term use, grounded primarily in similarity relations. Discussions of inference and argumentation were concerned chiefly with how predication is based on pattern recognition and analogical inference, and not with truth-functional or formally valid inference. Nevertheless, as we have seen, the early Chinese theoretical picture of correct term use as resting on norms for distinguishing similarities and differences could yield explanations of the semantic correctness of assertions which amount to implicit accounts of truth. Pre-Han thought thus exhibits a structural and thematic focus on dao, norms of term use, and the action-guiding role of language while still employing truth-like concepts and yielding explanations of semantic statuses comparable to truth.

126

C. Fraser

References Chmielewski, Janusz. 1966. “Notes on early Chinese Logic (VI).” Rocznik Orientalistyczny 30.1: 31–52. Fraser, Chris. 2012. “Truth in Mohist Dialectics.” Journal of Chinese Philosophy 39.3: 351–368. Fraser, Chris. 2013. “Distinctions, Judgment, and Reasoning in Classical Chinese Thought.” History and Philosophy of Logic 34.1: 1–24. Fraser, Chris. 2015. “Mohist Canons.” In E. Zalta, ed., The Stanford Encyclopedia of Philosophy (Fall 2015 Edition). http://plato.stanford.edu/archives/fall2015/entries/mohist-canons/. (Originally published in 2005.) Fraser, Chris. 2016. “Language and Logic in Xunzi.” In Eric Hutton, ed., Dao Companion to the Philosophy of Xunzi. Dordrecht: Springer. Graham, A.C. 1989. Disputers of the Tao. La Salle: Open Court Press. Graham, A. C. 2003. Later Mohist Logic, Ethics, and Science. Hong Kong: Chinese University Press. Reprint of 1978 edition. Hall, David, and Ames, Roger. 1987. Thinking Through Confucius. Albany: SUNY Press. Hansen, Chad. 1983. Language and Logic in Early China. Ann Arbor: University of Michigan Press. Hansen, Chad. 1985. “Chinese Language, Chinese Philosophy, and ‘Truth’.” Journal of Asian Studies 44.3: 491–519. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought. New York: Oxford University Press. Harbsmeier, Christoph. 1998. Science and Civilisation in China, vol. 7, Part 1: Language and Logic. Cambridge: Cambridge University Press. Knoblock, John, and Riegel, Jeffrey. 2000. The Annals of Lü Buwei. Stanford: Stanford University Press. Leong, Wai Chun. 2015. “The Semantic Concept of Truth in Pre-Han Chinese Philosophy.” Dao: A Journal of Comparative Philosophy 14.1: 55–74. Loy, Hui-Chieh. 2008. “Justification and Debate: Thoughts on Moist Moral Epistemology.” Journal of Chinese Philosophy 35.3: 455–471. McLeod, Alexus. 2015. Theories of Truth in Chinese Philosophy: A Comparative Approach. London: Rowman and Littlefield. Mengzi. 1995. A Concordance to the Mengzi 孟子逐字索引, ed. D.C. Lau, Ho Che Wah and Chen Fong Ching. ICS series. Hong Kong: Commercial Press. Mozi. 1948. A Concordance to Mozi 墨子引得. William Hung, ed. Harvard-Yenching Institute Sinological Index Series, supplement no. 21. Cambridge, MA: Harvard University Press. Munro, Donald. 1969. The Concept of Man in Early China. Stanford: Stanford University Press. Pulleyblank, Edwin. 1995. Outline of Classical Chinese Grammar. Vancouver: UBC Press. Robins, Dan. 2010. “The Later Mohists and Logic.” History and Philosophy of Logic 31.3: 247–285. Robins, Dan. 2012. “Names, Cranes, and the Later Mohists.” Journal of Chinese Philosophy 39.3: 369–385. Roetz, Heiner. 1993. “Validity in Chou Thought.” In H. Link and G. Paul, eds., Epistemological Issues in Classical Chinese Philosophy, Albany: SUNY Press. Saunders, Frank, Jr. 2014. “Semantics Without Truth in Later Mohist Philosophy of Language.” Dao: A Journal of Comparative Philosophy 13.2: 215–229. Van Norden, Bryan. 2007. Virtue Ethics and Consequentialism in Early Chinese Philosophy. New York: Cambridge University Press. Xunzi. 1966. A Concordance to Xunzi 荀子引得. William Hung, ed., Harvard-Yenching Institute Sinological Index Series, supplement no. 22. Cambridge, MA: Harvard University Press.

6  Truth in Pre-Han Thought

127

Chris Fraser  is Associate Professor in the Department of Philosophy at the University of Hong Kong. He is the author of The Philosophy of the Mozi: The First Consequentialists (Columbia University Press, 2016) and numerous scholarly articles and book chapters on Mohism, Confucianism, Daoism, and early Chinese philosophy of language, logic, and epistemology.

Chapter 7

Contradiction Xinyan Jiang

In classical Western logic, “contradiction” refers to “the conjunction of a proposition and its denial” (Honderich 2005: 175). The law of non-contradiction (also known as the law of contradiction) is the most fundamental principle of Aristotelian logic. As Aristotle has famously stated, “It is impossible to be and not to be at the same time” (Metaphysics 996b30-31; Apostle 1966, 42). More specifically, “The same thing cannot at the same time both belong and not belong to the same object and in the same respect” (Metaphysics 1005b19-23; Apostle 1966: 58–59). In other words, an object cannot possess opposite properties at the same time and in the same aspect.1 Such an expression of the law of non-contradiction is “an explicitly modal claim about the incompatibility of opposed properties applying to the same object (with the appropriate provisos)” (Horn 2018). For Aristotle, the law of

1  This is the ontological version of the law of non-contradiction. According to some scholars, “There are arguably three versions of the principle of non-contradiction to be found in Aristotle: an ontological, a doxastic and a semantic version. The first version concerns things that exist in the world, the second is about what we can believe, and the third relates to assertion and truth. The first version (hereafter, simply PNC) is usually taken to be the main version of the principle and it runs as follows: ‘It is impossible for the same thing to belong and not to belong at the same time to the same thing and in the same respect’ (with the appropriate qualifications) (Metaph IV 31005b19–20) … A table can be actually red and potentially not red, but not actually red and actually not red at the same time” (Gottlieb 2019). This version of the law of non-contradiction has been formulated as ¬◊(∃x)(Fx ∧ ¬Fx) (see Wedin 2004). The doxastic version of the law of noncontradiction (also known as its psychological formulation) is stated as “It is impossible for anyone to believe that the same thing is and is not, …” (Metaphysics. 1005b23–25); its semantic version or logical formulation can be summed up as “opposite assertions cannot be true at the same time” (Metaphysics 1011b13–20) (see both Gottlieb 2019 and Horn 2018).

X. Jiang (*) Department of Philosophy, University of Redlands, Redlands, CA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_7

129

130

X. Jiang

non-­contradiction is indemonstrable but self-evident. As long as a person is rational and capable of reasoning, he or she follows the law of non-contradiction. All oppositions are either contradictory opposites (e.g., white and non-white) or contrary opposites (e.g., white and black). The law of non-contradiction applies to both forms of oppositions since it holds that neither contradictories nor contraries may belong to the same object at the same time and in the same respect (Metaphysics 1011b17-19).2 In propositional logic, the law of non-contradiction applies to both contradictory and contrary propositions. According to the law of non-contradiction, two propositions are contraries if they can’t be both true but can be both false (e.g., “this dog is black” and “this dog is white”)3; two propositions are contradictories if they can’t be either both true or both false (e.g., “this dog is black” and “this dog is not black”). Simply put it, according to the law, if P and not-P are opposite properties,4 and s is an arbitrary thing, then a proposition like “s is P and not-P (at the same time and in the same aspect)”, the conjunction of “s is P” and “s is not-P,”5 can never be true.6 In ancient Chinese philosophy, although the law of non-contradiction is acknowledged in some way by some Chinese philosophers such as Han Fei 韓非and Moists,7 the majority of Chinese thinkers seem to even have never paid attention to it, not alone treated it as the most basic principle of logic. Moreover, some paradoxical propositions that well fit “s is P and not-P” are asserted and even taken for granted by some Chinese philosophers. However, the apparent inconsistency between certain Chinese paradoxical propositions and the law of non-contradiction will disappear, once the context of the former is understood. What Aristotle stresses by the law of non-contradiction is that s cannot be P and not-P at the same time and in same aspect, while what some Chinese philosophers are emphasizing by their paradoxical propositions is that s can be P and not-P at the different times and/or in different 2  As Laurence R. Horn puts it, “In accounting for the incompatibility of truth and falsity, LNC [the law of non-contradiction] lies at the heart of Aristotle’s theory of opposition, governing both contradictories and contraries. … LNC applies to both forms of opposition in that neither contradictories nor contraries may belong to the same object at the same time and in the same respect (Metaphysics 1011b17–19)” (Horn 2018). 3  The concept of contrary used there is from classical logic which presupposes existential import. 4  For the sake of simplicity, “not-P” in this chapter is used to represent both contradictory property and contrary property. Given that the law of non-contradiction does apply to both contraries and contradictions and what this chapter is concerned with is their relations to the law of non-contradiction only, using “not-P” to represent both contradictory property and contrary property will not sacrifice any logical consistence. 5  Given that “s is P and not-P” is a logical equivalent of “s is P and s is not-P”, for the sake of convenience, in this chapter I will use “s is P and not-P” as “s is P and s is not-P” wherever the latter is discussed. 6  In predicate logic, “s is P and not-P” or “s is P and s is not-P” can be formulated as follows: “Ps ∧ ¬Ps” (“s” refers to an arbitrary thing, and “P” and “not-P” refer to opposite properties); in propositional logic, it can be formulated as “A ∧ ¬A” (“A” represents “s is P,” and “¬A” represents “s is not-P”). Then, according to the law of non-contradiction, “Ps ∧ ¬Ps” or “A ∧ ¬A” is always false.” 7  Some scholars have argued that the law of non-contradiction has been explicitly recognized and formulated in classical Chinese philosophical writing. For more details, see Cheng 1971: 226–27.

7 Contradiction

131

aspects, even with different measurements in different contexts. Furthermore, contextualized “s is P and not-P” in Chinese philosophy reveals something very ­significant—dialectical relations are real and things always exist as the unity of opposites. Although, from a pure logical point of view, “s cannot be P and not-P” is true, it is an abstraction that does not take the context of the utterance and multidimension of things into consideration. On the contrary, a proposition like “s is P and not-P,” though paradoxical, reflects reality better since dialectical relations present in everything and everywhere. In short, Chinese dialectical thinking not only violates none of laws of Aristotelian formal logic such as that of non-contradiction, but also captures and highlights real dialectical relations in the universe and life that the latter does not. “Contradiction” has been translated into “maodun 矛盾” in Chinese. Ancient Chinese philosophers definitely had the idea of contradiction. Actually, the term “maodun” originated from a fable made by Han Fei. Here is what Han Fei has told: There was a man in the State of Chu who sold shields and lances. He praised his merchandise as follows: “My shields are so strong that nothing can pierce them.” He went on to praise his lances thus: “My lances are so strong, they will pierce anything whatever.” Someone commented: “What if you use one of your lances to pierce one of your shields?” The man was lost for words. “The shields that can’t be pierced by anything” and “the lances that can pierce anything” can’t be established at the same time 不可同世而立” (Han Feizi, “Nan Yi” (韓非子 • 難一): 265).8

Here he talks about two incompatible propositions that can’t be both true although can be both false. Such incompatible propositions are what are called “contraries” in logic. Furthermore, ancient Chinese philosophers such as later Moists also noticed what are called “contradictories” that cannot both be true neither both be false, i.e., one of which must be true. They said: “In debate, when one calls something ‘an ox’ and the other calls it ‘not an ox,’ they contend over the claims which are the contradictories of each other. Their claims cannot both be right, and not being both right, there must be one that is wrong, as, for example, if it were a dog.” (Mo Jing [Moists Canons], “Explanations A” (墨經 • 經說上), Mozi: 209).9 Later Moists have also said that to claim all sayings are absurd is absurd, since the claim itself is included in “all sayings” (Mo Jing, “Canon B” (墨經 • 經下), Mozi: 201). That is, a self-contradictory statement is false. Such textual evidence clearly shows that both Han Fei and Moists believe “s is P” or “s is not-P” but not “s is both P and not-P.” Their views completely coincide with the law of non-contradiction.10

8  Joseph Needham’s and Christoph Harbsmeier’s translation with my minor revision. See Needham and Harbsmeier 1998: 215. 9  My translation, but Fung Yu-lan’s (see Fung 1952: 257) and Joseph Needham’s translation (Needham 1998: 217) are referred. 10  These passages in Han Feizi and Mo Jing have been discussed by many scholars, e.g., Joseph Needham and Christoph Harbsmeier (Harbsmeier 1998: 214–17, 330–31), Fung Yu-lan (Fung 1952: 257), Cheng Chung-ying (Cheng 1965: 201–202), Zhang Chunpo and Zhang Jialong (Zhang 1997: 754–55).

132

X. Jiang

However, there are more ancient Chinese philosophers who seem to endorse “s is P and not-P”. As a matter of fact, even Moists have also acknowledged that it may be the case that s is P in one context while not-P in another context. For instance, in Mo Jing, there is mentioning that the same person can be young and old (at different stages of her life), and a person can be both older and younger brother in different contexts, etc.11 Fung Yu-lan comments on this and says: “General idea seems to be that things can have opposing characteristics, such as much and little, age and youth, white and black, center and sides, maintenance and loss, etc., depending upon the aspect we are considering them from” (Fung 1952: 262). Once the context in which Chinese philosopher claims “s is p and not-P” is made clear, we will see nothing contradicting the law of non-contradiction. Nevertheless, their way to reason is more relational and dialectical. That is why some scholars have classified such logical thinking as dialectical logic (Needham 1956: 199) or correlation logic (Zhang 1946: 182). As far as the School of Names or the Logicians and later Moists are concerned, Joseph Needham has commented as follows: Perhaps the most significant thing about them is that they show an unmistakable tendency towards dialectical rather than Aristotelian logic, expressing it in paradox and antinomy, conscious of entailed contradiction and kinetic reality. In this they strongly reinforced the tendencies which were characteristic of Taoism (Needham 1956: 199).

To explain how ancient Chinese philosophers make “s is P and not-P” logical, we must examine those contexts in which such a paradoxical proposition is uttered. First of all, let’s look at the context for propositions like Hui Shi’s “a white dog is black” (“白狗黑”) or “Each thing is the same as other things and yet each thing is different from other things” (“万物毕同毕異”).12 The first proposition says that a dog is both white and black, while the second one holds that each thing is both different from and same as other things. So, both are “s is P and not-P.” To make the sense of these propositions, we must first understand the Chinese view of the unity of opposites. In Chinese philosophy, it is a widely accepted view that everything consists of two opposite aspects. These opposites are referred as “yin” (陰) and “yang” (陽). Therefore, everything is the unity of yin and yang, i.e., every s is both P (yang) and not-P (yin). It has been said that “yang” originally refers to sunshine or what pertains to sunshine and light; and “yin” refers to the absence of sunshine, i.e., shadow or darkness. In later developments, the yang and yin came to be regarded as two cosmic principles or forces, the yang representing masculinity, activity, heat, brightness, dryness, hardness, etc., and the yin femininity, passivity, cold, darkness, wetness, softness, etc. (Fung 1948: 138). They are opposite but interdependent. They

 See Mo Jing [Moists Canons], “Explanations A,” Mozi: 213.  Hui Shi 惠施 (380–305 BC) was one of the most important members of the School of Names or School of Logicians. He was a good friend of Zhuangzi. His both propositions are recorded in the Zhuangzi. See Zhuangzi, “Tianxia,” (莊子•天下): 362 (“a white dog is black”) and 359 (“Each thing is the same as other things and yet each thing is different from other things”).

11 12

7 Contradiction

133

work together to produce all things in the universe.13 As Zhu Xi朱熹, one of the best known Neo-Confucian (1130–1200) says: “All things in the world can’t be changed by one, but can only by two. This is just as both yin and yang make change and production started”14 (Zhuzi Yulei 朱子語類 Book 98, vol. 7: 2512). The world as a whole consists of yang and yin; there are an infinite number of yang yin pairs in the universe, such as heaven vs. earth, high vs. low, large vs. small, and so on. Everything itself is a unity of opposites. “Generally speaking, yin and yang are two extremes. But within yin, there are both yin and yang; and within yang there are also both yin and yang. The way of Heaven produced men, while the way of Earth produced women. Male is classified as yang, but we cannot say that there is no yin in him; and female is classified as yin, but we cannot say that there is no yang in her” (Zhuzi Yulei 朱子語類 Book 94, vol. 6: 2374). It is relatively easy to see that the universe is made of yin yang pairs. But, how can it be said that every single thing or event also involves yin and yang? Actually, when we carefully reflect on everything that we know, we will see that everything always is two sided. For example, the color of a white dog consists of both white and black components, since it is not purely white (i.e., it is not 100% white), although its whiteness is much more than its blackness. Having a smartphone is a positive thing on the one hand, since it can bring so much convenience in communication and information; but it is also a negative thing on the other hand, since it contributes much to isolation and laziness of some individuals whose obsession of smartphones has minimized their real life interaction with others. Misfortune is a bad thing, but it also consists of both a bad side and a good side. Because its bad side is much stronger than its good side, in general it is a bad thing. However, a bad thing implies a good side, and we can say that misfortune implies good fortune. For a similar reason, we can say that good fortune implies misfortune, although its good side is much stronger than its bad side. “Misfortune is what fortune depends on, while fortune is where misfortune is hiding” (Laozi, ch. 58: 35). The famous Chinese idiom “When the old man on the frontier lost his mare, who could have guessed it was a blessing in disguise” (se weng shi ma, an zhi fei fu塞翁失馬, 安知非福) well illustrates how good fortune and misfortunate imply each other.15 As far as ““Each thing is the same as other things and yet each thing is different from other things” is concerned, it may mean that everything is different from other things in some aspects and same as other things in other aspects. Therefore, when different aspects of a thing are considered together, it can be both

13  Interestingly, Heraclitus (c. 530–470 B.C.) believed something similar to the yin yang, but such an idea did not prevail in ancient Greece in the way the doctrine of the yin yang does in China. 14  From now on, the translations in the chapter are all mine unless otherwise noted. 15  The idiom is derived from a story like this: once upon a time, there was an old man on the frontier. One day he lost his horse. This was a very bad thing for him. But very soon, the lost horse came back home and brought another horse with it. This was very good. Then his son got injured when he rode on one of these two horses. So, good fortune brought misfortune. But, later on young men were called to go to the army and join a war. Since his son was injured and he was not able to be a soldier, his son was safe. Therefore, misfortune again brought the old man good fortune. The story first appears in the Huainanzi, “Renjian Xun” (淮南子•人間訓).

134

X. Jiang

P and not-P without any logical inconsistency. No matter how similar two things or persons are, they must be different in some way. As an aphorism says, “There are not two leaves which are same.” However, if two things are very similar according to our common language, “sameness” between them strongly prevails over “difference.” Similarly, no matter how different two things or persons are, they must have something in common. But, “difference” between them overcomes “sameness.” In either case, “s is both P (sameness) and not-P (difference).” As long as we take both sides of a thing into account, we must admit that the thing is a unity of opposites, and “s is P and not-P” is true. In this sense, any proposition like “s is P” or “s is not-P” is one sided because it is made only from the perspective of one of two opposites of the same thing. Some Chinese philosophers have well acknowledged this. As Zhuangzi莊子, one of the best known Daoists, puts: “If we see things from the point of view of their difference, even our livers and galls are as far from each other as are the states of Chu and Yue.16 If we see things from the point of view of their sameness, all things are one” (Zhuangzi, “De Chong Fu” (莊子• 德充符): 58). Namely, from the point of view of difference, all things are always very different, while from the point of view of sameness, all things are the same. But, from both points of view, we see that all things are the same in some aspects and different in other aspects. Therefore, “s is P and not-P” in this sense shows how a thing is constituted. However, “s is P and not-P” understood in this sense does not mean that all things are indistinguishable, since it presupposes that s possesses P and not-P to different degrees or different s’s have P and not-P in different ways. First of all, although everything consists of two opposites, it cannot have them to the same degree at the same time. In the most cases, one of two opposites in a thing is much stronger than the other at any given time, i.e., one opposite dominates the other. It is the dominant one which determines the principal property of the thing. This is well illustrated by a saying of the most famous Daoist, Laozi. “Eternal non-being is one from which we see minimal being, while eternal being is one from which we see minimal nonbeing” (Laozi, ch.1: 1). He means that what we call eternal being is that which consists of maximal being and minimal nonbeing and what we call eternal nonbeing is that which consists of maximal nonbeing and minimal being.17 We may say that eternal being is different from eternal non-being, since the degree of the sameness between them is minimal. Similarly, we may claim that human beings are similar to each other but very different from trees, because the degree of the sameness among humans is much greater than that of the sameness between humans and trees. Secondly, different things have different ways to possess P and not-P.  For example, although all things are the same in some aspect, they are not the same in the same sense. Moists have given some argument in this aspect. 16  The two feudal domains Chu and Yue (in the period of the Warring States, 482–221 B. C.) were in fact far apart. 17  Since the original sentence is written in classical Chinese, it may be interpreted, and consequently translated, in various ways. The interpretation on that my translation is based here is adopted from Zhang Dainian (1982, 112).

7 Contradiction

135

In the history of Chinese philosophy, Hui Shi’s proposition “Each thing is the same as other things and yet each thing is different from other things” (万物毕同毕 異) was implicitly criticized by Moists. But the main disagreement between them did not lie in whether a thing can have opposite sides but whether all things can possess them in the same way. Moists clearly claim that each thing is made of both difference and sameness. “If there is difference between things, there is sameness between them; and vice versa” (Mozi, “Da Qu” (墨子 • 大取): 247). However, they don’t think that all things can be same and different in the exactly same way and therefore indistinguishable from others. In order to argue against Hui Shi’s view, Moists define four kinds of sameness and difference: Sameness (tong 同): there is that of identity when one thing has two names; there is that of part-and-whole relationship when different things are all parts of the whole body; there is that of coexistence when different things are in the same room but not parts of the same thing; there is that of a generic relation when there is something in common between things (Mo Jing, “Explanations A” (墨經 • 經說上), Mozi: 212).

Similarly, four kinds of difference are as follows: “Difference (yi 異): not being one but two things, not belonging to the same whole, not locating in the same room, and not sharing the common property” (Mo Jing, “Canon A” (墨經 • 經上), Mozi: 194). Furthermore, these have been explained further like this: “To be two things must mean that they are different, not belonging to the same whole must mean that they are not one body, not locating in the same spatial area must mean that they don’t coexist together, not sharing the common property must mean that they are not the same kind” (Mo Jing, “Explanations A” (墨經• 經說上), Mozi: 212).18 According to such understandings of “sameness” and “difference,” in Hui Shi’s proposition “Each thing is the same as other things and yet each thing is different from other things,” the sameness or difference among all things should not be interpreted as the same kind. For Moists, all things are the same in the sense that they share some common properties and co-exist in the same universe, but they are not the same as the one entity or inseparable parts of a whole.19 Similarly, although all things are different in some way, not all things differ in the exactly the same sense. Once different degrees and different kinds of the sameness and difference are taken into consideration, “Each thing is the same as other things and yet each thing is different from other things” does not have to entail that nothing can be distinguish from others. However, historically, Hui Shi seemed to regard that all things are equally different from each other and equally same as others, namely, each thing has both difference and sameness to the same degree. Furthermore, he did not distinguish different kinds of sameness and difference as Moists did. For those reasons, Hui Shi clams that all things are one body and cannot be distinguished from each other. If Hui Shi indeed is guilty of such a mistake, Chinese philosophers in general are not.  For more detailed discussions of the Moist views of sameness and difference, see Zhang (Dainian) 1982: 154–55, Sun 1993: 92–94. 19  For such an interpretation of Moist critique of Hui Shi, see Fung 1948: 126 and Zhang (Dainian) 1982, 156. 18

136

X. Jiang

For ancient Chinese philosophers, the opposites not only coexist inside a thing to different degrees, but their difference in power also constantly changes. That makes everything in eternal flux and eventually change into its opposite. However, none of those opposites that make up a thing can unlimitedly become stronger and stronger. At some point, a reversal will take place. When the stronger opposite reaches its extreme height, it will begin to decline. Namely, when the development of anything brings it to one extreme, a reversal to the other extreme takes place. This is so-called “wu ji bi fan 物極必反.” For example, “When the Sun stands at midday, it begins to set; when the Moon is full, it begins to wane” (“Fenggua: Tuanci” (豐卦• 彖辭) [“Commentaries on Hexagram No. 55”], Zhouyi (周易): 246).20 A white paper at some point will start to become less white and eventually no longer look white. A strong and healthy person at some point will become ill and eventually die. When the stronger opposite has become the weaker one, the nature of a thing will change. Therefore, everything will go toward its opposite. Given that such a reversal is a universal law, “in order to avoid losing something, one must complement with something of its opposite. Thus ‘Appendix III’21 says: ‘The man who keeps danger in mind is one who retains his position. The man who keeps ruin in mind is one who survives’” (Fung 1948: 172). All things are in constant change, all changes are caused by yin and yang, and everything involves its negation. These constitute the unchangeable law of the universe. For the Chinese, this is the most fundamental truth of the universe. Clearly, such Chinese beliefs go beyond what Aristotle has said about the law of non-contradiction, but it is perfectly compatible with that law. This Chinese belief in constant change may explain another type of Chinese paradoxical proposition that apparently seems inconsistent with the law of non-­ contradiction. One of examples of this kind of paradoxical statement is Hui Shi’s “The creature coming into life is the creature ceasing to be alive” (物方生方死).22 In this proposition, “coming into life” and “ceasing to be alive” are contraries. The proposition can be also formatted as “s is P (coming into life) and not-P (ceasing to be alive).” With Chinese belief in constant changes in the mind, it will be easy to understand why one may think that the creature coming into life is the creature ceasing to be alive without contradiction, since the creature coming into life can change into the creature ceasing to be alive. Also, given the constancy of changes, the identity of everything is not still. Therefore, the distinction between the creature coming into life and the creature ceasing to be alive can never be absolutely made. As some scholars have put: “At the deepest level of Chinese philosophical thinking, ‘to be or not to be is not the question because life is a constant passing from one stage of being to another, so that to be is not to be, and not to be is to be” (Peng and Nisbett 743). Nevertheless, this does not mean that two things such as the creature coming into life and the creature ceasing to be alive are the same. They are distinguished by  The translation is from Cary F. Baynes. See Richard Wilhelm and Cary F. Baynes 1967: 670.  This refers to “Xici Zhuan B” (系辭傳•下) in Yijing (Book of Changes). 22  This proposition is recorded in the Zhuangzi, “Tianxia” (莊子•天下). See Zhaungzi, 359. 20 21

7 Contradiction

137

different power relationships between two opposites within each of them. More specifically, within the creature coming into life, its property of coming into life strongly prevails over its property of ceasing to be alive, while within the creature ceasing to be alive its property of ceasing to be alive overwhelmingly dominates its property of coming into life. That is, at time1, s is P, while at time2 s is not-p, but the difference between time1 and time2 is so little that it almost can’t be noticed. What this kind of proposition like “The creature coming into life is the creature ceasing to be alive” emphasizes is that a thing can become its opposite so fast that it looks like two opposite things at same time. As one commentator puts it: “The creature coming into life turns into one ceasing to be alive in a breath: so we say that the creature is ceasing to be alive when it is coming into life (quoted in Zhuangzi Jizhu 莊子集 注 by Sha 1987: 360). Another says: Since things are constantly changing, […] we cannot distinguish between their birth and death despite they actually have birth and death. Like a film, although it is made of many separable pictures, we see a whole fact but not individual pictures of the film since pictures of the film is being moved constantly (Yang 1973: 300). Clearly, in “the creature coming into life is the creature ceasing to be alive,” Hui Shi does not imply that a creature is coming into life and ceasing to be alive at the exactly same time. Therefore, his proposition does not violate the law of non-contradiction. Once understood, it is perfectly compatible with Aristotle’s logic. But, why does Hui Shi claim that the creature coming into life can change into the creature ceasing to be alive “in a breath” so that the former and the latter almost exist simultaneously? Empirically, we feel that most creatures live for some time before they die, and we can distinguish between when they are born and when they are dead. What then does Hui Shi have in mind in regarding the two as essentially simultaneous? Strictly speaking, he does not talk about particular empirical facts. Looking from the cosmological point of view, he holds that all things are “one body.” For him, the universe is like that of Heraclitus. It is a whole and constantly changing, like a constantly flowing current. In such a universe, the creatures born change into the creatures ceasing to be alive, the creatures ceasing to be alive change into the creatures born, and then a new circle begins. The cycle never stops. No matter how long a creature lives, it only has a short life compared to the existence of the universe. Sub specie aeternitatis, all creatures live very short lives. So, Zhuangzi says of the legendary Chinese Peng Zu 彭祖who lived for more than 800 years: “Nobody lived longer than a dead child, yet Peng Zu had an untimely death 莫壽於殇子, 而彭祖為夭” (Zhuangzi, “Qiwulun,” (莊子•齐物論): 26). Any individual creature can only live in a flash in the sense that its life exists only in a moment of the constantly flowing current of the universe. Therefore, from the viewpoint of the whole universe, as soon as a creature is born, it is creasing to be alive. Indeed, it is fair to say that this conclusion expresses an important theme of Chinese philosophy: we should transcend the finite and see things from the point of view of the infinite. Such an idea is developed quite fully in Daoism, especially by Zhuangzi.

138

X. Jiang

The unity of opposites may also be understood in a different way than those discussed above. What has been explained previously is that “s is P and not-P” means that a thing has two opposite properties at the same time yet has them in different degrees, or becomes its opposite at different times. But, a paradoxical proposition like Hui Shi’s “The heavens are as low as the earth, mountains are on the same level as marshes” (天与地卑, 山与澤平)23 can’t merely be understood that way. Here, “the heavens” and “mountains” represent something very high, while “the earth” and “marsh” represent something very low. What the proposition says is that something very high is also something very low. If we let P refer to “highness”, not-P refer to “lowness,” and s refers to “the thing,” we have a proposition “s is P and not-­P,” but what it asserts is that a thing has a property and its opposite at the same time yet has them in different contexts. Here “opposition” is defined by criteria, relative to contexts. A proposition like “The heavens are as low as the earth, mountains are on the same level as marshes” means that a thing is high by one measure for highness relative to one context but low by another measure in second context. More specifically, the mountains in a basin may be higher than marshes nearby, but it is low as those marshes in yet higher land. Here we see two different measures chosen for height. There is no absolute highness and lowness. Whether something is high or low depends on what measure is used and in which context it is judged. Since the same thing can be “defined” quite differently in different contexts, such identity criteria are contextually chosen. As Fung Yu-lan points out: “Everything can be in one sense yang and in another sense yin, according to its relation with other things. For instance, a man is yang in relation to his wife, but yin in relation to his father” (Fung 1948: 169). In different contexts, criteria for yin and yang are different. In one context in which a man is compared to his wife he is yang, but in another context in which he is compared to his father he is yin. Therefore, the man is both  yin and  yang given different contexts. As mentioned early on, even the Moists have a similar view. They say: “A woman who is a daughter of her mother can be the mother of her children, so she can be older and yet younger. … A man can be a younger brother [for his older brother] and older brother [for his younger brother] at the same time (Mo Jing, “Explanations A” (墨經 • 經說上): 213). Similarly, a young man is yang relative to his sister, but yin relative to his father; a woman is yin relative to her husband, but yang relative to her son. In the case of the proposition “The heavens are as low as the earth, mountains are on the same level as marshes,” when mountains in a basin are compared to marshes nearby, they are high, while compared to marshes in higher land, they are low. Therefore, s is P by one measure but not–P by another. s is P and not-P in different contexts in which different criteria for P and not-P are used. This kind of reading of P and not-P clearly does not violate Aristotle’s law of non-contradiction but has not been a concern in Aristotelian logic. This kind of proposition “s is P and not-P” in Chinese philosophy emphasizes the relativity of distinctions between things inasmuch as the same thing can be defined differently in different contexts. Since we

23

 This proposition is recorded in the Zhuangzi, “Tianxia” (莊子•天下). See Zhuangzi: 359.

7 Contradiction

139

usually define a thing in a context relative to its own criterion, s is P does not exclude s is not-P. On the contrary, s is always P and not-P as long as different contexts of s are taken into consideration. “s is P and not-P” as such reminds us that there are no absolute distinctions between things and that contexts should be considered an integral part of identification.24 What has been discussed has demonstrated that Chinese dialectical way of thinking is logically consistent with Aristotle’s law of non-contradiction. Furthermore, the law of non-contradiction and Chinese dialectical thinking can complement each other and both serve our investigation of truth of the world and human life in different ways. However, that does not mean that Chinese philosophers and Aristotle have the same logical concern. After all, the former emphasizes that s is P and not-P in different aspects, different times, and different contexts. For Chinese philosophers, propositions like “s is P and not-P” reflect the natural world and human life better, although Aristotle’s law of non-contraction is true when the identity of a thing is taken out from its context and relations. As it has been rightly pointed out, “According to Chinese belief, the law of noncontradiction of formal logic works only in the realm of concepts and abstraction. Even in these cases, the rejection of conceptual contradiction based on formal logic can be mistaken, because concepts are reflections of things” (Peng and Nisbett 1999: 744). As a matter of fact, some Western thinkers such as Alfred North Whitehead and Joseph Needham have noticed exactly the same point. When commenting on Aristotelian logic, Whitehead says: We can never get away from questions:—How much,—In what proportions,—In what patterns of arrangement with other things. … Aristotelian logic, apart from the guardianship of mathematics, is the fertile matrix of fallacies. It deals with propositional forms only adapted for the expression of high abstractions, the sort of abstractions usual in current conversation, where the presupposed background is ignored (Whitehead 1933: 153).

Needham has given a similar critique: The so-called laws of identity, contradiction, and the excluded middle, according to which X must be either A or not-A, and either B or not-B, were constantly being flouted by the fact that A was palpably turning into not-A as one watched, or else showed an infinite number of gradation between A and not-A, or else indeed was A from some points of view and not-A from others. The natural sciences were always in the position of having to say “it is and yet it isn’t.” (Needham 1956: 201)

“s is P and not-P” is more real than “s is P” or “s is not-P,” given the complexity of reality and the presence of dialectical relations everywhere. Chinese mentality has been strongly influenced by dialectical thinking discussed above. Some recent empirical studies have also confirmed this conclusion. For example, five psychological experiments by Peng and Nisbett have found the following:

 Main ideas in sections II and III have also been discussed in my two published articles, “The Law of Non-Contradiction and Chinese Philosophy” (Jiang 1992) and “Chinese Dialectical Thinking— the Yin Yang Model” (Jiang 2013).

24

140

X. Jiang

Chinese participants preferred dialectical proverbs, whereas American participants preferred nondialectical ones; Chinese participants preferred compromising solutions to conflicts, whereas American participants preferred noncompromising ones; Chinese participants preferred dialectical arguments for a scientific proposition and a spiritual one, whereas American participants preferred logical arguments; Chinese participants moderated their views when confronted with opposing propositions, whereas American participants became more extreme (Peng and Nisbettet 1999: 750).25

Dialectical way of reasoning probably has contributed much to the shaping of Chinese character. First, given their beliefs in the unity of opposites and constant change, on the one hand, the Chinese are said to rarely despair in difficult times. They tend to think that difficulty cannot last very long, just as darkness cannot last very long before the coming of dawn.26 On the other hand, they also say of themselves that they are easily satisfied with an existing situation. Since they believe anything cannot be absolutely bad and anything can change by itself, there is no reason to take a hand in actively changing things. In general, because of the belief that the principle of dialectics is a universal law, the Chinese tend to take changes and reversals as a part of daily life. Second, keeping the relativity of criteria for distinctions in the mind has helped many Chinese prevent from being too arrogant or too desperate. For example, given the relativity of the nature of things, one can never regard himself or herself as the smartest and best person, and therefore one always needs self-improvement. The smartest or best people in one community are not necessarily the smartest and best in another. This idea is at the heart of many Chinese aphorisms such as “however strong you are, there is always someone stronger,” and “there are always mountains beyond the mountain, there are always heavens above the heavens.” They may sound paradoxical to the Western ear but have taken deep root in Chinese minds. Chinese dialectical thinking is rooted in philosophical understanding of the universe and human life. Regarding the reality as the unity of opposites and considering it eternally changing are deeply metaphysical. But, such high and abstract philosophical principles have indeed impacted the way of Chinese thinking and living. Therefore, it is really no exaggeration to say that it is part of Chinese characteristic that philosophical principles serve as guide to action for ordinary people. In this sense, Fung is quite right to say that “The place which philosophy has occupied in Chinese civilization has been comparable to that of religion in other civilizations” (Fung 1948: 1). Acknowledgement  I would like to thank the editor of the volume, Professor Fung Yiu-ming, for his valuable comments on earlier versions of this chapter. My appreciation also goes to Dr. Lawry Finsen and two anonymous reviewers for helpful suggestions. In addition, I am grateful to the University of Redlands for awarding me a faculty research grant during 2015–2016 for working on the chapter.

25 26

 For details of their five psychological experiments, see Peng and Nisbett 1999: 744–749.  For the similar idea, see Fung 1948: 19–20.

7 Contradiction

141

References Early Chinese Texts Han Feizi 韓非子. All references are to chapter titles and page numbers in Wang Xianshen 王先 慎 annot., Han Fei Zi Jijie韓非子集解, included in Zhuzi Jicheng諸子集成, vol. 5, reprint. Beijing: Zhonghua Shuju, 1954. Laozi 老子 or Daode Jing 道德经. All references are to chapter and page numbers in Wang Bi 王弼 annot., Laozi Zhu老子注, included in Zhuzi Jicheng諸子集成, vol. 3, reprint. Beijing: Zhonghua Shuju, 1954. Mozi 墨子. All references are to chapter titles and page numbers in Sun Yirang 孙诒让 ed. Mozi Xiangu墨子闲诂, included in Zhuzi Jicheng諸子集成, vol. 4, reprint. Beijing: Zhonghua Shuju, 1954. Zhuangzi 莊子. All references are to chapter titles and page number in Sha Shaohai 沙少海annot., Zhuangzi Jizhu莊子集注. Guiyang: Guizhou People’s Press, 1987. Zhuzi Yulei朱子語類, vol. 6 and vol. 7, Li Jinde 黎靖德 ed., reprint. Beijing: Zhonghua Shuju, 1986. Yijing易經. All references are to chapter titles and page number in 周易Book of Changes (original text, modern Chinese translation, and English translation combined in one volume), English trans. by James Legge, and modern Chinese trans. and annot. by Qin Ying. Changsha: Hunan Press, 1993.

Other Sources and Secondary Literature Apostle, H.G., ed. and trans. 1966. Aristotle’s Metaphysics. Bloomington: Indiana University Press. Cheng, Chung-ying. 1965. “Inquiry into Classical Chinese Logic.” In Philosophy East and West 15.3–4: 195–216. Cheng, Chung-ying. 1971. “Aspects of Classical Chinese Logic.” In International Philosophical Quarterly 11.2: 213–235. Fung, Yu-lan. 1948. A Short History of Chinese Philosophy. New York: Macmillan. Fung, Yu-lan 1952. History of Chinese Philosophy, vol. 1, 2nd edition. Princeton: Princeton University Press. Gottlieb, Paula. 2019. “Aristotle on Non-contradiction.” In Stanford Encyclopedia of Philosophy, https://plato.stanford.edu/entries/aristotle-noncontradiction/ (accessed on May 3, 2019). Honderich, Ted, ed. 2005. The Oxford Guide to Philosophy. Oxford: Oxford University Press. Horn, Laurence R. 2018. “Contradiction.” In Stanford Encyclopedia of Philosophy, https://plato. stanford.edu/entries/contradiction/ (accessed on May 3, 2019). Jiang, Xinyan. 1992. “The Law of Non-contradiction and Chinese Philosophy.” In History and Philosophy of Logic 13: 1–14. Jiang, Xinyan 2013. “Chinese Dialectical Thinking—The Yin Yang Model.” In Philosophy Compass 8/5: 438–446. Needham, Joseph. 1956. Science and Civilization in China, vol. 2. Cambridge: Cambridge University Press. Needham, Joseph, and Christoph Harbsmeier. 1998. Science and Civilization in China, vol. 7, part 1. Cambridge: Cambridge University Press. Peng, Kaiping, and Richard E.  Nisbettet. 1999. “Culture, Dialectics, and Reasoning About Contradiction.” In American Psychologist 54.9: 741–754. Sun, Zhongyuan孫中原. 1993. Mozi Tonglun (墨子通論). Shenyang: Liaoning Education Press. Wedin, Michael V. 2004. “Aristotle on the Firmness of the Principle of Non-contradiction.” In Phronesis 49: 226–65.

142

X. Jiang

Whitehead, Alfred North. 1933. Adventures of Ideas. New York: The Free Press. Wilhelm, Richard (trans. into German) and Cary F.  Baynes (trans. into English). 1967. The I Ching: Or, Book of Changes, 3rd edition. Princeton: Princeton University Press. Yang, Rongguo楊榮國. 1973. A History of Ancient Chinese Thought (中國古代思想史), 2nd edition. Beijing: People’s Press. Zhang Chunpo, and Zhang Jialong. 1997. “Logic and Language in Chinese Philosophy.” In Brian Carr and Indira Mahalingam, eds., Companion Encyclopedia of Asian Philosophy (562–575). London/New York: Routledge. Zhang, Dainian張岱年. 1982. An Outline of Chinese Philosophy (中國哲學大纲). Beijing: China Social Science Press. Zhang, Dongsun. 1946. Knowledge and Culture (知識與文化). Shanghai: Shangwu Yinshuguan. Jiang, Xinyan  is Professor of Philosophy at the University of Redlands. She was a faculty member of the Department of Philosophy at Peking University before leaving China. She has published in both Chinese and Western philosophy. Besides publishing numerous scholarly articles and several book chapters, she is also the author of two books and the editor of two anthologies. She was the founding chair of the Committee on Asian/Asian American Philosophers and Philosophies of the American Philosophical Association (1998–2002), a member of the Advisory Committee to the Program Committee of the Eastern Division of American Philosophical Association (2002–2005), and Deputy Executive Director (2007–2011) and Treasurer (2002–2011) of the International Society for Chinese Philosophy.

Chapter 8

Analogy

Alexei K. Volkov

1  Introduction One can define analogical inference as the process of making conjectures or conclusions about properties of (an) object(s) on the basis of known properties of other objects. Analogical inference arguably existed in a large number of pre-modern intellectual traditions, including that of early China. Several attempts to identify and to study the “analogical inference” in traditional China have been made in Asia and in the West starting from the early twentieth century. This paper provides a brief summary of these attempts and offers an analysis of one representative case.1 A note about the term “logic” is in order here: it can refer to (1) an “implicit logic” that amounts to socially sanctioned patterns of persuasive discourses, and to (2) an “explicit logic”, i.e. a discipline providing descriptions and/or interpretations (informal as well as formal) of the “implicit logic”. Historically, there were various ways to provide an explicit, and, in some cases, formal description of a given “implicit logic”; for instance, it can be suggested that Aristotelian syllogistics and Stoic logic were both designed as formalizations of the persuasive discourses adopted in pre-Hellenistic Greece, while the logical methods found in some ancient Chinese philosophical and scientific (in particular, mathematical) texts can be identified as cases of “implicit logic”. The lack of ancient Chinese works on “explicit logic” historically was the main difficulty for studies of Chinese logical reasoning; a number of attempts to find Western equivalents for Chinese logical methods were made at an early stage of 1  The pinyin transliteration system adopted in Mainland China and in European sinological works is used throughout this chapter. I use the traditional versions of Chinese characters for all Chinese and Japanese names and book titles.

A. K. Volkov (*) Center for General Education, National Tsing-Hua University, Hsinchu, Taiwan, Republic of China © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_8

143

144

A. K. Volkov

their studies. It can be suggested that these attempts were often unproductive from the outset, because their authors looked for interpretation of analogical inferences in terms of modern formal propositional or first-order predicate logic; below I will discuss examples of interpretation of this kind.

2  Studies of Chinese Analogical Reasoning: Past and Present In the late nineteenth century and in the first half of the twentieth century a number of Chinese scholars made attempts to provide interpretations to presumably “logical” excerpts found in ancient Chinese texts; among them were Sun Yirang 孫詒讓 (1848–1908), Yan Fu 嚴復 (1853–1921), Song Shu 宋恕 (1862–1910), Liu Shipei 劉師培 (1884–1919), Zhang Binglin 章炳麟 (1869–1936), Liang Qichao 梁啟超 (1873–1929), Wang Guowei 王國維 (1877–1927) and Zhang Dongsun 張東蓀 (1886–1973).2 Similar attempts were made by their Japanese counterparts (and, in some cases, precursors)3 Matsumoto Bunzaburō 松本文三郎 (1869–1944), Kanie Yoshimaru 蟹江義丸 (1872–1904) and Kuwaki Genyoku 桑木嚴翼 (1874–1946).4 These authors paid major attention to a comparison of their reconstructions of “Chinese logic” with the Greek (Aristotelian) logical tradition, in particular, syllogistics; the terms and excerpts related to analogical inference were either ignored or (mis)interpreted by them as referring to equivalents of forms of inference dealt with in Aristotelian logic.5 The first book-length publication in a Western language on the history of Chinese logic (Hu 1922) was authored by Hu Shi (Hu Shih) 胡適 (1891–1962).6 In his monograph Hu devoted two chapters to “logical excerpts” found in the Mo zi 墨子.7 In particular, Hu Shi rejected Zhang Binglin’s theory about Moist logic as using syllogisms that the latter advanced in his Guo gu lun heng 國故論衡 (Balance for  For a detailed study of their work see (Kurtz 2011).  See (Kurtz 2011: 319 ff) for more details. 4  For detailed descriptions of their works see, for example, (Kurtz 2011: 277ff) and two papers of Sakade Yoshinobu 坂出祥伸 he refers to on p. 454. Kuwaki, J. Kurtz suggests, can be identified as the actual “discoverer” of Chinese logic, see (Kurtz 2011: 298). 5  For instance, Liang Qichao interpreted the category of analogical inference yuan 援 mentioned in the Mo zi as sorites (Kurtz 2011: 323). On sorites in China see below. 6  It is generally assumed that Hu Shi’s monograph (Hu 1922) was based upon his doctoral dissertation prepared in Columbia University (New York) under directorship of J. Dewey (1859–1952) and completed in 1917; 2 years later Hu published a Chinese version of this work (Hu 1919). For a general description of Hu Shi’s work see (Kurtz 2011: 346–350); on the date of the formal conferral of Hu Shi’s doctoral degree (1927), see (Tang 1977: 56). 7  Chapter 3, “Cause, form, and deduction”, pp.  93–98; Chapter 4 “Induction”, pp.  99–108. The latter chapter is devoted to an analysis of the four forms of “induction” (i.e., analogical inference) found in the Mo zi; for their detailed discussion see (Fraser 2009) as well as the chapter on “Reasoning” in the present volume authored by Yang Wujin. On the textual history of the Mo zi see (Graham 1993). 2 3

8 Analogy

145

Weighing Discourses on National Past) published in 19108; Hu suggested instead that Moists used “[t]he principle of similarity which underlies all inference […].” (Hu 1922: 96) When elaborating on his claim, he, however, stated that “Neo-Mohist9 theory of deduction is […] essentially a theory of correct predication,” that is, a theory that allows to make statements about the properties of objects on the basis of knowledge concerning the classes to which these objects belong (Hu 1922: 97). In Europe one of the earliest attempts to discern Chinese “reasoning by analogy” was made by Paul Masson-Oursel (1882–1956); in his paper (Masson-Oursel 1917) he briefly mentioned the system of metaphors employed in the Shu jing 書經 (Book of Documents) and Shi jing 詩經 (Book of Poems). He identified a strong parallelism between natural and social phenomena found in the former classic (Masson-­ Oursel 1917: 61–62); as for the latter one, he suggested that the poems seemingly devoted to interaction between individuals or scenes of nature actually referred to “moral or political ideas” (Masson-Oursel 1917: 62–63). If his observation of the use of metaphors in poems does not look surprising, his note concerning the metaphoric interpretation of the Book of Documents may deserve a closer attention. However, in other publications Masson-Oursel especially focused on the so-called “Chinese sorites” (Masson-Oursel 1912),10 and cases of analogical inference found in Chinese texts occupied only a marginal position in his writings.11 An important contribution to the study of Chinese analogical reasoning was made by Henri Maspero (1883–1945). Even though his paper (Maspero 1928) is formally devoted to a critical evaluation of the reconstructions and interpretations of the Moist “logical” chapters by Sun Yirang (Sun 1895), Zhang Binglin (Zhang 1910), Liang Qichao (Liang 1920), Hu Shi (Hu 1919, 1922), and Alfred Forke (Forke 1922), Maspero offers a number of observations concerning analogical arguments in the Mo zi and other early Chinese philosophical texts. He begins with a reference to the works of Masson-Oursel on “Chinese sorites” and provides an example of a “sorites” that he himself identified in Chapter (juan 卷) 7 of the Mo zi 墨子 (Maspero 1928: 2); then he passes to the opening paragraph of Chapter 5 of the Mo zi. The original Chinese text reads as follows: 聖人以治天下為事者也, 必知亂之所自起, 焉能治之, 不知亂之所自起, 則不能治。譬 之如醫之攻人之疾者然, 必知疾之所自起, 焉能攻之;不知疾之所自起, 則弗能 攻。(Sun 2007: 99) 8  See (Hu 1922: 97). Hu does not mention Zhang’s attempts to compare Moist logical thought with Buddhist logic which occupied much more prominent place than Aristotelian syllogistics in the latter’s analysis. For a detailed presentation of Zhang’s studies of ancient Chinese logic see (Kurtz 2011: 301–312). 9  The term “Mohist” has been systematically used in literature in English; in this chapter I use the spelling “Moist” instead. 10  A sorites is an inference of the form A1⊃A2, A2⊃A3,…,An-1⊃An ⇒ A1⊃An; here and below “⊃” is the symbol of implication, and “F1, F2, .. Fm ⇒ F″ means “F can be deduced from {F1, …, Fm}”. 11  Masson-Oursel was not the first author who identified some excerpts in ancient Chinese treatises as “sorites”; to the best of my knowledge it was done for the first time some 80 years earlier by Guillaume Pauthier (1801–1873) (Pauthier 1832: 6, n. 2). However, when discussing “Chinese sorites” in (Masson-Oursel 1912), Masson-Oursel does not mention the work of Pauthier.

146

A. K. Volkov

The wise man who has charge of governing the empire should know the cause of disorder before he can put it in order. Unless he knows its cause, he cannot regulate it. It is similar to the problem of a physician who is attending a patient. He has to know the cause of the ailment before he can cure it. Unless he knows its cause he cannot cure it.12

Maspero’s interpretation of this excerpt reads as follows: I. Principal inference: 1. Conclusion: “Since governing the world well is the duty of the Saint, he has to know the origin of disorder.” 2. The middle element of the sorites not found in the conclusion: “and then he will be able to get rid of the disorder.” 3. Negative proposition: “If he did not know the origin of the disorder, he would not be able to get rid of it.” 4. Comparison: “It is exactly like a medical doctor”. II. Secondary inference concerning the comparison: 5. Conclusion: “(Since) he has to fight against a disease, he has to know the origin of the disease.” 6. The middle element of the sorites not found in the conclusion: “and then he will be able to fight against it.” 7. Negative proposition: “If he did not know the origin of it, he would not be able to fight against it.”13

Maspero thus subdivides the excerpt into two parallel “inferences”, the first of which features the actions of an ideal ruler, and the second, those of a (skillful) medical doctor; each of these inferences he identifies as “sorites”.14 He does not explain clearly how exactly these two “sorites” are supposed to be interpreted; as we shall see below, his interest is elsewhere, and he provides this example only to mention that if we follow Masson-Oursel’s approach, we also might consider these two “inferences” as two (interconnected) sorites.15

12 13

 Translation cited from (Mei 1929: 78).  The original French text reads as follows (Maspero 1928: 3): « I. Raisonnement principal:

1. Conclusion: “Comme c’est l’affaire du Saint de bien gouverner le monde, il doit connaître l’origine du désordre”. 2. Elément moyen du sorite non contenu dans la conclusion: “et alors il pourra écarter le désordre”. 3. Proposition négative: “S’il ignorait l’origine du désordre, il ne pourrait pas l’écarter”. 4. Comparaison: “C’est juste comme le médecin”. II. Raisonnement secondaire sur la comparaison: 5. Conclusion: “(Comme) il doit combattre la maladie, il doit connaître l’origine de la maladie”. 6. Elément moyen du sorite non contenu dans la conclusion: “et alors il pourra la combattre”. 7. Proposition négative: “S’il n’en connaissait pas l’origine, il ne pourrait pas la combattre”. » (Here and everywhere is this chapter translations into English are mine, unless stated otherwise.)  I do not discuss here whether this interpretation suggested by Maspero reflects adequately enough the subtleties of the original text. 15  Technically, since each inference contains only two premises and a conclusion, we are dealing 14

147

8 Analogy

If we note “to govern the world well” as A, “to know the origin of disorder” as B, and “to be able to get rid of the disorder” as C, then the first sorites identified (but not explicitly presented) by Maspero is composed of the “Negative proposition” (¬B ⊃ ¬C), “Middle element” (⊃ C), and “Conclusion” (A ⊃ B).16 The “middle element”, (⊃ C), syntactically is incomplete (the implication does not have an antecedent); to be a correct inference, the “middle element” should be (A ⊃ C), i.e., “if the Saint governs the world well, then the Saint will be able to get rid of the disorder”, and not just (⊃ C). The reconstructed sorites then will be a (logically valid) inference of (A ⊃ B) on the basis of (¬B ⊃ ¬C) (equivalent to (C ⊃ B)) and (A ⊃ C).17 This interpretation thus may seem plausible, but it does not explain why the excerpt contains the “secondary inference” concerning a medical doctor fighting against the diseases; did the ancient Chinese author provide it as a mere example? The inference concerning the doctor has exactly the same structure, and if one notes “to fight against the disease” as A’, “to know the origin of the disease” as B′, and “to be able to fight against the disease” as C′, then the conclusion (A’ ⊃ B′) can again be obtained on the basis of (¬B′ ⊃ ¬C′) and ([A’] ⊃ C′). Maspero opens a discussion of the “analogical reasoning” in the Mo zi and devotes a large section of his paper to interpretations of the neo-Moist chapter “Xiao qu” 小取 offered by Zhang Binglin, Liang Qichao, Hu Shi, and Alfred Forke (Maspero 1928: 5–36). When dealing with the patterns of reasoning listed in this chapter of the Mo zi,18 he suggests that one of them, mou 侔, can be exemplified with the aforementioned reasoning about a Saint (a sage ruler) and a physician; this time Maspero identified the parallelism between these two as follows (Maspero 1928: 19): The saint Since his duty is to govern well, he should know the reasons of the disorder, and then he is able to get rid of the disorder

The physician Since his duty is to heal well, he should know the reasons of the illness, and then he is able to get rid of the illness

The interpretation is now changed: if in the opening part of the paper the statement about physician was found logically sound yet superfluous, now the formal logical validity of both inferences is no longer Maspero’s concern, and the statement about physician starts playing a vital role: (1) it deals with matters supposedly well-known to the reader, and (2) provides a convincing example supporting the statement about the ideal ruler. At the end of his discussion of the “Xiao qu” section, Maspero arrives at the following conclusion: here with syllogisms, even though not Aristotelian ones, and not with just sorites. 16  Here and below I use the symbol “¬” for negation. 17  It remains unclear what was the definition of sorites used by Maspero (and, ultimately, by Masson-Oursel). The contemporaneous French manuals on logic provide rather informal and slightly diverging definitions; see, for instance, (Bain 1875: 311–312; Liard 1892: 51–52; de Fornel de La Laurencie 1906: 62–63). 18  Once again, see Yang Wujin’s chapter of the present volume for more details.

148

A. K. Volkov

To conclude, an example is the most frequently the center of the discussion: one may accept it or refute it, but it is always the example which is discussed, and not directly the proposition to be confirmed. As for the reasoning per se which is supposed to lead to this proposition, it is almost never presented. To replace it is the duty of the example, and this system [of argumentation], not rigorous from the viewpoint of Logic, is excellent from the viewpoint of the art of discussion, because an example is striking, and also because it allows to transfer a discussion of a complex question to another, simpler [matter] known to everybody. As I have said already, it is the undisputable self-evidence of the proposition confirmed with the example which, unless the adversary refutes it by arguing that it is not well applied, extends to the proposition under discussion to make it self-evident. One should add that if we take a less formal standpoint, then this combination, a confirmation and an example, form a kind of rudimentary reasoning which, properly speaking, is neither inductive nor deductive, but is analogical, drawing all its power from making closer to each other the analogies between the confirmed proposition and the example.19

Unfortunately, Maspero’s insightful observations were not duly appreciated by later authors. Up to mid-1960s, as Cheng Chung-Ying noticed, the work of Maspero (as well as those of Masson-Oursel and Forke) “succeeded in provoking very little interest or effort in the same direction of comparative studies” (Cheng 1965: 197, n. 11).20 Maspero’s path-breaking paper remained unmentioned even later, in (Graham 1978), and, surprisingly, in (Reding 1986, 2004). Harbsmeier mentions Maspero only as having “made a first attempt at a coherent philological interpretation of Mohist logic in any Western language” (Harbsmeier 1998: 327) without any further discussions of this “attempt”. Chmielewski in his essay review especially devoted to analogical reasoning (Chmielewski 1979) failed to mention Maspero’s interpretation; this is rather surprising given that his “Notes” (1962–1969) contained numerous references to Maspero’s paper of 1928. The latter author, Janusz Chmielewski (1916–1998) left his opus magnus on Chinese logic (Chmielewski 1962–1969; repr. 2009) unfinished; the published sections contain a number of paragraphs related to analogy and textual parallelism in

 “En résumé, l’exemple est le plus souvent le centre de la discussion: on l’accepte ou on le réfute, mais c’est toujours lui qu’on discute et non directement la proposition affirmée. Quant au raisonnement proprement dit dont cette proposition peut être considérée comme la conclusion, il n’est presque jamais exprimé. C’est l’exemple qui est chargé de le remplacer; et ce système qui, au point de vue de la Logique, manque de rigueur, est excellent au point de vue de l’art de la discussion, car l’exemple frappe, et d’autre part il permet de reporter la discussion d’une question compliquée à une autre plus simple et que tout le monde connait. Comme je l’ai déjà dit, c’est l’évidence indiscutable de la proposition affirmée dans l’exemple qui, si l’adversaire ne peut réfuter l’exemple en montrant qu’il ne s’applique pas bien, se communiquera à la proposition discutée, en sorte qu’elle deviendra évidente. I1 faut ajouter que, en se plaçant à un point de vue moins formel, cet ensemble, affirmation et exemple, constitue une sorte de raisonnement rudimentaire qui proprement n’est ni inductif ni déductif, mais est analogique, toute sa force reposant dans le rapprochement des analogies entre la proposition affirmée et 1’exemple.” (Maspero 1927: 35–36) 20  Apparently, Cheng himself did not value much Maspero’s interpretations of analogical reasoning, since he praised only “comparative studies” of the French scholar. In his comment on Maspero’s paper of 1928 Cheng wrote “Maspero’s own suggestions are highly open to criticism”, but did not specify what exactly did not satisfy him. 19

8 Analogy

149

classical Chinese texts.21 However, the most consolidated (yet very brief) presentation of his interpretation of analogical reasoning is found in his essay review of John S.  Cikoski’s paper (Cikoski 1975).22 Here Chmielewski depicts three “cases,” or categories, of “analogical thinking” he identified in ancient Chinese texts: (1) “purely literary” analogies that may have been used, for example, in poetry, and are devoid of any particular logical bearing (Chmielewski calls them “a-logical”); (2) “pseudo-logical” analogies that may have been used for persuasion, yet are “unsound from the logical point of view,” and finally (3) “those which derive in some way from ‘analogical thinking’ and which at the same time are instances of logically valid arguments […].”23 The case analyzed by Maspero, being a logically sound inference, apparently falls into the third category; another case of analogy of presumably this type studied by J.-P. Reding is presented below. Chmielewski states that he is particularly interested only in instances of logically valid reasoning (that is, inferences of the third category), and not in analogies of the first and the second categories; he acknowledges, however, that they may have a certain heuristic value. J.-P.  Reding’s paper (Reding 1986) and a chapter of his book (Reding 2004) largely based upon it provide a discussion of several cases of “analogical reasoning” in Chinese texts. The author begins with drawing a line between “illustrative analogies” used to convey certain ideas metaphorically, on one hand, and “argumentative analogies,” on the other; both of them, he claims, are different from ancient Greek “heuristic analogy.” (Reding 1986: 41–42) One can suggest that Reding combined “first and second categories of analogical thinking” identified by Chmielewski to form his class of “illustrative analogies”, while his “argumentative analogies” coincide with the “third category” of Chmielewski.24 Each argumentative analogy consists of two parts referred to by the author as “exposition” (an example) and its “application” (the statement supposed to be proved with the analogy).25 In 1986 Reding expressed certain caution when writing: “I will not hesitate to say that virtually every logical structure is expressible by means of an argumentative analogy, although, in practice, there may be limits” (Reding 1986: 44) but in 2004 he already claimed that “every argumentative analogy embodies a rule of inference.” (Reding 2004: 37) He provided several examples demonstrating logical validity of the “argumentative analogies” that he identified in pre-imperial Chinese

 See, for instance, (Chmielewski 1965a; 1969) reprinted in (Chmielewski 2009: 227–251, 317– 337), respectively. 22  See (Chmielewski 1979; repr. 2009: 339–352). The presentation of Chmielewski’s own vision of “analogical reasoning” is found on pages 71–78 of the original paper reprinted in (Chmielewski 2009: 345–342). 23  See (Chmielewski 1979: 76–77; 2009: 350–351). 24  In (Reding 1986) the author does not refer to (Chmielewski 1979), while in (Reding 2004) his reference to this work (Reding 2004: 31, n. 7) is not related to the “three categories” suggested by the Polish author. 25  (Reding 1986: 42; 2004: 33). 21

150

A. K. Volkov

texts. For instance, he discusses the following excerpt originating from the Meng zi 孟子26: [CYK1] 淳于髡曰:「男女授受不親, 禮與?」 [MZ1] 孟子曰:「禮也。」 [CYK2] 曰:「嫂溺則援之以手乎?」 [MZ2] 曰:「嫂溺不援, 是豺狼也。男女授受不親, 禮也; 嫂溺援之以手者, 權也。」 [CYK3] 曰:「今天下溺矣, 夫子之不援, 何也?」 [MZ3] 曰:「天下溺, 援之以道; 嫂溺, 援之以手。子欲手援天下乎?」 Ch’un-yü K’un [Chunyu Kun] said,27 “Is it prescribed by the rites that, in giving and receiving, man and woman should not touch each other?” “It is”, said Mencius. “When one’s sister-in-law is drowning, does one stretch out a hand to help her?”28 “Not to help a sister-in-law who is drowning is to be a brute.29 It is prescribed by the rites that, in giving and receiving, man and woman should not touch each other, but in stretching out a helping hand to the drowning sister-in-law one uses one’s discretion.”30 “Now the Empire31 is drowning. Why do you not help it?” “When the Empire is drowning, one helps it with the Way; when a sister-in-law is drowning, one helps her with one’s hand. Would you have me help the Empire with my hand?”32

The analogy advanced in this excerpt, according to Reding, can be represented with the Table 8.1.33 According to Reding’s interpretation, the part “Exposition” of the dialog of Mencius and Chunyu Kun amounts to the following set of four propositions: [1] (P ⊃ Q) (that is, if the rites are followed, man and woman should not touch each other), [2] ((R&¬S)  ⊃  T) (if a sister-in-law is drowning and not rescued by her brother-in-law, he is a brute/wolf),34 [3] (S ≡ ¬Q) (the sister in law is rescued if and

 The tags “CYKn” and “MZn” will be used below to refer to parts of this dialog.  J. Legge informs that the biography of Chunyu Kun, “a famous sophist”, is found in chapter 126 of the Shi ji 史記. (Legge 1875: 246) 28  The drowning person is a relative and not just any female. This element of the story usually is not discussed by modern authors. 29  J. Legge is probably closer to the original meaning of 豺狼 chai lang (lit. “ravenous wolves”/“jackals and wolves”, a popular metaphor for wicked/vicious individuals, often unprincipled/cruel rulers) when rendering this phrase as follows: “He who would not [so] rescue his drowning sister-in-law would be a wolf” (Legge 1875: 246). 30  Legge’s translation reads: “when a sister-in-law is drowning, to rescue her with the hand is a peculiar exigency.” (ibid.). 31  “Empire” can hardly be a good rendering of tian xia 天下 (“All-under-heaven”, the inhabited world) because Meng Ke lived prior to the establishment of the first Chinese Empire in 221 BC. Legge (ibid.) suggests “Kingdom”, which is not a good solution either since it limits the statements of the interlocutors to only one kingdom. 32  (Reding 1986: 42–43; 2004: 33). In both cases Reding quotes D.C. Lau’s translation (Lau 1983: 124–125). 33  (Reding 1986: 43; 2004: 33). 34  Here and below “&” is the symbol of conjunction. 26 27

8 Analogy

151

Table 8.1  Mencius’ analogy as analyzed by Reding Exposition the rites a constitutive principle of the rites (man and woman should not touch each other while giving and receiving) the sister-in-law is drowning the sister-in-law is rescued to be a brute to use one’s discretion, i.e. resort to weighing relative benefit and harm

Terms Application P the Way Q a constitutive principle of the Way (not making a compromise, for example) R the Empire is ‘drowning’ S the Empire is saved T to be a brute U resort to weighing

only if the “non-touch” rule is breached),35 and [4] ((¬Q & S) ⊃ U) (“If one seizes the hand of his sister-in-law to rescue her, then he should resort to ‘weighing’ (of relative benefit and harm)”). Reding states that the (presumed) Chunyu Kun’s conclusion [5] ((R&¬U) ⊃ T), that is, “if the Empire is drowning and you do not resort to weighing (deviate from your moral principles and make a compromise), then you are a brute” logically follows from the four propositions [1–4] listed above.36 But does this symbolic interpretation reflect the actual meaning of the dialog? The intention of Chunyu Kun is apparently not limited to a sophisticated demonstration that his venerable opponent is a brute (or, even worse, a ravenous wolf), as Reding suggests; he is apparently asking Meng Ke why the latter does not act as Chunyu Kun deem appropriate under the given (apparently, worsening) political/ social circumstances (“All-under-heaven is drowning”), even if these actions would go against the established rules of social behavior and conventional moral values. The reply of Meng zi is a small masterpiece of political sophistry and logical skill: he indicates that the “hand” introduced by his interlocutor in the context of rescuing one’s sister-in-law is inapplicable in the context of salvation of the world, and that “Dao” should be used instead. The example discussed above is crucial for understanding of the methods of argumentation based on analogical transfer, and this is why I will continue discussing it in the two subsequent sections; firstly, I will summarize its interpretations offered by several authors, from Zhu Xi 朱熹 (1130–1200) to modern historians of Chinese philosophy, and secondly, I will offer an interpretation which may shed light on a number of other cases of argumentation by analogy in pre-modern China.

 Here and below “≡” is the symbol of equivalence: X≡Y means (X⊃Y) & (Y⊃X).  See (Reding 2004: 34). Reding does not provide an actual inference of [5] from [1–4], but instead proves that [1]&[2]&[3]&[4] ⊃ [5] is a tautology. The inference would not be too difficult to obtain though: (¬U ⊃ ¬S) follows from [3] and [4], and it implies (R & ¬U ⊃ R & ¬S); then [2] can be used to obtain T, which means that [2],[3],[4],(R & ¬U) ⇒ T and therefore [2],[3],[4] ⇒ (R & ¬U ⊃ T) (interestingly enough, premise [1], seemingly the most important one, is not used in this inference). 35 36

152

A. K. Volkov

3  Interpretations of the Dialog of Meng zi and Chunyu Kun J. Legge quotes an excerpt of Zhu Xi’s (1130–1200) commentary on this dialogue and provides its translation: 言天下溺。惟道可以救之。非若嫂溺可手援也。今子欲援天下。乃欲使我枉道。求 合則先失其所以援之之具矣。是欲使我手援天下乎。37 The drowning kingdom can be rescued only by right principles,38 the case is different from that of a drowning sister-in-law who can be rescued with the hand. Now you, wishing to rescue the kingdom, would have me, in violation of right principles, seek alliance with the princes, and so begin by losing the means wherewith it might be rescued, − do you wish to make me rescue the kingdom with the hand?” (Legge 1875: 246–247).

Legge adds: “I do not see the point of the last question.” The last question, however, does make sense, if the comment of Zhu Xi is understood slightly differently, namely, [Meng zi] means [the following]: “[If] ‘all-under-Heaven’ is drowning, only the Dao [= correct teaching/way of action] can be the means to save it. [This situation] is not similar to the [case of] sister-in-law’s drowning [who] can be saved with hand(s). Now you want [me] to save ‘all-under-heaven’, and thus want to make me bend the/[my] Dao [i.e., modify my/ correct teaching/way of action]. [If I do so and] try to adjust [the Dao], then the [very] tool by means of which [‘all-under-haven’ can] be saved will be already lost! [When claiming that your analogy] is correct, [do you] want to make me save ‘all-under-heaven’ [with bare] hands”?!

That is, if the similarity with the case of drowning woman is accepted, it would mean that Meng zi should “bend his Dao”, i.e. alter what he believed was the right teaching/way of action, and thus lose his most powerful tool right away. “Saving the world with [bare] hands” then would be the only option left. One cannot be sure whether Zhu Xi fully understood the logical subtleties of Meng zi’s argument, but he certainly found an elegant way to interpret Meng zi’s mention of hand(s) in the last phrase. D.C. Lau devoted a paper to Meng zi’s use of analogy including an extensive discussion of the dialog of Meng Ke and Chunyu Kun (Lau 1963). Lau identifies the analogy used by Meng zi as follows: the opposition of “following rules about interactions of males and females vs formally unlawful salvation of a woman with hand” is similar to that of “following the world as it is (actually, taking an office in a corrupted government) vs salvation of the world with the Dao.” According to his understanding, in this dialog a salvation of the world with the Dao is allegorically equated to a conventionally unjustified action, “a salvation of a woman done against the rules”. Lau does not specify how exactly he understands this “Dao”, but his remarks on p. 180 can make one believe that he thinks of it as of the perfect social

 Meng zi ji zhu 孟子集注 (Meng zi with collected commentaries), 四庫全書薈要 (The Essentials of the Complete Collection of Books of the Four Treasuries), vol. 3024, juan 4, p. 11b. 38  Legge renders dao as “right principles”. 37

8 Analogy

153

order. The main problem he faces in his interpretation now apparently stems from the contradiction between his understanding of the Dao as the conventional moral/ social order, on the one hand, and the formal unlawfulness of saving a sister-in-law with hand as a metaphorical representation of it, on the other. Lau tries several explanatory strategies to deal with this paradoxical situation and ends his discussion abruptly in concluding that “Mencius shows clearly that in his view there can be no justification for compromising on the Way at all.” (Lau 1963: 182). The final phrase of the dialog thus remains unexplained, and the whole interpretation looks rather problematic. Lau’s difficulties with this dialog did not pass unnoticed by H.  Wilhelm (1905–1990) who, in his review of Lau’s translation of the Meng zi (1970), especially focused on it: The first question Meng-tzu [Meng zi] fends off rather skilfully by invoking the principle of ch’üan [quan] 權 (weighing of circumstances), an eminently Confucianist principle, a principle, however, to which Shun-yü K’un [Chunyu Kun] could not object. But as ch’üan [quan] means also power, Meng-tzu gives to Shun-yü K’un [Chunyu Kun] the handle for his second question, and when Meng-tzu [Meng zi] again counters with a principle, he runs straight into Shun-yü K’un [Chunyu Kun]’s snare. Thus the conversation ends with an exasperated rather than a victorious rhetorical question. For me, the message of this story is not that Meng-tzu [Meng zi] won another debate, but that Meng-tzu [Meng zi] was a man of courage who stood up for his principles even when this lost him a debate. (Wilhelm 1973: 489)

Wilhelm thus does not think that Meng zi won the debate; according to him, the last line of Meng zi is “an exasperated rather than a victorious rhetorical question.” However, the very fact that the dialog was retained by the compiler(s) of the treatise suggests that it was considered a victory of Meng Ke, at least by the contemporaneous readers. Moreover, the interpretations provided below may make one suggest that Wilhelm’s judgment of the debate as lost by Meng zi was itself problematic. In his monograph of 1992 Chad Hansen estimated Meng Ke’s analogical arguments as “inept”; he quotes D.C. Lau’s description (Lau 1963) of the impression that these arguments might make on the reader: I take the view that D. C. Lau characterizes as ‘not unusual for a reader of The Mencius’ that Mencius ‘indulged consistently in what appears to be pointless argument [and] that his opponents were always effectively silenced by non sequiturs.’ […] He deploys words almost randomly in facing his philosophical opponents. Mencius […] regards argument as wordplay that he has to endure from people of lesser cultivation. His only consistent purpose in argument is to fend off challenges, frequently with obfuscation and verbal smoke screens. To say, as his defenders do, that he depends on a weak form of argument by analogy is already to give him too much credit. I find his analogies atrociously inept and unconvincing. (Hansen 1992: 188)

As we remember, in his paper Lau did try to provide a sensible interpretation to Meng zi’s analogical arguments, and concluded that. […] Mencius was, indeed, a very skilful user of this method [of analogy], who never failed to throw light on philosophical issues that were discussed. This is an impression somewhat different from the ineffective debater that he is sometimes made out to be. (Lau 1963: 194)

154

A. K. Volkov

The position of Hansen is thus radically opposite to that of Lau. At least theoretically, one cannot completely exclude that Meng Ke’s analogical arguments were “verbal smoke screens,” but such a statement itself would have needed a rather convincing proof. It is possible that Hansen’s stand was provoked by problematic or outright flawed reconstructions offered by his precursors,39 but even if it were so, a sensible interpretation would immediately invalidate Hansen’s claim. Indeed, there exist at least two arguably plausible interpretations (if we don’t count that of Zhu Xi), offered by Liu Yameng and Chris Fraser. According to Liu Yameng, Meng zi’s strategy consists of “zeroing in on the structuring metaphor or analogy of his opponent and refuting the opponent’s point by problematizing and invalidating this very metaphor” (Liu 1996: 41): “He [= Meng zi] stresses, with his concluding rhetorical question in particular, a lack of analogous similarity between two sets of terms: the assumption of a public office by Mencius the individual /a world in crisis; and a person stretching out his hand toward a body of water /his sister-in-law struggling in the water. […]If ‘Mencius taking office’ is thus not to ‘the world in upheavals’ as ‘an individual lending a hand’ is to ‘someone in danger of getting drowned,’ or if the right analogy should rather be ‘dao is to the chaotic world’ as ‘a helping hand is to a drowning person,’ as Mencius’ response implies, Chun-yu Kun loses his case.” (Liu 1996: 42).

Liu praises Meng Ke for “technically sound strategy that demonstrates his argumentative prowess and rhetorical savvy” (Liu 1996: 43); he disagrees with A. Waley (Waley 1939) and D.C.  Lau (Lau 1963) who found Mencius’ argument “a very cheap debating point” which was “totally irrelevant” (Liu 1996: 55–56, n. 24). In his unpublished dissertation Christopher Fraser makes the following remark concerning the dialog of Meng zi and Chunyu Kun (Fraser 1999: 273–275): In his rejoinder, Mengzi denies the similarity between the proposed standard and the present case.40 Temporarily nullifying ritual to save a woman’s life is not an appropriate standard for how to handle a crisis of bad government.

However, Fraser also suggests that Meng zi did not deny the similarity completely, but only indicated that the suggested transfer should have been modified: Interestingly, Mengzi makes this point not by rejecting the drowning emergency as a standard altogether, but by claiming that in fact his present conduct is the genuinely analogous and appropriate response to the political circumstances. We should indeed try to save the world, just as we would try to save the sister-in-law, but we should do so by following ritual and dao, not by frantically extending a hand to unworthy rulers.

Fraser’s interpretation is certainly plausible and his treatment of the embarrassing “hand” is elegant indeed; the question is, to what degree it corresponds to the intended meaning of this metaphor.

39  Interestingly enough, neither Reding’s publications (Reding 1985, 1986) nor (Maspero 1928) were referred to in (Hansen 1992). 40  “Standard” is a technical term introduced by Fraser to discuss analogical reasoning conducted by a transfer from one “standard” (or “model”) situation to a situation under consideration.

8 Analogy

155

4  Meng Ke’s Analogical Argument: An Interpretation Let us return to the dialog. Until [CYK3], Meng zi apparently was not clear what Chunyu Kun was aiming at, and explained in detail the “emergency/contingency measures” in the situation presented by the latter. For sure, if in [MZ2] Meng zi had said something like “according to rules, one should not extend his hand to the drowning sister-in-law”, Chunyu Kun would have had a way to accuse him of immoral behavior (for instance, of not helping a family member). Metaphor introduced by Chunyu Kun in [CYK3] was supposed to take Meng zi off guard: Meng zi could not know in advance that Chunyu Kun was going to compare the d­ rowning woman with the world, and that everything he said earlier would be interpreted in terms of social actions. His strategy now is (a) to modify the analogical transfer suggested by Chunyu Kun in order to make it suitable for his own stand, and (b) to show that transfer suggested by his opponent is flawed, inacceptable, or at least incomplete. To address the former issue, he establishes an analogy that even his opponent probably would find difficult to reject: the world has to be saved with the correct Dao as a drowning person has to be saved with a hand. Now, how he can invalidate the hidden accusation of his opponent? He uses an element, namely the “hand,” which is not present in [CYK3] but was already mentioned in [CYK2] and [MZ2]. “Hand” is indeed the key item in the situation used by Chunyu Kun as example: it can be used for salvation of the drowning woman, and at the same time it is the element representing a transgression of social norms. In [CYK3] Chunyu Kun is focused only on “saving” (the “pulling up the world” is a metaphor itself, and both interlocutors understand what unmentioned actions they are talking about), while Meng zi focuses upon the means of salvation, the hand. His statement [MZ3] shows that he accepts that the world is to be saved, that is, he accepts the analogical transfer suggested by Chunyu Kun in general, but he does not accepts the way in which it should be done. Transferring the “hand” from one domain to another, from the discussion about drowning female into discussion about the salvation of the world would be logically incorrect, and when mentioning “saving the world with a hand” Meng zi thus refers to the inappropriate means in general. Meng zi’s message can be understood at least in two ways. First, when constructing a demonstration based on analogy, one should not transfer objects from the domain of “exposition” (to which the “hand” in this case originally belongs) to the domain of “application”, and hence “saving the world with one’s hand” is formally inappropriate. The illicit transfer of the “hand” from one domain to another may suggest that Chunyu Kun does the same when transferring a situation from one domain (“transgressing the established rules when saving a woman”) to another (“transgressing the established rules when saving the world”); the mention of the hand refers to an incorrect analogical transfer, and suggests that other elements of the transfer proposed by Chunyu Kun are equally wrong. The hand thus becomes a reference to the incorrect analogical inference, and mentioning it is not related to “using hand as the only remaining means” or “extending it to rulers”, as the abovementioned interpretations suggest. Second, Chunyu Kun apparently asks Meng

156

A. K. Volkov

Ke why the latter does not advocate some particular forms of political activities (which remain unspecified but apparently are well known to both interlocutors); one can guess that the activities to which the “hand” is referring may transgress either the established social order or the norms of behavior of a “gentleman”. The response of Meng Ke amounts to the claim that in the domain of “application” (i.e., social activities) the “salvation” should be conducted with relevant means, referred to as “[correct/my] teaching/way [of action]” (dao),41 and that the “saving the world with hand” exemplifies an incorrect application of the example to the situation discussed. If this understanding is correct, the interpretation suggested by Reding cannot be applied to explain the final statement of Meng zi, since the “hand” would refer to the (incorrect) analogical connection established between two domains, and not just within one of them: it would be therefore impossible to place this “hand” in Table 8.1, since it refers to an interrelationship between its two columns. One can add that this hand played one more crucial part: it saved Meng Ke who was about to lose the dispute.

5  Conclusions The dialog of Meng zi and Chunyu Kun discussed above exemplifies the style of argumentation found in a number of Chinese texts of the late first millennium BC – early first millennium AD. In (Volkov 1992) I provided an analysis of another dialogue from the Meng zi in which the disputers operate with a metaphor to discuss the properties of human nature, and of a dialogue from the Shuo yuan 說苑 (Garden of persuasion) compiled by Liu Xiang 劉向 (79-8 BC) in 17 BC on the basis of earlier sources found in Imperial Library42; in the latter dialog Hui Shi 惠施 (fl. 313 BC), the famous “sophist” of the Zhanguo period,43 used an elaborate analogy when discussing applicability of the method of argumentation by analogy itself. There exist other cases of argumentation by analogy found in contemporaneous philosophical and literary texts of this period still awaiting a study. Equally interesting is the use of analogical reasoning in scientific (in particular, mathematical) texts; their preliminary discussion can be found in (Volkov 1987, 1992, 1994, 2008, 2018). The problem of plausible interpretation of these excerpts in general case, however, remains unsolved. A number of attempts to suggest valid interpretative strategies, including those discussed above, have been made, but the lack of contemporaneous commentarial exegesis on the one hand, and of formal tools suitable for interpretation of analogical reasoning on the other often make these attempts inadequate, as my cursory overview demonstrated. One can, for instance, suggest

 In Reding’s table (see Table 8.1) “the Way” should be placed in a position corresponding to “a hand” (not shown in the table) and not to “the rites”. 42  See (Knechtges and Chang 2014: 952–955). 43  On Hui Shi, see (Fung 2009) and (Fraser 2017). 41

8 Analogy

157

that major difficulty encountered by a number of modern authors was directly related to their intention to apply propositional or predicate calculus to interpret Chinese logical inferences; unfortunately, first-order predicate logic apparently cannot be used for interpretation of such cases as the dialog between Meng Ke and Chunyu Kun discussed above. The reason is that the cases of analogical inference documented in this and other dialogues involve sequences of analogical transfers from one domain to another, some of which are rejected, and some, further developed in the course of interaction between the interlocutors. The analogy suggested in the dialog of Meng Ke and Chunyu Kun, as my analysis shows, may have generated various interpretations of the terms and statements introduced in the dialog at earlier stages; it remains unclear if all these interpretations were intended by the author(s) of the treatise, and whether the authors were even fully aware of them.44

References Bain, Alexander. 1875. Logique déductive et inductive. Trad. de l’anglais par Gabriel Compayré. Paris: Germer-Baillère. Cheng, Chung-ying. 1965. “Inquiries into classical Chinese logic.” Philosophy East and West 15.3/4: 195–216. (An overview of the studies of Chinese logic in China and the West accompanied by an annotated bibliography.) Chmielewski, Janusz. 1962–1969. “Notes on Early Chinese Logic.” Rocznik Orientalistyczny, [Part 1] vol. 26, no. 1 (1962): 7–21; [Part 2] vol. 26, no. 2 (1963a): 91–105; [Part 3] vol. 27, no. 1 (1963b): 103–121; [Part 4] vol. 28, no. 2 (1965a): 87–111; [Part 5] vol. 29, no. 2 (1965b): 116–138; [Part 6] vol. 30, no. 1 (1966): 31–52; [Part 7] vol. 31, no. 1 (1968): 117–136; [Part 8] vol. 32, no. 2 (1969): 83–103. (A series of papers devoted to the early history of Chinese logic reprinted fac-simile in Chmielewski 2009: 175–337). Chmielewski, Janusz. 1979. “Concerning the Problem of Analogical Reasoning in Ancient China (Review Article).” Rocznik Orientalistyczny, 40.2: 65–78. [Reprinted fac-simile in Chmielewski 2009: 339–352]. (A review of Cikoski 1975 containing Chmielewski’ interpretation of analogical reasoning in China). Chmielewski, Janusz. 2009. Language and Logic in Ancient China. Collected Papers on the Chinese Language and Logic. Warszawa: Polska Akademia Nauk Publishers. (Reprinted publications of Janusz Chmielewski including his works on Chinese logic.) Cikoski, John S. 1975. “On Standards of Analogical Reasoning in the Late Chou.” Journal of Chinese Philosophy 2.3: 325–357. (An attempt to interpret Chinese analogical thinking with modern mathematical concepts, such as boolean algebras and homomorphisms.) De Fornel de La Laurencie, Martial-Marie-Paul. 1906. Logique générale et appliquée. Paris: Delagrave. Forke, Alfred (tr.). 1922. Mê Ti, des Sozialethikers und seiner Schüler philosophische Werke (The philosophical works of the social ethicist Mo Di and his disciples). Berlin: Vereinigung Wissenschaftlicher Verleger. (A German translation of the Mo zi.)

44  V. Spirin’s attempts (1976, 1991) to study textual parallelisms in Chinese texts as forming particular patterns, theoretically, may be useful for a study of the use of parallelisms in Chinese argumentation (Volkov 1991); see also A. Krushinskiĭ’s (1988) short paper for an attempt to design an appropriate formal language for dealing with Chinese analogical inferences.

158

A. K. Volkov

Fraser, Christopher J. 1999. Similarity and standards: Language, cognition, and action in Chinese and Western thought. PhD Thesis. Hong Kong: The University of Hong Kong. (A study of the theoretical framework underlying classical Chinese philosophy of language, theory of action, and epistemology.) Fraser, Christopher J. 2009. “The Mohist School.” In Bo Mou, ed., History of Chinese Philosophy (137–163). London/New York: Routledge. (A succinct yet comprehensive introduction to the philosophy of Mo zi.) Fraser, Christopher J.  2017. “School of Names.” In Edward N.  Zalta, ed., The Stanford Encyclopedia of Philosophy (Spring 2017 Online Edition). https://plato.stanford.edu/archives/ spr2017/entries/school-names/. (An introduction to a group of Warring States period (479–221 BC) thinkers, including Hui Shi, who shared an interest in language and disputation and were often referred to as “sophists” or “logicians.”) Fung, Yiu-Ming. 2009. “The school of names.” In Bo Mou, ed., History of Chinese Philosophy, vol. 3 (164–188). London/New York: Routledge. (An introduction to a group of thinkers known as “School of Names” and often referred to as “sophists” or “logicians.”) Graham, Angus Charles. 1978. Later Mohist Logic, Ethics and Science. Hong Kong/London: The Chinese University Press/SOAS, University of London. (The standard introduction to Moist semantics and logic.) Graham, Angus Charles. 1993. “Mo tzu 墨子.” In Michael Loewe, ed., Early Chinese Texts (336–341). Berkeley: The Society for the Study of Early China and the Institute of East Asian Studies, University of California, Berkeley. (A bibliographical notice on the treatise Mo zi 墨 子 providing information about its contents, date of composition, text history, main editions and translations.) Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. Oxford etc.: Oxford University Press. (A comprehensive study of the major philosophical schools of pre-Imperial China.) Harbsmeier, Christoph. 1998. Language and Logic. In Joseph Needham and Christopher Cullen, eds., Kenneth Robinson (volume editor), Science and Civilisation in China, Vol. 7, Pt. 1. (The volume provides a detailed overview of logical reflection in Ancient and early medieval China including a chapter on Moist logic). Hu, Shi 胡適. 1919. An Outline of the History of Chinese Philosophy 中國哲學史大綱. Shanghai: Shangwu yinshuguan. (A pioneering comprehensive study of ancient Chinese philosophy partly based on the author’s doctoral research conducted prior to 1917.) Hu, Shih [Hu, Shi胡適]. 1922. The Development of the Logical Method in Ancient China. 先秦名 學史. Shanghai: The Oriental Book Company. (A pioneering work on the history of logic in China, including an interpretation of the “logical excerpts” found in the Mo zi 墨子.) Knechtges, David R., and Chang, Taiping. 2014. Ancient and Early Medieval Chinese Literature. A Reference Guide. Part 2. Leiden/Boston: Brill. Krushinskiĭ, Andrei Andreevich. 1988. “Teoretiko-kategornyĭ podkhod k semantike drevnekitaĭskikh klassicheskikh tekstov (An Approach to Semantics of the Classical Ancient Chinese Texts Based on Category Theory).” In Proceedings of the 19th Annual Scholarly Conference “Society and State in China,” 1 (73–77). Moscow: Nauka Publishers. (An attempt to interpret Chinese analogical inference with the help of the category theory.) Kurtz, Joachim. 2011. The Discovery of Chinese Logic. Leiden/Boston: Brill. (A comprehensive study of the introduction of Western logic to China in the 17th–19th centuries and of the discovery of ancient Chinese logic in the 19th and 20th century.) Lau, Din Cheuk. 1963. “On Mencius’ Use of the Method of Analogy in Argument.” Asia Major, n.s., 10: 173–194. (A study of the analogical arguments in Meng Ke’s treatise.) Lau, Din Cheuk. 1970 (repr. 1983). Mencius. Harmondsworth: Penguin Books. Legge, James. 1875. The Chinese Classics Translated into English. Vol. 2: The life and works of Mencius. London: Trübner & Co. Liang, Qichao 梁啟超. 1920. The “Moist Canons,” Collated and Annotated 墨經校釋. Shanghai: Shangwu yinsuguan.

8 Analogy

159

Liard, Louis. 1884 (repr. 1892). Logique (3e édition). Paris: G. Masson. Liu, Yameng. 1996. “Three Issues in the Argumentative Conception of Early Chinese Discourse.” Philosophy East and West 46.1: 33–58. Maspero, Henri. 1928. “Notes sur la logique de Mo-tseu et de son école.” T’oung Pao, Série 2, 25.1/2: 1–64. (A critical evaluation of interpretations of the Moist “logical” chapters by Chinese and Western authors containing a number of path-breaking observations concerning analogical arguments in this and other early Chinese philosophical texts.) Masson-Oursel, Paul. 1912. “Esquisse d’une théorie comparée du sorite.” Revue de Métaphysique et de Morale, 20.6: 810–824. Masson-Oursel, Paul. 1917. “Études de logique comparée (Suite).” Revue Philosophique de la France et de l’Etranger, 84: 59–76. Mei, Yi-Pao. 1929. The Ethical and Political Works of Motse. London: Probsthain. Pauthier, Guillaume. 1832. Le Ta-Hio, ou La grande Étude: ouvrage de Confucius et de son disciple Tseng-Tseu. Paris: l’Imprimerie d’Éverat. Reding, Jean-Paul. 1985. Les fondements philosophiques de la rhétorique chez les sophistes grecs et chez les sophistes chinois. Berne: P. Lang. (A comparative study of the Chinese philosophical “School of Names” ming jia名家 and Greek Sophists). Reding, Jean-Paul. 1986. “Analogical Reasoning in Early Chinese Philosophy.” Asiatische Studien/ Études asiatiques 40.1: 40–56. (A discussion of “argumentative analogies” in Chinese texts interpreted as mappings from domain of “exposition” to domain of “application”.) Reding, Jean-Paul. 2004. “The Origin of Logic in China.” In Reding, Jean-Paul, ed., Comparative essays in early Greek and Chinese Rational Thinking (31–48). Aldershot & Burlington: Ashgate. (A further development of ideas presented in Reding 1986.) Spirin, Vladimir Semenovich. 1976. Postroenie drevnekitaĭskikh tekstov [Structure of ancient Chinese texts, in Russian]. Moscow: Nauka Publishers. (An attempt to study textual parallelisms in ancient Chinese texts forming particular structures.) Spirin, Vladimir Semenovich. 1991. “Composition des texts chinois anciens”. Translation by Karine Chemla and Alexei Volkov. In Karine Chemla, Alexei Volkov, Vera [Dorofeeva-] Lichtmann, eds., Modèles et structures des textes chinois anciens. Les formalistes soviétiques en sinologie (Extrême-Orient Extrême-Occident, no. 13): 31–58. Paris: PUV. (An abridged French translation of the first part of Spirin 1976.) Sun, Yirang 孫詒讓. 1895 (repr. 2007). Leisurely Glosses on the Mo zi 墨子閒詁. Suzhou. [Reprinted by Beijing: Zhonghua shuju Publishers.] Tang, Degang 唐德剛. 1977. “70% traditional, 30% Western. Memories of Mr Hu Shi and Oral History. Part 3. 七分傳統·三分洋貨. 回憶胡適之先生和口述歷史之三.” Chuanji wenxue 傳 記文學 31.4: 55–62. Volkov, Alexei K. 1987. “O metode analogii v kitaĭskoi matematike” (On the Method of Analogy in Chinese Mathematics, in Russian). In Proceedings of the 18th Annual Scholarly Conference “Society and State in China,” 1: 113–117. Moscow: Nauka Publishers. (An attempt to provide a formal interpretation of the logic used by the compilers of ancient Chinese mathematical treatises.) Volkov, Alexei. 1991. “Recherches sur les structures des textes chinois anciens en URSS” (essay review). In Karine Chemla, Alexei Volkov, Vera [Dorofeeva-]Lichtmann, eds., Modèles et structures des textes chinois anciens. Les formalistes soviétiques en sinologie (ExtrêmeOrient Extrême-Occident, no. 13): 11–30. Paris: PUV. (A general overview of the works on the structures of ancient Chinese texts published by V.S. Spirin, A.M. Karapetyants, A.I. Kobzev, V.V. Dorofeeva-Lichtmann and other Soviet sinologists in 1960s–1980s.) Volkov, Alexei. 1992. “Analogical Reasoning in Ancient China: Some Examples.” In Karine Chemla, ed., Regards obliques sur l’argumentation en Chine (Extrême-Orient ­Extrême-­Occident, no. 14): 15–48. Paris: PUV. (A study of analogical reasoning in Chinese philosophical and mathematical texts.)

160

A. K. Volkov

Volkov, Alexei. 1994. “Transformation of Objects in Ancient Chinese Mathematics and Their Evolution.” In Viviane Alleton and Alexei Volkov, eds., Notions et perceptions du changement en Chine, 36 (133–148). Paris: Collège de France, Mémoires de l’Institut des Hautes Etudes Chinoises. (A study of analogical reasoning in Chinese mathematical texts compiled prior to the late first millennium CE.) Volkov, Alexei. 2008. “Raisonnement par analogie dans les mathématiques chinoises.” In Marie-­ José Durand-Richard, ed., Le statut de l’analogie dans la démarche scientifique, Perspective historique [The Status of the Analogy in Scientific Work: A Historical perspective]: 61–95. Paris: l’Harmattan. (A further elaboration of the ideas presented in Volkov 1994.) Volkov, Alexei. 2018. “Analogical reasoning in early Chinese mathematics.” In Ivan Gros and Thierry Lucas, eds., LOGOS ET ANALOGIA: La pensée analogique entre Orient et Occident (137–157). Louvain-la-Neuve: Academia-l’Harmattan. (A discussion of application of analogical reasoning in ancient Chinese mathematics.) Waley, Arthur. 1939. Three Ways of Thought in Ancient China. London: Allen and Unwin. Wilhelm, Hellmut. 1973. [Review of Lau 1970 (1983)]. Bulletin of the School of Oriental and African Studies, 36.2: 489–490 Zhang, Taiyan章太炎 [= Zhang, Binglin 章炳麟]. 1910. Balance for weighing discourses on national past 國故論衡. [Shanghai:] Guoxue jiangxihui 國學講習會/Kyoto: Shūkōsha 秀光 社. Alexei K. Volkov,  is Professor of the Center for General Education and the Graduate Institute of History of the National Tsing Hua University, Taiwan. His research focuses on the history of mathematics and mathematics education in East and Southeast Asia. He has published a number of papers and book chapters on these topics, including “Analogical Reasoning in Ancient China: Some Examples,” Extrême-Orient Extrême-Occident, 14 (1992), “Raisonnement par analogie dans les mathématiques chinoises,” in M.-J.  Durand-Richard (ed.), Le statut de l’analogie dans la démarche scientifique, Perspective historique (Paris: l’Harmattan, 2008), and “Argumentation for State Examinations: Demonstration in Traditional Chinese and Vietnamese Mathematics,” in K. Chemla (ed.), The History of Mathematical Proof in Ancient Traditions (Cambridge University Press, 2012).

Chapter 9

Reasoning (Pi 譬, Mou 侔, Yuan 援, Tui 推) Wujin Yang

In ancient China, reasoning or argument was called by the Mohist School as shuo 說. The Mohists said: “Reasoning is the means by which clarification is effected” (Johnston 2010: 435) (shuo, suoyi ming ye. 說, 所以明也。) and “Reasoning is the means of bringing out reason” (yi shuo chu gu. 以說出故) (Johnston 2010: 621).1 The Mohists advanced four kinds of reasoning pattern: “illustrating” (pi 譬), “parallelizing” (mou 侔), “adducing” (yuan 援), “inferring” (tui 推). In addition, “Inferring” can also be divided into two kinds of “stopping” (zhi 止), that is, inductive and deductive “stopping” inferences. It is commonly thought that these patterns involve reasoning by analogy to an absurd conclusion as a form of reductio ad adsurdum. In my opinion, generally, these patterns are used to make inference; but, specifically, some of them are used for making a proof and the others for making a refutation. Among them, “illustrating” and “parallelizing” lay stress on the function of proof, i.e., asserting a proposition true; but “adducing” and “inferring” lay stress on that of refutation, i.e., asserting a proposition false.

1  The Nature and Role of the Mohist Reasoning According to Mohists’ opinion, a person’s capacity to reason can surmount the bounds of time and space: it has the function of “knowing the future from the past and the present” and “knowing the unseen by the seen” (Sun 1987: 51).  Most translations in this paper will be from Johnston 2010, with certain adaptions. For example, here I translate 說 as “reasoning” rather than “explanation.” All the departures from Johnston’s translation will be noted. 1

W. Yang (*) Philosophy School, Renmin University of China, Beijing, China e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_9

161

162

W. Yang

In what sense reasoning is not bound by space? The Mohists said: “The square (space) cannot become the obstruction of reasoning” (fang buzhang, shuo ye. 方不 彰, 說也。).2 That is to say, reasoning is not bound by space. The thought that reasoning is not bound by space and it can surmount the obstacle of space is shown by the following passage (Johnston 2010: 559)3: What is outside is what you know. What is in the room is what you don’t know. Someone says: ‘The color of what is in the room is like the color of this.’ This is to hear that what you don’t know is like what you do know. One says that white is like white, black is like black. It is like this is its color, like white necessarily being white. Now you know its color is like white, therefore you know it is white. Names are the means whereby what is not known is made clear and corrected and not the means whereby doubt is cast on what is clear. It is like using ruler as a measurement for an unknown length. What is outside you know by direct experience; what is in the room you know by reasoning (shuo 說).

We know by seeing that the color of an object outside is white. This is knowledge by experience. Someone tells us that the color of an object in the room is like the color of an object outside. This is knowledge by being told. Now, we can reason that the color of an object in the room is white. This is knowledge by reasoning. To know by reasoning is “to know the unseen in term of the seen”, viz. surmounting the obstacle of space. The Mohists thought that the relation of reasoning is that of “necessarily coming about” (Zhang 2006: 17). The relation between premises of cause/reason (gu 故) and concluding sentence (ci 辭) is expressed by: “If we get the premises, then we will have the conclusion” (所得而後成) (墨經經上) (Mohist Canon A1) and “having A and then necessarily having B” (有之必然) (墨經經說上) (Canon Explanation A1). The Mohists said: “A cause is that which obtains before something comes about” (Johnston 2010: 375). So, premises of cause/reason as grounds are what conclusions are drawn from. Again, they said: “When there is a minor cause/reason, something is not necessarily so; when there is not, something is necessarily not so” And “When there is a major cause/reason, something is necessarily so; when there is not, something is necessarily not so” (Ibid). In the Mohist Canons, “cause/reason” was divided into “minor cause/reason” and “major cause/reason.” The minor cause/reason is a necessary condition for a proposition to hold, and the major cause/reason is a necessary and sufficient condition. The Mohists take the relation between premises and conclusion of reasoning to be “the major cause/reason” of “having A and then necessarily having B” (有之必 然) (Zhang 2006: 18) which is completely in accord with Aristotle’s view that the conclusion “necessarily comes about” given the premises. Aristotle once said: “Now a deduction is an argument in which, certain things being laid down, something other than these necessarily comes about through them”(Barnes 1984: 167).

 Here, I don’t agree with the translation of Ian Johnston. It is an error to translate “square” (fang方) into “method.” 3  Replacing Johnston’s “being told” by “reasoning.” 2

9  Reasoning (Pi 譬, Mou 侔, Yuan 援, Tui 推)

163

This relation is also expressed in the Mozi by “a thing is so if it is this” (shi er ran 是而然) (Johnston 2010: 625).

2  The Reasoning of “Illustrating” (Pi 譬) The Mohists defined that: “Illustrating is that, in order to make someone else know one thing, you refer to a different thing known by him already” (pi ye zhe, ju tawu er yi ming zhi ye. 辟也者, 舉他物而以明之也。) (墨子小取) (Mozi, Xiaoqu 2). This pattern of reasoning is found in works much earlier than the Mohist Canons, like the Book of Odes around 1000  B.C.  The well-known sophist Hui Shi (380–305  B.C.) was famous for his talent in using this sort of reasoning in his arguments. The feature of “illustrating” is that two different things A and A’ are used in the reasoning. The reason why one can get to know A by appealing to a different A’ lies in the similarity between A and A’, as mentioned above. To be more specific, consider an example from the book Gongshu 公輸 of the Mozi. Mozi met the King of the State Chu 楚. In order to convince the King that it is not right for the rich Chu to invade the poor State Song 宋, he used a more obvious example. Namely, it is not right for rich people to leave their property behind and go robbing poor people. Since the King sees the injustice of the latter, he realizes that of the former, too. Clearly, in this example, from the similarity between (a) the wealthy state invading the poor state and (a’) rich people robbing poor people, and (a’) is unjust, one can then infer the injustice of (a) too. Notice that the purpose of “illustrating” is to make someone else know the right judgment, not to make oneself know. In this sense, it is a typical illustration of the interactive argumentative slant of the Mohist Canons (Liu and Yang 2010: 112). “Illustrating” is different from metaphor. The latter is a figurative speech which invites one to understand some characteristics of one thing through its hidden similarity with another thing, and it is not used for making true or false judgment. In contrast, the former is a kind of argument, that is to say, it needs to support a conclusion through a series of explanatory statements. To judge if a discourse is a demonstration of “illustrating,” the most critical point is to see whether it is used to provide evidence to support a concluding view but not merely used to express a state of affairs. Let us have a look at the following example (Johnston 2010: 717–719). Master Peng Qingsheng 彭輕生, a disciple of Mozi, said: “What has gone can be known; what is to come cannot be known.” Master Mozi asked: “Suppose your parents encountered difficulty when they were fifty kilometers away and there was a limit of one day. If you reached them, they would live. If you did not reach them, they would die. Now if, on one side, there was a strong cart and an excellent horse, and, on the other side, there was a worn-out old horse and a cart with wheels at the four corners, and you had to make a choice, which would you make use of?” He

164

W. Yang

replied: “I would make use of the excellent horse and the strong cart, so I could arrive quickly.” Master Mozi said: “How is not there knowledge of what is to come?!” Here, Mozi carried on a very good demonstration of “illustrating.”

3  The Reasoning of “Adducing” (Yuan 援) The Mohists defined that: “Adducing means: if it is so in your case, why may it not be so in mine too?” (yuan ye zhe, yue: zi ran wo xi du bukeyi ran ye? 援也者, 曰:子 然我奚獨不可以然也?) (墨子小取) (Mozi, Xiaoqu 2). “Adducing” is carried out in the following steps: one first quotes an opinion that the opponent accepts, then one argues that the opponent’s opinion and one’s own are the same or belong to the same kind. Then, it naturally follows that one’s opinion should be accepted as well if the opponent insists on his opinion. Gongsun Long 公孫龍, who is one of the most important scholars of the School of Names, was good at using the reasoning pattern of “adducing” to argue. When he defended his thesis “a white horse is not a horse,” he adduced the thesis “Chu’s man is not a man” in the story of Confucius. That is (Sun 2014: 9): One day, the king of the state of Chu went hunting with the bows of Fanrou 繁弱and the arrows of Wanggui 亡歸 in Yunmeng zhi pu 雲夢之圃, then he lost a bow. The followers requested to look for the missing bow. The King said: “No. The king of the State of Chu lost a bow and some Chu’s man got it, so it is unnecessary to look for the missing bow.” Confucius said: “The king of the State of Chu is very kind, but not perfect. He might say one man lost a bow and then another man got it. Why should he say a Chu’s man?” Gongsun Long said: “So, Confucius thought that a Chu’s man is different from a man.”

Gongsun Long asked why it would be a problem for him to say “a white horse is not a horse” if we accepted what Confucius said: “Chu’s man is not a man.” The argument used here is “adducing.” Again, the basis of “adducing” is the earlier-discussed notion of a kind. In the above example, “Chu’s man is not a man” and “a white is not a horse” are of the same kind. So, if one of them is true, the other should be true as well. In fact, both “illustrating” and “adducing” can be formalized into the following schema: (i) The objects/events of statements A and A’ are of the same kind (i.e., A has the kind-defining property P if A’ has that property P). (ii) A has the property P. (iii) Therefore, A’ has the property P. As we can see from the above examples, the premise (i) is often omitted from the reasoning as being common knowledge. What is left then is a transition from one object/event to another one of the same property or from one statement to another one of the same kind. But of course, establishing the sameness in kind is an essential feature in practice.

9  Reasoning (Pi 譬, Mou 侔, Yuan 援, Tui 推)

165

4  The Reasoning of “Inferring” (Tui 推) The Mohists defined that: “Inferring is using what is the same in that which he refuses to accept and that which he does accept in order to propose the former” (tui ye zhe, yi qi suo buqu zhi tong yu qi suoqu zhe, yu zhi ye.推也者, 以其所不取之同 於其所取者, 予之也。) (墨子小取) (Mozi, Xiaoqu 2). Consider the following scenario. If someone proposes a statement you disagree with, what you need to do is to choose a statement that belongs to the same kind as what he proposed (and which he should therefore accept), but in fact, he cannot accept it. In that case, he has to give up his initial statement. This describes precisely how “inferring” proceeds. The logical reasoning pattern here is this: (i) The objects/events or statements A and A’ are of the same kind (i.e. A has property P if A’ has property P). (ii) A’ does not have the property P. (iii) Therefore, A does not have the property P. Let us look at an example in the book Gongmeng 公孟 of the Mozi. Gongmengzi 公孟子 does not think gods or ghosts exist, but he claims that junzi 君子 (gentleman) should learn how to pray. Mozi then says: “What Gongmengzi said is just like saying you have to learn how to treat your guests well, but there is no guest at all. This is also like having to make a fish net, but there is no fish.” The absurdity of the last two cases is clear, so we conclude that what Gongmengzi said was wrong. In this example, what Gongmengzi said about gods and what Mozi said about guests or fish are of the same kind. Clearly, Gongmengzi would not agree with the statement about guests and fish, so his statement about gods can also be rejected. The Mohists further divided “inferring” into two kinds of “stopping” (zhi 止), which, I think, are inductive and deductive inference.

4.1  Inductive Inference of “Stopping” The Mohists defined that: “Stopping: If another, by enumerating this is so, just considers everything is so, then I will ask him by enumerating that is not so” (bi ju ran zhe, yiwei ci qi ran ye, ze ju buran zhe er wen zhi. 彼舉然者, 以為此其然也, 則舉 不然者而問之。).4 To “stop” someone in argumentation is to make one’s opponent stop to argue or is to refute him. When one’s opponent enumerates “what is so” (ran zhe 然者) and then infers inductively that everything is so, then one can refute him

4  This translation differs importantly from Johnston’s translation: “Stopping: If another, by enumerating this is so, just considers these are so, then I will ask him by enumerating that is not so” (Johnston 2010: 463).

166

W. Yang

by enumerating cases of “what is not so” (buran zhe 不然者). For example, if one’s opponent gives a number of examples of black people and infers that all people are black, one can refute him by giving examples of people who are not black. The inductive “stopping” inference is a method of refuting an inference with the following structure: The reasoning being refuted: (i) Si is P. (or: some Si are P.) ( ii) Therefore, all S are P. Proof: (i) Sj is not P. (or: some Sj are not P.) (ii) Si and Sj are of the same kind as that for all S. (iii) Therefore, not all S are P If one’s opponent draws a general conclusion from particular premises, then one can refute him by enumerating counter-examples. Because in induction by enumeration the most likely source of error is “sampling error.” Giving a counter-­ example is the most effective way to defeat your opponent.

4.2  Deductive Inference of “Stopping” Again, the Mohists defined that: “Stopping: Another, on the basis all being so, says this is so. I, on the basis of that not all being so, call in question this being so” (bi yi ci qi ran ye, shuo shi qi ran ye. wo yi ci qi buran ye, yi shi qi ran ye. 彼以此其然也, 說是其然也。我以此其不然也, 疑是其然也。).5 When one’s opponent deduces “this is so (shi qi ran ye 是其然也)” on the basis of the claim that everything is so, then one can refute him by pointing out “not all things are so” (ci qi buran ye 此其 不然也). The deductive “stopping” inference is a method to refute an inference with the following structure: The reasoning being refuted: (i) All S are P. (ii) Therefore, Si is P. Proof: (i) Not all S are P. (ii) Si is of the same kind as that for all S. (iii) Therefore, probably Si is not P. 5  This translation differs importantly from Johnston’s translation: “Stopping: Another, on the basis of these being so, says this is so. I, on the basis of these not being so, call in question this being so” (Johnston 2010: 466–7).

9  Reasoning (Pi 譬, Mou 侔, Yuan 援, Tui 推)

167

The particular conclusion of the opponent’s inference is inferred from its general premise. So, the best way to refute him is to challenge the truth of the general premise. For example, in the chapter Condemning Offensive Warfare III of the Mozi, it says that (Johnston 2010: 189): Nowadays those rulers who favour offensive warfare also embellish their arguments in order to refute Mozi by saying: “Do you take offensive warfare to be unrighteous and not to be beneficial in affairs? In former times, Yu 禹 reduced the You Miao 有苗, Tang 湯 overthrew Jie 桀, and King Wu 武 overthrew Zhou 紂, yet they are all established as being sage kings. How do you account for this?” Mozi replied: “You have not considered the category of my words nor have you understood their basis. What they did is not called ‘attack’, it is called ‘punishment’.”

According to Mozi’s advocating that “all the offensive warfare are unjust,” those rulers who favour offensive warfare deducted the specific cases of “Yu reduced the You Miao and Tang overthrew Jie,” etc. are also unjust. This inference contains a hypothesis, that is to say, “all war are the offensive warfare.” Mozi denied it, and then suspected the opponent’s conclusions. Special need to point out is that Mohist logic mainly studied the patterns of argumentation used for proof and refutation in practical problems, rather than forms of inference in general. This is the main difference between Mohist and Western logic. For example, the inductive “stopping” inference mentioned above has the form of a syllogism, but the Mohists never studied this form in depth; they were only interested in its use in argumentative practice. Shen Youding once said that this “is an application of the syllogism of the 3rd figure” (Shen 2000: 28). For example, suppose one’s opponent infers “all people are selfish” from several examples such as “Jack is selfish.” Then, we can refute him by saying “John is not selfish.” Why? Because from the premises “John is not selfish” and “John is a person,” by applying the Bokardo of the syllogism of the 3rd figure, we can infer that “some persons/ people are not selfish,” which refutes the opponent’s general proposition “All people are selfish.” Since the two propositions “some persons/people are not selfish” and “all people are selfish” are mutually contradictory to each other, one of them must be false. However, the Mohists neither gave this kind of explanation nor abstracted the form of the syllogism in general. Whereas Aristotle emphasized the general forms of inference and proposed a system of syllogistic reasoning, the Mohists studied forms of reasoning and argumentation only by enumerating particular cases (Yang 2004: 45).

5  The Reasoning of “Parallelizing” (Mou 侔) The Mohists defined that: “Parallelizing is to compare or align [two] utterances which are both maintained” (mou ye zhe, bi ci er ju xing ye. 侔也者, 比辭而俱行 也。) (墨子小取) (Mozi, Xiaoqu 2). That is to say: “parallelizing” is comparing

168

W. Yang

sentences of the same kind to explain how one can pass from one to the other while keeping the same pattern. The logical reasoning pattern of “parallelizing” is this: If A is B, then CA is CB. Here, A expresses the subject, B expresses the predicate, and C expresses some attribute attached to A or B. For example, the Mohists said that: “to know a pooch (gou 狗) yet to say of yourself that you do not know a dog (quan 犬) is a mistake. The explanation lies in duplication” (Johnston 2010: 517). The relation between pooch and dog is the sameness in terms of duplication. The relation between them cannot be changed when we use some attribute “know” to attach to them. So, we can carry out the reasoning of “parallelizing” of “a thing is so if it is this.” In the writings of the later Mohists, we find that: “A white horse is a horse. To ride a white horse is to ride a horse. A black horse is a horse. To ride a black horse is to ride a horse. Huo 獲 is a person. To love Huo is to love a person. Zang 臧 is a person. To love Zang is to love a person. These are examples of there being this and it is so” (Johnston 2010: 627). These are examples of “a thing is so if it is this.” For each parallelizing, if we affirm the premise, then we can affirm the conclusion. If the premise is true, then the conclusion is true. “Shi 是” stands for the affirmation of the premise and “ran 然” stands for the affirmation of the conclusion. There is a relation of “necessarily coming about” between the premise and the conclusion. That is to say, the reasoning of parallelizing, “a thing is so if it is this,” also applies to the relation of being proper inclusion and the relation of attribution. It is noted that the Mohists called the reasoning mentioned above “parallelizing” (mou 侔), (Graham 1978: 483) which is not workable for all cases. Although it works for the above situations, it does not work in other situations. The reason why it works in these cases lies in the fact that the two tokens of the same term “C,” say, “ride (cheng 乘)” or “love (ai 愛),” appeared in “CA” and “CB” are constant and without ambiguity. Because Mohist logic needs to serve practical argumentation and needs to explain and solve concrete problems in the practical ethics of social and political contexts, it cannot be like Western logic, which emphasizes extensional relations. Sometimes, the Mohists have to analyze and grasp the relations among terms in a logical inference from the intensional point of view. For example: “A robber is a person. Killing a robber is not killing a person” (Johnston 2010: 627). Obviously, the Mohists also admit that a robber is a person, that is to say, “robber” is properly included in “person” when considered extensionally; yet they insist that a person may not be killed but a robber may be, when considered from an ethical point of view within a social and political context. So, for the Mohists, it is very important to use their logic to explain and understand the political and ethical views of people in ancient China. If we analyze only from an extensional perspective, the statement “Killing a robber is not killing a person” is obviously absurd. However, if we also analyze from an intensional perspective, the statement stated above is reasonable. Considering the

9  Reasoning (Pi 譬, Mou 侔, Yuan 援, Tui 推)

169

meaning of “killing a robber” (shadao 殺盜), the action of killing a robber is not directed at killing a “person” who happens to be a robber, it is directed at killing a “robber.” “Robber” belongs to “person” extensionally but intentionally is not only a person, namely, “robber” has essential properties that “person” lacks. We can analyze the class an object belongs to from different aspects. When compared with a student and a worker, a robber also has the properties of a person; but, with respect to the properties of being corrupt and looting, a robber’s performance has the property of unlawfulness that other people’s performance may not have. That is the difference (Yang 2011: 117–118).

References Barnes, Jonathan, ed. 1984. The Complete Works of Aristotle. Princeton: Princeton University Press. Graham, A.C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: The Chinese University Press. Johnston, Ian. 2010. The Mozi. A Complete Translation. Hong Kong: The Chinese University Press. Liu, Fenrong 劉奮榮, and Wujin Yang 楊武金. 2010. “A Brief History of Chinese Logic.” Journal of Indian Council of Philosophical Research XXVII.1: 101–126. Shen, Youding 沈有鼎. 2000. “Some Problems about Syllogism” (有關三段論的幾個問題). In Research Division of Logic, Institute of Philosophy, Chinese Academy of Social Sciences, eds., Describing for Comparison: Shen Youding and His Academic Road (摹物求比:沈有鼎及 其治學之路), 6–30. Beijing: Literature in Social Sciences Press. Sun, Zhongyuan 孫中原. 1987. A History of Chinese Logic (中國逻辑史). Beijing: People’s University Press. Sun, Qiqin 孫啟勤. 2014. The English Version of Gongsunlongzi. Hangzhou: Zhejiang Gongshang University Press. Yang, Wujin 楊武金. 2004. A Study of Mojing Logic (墨經邏輯研究). Beijing: China Social Science Press. Yang, Wujin 楊武金. 2011. “Valid Reasoning in Ancient China from the Perspective of Modern Logic.” Studies in Logic 4.3: 115–125. Zhang, Zhongyi 張忠義. 2006. Research of Chinese Logic to Certain Gain (中國邏輯對 “必然地 得出” 的研究). Beijing: People’s Daily Press. Wujin Yang  graduated from the Renmin University of China in 2001 with a PhD. From 2001 to 2003, he conducted postdoctoral research at the Chinese Academy of Social Science’s Institute of Philosophy. He currently serves as a philosophy professor and doctoral tutor at the Renmin University of China. His main areas of research are the history of Chinese logic, Mohism and critical thinking. His publications include the Logical Basis of Dialectics, Reading the Mo Jing, Logic and Critical Thinking, and Capability and Literacy in Logical Thought.

Chapter 10

Argumentation (Bian 辯) Lisa Indraccolo

Argumentation1 or disputation (bian 辯) is one of the two main “rhetorical”2 techniques employed by ancient Chinese debaters in the late pre-imperial and early imperial period (fifth century B.C. to second century A.D.). Evidence of the practice and the performance of this technique has mostly been transmitted under the form of fictionalized dialogues and more or less elaborated narrative anecdotes that are recorded in the received literature. Moreover, a few cases of meta-reflection on the (mostly negative) effects that the (mis)use of such techniques produces on society have to be added to the overall picture. This kind of materials are preserved especially in a group of texts that have allegedly been compiled, at least in part, during the “Golden Age” of Classical Chinese rhetoric, the Warring States period (475–221 B.C.). This type of argumentative texts largely – though not invariably – belong to the category of so-called “Masters literature” (zishu 子書). “Masters literature” is a label that has been established a posteriori to classify heterogeneous collections of anecdotes, dialogues, sayings, and maxims traditionally attributed to and named after semi-legendary “Masters” of thought, such as, for instance, Mengzi 孟子 (Master Meng, more commonly known as Mencius in the Western world), Zhuangzi 莊子 (Master Zhuang), Xunzi 荀子 (Master Xun), and so on.

I would like to thank the editor and an anonymous reviewer for their helpful comments and suggestions on a previous draft of this contribution.  On the definition of bian as “argumentation,” see Garrett 1993: 107–109; Lu 1998: 84–90, esp. 88–89; 191–192; 214–215. 2  On the issue of the use of the term “rhetoric” in a non-Western context and the limits of its applicability, see Indraccolo and Behr 2014. 1

L. Indraccolo (*) Institute of Asian and Oriental Studies, University of Zurich, Zurich, Switzerland e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_10

171

172

L. Indraccolo

In Classical Chinese polemico-philosophical texts, argumentation is often understood as one of the two poles in a pair of binary opposites together with its dialectical counterpart, the technique of persuasion (shui 說).3 Despite the fact that no proper handbook nor manual on Classical Chinese rhetoric has been handed down, it is still possible to claim with a certain degree of certainty that argumentation and persuasion are two opposite, though complementary and closely interconnected skills that a persuader has to acquire and master as part of their repertoire. As Kroll has underlined, “it appears that Shang Yang, Hsun Tzu [Xunzi], and, evidently, Meng Tzu [Mencius], among others, happened both to persuade and to dispute. The arts of ‘disputation’ (pien) [bian 辯] and ‘persuasion’ (shui) [shui 說] were ascribed to the same person as mutually connected skills” (Kroll 1985–86: 126). Argumentation and persuasion are two sides of the same coin. A persuader or debater has to choose the most appropriate and potentially effective technique according to an accurate analysis of the circumstantial and contextual characteristics of the specific situation in which they find themselves  to act, and especially according to the hierarchical role (argumentation is typically used among peers, persuasion is almost invariably addressed to a superior in rank) and the inner dispositions and inclinations of the addressee. Bian literally means “to distinguish,” “to discriminate,” or “to argue.” More precisely, according to the etymological dictionary Shuowen jiezi 說文解字, compiled by Xu Shen 許慎 (58–147  B.C.), bian can assume four different meanings: “(1) ‘bian’ 辯 (argumentation, disputation); (2) ‘bian’ 辨 (making distinctions); (3) ‘bian’ 變 (making changes); (4) ‘zhi’ 治 (achieving justice and order).”4 The term bian and the activity that it describes can assume slightly different nuances of meaning and be morally charged according to the interpretations proposed by different trends of thought or individual thinkers. In most cases, however, bian has a decidedly negative connotation and is harshly criticized and morally condemned by contemporary and later scholars. Argumentation is described as a form of hair-splitting, “sophistic” argument. The main criticism directed against the use of the technique of argumentation is the apparent disinterest of the persuaders making use of this technique in achieving any superior ethical goal, and the lack of scruples with which they are prone and ready to support even the most incredible propositions and absurdities for the only sake of winning an argument, with the result of spreading false beliefs and bizarre sayings that might exercise a negative influence on the ruler or the common people. Practicing bian eventually is a pointless and egotistic activity that entails no higher ethical goal nor purpose beyond the sheer pleasure debaters take in winning an argument. Therefore, those who practice bian are accused of being fond of and deliberately indulging in the deplorable activity of arguing for arguing’s sake. One of the most tenacious detractors of the technique of argumentation is, for instance, the late Warring States text Han Feizi 韓非子 (Master Han Fei). The text

 Crump 1964, esp. 6–7, 100; Kroll 1985–86: 126.  Lu and Frank 1993: 452; Lu 1998: 88.

3 4

10  Argumentation (Bian 辯)

173

repeatedly complains about the pernicious influence exercised by shrewd and devious disputers crowding the court, who dwell on and indulge in “thorny speeches of the same kind of the ‘white horse argument’” (「棘刺白馬之說」in Ch. 32 ‘Wai chushuo zuo shang’ 外儲說左上) to captivate the ruler’s attention, distracting him from more urgent and concrete government issues. Argumentation becomes synonym for a perverse and distorted kind of debate, a hair-splitting argument, or an artful, deceitful or cunning type of reasoning, shrewdly employed in a deliberately misleading fashion with no other intention than outtalking an opponent at any cost. However, a fundamental premise must be made here. Most of these apparently sincere and heartfelt criticism and outraged “cris de couer” are rather to be considered as manifestations of the so-called “rhetoric of anti-rhetoric,” that is “a sophisticated rhetorical mechanism […] employed to convey a message that attacks ‘rhetoric’,”5 typically used to attack an adversary in the political or debate arena. This phenomenon is a well attested trope in Western Classical rhetoric, and the mastery of this technique is commonly acknowledged as being part of a persuader’s repertoire. However, the same process is at play also in early Chinese argumentative texts, where anti-rhetorical accusations are widely employed to denigrate opponents, and accuse them of ungentlemanly behaviour or lack of moral commitment. Argumentation is often sharply criticized in “Masters texts” as a method unworthy of a true scholar by persuaders that eventually end up using very similar techniques. Ironically, even those received texts associated with the teaching of masters of thought considered epitomes of “virtue,” such as, for instance, the Mengzi, make widespread use of the same argumentative techniques, tropes, and rhetorical devices employed by the debaters they so vehemently attack, in particular the so-called “Chinese Sophists”6 or Logicians (Mingjia 名家, literally “experts on names”).7 Logicians are a more or less loosely defined group of thinkers, allegedly active during the Warring States period. They are traditionally described in the received literature as exceptionally skilled and shrewd debaters and diplomats who are especially fond of language jokes, paradoxes, and dilemmas. Infamous for their sharp tongues, a malicious love for hair-splitting arguments, and a general lack of ethical commitment, they excel in the technique of bian. As such, they are often called dialecticians or debaters (bianzhe 辯者, bianshi 辯士, or more rarely, yizhe 議者). The dialectician par excellence is the Classical Chinese pseudo-historical figure of Gongsun Long 公孫龍, who is famous for excelling in debating the “white horse not horse” argument (bai ma fei ma 白馬非馬), among others. The “white horse” argument is the most famous paradoxical statement in the history of Chinese philosophy and represents an ideal case of argumentation, which will be discussed in more detail below. Despite being predominantly associated with the Logicians’ trend of thought, bian is not practiced exclusively by this relatively small group of debaters. A defini Valesio 1980: 41–42.  Forke 1901–02. 7  On the obsoleteness and unsuitability of the term “School” to translate jia 家 in this context, see Petersen 1995; Ryden 1996; Nylan 2000; Csikszentmihalyi 2002; Csikszentmihalyi and Nylan 2003; Smith 2003. 5 6

174

L. Indraccolo

tion of bian is provided also in the so-called Mohist Canons (Mojing 墨經), a cluster of chapters (40–45) preserved in the Mozi 墨子 (Master Mo) and compiled more or less around 300 B.C. that deal with highly technical subjects, such as logic, engineering, mathematics, optics, and the sciences at large. In particular, Ch. 40 ‘Canon A’ (Jing shang 經上) and Ch. 41 ‘Canon B’ (Jing xia 經下) are constituted by collections of dry and at times rather enigmatic sentences or “canons” (jing 經), which find their corresponding illustrative explanations (shuo 說)8 respectively in Ch. 42 ‘Explanation A’ (Jingshuo shang 經說上) and Ch. 43 ‘Explanation B’ (Jingshuo xia 經說下). Canons (C) and explanations (E) A70 to A75 deal especially with technical jargon closely related to the art of rhetoric, including definitions for bian and shuo 說 (“explanation,” “demonstration”). The definition of bian is given in Canon A74 as follows9: C 辯爭攸10 也。辯勝當也。 E 辯: 或謂之牛, 或謂之非牛, 是爭彼也。是不俱當, 不俱當必或不當。不若當 犬。 C: Bian (“discrimination”) is contending over opposite claims. Winning in discrimination depends on fitting the case. E: Bian (“discrimination”): When one calls it “ox” and the other calls it “non-ox” it is a case of contending over opposite claims. This is a case in which they do not both fit; if both do not fit, then necessarily one (of the two) does not fit. It is not like the case of “to fit into the category of ‘dog’.”11 Textual sources associated with Mohism provide a far less negative reading of bian than is typically conveyed in the received literature. In the Mozi, bian is described as a fundamentally healthy and eventually even ethical technique to be used in “dialectics  – in the sense of making accurate discriminations,” (Raphals 1992: 64) or the art of “arguing over alternatives.”12 Accordingly, also Mary Garrett defines bian as a “disagreement over theses that are contradictories […]” (Garrett 1993: 107). Keeping in mind the above mentioned four possible meanings of bian listed in the Shuowen jiezi, here the accent is put on the second meaning of bian – as the pragmatic capacity to discriminate among things, and to operate appropriate distinctions between opposite statements – rather than on its first meaning, as argumentative power or efficacy in debate. From the Mohist perspective, the final goal of bian is not to win an argument at any cost by outtalking an opponent in a “battle of words” (Carson 1983), which is the most common understanding of the term in

8  On the polysemic nature of the character 說 and its two different readings and corresponding meanings, respectively as shui (“persuasion”) and shuo (“explanation,” “demonstration” or “saying”), see Lau 1963: 185, note 9; Garrett 1993: 109–110; Schwermann 2011; Indraccolo 2014. 9  Translations included in this article are mine unless otherwise stated. 10  The original character here is corrupted to you 攸. On this issue, see Graham 1978: 318; Johnston 2010: 436, note 121. 11  Cfr. Graham 1978: 318–319; Johnston 2010: 436–437. 12  Chong 1999; Hansen 1983: 120–121; Graham 1978:120–123.

10  Argumentation (Bian 辯)

175

Classical Chinese literature. The Mohist use of the term is somewhat more restrictive and partial. According to Angus C. Graham, “the Mohist canons confine the pien ‘discrimination’ (辯) to disputes of the form ‘X is or is not Y’, in which one alternative must be right and the other wrong […]” (Graham 1959: 91). As the opening lines of Ch. 45 ‘Xiaoqu’ 小取 (Lesser Pick) explain, bian is “about making clear the distinction between right and wrong (true and false), and investigating the pattern of order and disorder. It is about clarifying instances of sameness and difference, examining the principles of name and entity, determining what is beneficial and harmful, and resolving what is doubtful and uncertain.”13 As such, bian has a strong moral connotation. By practicing bian, it is possible to operate necessary, informed distinctions between mutually opposite claims, and thereby to unequivocally establish and distinguish between right and wrong. Bian is a powerful means to know the world thoroughly and to develop a critical conscience. On the basis of the knowledge thereby acquired, it is possible to act upon the world in an ethically conscious, proactive way in order to enact meaningful changes in government practice and society at large. This decidedly positive vision of bian and its potential as a moral and moralizing technique is somewhat exceptional in the early Chinese contemporary intellectual scenario. It is obviously in stark contrast against the more widespread, largely negative understanding of bian as sheer debating prowess,14 and the morally disengaged attitude manifested especially by and typically ascribed to debaters associated with the Logicians’ trend of thought that were known to be exceptionally skilled in this technique. In Classical Chinese literature, cases of argumentation typically assume the form of a dispute or a dialectical skirmish between two (or occasionally more) debaters that enjoy the same social status. Examples of bian are, for instance, the several exchanges between Zhuangzi 莊子 and is friend and debate sparring partner Hui Shi 惠施, but also, more interestingly, the highly cryptic argumentative materials included in a series of dialogues preserved in the composite collection Gongsun Longzi 公孫龍子 (Master Gongsun Long). This is the case of the most famous “white horse” argument and the “hard and white stone” (jianbai shi 堅白石) debate, recorded respectively in the chapters ‘Disquisition on White and Horse’ (Baima lun 白馬論) and ‘Disquisition on Hard and White’ (Jianbai lun 堅白論), and the three incomplete and partially corrupted exchanges preserved in the chapter ‘Disquisition on Continuity and Change’ (Tongbian lun 通變論). These texts are almost unique in the scenario of Classical Chinese literature. They are deprived of any fictionalizing element or extra-textual information concerning the backstory, the identity of the characters involved, and the setting of the dialogic action. As such, they provide a pure theoretical framework for discussion, and possibly represent a sort of ­rhetorical exercise. Such instances of argumentation are especially worth attention,

 Johnston 2010: 621; cfr. Graham 1964: 2. See also Raphals 1992: 64–65.  “Pien of the Canons is somewhat narrower than argumentation in general, its usual meaning in pre-Han literature.” (Graham 1978: 319). 13 14

176

L. Indraccolo

as they suggest that these kinds of materials were most probably not meant to be read alone in silence without an external aid, nor enjoyed as written texts as such. We can hypothesize that they might have been used as written drafts, teaching materials or aide-memoires to help the study and the memorization of a sample argument, and to promote its subsequent original re-elaboration and adaptation to a different context. A case of argumentation is a polemical piece, a heated and fast-paced dialogue characterized by a thrust-and-counterthrust structure that can be occasionally framed by an introductory narrative, ranging from a one liner to a short paragraph, contextualizing the dialogue within a fictitious debate scene. Usually, the two characters involved are 1) a persuader or debater that supports a somewhat disconcerting, unusual or subversive perspective on reality and things, and 2) his opponent or antagonist, who typically represents common sense and more or less rigorously applies the law of inference to the argument in question. Argumentation can deal with highly speculative and abstract problems, proposing insightful reflections on theoretical questions that are apparently far removed from the concerns of everyday life. However, such speculations are only apparently detached from reality. Most cases of argumentation are more than pure divertissement for its own sake. They often entail a deeper meaning and allude to concrete social and political situations. As several examples embedded in narrative anecdotes preserved in the received literature show, these only apparently sterile argumentative exercises could actually be used to address and tackle thorny political issues in an indirect way, for instance to solve delicate diplomatic questions or matters of foreign policy. In a case of argumentation, rhetorical strategies commonly employed  in the development of an argument typically include the sorites or chain-reasoning, the syllogism supported by the law of inference, the reductio ad absurdum, dilemmas, “‘inference by analogy,’ ‘thesis and antithesis,’ ‘paradox and dilemmas,’ ‘comparing things and joining objects of the same kind,’ ‘the method of [discussing] advantages and disadvantages’.”15 Moreover, illustrative examples cast in the form of analogies and metaphors that are mostly drawn from the natural world and the domain of manual work and handicraft can be used by persuaders to increase argumentative force. In a case of argumentation, the main goal of a persuader is to do and say whatever it takes to outtalk his opponent and leave him speechless, thereby winning the debate. To achieve this goal, the persuader shrewdly employs any means at his disposal to puzzle his adversary, and to provoke in him the strong and outraged reaction he needs to put forward his own argument, so as to proceed to the following step in his reasoning. Thus, he pressures and dazzles his opponent with stringent logic, or deliberately confuses him with apparently nonsensical language jokes or paradoxical statements. The persuader consistently turns the opponent’s answers against him by twisting his words and totally subverting or pretending to ­misunderstand their meaning. Eventually, the opponent, frustrated and bewildered, fails to notice the potential flaws or evident fallacies in the persuader’s reasoning. He usually ends up falling into plain contradiction and, unable to object, is forced to 15

 Kroll 1985–86: 124–25. See also Garrett 1993: 108.

10  Argumentation (Bian 辯)

177

explicitly admit – if not accept – the validity of the persuader’s point of view and the soundness of his argumentation. As it has already been mentioned above, a typical case of argumentation par excellence is the “white horse” argument preserved in the ‘Baima lun.’ This dialogue discusses the most famous paradoxical statement in the history of Classical Chinese philosophy, “white horse is not horse” (bai ma fei ma 白馬非馬). However, apart from being a refined logical exercise, the argument can metaphorically be interpreted also as an exhortation to re-establish a univocal correspondence between names and the realities they designate. Read in these terms, the whole argument might allude to the disrupt of the fundamental relationships that frame, inform and order civil society, conveying also an implicit underlying criticism directed towards a contemporary situation of social and political disorder, and the consequent corruption and decay of moral values. The debate takes place in an unknown time and place between two anonymous characters, a persuader and his opponent. The persuader (P) supports a rather unconventional and somewhat bizarre view that justifies the admissibility of the claim that, under certain circumstances, a white horse would not be a horse. The opponent (O) instead represents common thought and consistently applies the law of inference throughout the whole dialogue, supporting the idea that a white horse must necessarily be a horse. The first exchange already explains where the main source of confusion lies, and provides the grounds on which the persuader is capable of supporting such an outrageous claim: O (曰):「白馬非馬可乎。 」 P曰: 「可。」 O曰: 「何哉。」 P曰:「馬者所以命形也。白者所以命色也。命色者非命形也故曰白馬非馬。 」 O: “To say that ‘white horse is not horse,’ is that admissible?” P: “It is admissible.” O: “How is that possible?” P: “‘Horse’ is what denotes shape, ‘white’ is what denotes color. What denotes color is not (the same as) what denotes shape. Therefore, I say that ‘white horse’ is not (the same as) ‘horse’.” It is clear that the persuader is not talking about concrete horses existing in the real world, but he is rather arguing that to say “white horse” or “horse” is not exactly the same thing, as the intension and the extension of the two terms are different. “White horse” provides substantial additional information about the color of the horse's coat, which immediately circumscribes the subset of potential horses that might match the definition “white horse.” “Horse” instead is a generic term that only conveys the most basic information about the shape of the animal taken into consideration. The opponent attacks the persuader and accuses him of denying the most obvious and evident fact that horses need to be of a certain color. According to the opponent, accepting the persuader’s claim would correspond to saying that only colorless horses could be considered actual horses, which is obviously nonsense since there

178

L. Indraccolo

is no such thing as a colorless horse in the world. Once again, the persuader underlines that he is not talking about real horses of flesh and bones. As he candidly admits, horses in the real world necessarily have to be of a certain color, which is also the reason why it is possible to have white horses. Still, at the same time from his point of view it is equally undeniable that “white horse” is not the same as “horse.” To further illustrate this point, the persuader uses an ingenuous example that shows how this is the case also for “dun horse” and “black horse.” P曰: 「求馬黃黑馬皆可致。求白馬黃黑馬不可致。使白馬乃馬也是所求一 也。所求一者白者不異馬也。所求不異如黃黑馬有可有不可何也。可與 不可其相非明。故黃黑馬一也而可以應有馬,而不可以應有白馬。是白馬 之非馬審矣。」 P: “If you were looking for a horse, a dun horse or a black horse could be delivered to you. If you were looking for a white horse, a dun horse or a black horse could not be delivered to you. If you now consider a white horse as a horse, it means that what you were looking for was one and the same thing. If what you were looking for was one and the same thing, then white would not be different from horse. If what you were looking for was one and the same thing, then it would be like the case of the dun horse and the black horse that are at the same time admissible in one case and inadmissible in the other case, how can this be possible? It is clear that admissible and inadmissible mutually exclude each other. Therefore, a dun horse and a black horse are the same insofar as they can both correspond to there being a horse, but cannot correspond to there being a white horse. It is indeed true that ‘white horse’ is not (the same as) ‘horse’.” Finally, the persuader masterfully applies one of the characteristic argumentative methods associated with the technique of argumentation, the reductio ad absurdum. He suggests that if indeed white horse is the same as horse – as the opponent claims, then we have to assume that a dun horse is necessarily the same as horse as well. But if that was the case, we would have to admit that white horses are horses and, at the same time, dun horses are equally horses. Therefore, applying the law of inference that the opponent has been diligently employing so far, one might also say that white horses are dun  horses, which is obviously absurd. Thereby, the persuader forces the opponent into open contradiction, and the latter eventually has to admit nolens volens that, under certain conditions, “white horse” indeed is not (the same as) “horse.” The issue at stake here is of course much more urgent and crucial than to simply disambiguate animals of a certain specific color or, more generally, expressing a characteristic that somehow further defines them, from the broader set that entails all the possible generic animal variants belonging to the same animal kind. What the persuader is trying to show here is the fundamental need to establish univocal and unambiguous definitions that match their corresponding realities in order to avoid any possible source of confusion. In this rather unconventional way, the persuader is eventually supporting the doctrine of the “rectification of names” (zhengming 正 名) that claims that it is necessary to ensure that names match their corresponding

10  Argumentation (Bian 辯)

179

actualities, and that such names veritably express all the qualities and characteristics that define such actualities and make them be what they actually are. Finally, as it has been shown, argumentation represents a fundamental rhetorical technique in the scenario of Classical Chinese philosophical literature. Persuaders were expected to master this technique, which  had to be acquired together with persuasion as a set of complementary and mutually interconnected rhetorical skills. In the received literature, persuaders and debaters that engage in argumentation are harshly criticized for their fondness of paradoxes and language jokes and their apparent lack of moral commitment to a higher cause. However, this kind of criticism has rather to be considered as a well-established rhetorical convention or literary cliché, an astute stratagem employed to criticise one’s adversaries, rather than a sincere manifestation of disdain and moral condemnation. A case of argumentation usually assumes the form of a dialogue between two or more characters supporting opposite claims and assuming contrary points of view, corresponding respectively to an unusual or subversive take on reality in the case of the persuader, and a more conventional and familiar perspective that complies with common sense in the case of the opponent. Through the use of a more or less fixed and predictable set of argumentative methods and rhetorical stratagems and devices (such as chain reasoning, reductio ad absurdum, paradoxes and language jokes, among others), a case of argumentation might at times entail a deeper meaning, and make a fundamental point or convey a moral lesson disguised as a highly sophisticated language play, as for instance happens in the “white horse” argument.

References Carson, Michael. 1983. “The Battle of Words in Ancient China.” Asian Culture Quarterly 11.2: 48–61. Chong, Chaehyun. 1999. “The Neo-Mohist Conception of Bian (Disputation).” Journal of Chinese Philosophy 26.1: 1–19. Crump, James I., Jr. 1964. 戰國策 Intrigues: Studies of the Chan-kuo Ts’e. Ann Arbor: The University of Michigan Press. Csikszentmihalyi, Mark. 2002. “Traditional Taxonomies and Revealed Texts in the Han.” In Livia Kohn and Harold D. Roth, eds., Daoist Identity – History, Lineage, and Ritual, 81–101. Honolulu: University of Hawai’i Press. Csikszentmihalyi, Mark, and Michael Nylan. 2003. “Constructing Lineages and Inventing Traditions through Exemplary Figures in Early China.” T’oung Pao 83: 59–99. Forke, Alfred. 1901–02. “The Chinese Sophists.” Journal of the China Branch of the Royal Asiatic Society 34: 1–100. Garrett, Mary M. 1993. “Classical Chinese Rhetorical Conceptions of Argumentation and Persuasion.” Argumentation and Advocacy 29.3: 105–115. Graham, Angus C. 1959. “‘Being’ in Western Philosophy Compared with Shih/Fei and Yu/Wu in Chinese Philosophy.” Asia Major New Series 7.1: 79–112. Graham, Angus C. 1964. “The Logic of the Mohist Hsiao’chü.” T’oung Pao 51.1: 1–54. Graham, Angus C. 1978. Later Mohist Logic, Ethics and Science. London/Hong Kong: School of Oriental and African Studies/The Chinese University Press.

180

L. Indraccolo

Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: The University of Michigan Press. Indraccolo, Lisa. 2014. “The Difficulties of shuì/shuō 說: Persuasions, Explanations and Sayings in the Hán Fēizǐ 韓非子 and Beyond.” Paper presented at the Sinologisches Kolloquium (Sinological Colloquium), Rheinish Friedrich Wilhelm-University of Bonn, Germany, 22nd January 2014 (unpublished). Indraccolo, Lisa, and Wolfgang Behr. 2014. “Introduction.” In “Masters of Disguise? Conceptions and Misconceptions of ‘Rhetoric’in Chinese Antiquity” Special Issue, Asiatische Studien/ Études asiatiques 68.4: 889–913. Johnston, Ian. 2010. The Mozi: A Complete Translation. New York: Columbia University Press. Kroll, Jurij L. 1985–86. “Disputation in Ancient Chinese Culture.” Early China 11–12: 119–145. Lau, Din Cheuk (D.C.). 1963, “On Mencius’ Use of the Method of Analogy in Argument.” Asia Major 10: 173–194. Lu, Xing. 1998. Rhetoric in Ancient China, Fifth to Third Century B.C.E. – A Comparison with Classical Greek Rhetoric. Columbia: University of South Carolina Press. Lu, Xing, and David A. Frank 1993. “On the Study of Ancient Chinese Rhetoric/Bian 辩.” Western Journal of Communication 57.4: 445–463. Petersen, Jens Østergaard. 1995. “Which Books Did the First Emperor of Ch’in Burn? On the Meaning of Pai Chia in Early Chinese Sources.” Monumenta Serica 43: 1–52. Nylan, Michael. 2000. “Textual Authority in Pre-Han and Han.” Early China 25: 205–258. Raphals, Lisa. 1992. Knowing Words – Wisdom and Cunning in the Classical Traditions of China and Greece. Ithaca/London: Cornell University Press. Ryden, Edmund. 1996. “Was Confucius a Confucian? Confusion over the Use of the Term ‘School’ in Chinese Philosophy.” Early China News 9.5–9: 28–29. Schwermann, Christian. 2011. “Gattungsdynamik in der traditionellen chinesischen Literatur: Von der ‘Erläuterung’ (shuō) zur ‘Erzählung’ (xiǎoshuō)” [Generic Dynamics in Traditional Chinese Literature. From the ‘Explanation’ (shuō) to ‘Narrative’ (xiǎoshuō)]. In Stephan Conermann and Amr El Hawary, eds., Was sind Genres? Nicht-abendländische Kategorisierungen von Gattungen [What are Genres? Non-Western Categorizations of Genres]. Berlin: EB Verlag, 47–85. Smith, Kidder. 2003. “Sima Tan and the Invention of Daoism, ‘Legalism’ et cetera.” The Journal of Asian Studies 62.1: 129–156. Valesio, Paolo. 1980. Novantiqua – Rhetorics as a Contemporary Theory. Bloomington: Indiana University Press. Lisa Indraccolo  is Postdoctoral  Research Fellow in Chinese Studies, Institute of Asian and Oriental Studies, University of Zurich, Switzerland. She earned her PhD from Ca’ Foscari University, Venice, with a thesis on the early Chinese “sophistic” thinker Gōngsūn Lóng. Her main publications include studies on “Masters texts” (zǐshū)  and especially texts of the Chinese Logicians (míngjiā), the use of paradoxes in Classical Chinese philosophical texts, and early Chinese rhetoric, with focus on the two techniques of argumentation (biàn) and persuasion (shuì) in pre-imperial and early imperial Chinese literature. Her main research interests include early Chinese thought with focus on Warring States philosophical literature; Classical Chinese rhetoric, paradoxes and language jokes; early Confucian political thought; and conceptual history of premodern China from a comparative perspective.

Chapter 11

Reason (Gu 故) and Principle (Li 理) Yiu-ming Fung

1  Different Senses of “Gu” In classical Chinese, the word “gu” 故 is used in different ways. It can be used as a noun, an adjective or a conjunction and thus can be used with different senses. For example: 1. Event, incident (shigu 事故 or biangu 變故): 父母俱存,兄弟無故,一樂也。(孟子:盡心上) (Refer to http://ctext.org/mengzi) It is one of his delights that his father and mother are both alive and his brothers do not have any incident. (Mencius XI: 20) (My translation) 2. Old, past (jiu 舊 or gujiu 故舊): 農民無所聞變見方,則知農無從離其故事,而愚農不知,不好學問。(商君書: 墾令) (Refer to http://ctext.org/shang-jun-shu) If the farmers have no opportunity to hear of changes or see different places, clever farmers will have no opportunity to discard old ways, and stupid farmers will not become clever, nor will they become fond of study. (Book of Lord Shang II: 14) (Translation: Duyvendak 1928, with my minor modification) 溫故而知新,可以為師矣。(論語:為政) If a man keeps cherishing his old knowledge, so as continually to be acquiring new, he may be a teacher of others. (Analects II: 11) (Translation: Legge 1861a, refer to http://ctext.org/analects)

Y.-m. Fung (*) Department of Philosophy, Tunghai University, Taichung, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_11

181

182

Y.-m. Fung

3. Death (dagu 大故 or siwang 死亡): 今也不幸至於大故,吾欲使子問於孟子,然後行事。(孟子:滕文公上) (Refer to http://ctext.org/mengzi). Now, I am so unfortunate as to have the great incident [of losing my father]; I wish to send you to ask the advice from Mencius, and then to deal with the matter [of funeral]. (Mencius V: 1) (My translation). In addition to the senses mentioned above, there are some other uses of the word with logical significance. One of them is used as a conjunction to connect an expression of a causal factor or an expression of a logical condition (including necessary and sufficient conditions) with an expression of a phenomenon and to treat the former as an explanan and the latter as an explanadum for the whole series of connected sentences. It can be formed as a causal explanation if the explanan is about a causal factor of the explanadum. It can also be formed as a logical explanation if the explanan is about a logical reason of the explanadum. Just like the English words “therefore,” “hence” and “thus,” the Chinese words “gu” and “guyue” 故曰 (so I say that) can be used to make a causal explanation or a logical explanation. In regard to using “gu” as a conjunction, Kong Yingda 孔穎達 in his Commentary of the Book of Rites, Liji zhengyi 禮記正義 (The Correct Meaning of the Book of Rites), said that: “‘gu’ is a term connecting the former and the latter sentences.” (故, 承上起下之辭。) (曲禮上) and “‘gu’ is following the former sentence and leading to the latter one; so it is said: ‘guyue’.” (「故」者,因上起下,所以言「故曰」。) (曲禮下) (Legge 1885. Refer to http://ctext.org/wiki) For example1: 1  One of the readers thinks that: “Kong Yingda is cited as saying that gu connects the former and latter ci. But the examples are explicated as presenting causal relations.” So, the reader judges that there may be a confusion. However, I think, there is a debate in contemporary Western philosophy on “causal relata.” On the one hand, following David Hume’s view, Donald Davidson (1967) argues that the relata are events which should be understood as unstructured particulars or objects. On the other hand, Jaegwon Kim (1976) understands events that are the basic causal relata to be instantiations of properties by objects at certain times; D. H. Mellor (1995) takes causes and effects to be facts, which are the truth-makers of true contingent sentences. In our ordinary language, when we say that: “This short circuit caused this fire,” it can be explained as “a caused b” and thus, following Hume or Davidson, the relata can be treated as two events as particulars expressed by two (individual) terms. If we say that: “The fact that there was a short circuit caused it to be the case that there was a fire,” it can be explained as “P caused Q” or “P thus Q” and thus, following Kim or Mellor, the relata can be treated as two events as facts expressed by two propositions or sentences. So, when one says that: “Event a causes event b,” it is a singular statement expressing a causal relation between two events as objects. If we use “gu” (or “thus”) to express this relation, we could use “P” (a sentence symbol) to stand for “there is an event a” and “Q” (another sentence symbol) for “there is an event b” and use “gu” (or “thus”) to connect these two sentences in “P, thus Q” or “P, gu Q” as exemplified in the following example: (C2) Women give up their work of spinning and weaving and engage in literary grace, thus people will be cold. Men get off their job of tilling and planting and engage in engraving, thus people will be starved. (Mozi VI: 7) As argued by Donald Davidson, an event is expressed by a term as an individual or object, not by a sentence as a fact. We can say that “event a causes event b,” but we cannot say that “event a,

11  Reason (Gu 故) and Principle (Li 理)

183

(C1)

故倉無備粟,不可以待凶饑。(墨子:七患) Therefore, if it were no reserves in the warehouse, it would have no way to deal with famine. (Mozi V: 4) (Translation: Y. P. Mei 1929, refer to http:// ctext.org/mozi) 女 子 廢 其 紡 織 而 脩 文 采 ,故 民 寒 。 男 子 離 其 耕 稼 而 脩 刻 鏤 ,故 民 (C2)  饑。(墨子:辭過) Women give up their work of spinning and weaving and engage in literary grace, thus people will be cold. Men get off their job of tilling and planting and engage in engraving, thus people will be starved. (Mozi VI: 7) (Translation: Y. P. Mei 1929, refer to http://ctext.org/mozi) (C1) uses “gu” to assert a causal relation which is expressed in a form of counterfactual of past or present event: no reserves causes no way to deal with famine, while (C2) uses “gu” to assert a causal relation between the former event and the latter event. Sometimes, “gu” can also be used to assert a logical condition whose antecedent is treated as a reason for an action described by the consequent. In some contexts, a reason for action is also a mental cause of the action. For example: (R1)

(R2)

先王惡其亂也,故制禮義以分之。(荀子:王制/禮論) The ancient kings abhorred such disorder. Thus they instituted regulations, ritual practices, and moral principles in order to create proper social class divisions. (Xunzi IV: 3/XIX: 1) (Knoblock 1999: 215, 601) 故欲過之而動不及,心止之也。(荀子:正名) Thus, when desires run to excess, action do not reach that point because the mind stops it. (Xunzi XXII 11) (Knoblock 1999: 729)

Both (R1) and (R2) are about a reason for action respectively. But the reason is not mere reason without mental causation. They are rational causes between a mental event and an action (including taking action with and without physical movement). Here, the mental event is not only a reason for taking the action, but also a cause of the action.2

thus event b.” It is because “gu” (or “thus” as a sentential connective) is used to connect propositions or sentences, not terms of individuals or objects. Moreover, in using the sentential connective “gu” to connect two sentences, the relata of the connection may be about the causal relation of the happenings of both a cause and an effect, or about the logical relation (say, sufficient or necessary condition) of the happenings of two events which are without causal relation. (For example, in “If you pass the exam I will give you a present.” the promise in the conditional sentence is about the sufficient condition of the antecedent for the consequent. But this logical condition is not causal.) Based on the above explanation, I think there is no confusion in connecting sentences or propositions with gu for causal relation if the causal relata or the logical relata in all my examples are ci (sentences or propositions) of facts and not terms of individuals or objects. 2  Donald Davidson thinks that “the primary reason for an action is its cause.” So, explanation in terms of reasons can be understood as a species of causal explanation. See Davidson (1963: 686).

184

Y.-m. Fung

As we know, conditionals of causal relation and rationalization can be formed as a part of a logical reasoning. For example: (A1) Water is formed when two hydrogen atoms and one oxygen atom combine. (A2) Two hydrogen atoms and one oxygen atom combine in context C. (A3) Therefore, water is formed in context C. (B1) If I have examination tomorrow morning, then I cannot go to the movies tonight. (B2) I have examination tomorrow morning. (B3) Therefore, I cannot go to the movies tonight. The former reasoning from (A1) to (A3) is an argument with a premise of an implicit causal statement, while the latter one from (B1) to (B3) is an argument with a premise of rationalization in terms of a sufficient condition (provided that I need enough time to make preparation before examination). In ancient Chinese texts, there are a lot of examples of both kinds of reasoning. For example, in the Mencius (III: 2) (孟子:公孫丑上), there is a reasoning with a premise about mental causation: (C1) (C2) (C3) (C4)

夫志,氣之帥也。 The will (zhi 志) is the leader of vital force (qi 氣). (Definition 1) 氣,體之充也。 The vital force pervades and animates the body. (Definition 2) 夫志至焉,氣次焉。 The will is first and chief, and the vital force is subordinate to it; (Premise 1) 故曰:持其志,無暴其氣。 Therefore, I say, maintain firm the will, and do no violence to the vital force. (Conclusion)

Based on the definitions (C1) and (C2) and the premise (C3), it seems that, for Mencius, there is a causal relation between the will such as decisions and intentions and the vital force such as desires and emotions. I think (C3) implies that vital force will lose control if the will does not maintain in a state which is not affected by external conditions. So, if we divide (C4) into two parts: the first part is the sentence “maintain firm the will” as another premise and the second part is the sentence “do no violence to the vital force” as a conclusion, it can be understood as an argument based on a conditional of causal relation. In comparison with the material or informal inference mentioned above, the following passage in the “Discourse on the White-Horse” (Baima Lun 白馬論) of the Gongsun Longzi 公孫龍子 is an example of standard logical argument: (D1) (D2) (D3) (D4)

馬者所以命形也; “Horse” is used to name something of shape [i.e. the shape of horse]. 白者所以命色也; “White” is used to name something of color [i.e. the color of white]. 命色者非命形也; To name something of color is not to name something of shape. 故曰:白馬非馬。

11  Reason (Gu 故) and Principle (Li 理)

185

Therefore, I say: [the name] “white-horse” is not [the name] “horse.” (or: Therefore, white-horse (what is named by “white-horse”) is not horse (what is named by “horse”). I think this is a well-formed argument though some modification based on background knowledge is required to make it valid.3 The following argument is also valid if we accept a background condition that the unconsciousness of the harmfulness of an event will lead to the lessening of larger advantages: 愛小利而不慮其害。故曰:顧小利則大利之殘也。(韓非子:十過) It was because of his love of small advantages and unconcern about its harmfulness. Hence the saying: “To esteem small advantages hampers large advantages.” (Hanfeizi X: 3) (Liao 2015: 150–1)

Just like the above example, usually, reasonings or inferences in ancient Chinese texts explicitly or implicitly (i.e. with skipped words but context-dependent) use the conjunction “gu” to connect the premise(s) and the conclusion. But most of them are not expressed in a form of completed argument. Some of them can be understood as a valid argument if the textual context indicates that some background definition, assumption or factual statement can be used as a supplementary premise to support the argument. In this sense, we can call this kind of argument “informal argument.” It seems that this kind of argument is similar to Robert Brandom’s idea of “material inference.” Brandom follows Wilfrid Sellars to claim that the correctness of some non-formal inferences essentially involve the conceptual contents of its premises and conclusions. Sellars calls this kind of inference “material inference.” For example, the inference from “Pittsburgh is to the West of Philadelphia” to “Philadelphia is to the East of Pittsburgh” and that from “Lightning is seen now” to “Thunder will be heard soon” are arguments of such kind. They are good material inferences, according to Brandom, not because their correctness is based on formal norm or standard, but because it is rooted in the content of their non-logical vocabulary (Brandom 2000: 85). Jerry Fodor and others argue against the popular view that some informally valid arguments can be explained formally as supplemented with an implicit premise of definition. The popular view is that, to use Robert Brandom’s example, from “Pittsburgh is to the West of Philadelphia” to “Philadelphia is to the East of Pittsburgh,” we can find some hidden premises or definitions to elaborate a full-­ fledged argument of validity. However, Sellars and Brandom claim that the validity of a material inference is not dependent on formal rules. Both think that logical syntax is not necessary for the justification of informal arguments; Brandom even maintains that the material inference has primacy in logical thinking. Nevertheless, some other philosophers do not agree with Brandom’s view. They think that definitions or meaning postulates (of which definitions could be regarded as a special case) play an important role in bridging the gap between formal and informal inferences. This disagreement of views is still an on-going issue of debate (Fodor et al. 1980: 263–367; Pitt 1999: 139–159).

 The detail argument about this passage can be found in Fung (2007: 524–5).

3

186

Y.-m. Fung

In addition to using “gu” as a conjunction for informal inference, “gu” can be used as a name for conditions, including conditionals of causal and logical relations: For example, the question “What is its reason/cause?” (其故何也) in the Zhuangzi (II: 10/XII: 7) (莊子:齊物論/天地) is to ask for a reason to explain the existence of an event or the happening of an action. In the Book of Rites, the expression “wu gu” 無故 in the passage “Without some reason/cause for it, a ruler did not kill an ox, nor a great officer a sheep, nor a lower officer a pig or a dog.” (Liji XIII: 3) (My translation) (君無故不殺牛,大夫無故不殺羊,士無故不殺犬、豕。) (禮記:玉藻) (Refer to http://ctext.org/liji) is used to describe an action without reason or cause. Besides using the word “gu” in this non-theoretical context, Later Mohists also provide a general definition and specific sub-definitions of “gu” as follows: (Df1)

故,所得而後成也。(墨子:經上) The gu (reason/cause) of something is what it must get before it will come about. (Mozi, Canon I: 1) (Translation: Graham 1978, refer to http://ctext. org/mozi/cznon) 故:小故,有之不必然,無之必不然。體也,若有端。大故,有之必然,無之 (Df2)  必不然,若見之成見也。(墨子:經說上) “Minor reason”: having this, it will not necessarily be so; lacking this, necessarily it will not be so. It is a part, like having a starting point. “Major reason”: having this, it will necessarily [be so]; lacking [this, necessarily it will not] be so. Like the appearing bringing about the seeing. (Mozi, Canon Explanation I: 1) (Translation: Graham 1978, refer to http:// ctext.org/mozi/cznon) Here, “minor reason” means “necessary condition” while “major reason” means “both sufficient and necessary condition.” The Later Mohists’ example is: having a part (ti 體) is a necessary condition of having a whole or compound (jian 兼) which is constituted of the part. Just like the case that having a point (duan 端) is necessary for constituting a line (chi 尺). In contrast, for example, to make the act of seeing (jian 見) to be a phenomenon of seeing (cheng jian 成見), it must be based on some necessary conditions (such as eyesight, light and the object being seen) which are combined together to form a sufficient condition. In other words, if the propositions P, Q, R (say, having eyesight, light and the seen object as external conditions) express the necessary conditions of the proposition S (say, having a phenomenon of seeing) and the combination of them is able to make S true, then P, Q, R are the necessary condition of S respectively and the sufficient condition of S collectively. Besides using “gu” to mean “reason” in terms of “necessary condition” or “both sufficient and necessary condition,” sometimes, it can also be used to mean “explanatory reason” which presupposes a physical cause or a mental reason. In the following example, the former token of “gu” presupposes a cause while the second one presupposes a reason: 凡物之然也,必有故。… 水出於山而走於海,水非惡山而欲海也,高下使之然也。稼生 於野而藏於倉,稼非有欲也,人皆以之也。(呂氏春秋:審己) (Refer to http://ctext.org/ lv-shi-chun-qiu)

11  Reason (Gu 故) and Principle (Li 理)

187

That all things/events become so must have their reasons. Water comes from the mountain but goes to the sea. It is not because water is in favor of the sea, not the mountain, but because the water from up to down makes it to be the case. Farming happens in the field but is stored in the warehouse. It is not because farming has desire, but because all people use it. (The Annals of Lu Buwei XLIV: 1) (My translation)

In this passage, “the water from up to down makes it to be the case” is the gu as a cause to explain the phenomenon that “water comes from the mountain but goes to the sea.” On the other hand, “all people use it” is the gu as a reason to explain the event that “farming happens in the field but is stored in the warehouse.” To determine whether “gu” is used to mean cause or non-causal reason is context-­ dependent. To ask for why does something happen is probably an inquiry of causal explanation. For example: 一時遇風,同時得病,其病各異,願聞其故。 (黃帝內經:靈樞:五變) (Refer to http://ctext. org/huangdi-neijing) Once getting cold, both [people] have disease. But their diseases are different. I would like to know their reasons [from you]. (The Inner Canon of the Yellow Emperor II: Lingshu (Spiritual Pivot) XLVI: 2) (My translation)

I think this is a question for a causal explanation. That is: based on which medical principles or causal regularities we can explain the happening of different diseases? If a principle is used as a standard to regulate something, sometimes, the word “gu” can be used as a non-causal reason to explain why something is in accord with the principle or why the principle is violated. For example: 是故子墨子之有天之[意],辟人無以異乎輪人之有規,匠人之有矩也。今夫輪人操其 規,將以量度天下之圜與不圜也,曰:中吾規者謂之圜,不中吾規者謂之不圜。是以圜與 不圜,皆可得而知也。此其故何?則圜法明也。 (墨子:天志中) Therefore the will of Heaven is like the compasses to the wheelwright and the square to the carpenter. The wheelwright tests the circularity of every object in the world with his compasses, saying: “That which satisfies my compasses is circular. That which does not is not circular." Therefore whether an object is circular or not is all known, because the standard of circularity is all established. The carpenter also tests the squareness of every object in the world with his square, saying: “That which satisfies my square is square; that which does not is not square.” Therefore whether any object is square or not is all known. Why so? Because the standard of squareness is established. (Mozi XXVII: 9) (Translation: Y. P. Mei 1929, refer to http://ctext.org/mozi)

Here, in general, to ask for a gu is to ask for a standard or criterion for explaining whether an event or action is appropriate in terms of its being in accordance with the standard or criterion. That is, whether there is a reason in terms of normative principle or causal regularity to support the explanation.

2  Different Senses of “Li” In ancient Chinese texts, the word “li” 理 also have different uses and senses. If it is a verb, it can be used to mean (A) the action of management, adjustment or arrangement. If it is a noun, it can be used to mean (B) the texture or pattern of a material

188

Y.-m. Fung

object such as wood and jade, (C) the property of an object or event, (D) the structure or order of physical events or human affairs, (E) the principle or theory of natural phenomena or human events, and (F) the reasonableness of taking verbal and non-verbal actions. In regard to these different uses and senses, let’s see the following examples: 聖人忘乎治人,而在乎自理。(文子:道原) (Refer to http://ctext. (A1)  org/wenzi) The sage forgets to administer others, but cares to manage himself. (Wenzi I: 1) (My translation) 治其形,理其心。(莊子:則陽) (Refer to http://ctext.org/zhuangzi) (A2) To manage his figure and to organize his mind. (Zhuangzi XXV: 6) (My translation) (B1)

物成生理,謂之形。(莊子:天地) (Refer to http://ctext.org/zhuangzi) When a thing is produced, it generates a pattern/texture. It is called “form/ shape.” (Zhuangzi XII: 8) (My translation) 果蓏有理,人倫雖難,所以相齒。(莊子:知北遊) (Refer to http://ctext.org/ (B2)  zhuangzi) Fruits and melons have texture/structure. Even human relations are complicated, people are thus arranged by seniority in order/structure. (Zhuangzi XXII: 5) (My translation) 凡理者,方圓、短長、麤靡、堅脆之分也。(韓非子:解老) (Refer to (C1)  http://ctext.org/hanfeizi) In general, properties are what distinguish the square from the round, the short from the long, the coarse from the fine, and the hard from the brittle. (Hanfeizi XX: 25) (My translation) 短 長、大小、方圓、堅脆、輕重、白黑之謂理。(韓非子:解老) (C2)  (Refer to http://ctext.org/hanfeizi) The short and the long, the large and small, the hard and the brittle, the light and the heavy, and the white and the dark, are called “property.” (Hanfeizi XX: 25) (My translation)

(D1)

文理密察。(禮記:中庸) (Refer to http://ctext.org/liji) The cultural structure/order is closely investigated. (Liji XXXI: 32) (My translation) 依乎天理,批大郤,導大窾,因其固然。(莊子:養生主) (Refer to http:// (D2)  ctext.org/zhuangzi) Observing the natural structure/course, [my knife] slips through the great crevices and slides through the great cavities [of the cow]. It is to take action in accordance with its original case. (Zhuangzi III: 2) (My translation)

11  Reason (Gu 故) and Principle (Li 理)

189

(E1)  季真之莫為,接子之或使,二家之議,孰正於其情?孰偏於其理?(莊子:則 陽) (Refer to http://ctext.org/zhuangzi) In regard to Ji Zhen’s idea of “nothing to make things happen” and Jiezi’s idea of “something to make things happen,” which one of these two scholars’ argumentations is right to the real/reality? And which one is deviating from the principle/truth? (Zhuangzi XXV: 12) (My translation) 口之於味也,有同耆焉;耳之於聲也,有同聽焉;目之於色也,有同美焉。 (E2)  至 於 心 ,獨 無 所 同 然 乎 ?心 之 所 同 然 者 何 也 ?謂 理 也 ,義 也。(孟子:告子上) Men’s mouths agree in having the same relishes; their ears agree in enjoying the same sounds; their eyes agree in recognising the same beauty  shall their minds alone be without that which the similarly approve? What is it then of which they similarly approve? It is, I say, the principles of our nature, and the determinations of righteousness. (Mencius XI: 7) (Translation: Legge 1861b, refer to http://ctext.org/mengzi) 上失其道而殺其下,非理也。(孔子家語:始誅) (refer to http://ctext.org/ (F1)  kongzi-jiayu) The superior deviates from dao and kills his inferior, it is not reasonable/ rational. (Kongzi Jiayu II: 2) (My translation) 夫勸人而弗能使也,禁人而弗能止也,其所由者非理也。(淮南子:人閒 (F2)  訓) (Refer to http://ctext.org/huiananzi) One persuades others to do something but cannot make them to do, or prohibits others to do something but cannot stop them to do, it is because what he leads them to do or not to do is not reasonable/rational. (Huiananzi XVIII: 31) (My translation) Some ancient thinkers use the term “wuli” 物理 or “wu zhi li” 物之理 to mean “physical principle” or “property of physical object” and think that the physical principle is factual and objective, so they call it “tianli” 天理 (natural principle). For the non-physical or normative principle, sometimes they call it “shili” 事理 (principle of affairs), “daoli” 道理 (theoretical principle or principle of dao) or “yili” 義 理 (principle of appropriateness or principle of righteousness). For ancient thinkers, all argumentations (bian 辯) and disputations (yi 議) are required to provide reasons to support their relevant views. Usually, a good reason should be expressed as a principle. It means that the reason is able to apply to all examples of the same kind/class – lei 類. Nevertheless, not all grounding principles are accepted as nomological and as being able to make views or beliefs true. So, for some ancient Chinese thinkers, an accepted grounding principle is required not only to be consistent, necessary, universal, but also to be right to the reality or useful in a moral or socio-political context. In regard to the physical principle, there are two ways to express its regularity: one is to treat a physical event under explanation as subsumed under a general and

190

Y.-m. Fung

ultimate principle, that is, to treat it as following a dao 道 or as according to a regulative rule of the Yin-Yang model. For ancient Daoism and the School of Yin-Yang (Yin-Yang Jia 陰陽家), their dao is not a principle which is verifiable, confirmable or falsifiable in terms of empirical testability, rather, it is a kind of metaphysical principle which is used to explain physical phenomena in their proto-scientific reasoning. The most popular examples of explanation and reasoning in this way are those in ancient Chinese medicine, especially in the Suwen 素問 (Basic Questions) of the Huangdi Neijing 黃帝內經 (Inner Canon of the Yellow Emperor) of the Han dynasty. The most general principle stressed in this text is “the ideal state of thing is the harmony of yin and yang” or “the circulation of yin-yang is well-order.” For example: 故陰陽四時者,萬物之終始也,死生之本也。逆之則災害生,從之則苛疾不起,是謂「得 道」。(素問:四氣調神大論) (Refer to http://ctext.org/huangdi-neijing) Therefore, yin-yang and four seasons are the circulation rule of all things, the original ground of death and birth. If it is violated, disasters will happen; if it is observed, there will be no serious disease. This is called “obtaining dao.” (Suwen II: 2) (My translation)

Another way to express the regularity of a physical principle is based on an operative concept of fa 法, faze 法則 or fayi 法儀 (standard, rule or model). For example, according to Later Mohists, based on an appropriate fa, we can make similar things and explain why something is well done or not. They claim that: 法,所若而然也。(墨子:經上) The fa (standard) is that in being like which something is so. (Mozi, Canon I: 71) (Translation: Graham 1978, refer to http://ctext.org/mozi/canon) 法:意、規、員[圓]三也,俱可以為法。 (墨子:經說上) (Refer to http://ctext.org/ mozi/canon) The idea, the compasses, a [concrete] circle, all three can serve as standard. (Mozi, Canon Explanation I: 71) (My translation)

If “gu” is used in a conditional to express a physical cause, it is objective or factual and thus can be treated as a causal reason to explain the happening of a physical event. However, if “gu” is used in a conditional to express a non-causal reason, it may not be a good reason to explain an event or action. A good reason for taking action, including verbal and non-verbal actions, is a reason with the characteristic of reasonableness in the sense that the reason or what is described by the antecedent of the explanatory conditional has the modal power of necessity and universality in some sense. That is, it entails or presupposes a principle which is applicable to similar actions or to events of the same kind/class. So, to accept a gu as a non-causal reason has to be supported by a li which is applicable to all the actions or events of the same kind/class. In this regard, as stressed by Later Mohists, the concepts of gu, li and lei are connected in a logical web to constitute a rational base for making practical reasoning. In regard to the non-physical reason, especially in an ethical context, most grounding principles are not recognized as without controversy and thus are debatable. This may be one of the reasons why Zhuangzi thinks that “the truth and falsity claimed by Confucianism and Mohism” (儒墨之是非) are uncertain. Of course, all schools assert that their views are true and their grounding principles are right and

11  Reason (Gu 故) and Principle (Li 理)

191

thus reject others’. So, for example, Confucians and Mohists’ moral principles of righteousness (yi 義) are contrary to each other. For Confucius, yi is grounded on ren 仁 (benevolence) and realized in li 禮 (rites) and li 利 (benefit or utility) is not the first priority in consideration. For Xunzi, “ritual and its moral principles are used for cultural construction; social relationship and its classifying order are used as grounding principle.” (Xunzi XIII 7) (My translation) (禮義以為文,倫類以為理。) (荀子:臣道) (Refer to http://ctext.org/xunzi) On the other hand, for Mohists, moral principle is based on mutual love (jian xiang ai 兼相愛) which results in mutual benefit (jiao xiang li 交相利). As the criterion of yi, mutual love which is required by the will of the Heaven is both a necessary and sufficient condition of mutual benefit. Generally speaking, yi plays a role of moral principle as li or yili 義理 which can be treated as the grounding principle in moral reasoning though different schools have different ideas and criteria of yi and li. Based on the theory of each school, all accepted statements or views must be in accordance with yi or yili (zhongli 中理 or dangli 當理). So, in making statement or argumentation, the basic requirement is zhongli or dangli. That is: 凡君子之說也,非苟辨也;士之議也,非苟語也。必中理然後說,必當義然後議。(呂氏 春秋:懷寵) (Refer to http://ctext.org/lv-shi-chun-qiu) All the doctrines made by the gentleman are not indifferent/insignificant disputations; all the judgments made by the learned person are not indifferent/insignificant sayings. To accord with a principle in advance is necessary for making a doctrine; to validate a judgment with a moral principle in advance is necessary for uttering it. (The Annals of Lu Buwei XXXV: 1) (My translation) 理也者,是非之宗也。(呂氏春秋:離謂) (Refer to http://ctext.org/lv-shi-chun-qiu) Li is the original ground of truth and falsity. (The Annals of Lu Buwei CIV: 2) (My translation)

As the rational base of a judgment, li is the original ground for justifying whether a judgment is truth or false. Later Mohists make this point clearly in saying that: 誹之可否,不以眾寡。說在可非。(墨子:經下) (Refer to http://ctext.org/mozi) Whether a rejection is acceptable is not based on how many people make it. The explanation is: it is accepted to reject. (Mozi, Canon II: 179) (My translation) 論誹誹之可不可:以理之可,誹雖多,誹其誹,是也。其理不可,誹雖少,誹[其誹],非也。( 墨子:經說下) (Refer to http://ctext.org/mozi) On whether it is acceptable or not to reject a rejection: if it is acceptable by virtue of li, even though many people accept the rejection, to reject their [accepted] rejection is right/true; if it is not acceptable to reject a rejection, even though few people accept the rejection, to reject their rejection is wrong/false. (Mozi, Canon Explanation II: 179) (My translation)

Here, the implicit argument in terms of zhongli (fitting the principle) or dangli (validation by the principle) usually has the following forms for moral reasoning (from (M1) to (M3)) and for factual reasoning (from (F1) to (F3)), respectively: (M1) If an action A1 or a statement of the action S1 is in accordance with their relevant yi or yili, it is appropriate or right. (M2) Action A1 or statement S1 is in accordance with their relevant yi or yili. (M3) Therefore, action A1 or statement S1 is appropriate or right.

192

Y.-m. Fung

(F1) If a physical event P1 or a statement of the event S1 is in accordance with their relevant yi or yili, it is real or true. (F2) Event P1 or statement S1 is in accordance with their relevant yi or yili. (F3) Therefore, event P1 or statement S1 is real or true. In general, if one accepts a moral principle (li) as a ground for making a judgment, the principle must be explicitly expressed or implicitly presupposed in the place of the antecedent of a conditional statement as a sufficient condition (gu). However, whether a particular event or action is an example covered by the general principle is dependent on whether it belongs to the class of examples all of which are subsumed under the general principle, or whether it is similar in kind to the paradigmatic example of the general principle. In this regard, besides gu and li, lei also plays a crucial role for making judgment and reasoning.

3  Lei: Kind or Class? The word “lei” 類 in classical Chinese also has various uses and senses. It can be used to mean “class,” “group,” “kind,” “category” or “similarity.” Since the eleventh chapter of this volume focuses mainly on “classes and individuals,” readers can refer to it for detailed information about the different senses of “lei.” Here, I would like to mention that, “lei” used as a crucial concept in making judgment and reasoning is controversial about its exact meaning: is it a class or a kind, a universal or a particular, a collective whole of mass-stuffs or an abstract individual? In the literature of reference theory and philosophy of biology, it is still a hot debate on the concept of kind without conclusive answer to its ontological status. (Bird and Tobin 2018) Here, the question relevant to our concern in this chapter is: what is the exact meaning of the word “lei” in the rational thinking and reasoning of the ancient Chinese thinkers? Most of the scholars in Chinese philosophy and sinology in Chinese communities incline to treat lei as class rather than kind in understanding most logical reasonings in ancient China. In contrast, some representative Western scholars in Chinese philosophy such as A. C. Graham and Chad Hansen claim that it is mainly used to mean kind rather than class (Graham 1989: 148 and 417; Hansen 1992: 244–5). Graham and Hansen think that the criterion for identifying lei in most ancient Chinese texts is by the similarity between individuals or stuffs, not by the property to fix the membership between a class and its members. So, for them, lei is kind rather than class. They also think that lei as kind is not abstract but concrete and thus the relation between a kind in terms of a collective whole and a thing or stuff in terms of the distributive units of the whole is mereological, that is, their relation are that between the whole and its parts rather than the class and its members. Based on this presumption, they even claim that the relation in terms of a part-whole scheme implies a nominalist sense of naming and reference. So, they both reject abstract interpretation such as that of ontological realism.

11  Reason (Gu 故) and Principle (Li 理)

193

Unfortunately, both Graham and Hansen haven’t defined and explained “what a kind is.” In contemporary Western philosophy, it can be treated in an abstract sense as a class, an individual with essence, a universal, or in a concrete sense as a collection of individuals with some natural physical-chemical structure (for natural kinds) or with an origin of biological evolution (for species). But, no theories, as I know, treat kinds as mass-stuffs or stuff-kinds as claimed by Hansen. Moreover, for cultural and functional kinds, some Western philosophers even maintain that they are neither abstract nor concrete, but hybrid-entities. Here, I don’t want to enter into the discussion of this problem, but I want to mention that it is still a hot topic for debate and thus it is not easy, if not impossible, to choose one of these conceptions of kind to interpret all the occurrences of the word “lei” used for reasoning in ancient Chinese texts. I also want to argue that Hansen’s peculiar interpretation by virtue of “mass-stuff” or “thing-kind” for all uses of the term (Hansen 1992: 259–60, 408) is not only without textual evidence, but also without theoretical justification. Why Graham and Hansen’s interpretation is not acceptable? I think there are many difficulties in interpreting lei as their conception of kind with respect to judgment and reasoning in ancient Chinese texts. Here, for example, I want to mention one of the difficulties for Hansen’s mass-stuff interpretation. Based on his mass-­ noun hypothesis, Hansen thinks that a compound term refers to a compound stuff (jian 兼) while a single term refers to a corporeal unit (ti 體) or an incorporeal unit. His idea of compounding implies that tis, corporeal units, and incorporeal units are simple, basic and primitive. He treats ti as a mere stuff. However, he also says that, “red is the stuff that covers apples and the sky at sunset.” It means that [red-] apple-­ stuff is included in red-stuff. It is obvious that, here, he treats ti (say, red) as including compounding stuffs (say, red-apple) in its extension. (Hansen 1983: 35) On the other hand, in regard to Gongsun Long’s argument of “search horse,” Hansen uses an algebraic form to formulate the structure of the argument as “A is B; A is not-C; therefore B is not-C” (Yellow-horse is [identical with part of] horse; yellow-horse is not [identical with part of] white-horse; therefore horse is not [identical with part of] white-horse.) and regards it as “a simple and common deductive fallacy” (Hansen 1983: 161). In his interpretation of the above passage, he also treats the so-called horse-stuff as covering the so-called yellow-horse-stuff. But this kind of treatment (including treating “red” as covering “red-apple” and “horse” as covering “yellow-­ horse” in their extension respectively) is clearly inconsistent with his treatment of horse as a ti (a part) and yellow-horse or white-horse as a jian (a whole). That is: the former relation between ti and jian is “dividing down” while the latter relation between ti and jian is “dividing up.” Besides, most importantly, Hansen’s elaboration is incorrect. It is because the concluding sentence “white-horse is not horse” as the original wording in the text is different from Hansen’s “horse is not white-horse” and thus doesn’t fit to his term logic. So, I think there is no reason for Hansen to treat the above reasoning as an argument by means of term logic. In regard to the problem of dividing a large unit of mass-stuff into a small one, Hansen’s mass-like semantics and ontology is not coherent. If we, for the sake of argument, agree with Hansen to treat individual as not primitive in common-sense knowledge, but as derivative and as a special case of mass-stuff cutting, then,

194

Y.-m. Fung

according to Hansen’s view, we will have trouble to identify the ontological status of an individual. For Hansen, red as color-stuff is treated as covering [red-]apples and the sky at sunset, horse as shape-stuff should also be treated as covering different horse-shape-objects which have different qualities, including colors. So, red as a kind of color-stuff and horse as a kind of shape-stuff cannot be treated as two single mass-stuffs. It is because both mass-stuffs cover concrete objects or individuals in their extension respectively. In this regard, I think, his examples of ti cannot be treated as a single unit in comparison with jian as a compound unit and, more importantly, the referent of a single term, like “red,” “white,” “horse” or “stone,” cannot be treated as “ti,” a corporeal unit. If compounding is derivative in terms of making a compound (jian), such as white-horse and red-apple from two single units (ti), such as white-stuff, horse-stuff, red-stuff and apple-stuff, it is impossible for a ti to cover a number of concrete objects in its pool. So, compounding is not to make compound in a way of mass-sum (say, ox-horse) or mass product (say, white-horse), as claimed by Hansen, Instead, in most cases, a compound term is used to modify a relevant single term to narrow down its generality. If my analysis is right, then, to treat the term “lei” in classical Chinese as identical with the term “kind” in English and to interpret “kind” as meaning thing-kind in terms of mass-stuff is not a correct view. So, I think, at least in one of its senses, “lei” has the meaning identical with that of the term “class” in English. For example: 1. 小方,大方之類也;小馬,大馬之類也;小智,非大智之類也。(呂氏春秋:別類) (Refer to http://ctext.org/lv-shi-chun-qiu) A small square is in the [same] class of a big square. A small horse is in the [same] class of a big horse. But a shallow wisdom is not in the [same] class of a great wisdom. (The Annals of Lu Buwei CL: 1) (My translation) 2. 小圜之圜,與大圜之圜同。(墨子:大取) (Refer to http://ctext.org/mozi) The circle of a small circle is the same as the circle of a large circle. (Mozi, Daqu) (Major Illustrations or Choosing the Greater) (My translation) 3. 是以水流而涕從之者,其行類也。夫涕之與泣者,譬如人之兄弟,急則俱死, 生則俱生,其志以早悲,是以涕泣俱出而橫行也。夫人涕泣俱出而相從者,所 屬之類也。(黃帝內經:素問:解精微論) (Refer to http://ctext.org/ huangdi-neijing) The reason why snivel comes out together with tears is that they belong to the [same] class of shedding. The tears and snivel are just like brothers: They disappear together under critical condition and appear together under normal condition. When one’s zhi 志 (mind) appears sorrowful first, tears and snivel come out together. Tears and snivel come out simultaneously because they belong to the same class. (Yellow Emperor’s Canon of Medicine: Plain Conversation 81.6) (Li Zhaoguo’s translation 2005: 1291 with my minor revision) 4. 夫經言有餘者寫之,不足者補之。今熱為有餘,寒為不足。夫瘧者之寒,湯火 不能溫也,及其熱,冰水不能寒也,此皆有餘不足之類。(黃帝內經:素問:瘧 論) (Refer to http://ctext.org/huangdi-neijing) The Canon says that youyu 有餘 (excess) should be reduced and buzu 不足 (insufficiency) should be supplemented. [In malaria,] fever is youyu (excess) and

11  Reason (Gu 故) and Principle (Li 理)

195

cold is buzu (insufficiency). In malaria cold cannot be warmed even by hot water and fire. When fever runs, it cannot be cooled down even by icy water. These phenomena belong to the class of youyu (excess) and the class of buzu (insufficiency), respectively. (Yellow Emperor’s Canon of Medicine: Plain Conversation 35.17) (Li Zhaoguo’s translation 2005: 441 with my minor revision) 5 . 故五害之屬,傷殺之類,禍福同矣。(管子:度地) (Refer to http://ctext. org/guanzi) Therefore, the five disasters [i.e. (1) flood, (2) drought, (3) damage caused by wind, fog, hail and frost, (4) epidemic disease, (5) harmful insects] as subclasses [shu 屬] are in the class [lei 類] of casualities, they are same by virtue of misfortune and fortune. (Guanzi LIV: 2) (My translation). I think the word “lei” in these five examples cannot be interpreted as “kind” in terms of mass-stuff. Because, first, to classify two kinds or classes of wisdom, something which is not for dividing up or down, is not to distinguish between two kinds of mass-stuffs, let alone the problem of ambiguity of “wisdom” between “shallow wisdom” (smartness in trivial matters) and “great wisdom.” Second, even though a small circle and a large circle may be treated as two mass-stuffs, to regard them as the same circle cannot be meant that they have the same amount of massstuff. Third, the reason why tears and snivel come out simultaneously is that these two phenomena belong to the same class of shedding. Fourth, to make a distinction between youyu (excess) and buzu (insufficiency) and to assign getting fever and getting cold to fall under these two respectively is talking about membership, not about relationship between thing-kind and its thing-stuff (or mass-kind and its mass-stuff). Fifth, to make a classification of five disasters and to attribute them to the class of casualties is to treat the former as the subclass of the latter, talking about a class (lei 類) and its subclass (shu 屬). If the word “lei” in some contexts can be used to mean class, and, in regard to judgment and reasoning, to know that various things, events or cases are under the same class is crucial for the truth of the judgment and the validity of the reasoning. So, for most ancient Chinese thinkers, to know lei is most important for making their judgment and reasoning. Some examples of “zhi lei” 知類 (knowing/understanding class) are: 1. 指不若人,則知惡之;心不若人,則不知惡,此之謂不知類也。」(孟子:告子 上) (Refer to http://ctext.org/mengzi) When a man’s finger is not like those of other people, he knows to feel dissatisfied, but if his mind be not like that of other people, he does not know to feel dissatisfaction. This is called “ignorance of the class [to which both events belong].” (Mencius XI: 12) (Translation: Legge 1861b) 2. 故法者,國之權衡也,夫倍法度而任私議,皆不知類者也。(商君書:修權) (Refer to https://ctext.org/shang-jun-shu) Therefore, law [fa 法] is the standard scales of the nation. If we violate the models and measures and use private appraisal, we do not know the class [of all the cases which are under the standard of law]. (Book of Lord Shang XIV: 2) (My translation)

196

Y.-m. Fung

3. 義不殺少而殺眾,不可謂知類。(墨子:公輸) (Refer to http://ctext.org/mozi) To hold the principle of righteousness (yi 義) that forbids the killing of few but allows that of many [such as to attack the nation Song 宋] cannot be said to be knowing of the [same] class. (Mozi L: 1) (Translation: Y. P. Mei 1929, with my minor modification). For Later Mohists, the necessary condition for identifying the difference in classes or of kinds for two things or events is that they do not share a relevant or designated property or attribute. They think that “to know lei 類 (class/kind)” and “to understand gu 故 (reason/cause)” are necessary for making inference. If one uses a wrong example to illustrate the sameness or difference between two terms, i.e. giving a wild example (kuang ju 狂舉), one cannot construct an acceptable parallel (mou 侔) inference. Because, if one misidentifies different things or events as of the same class, what one can provide as a relevant reason to support the assertion of an event of one class is not the same reason to support the assertion of another event of another class in a parallel inference, say, “A thief is a man; killing a thief is killing a man.” In this example, “killing a thief” and “killing a man” are two events of two different classes in the sense that the term “killing” is ambiguous: “killing” in terms of “execution” versus “killing” in terms of “murder.” If these different things or events seem to be in the same class but actually not, the different referents of the relevant terms cannot be identified as covered by the same concept though the terms look like expressing the same concept as required in the inference. So, in the Mozi, Canon II: 2, Later Mohists stress “the difficulty of the generalization of class” (tuilei zhi nan 推類之難). In Canon and Explanation I: 86 and I: 87, although they give a very sophisticated analysis of the different concepts of “similarity” and “difference” for the identification of various relations including the sameness and difference of classes,4 they still warn us to strictly abide by the principle of “different classes are not comparable” (yi-lei bu bi 異類不吡[比]) (Mozi, Canon II: 6).5 In the second paragraph of Xiaoqu 小取 (Minor Illustrations or Choosing the Lesser) of the Mozi, Later Mohists defines “parallel inference” (mou 侔) as a kind of inference based on comparing two propositions both of which can be maintained (比辭而俱行). However, in the third paragraph, they give a serious warning to people who want to use this or a similar kind of inference: 夫物有以同而不率遂同。辭之侔也,有所至而正。其然也,有所以然也;其然也同,其所 以然不必同。其取之也,有所以取之。其取之也同,其所以取之不必同。是故辟、 侔、援、推之辭,行而異,轉而危,遠而失,流而離本,則不可不審也,不可常用也。故言 多方,殊類,異故,則不可偏觀也。(墨子:小取) (Refer to http://ctext.org/mozi) Of things in general, if there are respects in which they are the same, it does not follow that they are altogether the same. The parallelism of propositions is valid only as it reaches. If

4  Please refer to Chap. 13 of this volume for a detailed description of the four different concepts of “sameness” and the four different concepts of “difference.” 5  Some incomparable examples mentioned by Later Mohists are: “Which is longer, a piece of wood or a night? Which do you have more of, knowledge or grain? Which is the most valuable, aristocratic rank, one’s own parents, right conduct, a price? ….” (木與夜孰長?智與粟孰多?爵、親、 行、賈四者孰貴?…) (Explanation II 6) (Graham 1978: 357) Other examples can be found in Canon and Explanation II: 2 (Graham 1978: 349).

11  Reason (Gu 故) and Principle (Li 理)

197

something is so of them there are reasons why it is so; but though its being so of them is the same, the reasons why it is so are not necessarily the same. If we accept a claim we have reasons for accepting it; but though we are the same in accepting it, the reasons why we accept it are not necessarily the same. Therefore propositions of analogy, parallelism, citing a similar example, and extending to an isomorphic example, become different as they proceed, become dangerous when they change direction, fail when carried too far, become detached from their base when we let them drift, so that we must on no account be careless with them, and must not use them too rigidly. Hence language functions in many ways, is of distinct classes, and with different reasons, which must not be looked at only from one side. (Mozi, Xiaoqu: 3) (Graham 1978: 482–5 with minor revision)

Here, my understanding of the message is that: when we make a parallel inference or the three other kinds of inference (pi 譬 analogy, yuan 援 citing a similar example, and tui 推 extending to an isomorphic example), we should be very careful about the problems of the various ways of usage, of the distinct classes or kinds, and of the different reasons or causes, in using language, if we are required to use the same concept for the various occurrences of a term in the inference. Although Later Mohists cannot justify which parallel argument is valid and which other one is invalid by means of a formal proof, they still can roughly distinguish the logical status of various arguments by some concrete and paradigmatic examples which are very easy for people to recognize or confirm their acceptability or validity by people’s linguistic intuition or semantic sensibility. But this method cannot be used to identify all the valid arguments and sometimes cannot be used to exclude an invalid argument which comes from a wild example or crazy demonstration (kuang-ju 狂舉). So, to escape from the difficulty coming from a wild example, in addition to this first level illustration, i.e. recognition or confirmation by paradigmatic examples, Later Mohists stress the necessity of comparing the paradigmatic examples with some anomalies, i.e. the second level reflection or reconfirmation. Equipped with the mechanism of these two levels of examination, Later Mohists are able to offer a comparable scheme similar to a sort of model theory of proof, i.e. a procedure of proof based on some comparable schematic examples which can be used as normal interpretations to support an argument.6 Since the linguistic ­sensibility of the first level illustration and the reflective thinking of the second level comparison are grounded on the semantic content and pragmatic convention of linguistic expressions, in this sense, the logical demonstration suggested by Later Mohists is not formal or syntactic, but semantic and pragmatic. 6  A model of a sentence (form) S is an interpretation of S which can make S true or false. For example, if S has the form (∀x)(Fx→Gx) and Fx and Gx are interpreted as “… is a man” and “… is an animal” respectively, the sentence “Every man is an animal” or “Whoever is a man is an animal” is a true interpretation or true model of S. If Fx and Gx are interpreted as “… is a man” and “… is a monkey” respectively, the sentence “Every man is a monkey” or “Whoever is a man is a monkey” is not a true interpretation or true model of S. Since a particular model or interpretation of an argument form can be true for the premise(s) and false for the conclusion, it can be used as a counter example to prove the invalidity of an argument. However, a limited number of models normally cannot be used to prove the validity of an argument and its form in the language of predicate logic though a counter example has not been found. Detailed information about model theory can be found in Hodges 2009.

198

Y.-m. Fung

In general, lei is related to gu and li (or with the explicit form of fa) in making judgment (yi 議), reasoning (lun 論) and disputation (bian 辯). So, using an appropriate lei in an argument is crucial for making it valid. But, what are their logical roles in this regard? I will try to give an answer in the following section.

4  The Logical Roles of Gu, Li and Lei Xunzi often stresses the importance of gu and li in making views and argumentations. So, he thinks, the acceptance of a view or argumentation is based on that “the view is held with gu and is expressed in accordance with li” (Xunzi VI: 2–6) (其持 之有故,其言之成理。) (荀子:非十二子). To ask for a gu is to ask for a reason for justifying a view or judgment. But not all reasons are acceptable. Only those grounded on an objective or intersubjective li can be accepted as a good reason. In other words, the antecedent of the conditional which is used to express a gu is required to ground on an objective or intersubjective li, not on a li generated from personal or subjective opinion, otherwise it cannot be accepted as a good reason. In addition to the problem whether a view or judgment under debate is supported by a gu which is grounded on an objective or intersubjective li, whether the view or judgment under debate is covered or constrained by the li is another problem. For example, “to enter a country and join with its traitors is not an action in the class of righteousness” (Mozi XXXIX: 8) (入人之國而與人之賊,非義之類也。) (非儒 下). It means that to attack a country is not an action of the class to which the principle of righteousness is applicable. On the other hand, for the Mohists, to kill a dictator is an action of the class to which the principle of righteousness is applicable. In general, in argumentation or disputation, the first thing is to make a sentence (ci 辭) to express a view or judgment, the second thing is to ask for a supporting reason (gu) which is expressed in a conditional, and the third thing is to assert an acceptable principle (li) as the ground of the supporting reason. However, sometimes, the acceptability of the principle is controversial. In this case, a further disputation or argumentation on the principle is required. Once the problem is solved, people can distinguish what is the right li from the wrong one and thus would treat the right one as objective or intersubjective. Let’s illustrate the logical roles of these three concepts in making judgment and reasoning with the following example based on the Mohists’ idea of “fei gong” 非 攻 (condemning offensive warfare): 1. To attack a country (such as a big country to seize a small one by military attack) is not allowed. (judgment: ci) 2. To attack a country (such as a big country to seize a small one by military attack) causes damage to the mutual utilities of the society, so, it is not allowed. (reason: gu)7 7  If an argument “A├ B” is valid or “A logically implies B,” it means that the assertion of the conditional sentence (A→B) is true or that the sentence “A materially implies B” is true.

11  Reason (Gu 故) and Principle (Li 理)

199

3. To cause damage to the mutual utilities of the society is an action in the class of not righteous. (class: lei) 4. The principle of righteousness is “mutual love entails mutual utility.” (principle: li) But, why, in ancient times, King Yu 禹 attacked You-miao 有苗, King Tang 湯 attacked Jie 桀, and King Wu 武 attacked Zhou 紂, yet all of them were later honored as sage kings? It seems that both actions violate the principle (4). It seems also that this kind of reasoning sometimes encounters count-example and thus is not reliable. When there seems a conflict between the Mohists’ ethical principle such as jian-ai 兼愛 (universal love) and some (counter-) examples in a parallel reasoning, such as “a thief is a man; killing a thief (or a dictator) is not killing a man (or dictator).” (Mozi, Minor Illustrations (Xiaoqu 小取): 5), how can they solve the problem? Hansen’s answer is to appeal to an interpretation of this kind of example as in an intensional or opaque context, and thus he regards this as the main reason for Later Mohists to reject the inference rule. So he makes the following point in a long passage: (Hansen 1983: 132–3) The end purpose of the attack on the reliability of “sentence matching” is apparently to disfuse attacks on the Mohist view that “to kill a thief is not to kill a man.” They saw their opponents’ attack on this view as being based on the application of the inference rule in question to the identity sentence [my italic], “Thief is man,” which Mohists accepted. We have a way of formulating their view that has been common to the earlier examples. “Thief is man” is based on the ming “name” shih “stuff” level of semantics. Kill thief and kill man are phrases conveying actions (intents or purposes). The identity of the formal designation of a term in such a phrase does not make the intentions expressed by the phrases identical. This way of explaining the failure of the inference is confirmed by the last two examples, which overly introduce propositional attitudes (hating and wishing). One must note in passing that if the Mohist did indeed have some notion of the intention, sense, or idea of a term as opposed to its extension or denotation, this would have been a natural place for it to have been employed. But the Mohist does not offer any such analysis. The first example in this segment is a classic fallacy of division. It seemingly has nothing to do with purposive i “intent.” The exposition does not give any special diagnosis of the breakdown of the inference, it simply observes that the inference fails in such cases. Sentence matching is, accordingly, not a totally reliable rule, and Mohists should not be criticized for violating it in asserting that “kill thief’ is not “kill man” while accepting “thief is man.”

I don’t think Hansen is right to interpret or identify the sentence “[A] thief is [a] man,” as an equation or identity sentence. I think there is no real problem about the failure of substitution between “killing [a] thief” and “killing [a] man” in the so-­ called intensional context, and thus no real problem of referential opacity. It is because “thief” and “man” is not co-extensive for substitution and, for the Mohists, the two tokens of “killing” in “killing a thief” and “killing a man” are not expressed as the same concept. Moreover, Mozi has made a distinction to explain why killing people or killing people through military attack is unjust or not righteous and killing a thief or executing a dictator is just or righteous in the following dialogue: 今遝夫好攻伐之君,又飾其說以非子墨子曰:「以攻伐之為不義,非利物與?昔者禹征 有苗,湯伐桀,武王伐紂,此皆立為聖王,是何故也?」子墨子曰:「子未察吾言之類,未明 其故者也。彼非所謂攻,謂誅也。(墨子:非攻下) (Refer to http://ctext.org/mozi) Those warlike lords attempt to defend their views by reproaching Master Mozi: “Do you regard offensive warfare as unrighteous and unprofitable? In ancient times, King Yu attacked

200

Y.-m. Fung

You-miao, King Tang attacked Jie, and King Wu 武 attacked Zhou, yet all of them were later honored as sage kings. How can you explain this?” Master Mozi said: “You neither clearly aware the class (lei) of what I said to refer to nor understand the reason (gu) behind it. What the sage kings did was not ‘to attack’ (gong 攻) but ‘to punish/execute’ (zhu 誅).” (Mozi XIX: 4) (Wang and Wang 2006: 159–161, with my minor revision)

Mozi’s argument is that: All actions or events A1, A2, A3 … and all other actions or events B1, B2, B3 … are military attacks. But the first group of actions or events is different from the second group of actions or events because the former is classified as gong 攻 (attack) while the latter is classified as zhu 誅 (execution). Since gong is not in accordance with the principle (li 理) of yi 義 (righteousness) while zhu is, this is the gu 故 (reason) why Mozi approves the latter instead of the former. For the Mohists, military attack for invasion and military attack for taking over a dictator are not instances in the same class of righteous actions. If killing a dictator by execution is in accordance with the principle of righteousness, it would be identified as a member in the class or domain of righteous actions. So, unlike the instance of killing a dictator by execution, the instance of invasion and killing people by military attack cannot be treated as belonging to the class of righteous actions. Both Early and Later Mohists claim that people should be strictly taking their action in accordance with this principle of righteousness and very cautious in making parallel inferences. Here, Mozi distinguishes the concept of “attack” (gong 攻) from that of “execute” (zhu 誅) though both terms seem to refer to the same kind of actions with military force (Against Military Attack III: 5). He thinks that the former as an aggression with military force is unjust or not of the same class of righteousness (fei yi zhi lei 非義之類) (Against the Confucians B9, Johnston, 360–1) while the latter as a punishment with military force is just or righteous. So, if we make an acceptable statement with the latter concept, it does not follow that we can make a similar permissible statement with the former concept. Based on this reason, Mozi criticizes those who “do not know the class to which my words belong” and “do not understand the reason on which my words are based” (子未察吾言之類,未明其 故。). (Mozi XIX: 5) In other words, the argument criticized by Mozi is based on a wild example. The distinction between the instance of attack for invasion and the instance of attack for execution is that the former is based on the principle (li) of righteousness (yi 義) which entails mutual or universal utilities, while the latter is based on the principle of private or personal benefit (sili 私利). The different causes or conditions (gu) of them can be explained in terms of the different consequences led by them which exemplify that it is an action for private or personal benefit or for mutual or universal utilities. Based on the different reasons and causes, we can distinguish them as belonging to different classes (lei) of actions. Mozi thinks that no matter how serious the crime made by a thief or warlike lord they should be punished or executed. But punishment or execution is not the same thing as murder or killing an innocent. So Later Mohists are consistent in claiming that “killing a thief or dictator [by execution]” is not “killing a man [by murder].” The two events are not of the same class. The two tokens of the term “killing” here are not the same concept and also are not appearing in an intensional context; they

11  Reason (Gu 故) and Principle (Li 理)

201

are different concepts in an extensional context. It is obvious that the parallel inference with validity requires that the adding terms, such as “killing” and “loving,” in the concluding sentence should be used as the same concept or what they refer to are events of the same class. Since the relevant terms used in the above sentences cannot meet this requirement, so the sentences embedded with these terms cannot constitute a valid argument as that in the paradigmatic example of “A white horse is a horse; to ride a white horse is to ride a horse.” in which the two tokens of “ride” mean the same. However, this (about the two senses of “killing”) and other pairs of sentences are not used by Later Mohists as counter-examples to discredit the validity of the parallel inference but as anomalies to remind scholars to be careful in using parallelism, to have ample attention to the semantic ambiguity and pragmatic deviation of terms which appear in the relevant sentences, and to suggest them not to use these terms in forming their parallel inference.

References Bird, Alexander, and Emma Tobin. 2018. “Natural Kinds.” In Edward N. Zalta, ed., The Stanford Encyclopedia of Philosophy. URL https://plato.stanford.edu/archives/spr2018/entries/ natural-kinds/. Brandom, Robert. 2000. Articulating Reasons: An Introduction to Inferentialism. Cambridge, MA: Harvard University Press. Davidson, Donald. 1963. “Actions, Reasons, and Causes.” The Journal of Philosophy 60.23: 685–700. Davidson, Donald. 1967. “Causal Relations.” The Journal of Philosophy 64.21: 691–703. Duyvendak, J.J.L. 1928. The Book of Lord Shang. London: Arthur Probsthain; reprinted (1963), Chicago: University of Chicago Press. Fodor, J.A., et al. 1980. “Against Definitions.” Cognition 8: 263–367. Fung, Yiu-ming. 2007. “A Logical Perspective on ‘Discourse on White-Horse’.” Journal of Chinese Philosophy 34.4: 515–536. Graham, A.C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: The Chinese University Press. Graham, A.C. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle: Open Court. Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. New York: Oxford University Press. Hodges, Wilfrid. 2009. “Model Theory.” In Edward N.  Zalta, ed., The Stanford Encyclopedia of Philosophy (Fall 2009 Edition). URL http://plato.stanford.edu/archives/fall2009/entries/ model-theory/. Kim, Jaegwon. 1976. “Events as Property Exemplifications” In Brand M. and D.  Walton, eds., Action Theory, 159–177. Dordrecht: Reidel; reprinted in Kim 1993. Supervenience and Mind, Cambridge: Cambridge University Press: 33–52. Knoblock, John (translation). 1999. Xunzi. Hunan People’s Publishing House and Beijing: Foreign Languages Press. Legge, James (translation). 1861a. The Analects of Confucius. In The Chinese Classics with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes: Volume 1.

202

Y.-m. Fung

Hong Kong: at the author’s; London: Trübner & Co. Chinese Text Project. http://ctext.org/ analects. Legge, James (translation). 1861b. The Works of Mencius. In The Analects of Confucius. In The Chinese Classics with a Translation, Critical and Exegetical Notes, Prolegomena, and Copious Indexes: Volume 2. Hong Kong: at the author’s; London: Trübner & Co. Revised second edition (1895), Oxford: Clarendon Press. Reprinted by Dover Books in 1990 Dover Publications. Chinese Text Project, http://ctext.org/mengzi. Legge, James (translation). 1885. The Book of Rites. In Sacred Books of the East, vol. 28, part 4: The Li Ki. Chinese Text Project. http://ctext.org/liji. Li Zhaoguo (translation). 2005. Yellow Emperor’s Canon of Medicine: Plain Conversation. Xi’an/ Beijing: World Publishing Corporation. Liao, W.K. (translation) 2015. Han Fei Zi. Beijing: The Commercial Press. Mei, Y.P. 1929. The Ethical and Political Works of Motse. London: Probsthain. Chinese Text Project. http://ctext.org/mengzi. Mellor D.H. 1995. The Facts of Causation. London: Routledge. Pitt, David. 1999. “In Defense of Definitions.” Philosophical Psychology 12.2: 139–159. Wang, Rongpei, and Hong Wang. (translation) 2006. Mozi. Hunan People’s Publishing House. http://ctext.org of Chinese Text Project, edited by Dr. Donald Sturgeon. Yiu-ming Fung  is Emeritus Professor at the Hong Kong University of Science and Technology and currently Visiting Professor in the Department of Philosophy at the Tunghai University, Taiwan. He is an analytic philosopher with special interest in Chinese philosophy and comparative philosophy. Fung is the author of several books, including The Methodological Problems of Chinese Philosophy (1989); Chinese Philosophy in the Ancient Period, 4 volumes (1992); The Gongsun Longzi: A Perspective of Analytic Philosophy (1999); and The Myth of Transcendent Immanence: A Perspective of Analytic Philosophy on Contemporary Neo-­Confucianism (2003). He has also published more than 100 research papers both in Chinese and English.

Chapter 12

Classes (Lei 類) and Individuals Wujin Yang and Wanqiang Zhang

The concepts of “class” and “individual” play special roles in ancient Chinese logical theory of classification. From the view of etymology, in the Chinese dictionary named Shuo Wen Jie Zi 說文解字, the word “class” is generally interpreted as: “Things of a species are similar, the most salient example is the dog” (zhong lei xiang si, wei quan wei shen. 種類相佀[似], 唯犬為甚。). It means that if things are of the same class, they must have some similar properties. For example, the canine is one of the typical cases. In the same Chinese dictionary, the word “individual” is understood as private object or singleton and its name as “private (name).” In general, “name” (ming 名) means concept or word in classical Chinese. In answering the question how to classify names, the Mohists suggested that names could be classified into three types, namely “the extended name” (da ming 達名), “the class name” (lei ming 類名) and “the private name” (si ming 私名) in the Mohist Canons (Mojing 墨經). Similarly, in the chapter Zhengming 正名 (the Rectification of Names) of the Xunzi 荀子, Xunzi also classified names into different kinds, including “the great general name” (da gong ming 大共名), “the general name” (gong ming 共名), “the great particular name” (da bie ming 大別名), “the particular name” (bie ming 別名), and “the name without particularity” (wu bie zhi ming无别之名), etc. In the following, first, we will discuss the theory of classifying “names” in the Mohist Canons and that in the Xunzi, especially in the chapter Zhengming. Second, we will examine the concepts of “class” and “individual” in ancient Chinese logical thought, with special reference to the meaning and feature of the Chinese word “lei.” W. Yang (*) School of Philosophy, Renmin University of China, Beijing, China e-mail: [email protected] W. Zhang Department of Philosophy, Xidian University, Xi’an, China e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_12

203

204

W. Yang and W. Zhang

1  T  he Concepts of “Class” and “Individual” in the Mohist Canons According to Mohists, “names” are used to refer to various kinds of things. In the Canon Explanation A, the Mohists said: “To speak: Speaking is said of the month’s ability to bring forth names. A name is like a picture of a tiger. Speaking is to say. Speech is what is achieved through names” (Johnston 2010: 401) (言也者, 諸口能 之, 出名者也。名若畫俿也。言, 謂也。言猶名致也。) (經說上). Here, a name in a speech is used to name or to refer to an object or a collection of objects. As described by the Mohists: “What something is called by is its name. What is called by the name is the object” (Canon Explanation A 81) (所以謂, 名也; 所謂, 實也。) (經說上). This is about the relation between the name and what is named. Besides, “names” can be divided into different categories based on different criteria. According to the scope of a concept’s extension, as mentioned above, “names” can be divided into three types, namely, “the extended name,” “the class name” and “the private name” as described in the following passage: 名:達、類、私。(經上) Ming (name): the extended (da 達); the class (lei 類); the private (si 私). (Mohist Canon A 79) 名:物, 達也, 有實必待之名也命之。馬, 類也, 若實也者, 必以是名也命之。臧, 私也, 是名也止於是實也。聲出口, 俱有名, 若姓字麗。(經說上) Ming (name): “Thing” is an extended name. Any object necessarily requires this name. “Horse” is a class name. For similar objects [of horse], we necessarily use this name. “Jack” is a private name. This name stays confined in this object. All the words issuing from the mouth all are names, like the pairing of surname and style-name. (Canon Explanation A 79)

The extended name is of the largest extension, summing up all such objects that exist in the universe. For example, “thing” (wu 物) is an extended name. The scope of the class name is smaller than that of an extended name in the sense that the former only covers all the objects belonging to a particular collection or set. The private name is of the narrowest extent of what a name might express, pointing only to a specific entity. Using as a general concept, the class name covers all the objects which belong to one class of same nature and property. For instance, the name “horse” refers to all horses which have the same nature or property. So, the class identified by a class name is an abstraction and generalization of individuals of the same or similar property which reflects a kind of abstract thinking of the ancient Chinese philosophy. According to the name which refers to the whole or the part of one thing, “names” can be divided into “the collective name” (jian ming 兼名) and “the part name” (ti ming 體名). According to the Mohists, “a part” can be defined as “a division of a collection” (Mohist Canon A 2) (體, 分於兼也。) (經上). “Not being outside the whole is the sameness of being a part”(Mohist Canon A 87) (不外於兼, 體同也。) (經說上). For example, a person’s hands and feet are the main parts of his whole

12  Classes (Lei 類) and Individuals

205

body. There is one significant sentence to illustrate the part-whole relationship of names in Canon Explanation B 67. That is: “The statement that ox-horse is not-ox and the statement that ox-horse is ox are both right and the same. The explanation lies in the collective meaning” (牛馬之非牛與可之同, 說在兼。) (經說下). In this sentence, “ox-horse” is “a collective name” which refers to the collective whole of all oxen and all horses. Here, “ox” or “horse” is only used as “a part name” which refers to one kind of animal, ox or horse, respectively. Moreover, when the Mohists expounded the relationship of sameness and difference between things, they also talked about classes and individuals. They gave a definition to the sameness through the difference between things. The Mohists said, “Being the same is being different but both in this aspect being one” (Mohist Canon A 39) (同, 異而俱於之一也。) (經上). For example, “Two people, yet both see that this is a pillar” (經說上). It means that sameness is the common aspect of different things. In fact, the Mohists insisted that there were different types of sameness, such as different names referring to the same object (e.g. “Confucius” 孔子 and “Zhongni” 仲尼 refer to the same man, namely, Confucius.); different parts of the same thing (e.g. one’s hands and feet are different parts of the same body.); many people in the same room; different things with the same property (e.g. a white horse and a white feather have the same color); different things having a common origin or causing a common result, etc. According to the Mohists, all these mentioned above are various senses of sameness between things. Besides, the Mohists also insisted that: “Two objects are necessarily different” (Canon Explanation A 88) (二 必異) (經說上). It means that the difference between two distinct things is regarded as inevitable and self-evident. In fact, a private name just refers to one thing which is necessarily different from other things. The Mohists also gave a definition of sameness in terms of things being included in a class (lei tong 類同) and a definition of difference in terms of things being not included in a class (bu lei zhi yi 不類之異) in the Canon Explanation A 87: “The sameness of things in terms of their being included in a class” is “being the same in some respects” (you yi tong 有以同), and “the difference of things in terms of their not being included in a class” is “not having what is the same in some respects” (bu you tong不有同). If two distinct things have the same or similar property, they could be treated as belonging to one class. For example, both ox and sheep have horns, so they can be regarded as belonging to the same class of animal which is called “animal with horns.” If two things don’t have the same or similar property, they could not be treated as belonging to the same class. Obviously, “the sameness in terms of things being included in a class” and “the difference in terms of things not being included in a class” are related to whether things have the same or similar property. The Mohists gave a further explanation to “the sameness in terms of things being included in a class” in the Mohist Canon B 64. That is: “When things put together under one criterion they exhaust a class, like putting all squares together. The explanation lies in squareness.” (一法者之相與也盡類, 若方之相合也, 說在 方。) (經下). This explanation means that things which are identified by the same

206

W. Yang and W. Zhang

standard or criteria are of the same class. For example, square wood, square stones, and square casks are different matters in nature, but they have the same square shape, so they could be treated as belonging to one class of square things. Besides, the Mohists also gave a further explanation to “the difference in terms of not being included in a class” in the Mohist Canon B 65. That is: “‘wild raising’ is inadmissible as a means of knowing differences. The explanation lies in their being inadmissible” (狂舉不可以知異, 說在有不可。) (經下). This explanation means that if we want to define the difference in terms of things not being included in a class, we must follow the principle that “anything of one class has the same property, while anything of another class does not.” For example, if we use “ox has incisors” and “horse has a tail” to distinguish ox and horse and thus say that they are not belonging to the same class, it may violate the above principle and thus is an incorrect illustration or “wild raising.” Because both ox and horse have incisors and tail. On the other hand, if we use “ox has horns” and “horse does not have horns” to distinguish ox and horse and thus say that they are not of the same class of animal, it is also a wrong illustration or “wild raising.” Because ox and horse may have the same property in other respects, such as they have four legs. According to Mohists, the difference between ox and horse is just that they are distinct individuals of the same class of animal. In the Mohist argumentation, they put much emphasis on the distinction between different classes. One of the salient cases is that: they distinguished clearly that “attack” (gong 攻) and “punishment” (zhu 誅) are not of the same class of behavior. The former is not righteous, while the latter is. For example, Da Yu 大禹 punished the You Miao 有苗, Shang Tang 商湯 punished Xia Jie 夏桀, and King Wu 武王 punished King Zhou 紂王 in ancient Chinese history should be considered as “punishment,” while the conquest wars of princes in Spring and Autumn Period are interpreted as “attack.” So, we should condemn the behaviors of “attack” because the behaviors are without righteousness; but we should not condemn the behaviors of “punishment” which are righteous. Obviously, to distinguish “attack” and “punishment” is to ensure full compliance with the principle that “anything of one class has the same property, while anything of another class does not have the same property.” Namely, any behavior of “attack” is not righteous, while that of “punishment” is righteous. So, According to the Mohists, we should condemn the behaviors of “attack” and support those of “punishment.” In summary, in the Mohist text, the Mohists provided many examples for classifying different types of names and distinguishing different conceptions of sameness and difference between things, especially the sameness in terms of things being included in a class and the difference in terms of things not being included in a class. The concepts of “class” and “individual” in the Mohist Canon are all illustrated on the basis of different circumstances of using names to refer to objects. The Mohists emphasized that object precedes over the name, naming must be grounded on the object, and observing name should be through entity.

12  Classes (Lei 類) and Individuals

207

2  The Concepts of “Class” and “Individual” in the Xunzi In the chapter Zhengming, Xunzi expounded his understanding and classification of names, and, particularly, explained the concepts of “class” and “individual.” In his opinion, most of the names have political and legal meanings in forming professional vocabulary for the socio-political system in ancient China, such as official names and title names of nobility, etiquette names and names of rites, penal names and names of law, and other disperse names (san ming 散名) which are used to denote all kinds of things or events. He thought that names could be divided into two types according to the composition of words: one is the type of simple names (dan ming 單名) which are constituted of one character, such as the name “ox” (niu 牛) or “horse” (ma馬). The other is the type of compound names (jian ming 兼名) which are constituted of two or more characters, such as the name “white horse” (baima 白馬) or cockfighting (douji 鬥雞). He also said, “When a simple name is sufficient to express the meaning, a simple name is used. When a simple name is insufficient, then a compound name is used.” (單足以喻則單; 單不足以喻則兼。) (Chan 1969: 126).1 Moreover, if we notice the different kinds of objects could be denoted by different names in the Xunzi, we know that names could be divided into different categories: the class name is used to refer to a class of things (e.g. “ox” and “horse”, etc.), the proper name is used to refer to any individual entity (e.g. “Mozi” and “Confucius,” etc.), the name of natural things is used to refer to natural things (e.g. “wood” and “stone,” etc.), and the name of values in society is used to refer to norms and values in human society (e.g. “ruler” and “minister,” etc.). Therefore, we think that names in the Xunzi include almost all terms of individuals and classes, objects and properties, abstract and concrete entities, as well as facts and values of lexical items (Chen 2008: 27). Moreover, Xunzi divided names into “general names” and “particular names” based on the scope of objects denoted by names. In the chapter Zhengming, Xunzi said that (Chan 1969: 126): 萬物雖眾, 有時而欲遍舉之, 故謂之物。物也者, 大共名也。推而共之, 共則有共, 至 於無共然後止。有時而欲偏舉之, 故謂之鳥、獸。鳥、獸也者, 大別名也。推而別之, 別則有別, 至於無別然後止。(荀子正名) Although the myriad things are innumerable, sometimes we want to speak of them as a whole, so we call them “things.” “Things” is the great general name. We carry the process further and generalize. In generalizing, we find more things to generalize. We go on and will not stop until there is nothing more general. Sometimes we want to speak of one section of things (e.g. animals), and so we call them “animals.” “Animals” is the great particular name. We carry the process further and particularize. In particularizing, we find more things to particularize. We go on and will not stop until there is nothing more particular.

1  Here, we need to pay more attention to the difference between Xunzi’s compound name and the Mohist collective name, because both of them were called “jianming” 兼名 in classical Chinese language. For example, the name “white horse” is a compound name in the Xunzi, while not a collective name in the Mohist Cannons.

208

W. Yang and W. Zhang

According to Xunzi’s view, generally, the general name is used to refer to things with different levels of scope or extension. The highest level is “the great general name” (the greatest genus concept) which means “there is nothing more general.” He gave one example of the great general name, that is the name “things” (wu 物) which refers to every entity in the world and is equivalent to the extended name (da ming 達名) in the Mohist Canons. The less general one is “the particular name” which refers to something with different levels of generality: the highest level is “the great particular name” which is used to distinguish different kinds or classes of things and is similar to the class name in the Mohist Canons; The lowest level is the name without further specification which is used to refer to individuals and is equivalent to the proper name. The example of “the great particular name” is “birds” and “animals,” and the example of “the name without further specification” is “Mozi” and “Confucius,” as given by Xunzi. Obviously, if we classify names according to the scope of a concept’s extension based on Xunzi’s theory, we will find that “the great general name” has the maximum extension and “the particular name without further specification” has the minimum extension. What is between them is other kinds of “general name” and “particular name.” This means that the generality and particularity of a name are relative to the scope of extension of other names. For example, in comparison, the name “animal” as a “general name” is more general than the name “human” while as a “particular name” is less general than the name “biological being.” Therefore, in Xunzi’s theory, there is a hierarchy among different class names and individual names.

3  T  he Concept of “Lei” in Ancient Chinese Philosophy of Logic In fact, in addition to the Mohist Canons and the Xunzi, almost all ancient Chinese philosophers or thinkers often used the concept “class” (i.e. “lei”) to form or express their argumentations. The concept of “class” plays important roles in ancient Chinese logical thought, in the sense that it is one of epistemological and semantic bases for making and developing argumentations in ancient Chinese philosophy (Wu 1980: 51–75). We think that to grasp the concept of “class” in ancient Chinese logical thought needs to understand, in advance, the literal meaning of the character “lei” 類 in classical Chinese language and the general usage of the concept in the history of Chinese thought. From the view of etymology, the word “lei” first appeared in the ritual which was used as a proper name referring to the ceremony of emperor’s expedition or ascending the throne. For example, the character “lei” in the sentence “si lei shang di” (肆類上帝) recorded in Yao Authority of the Book of History (Shangshu Yaodian尚書堯典) could be translated into that the king Yao 堯 held ritual for the Heaven. So, the Chinese character has both moral and ritual significance. It also has the meaning in terms of ethnicity which represents human race from the same ancestors, such as the character in the proverb “Someone is not my race, his heart would be different” (fei wo zu lei, qi xin bi yi. 非我族類, 其心必異。).

12  Classes (Lei 類) and Individuals

209

Moreover, the word “lei” could be used to mean the same class (tong lei 同類), say, similarity or resemblance in human. In this sense, the character refers to family name shared by all members of the same family (Wu 1981: 64–84). In summary, we think that the character “lei” in pre-Qin times could be distinguished into different categories, such as natural classes in biological sense (e.g. “ox” and “horse” are not the same class of animal.), classes in ethic sense and moral sense (e.g. “gentlemen” and “mean men” are not the same class of human in the sense that if one has a kind of value, he or she should have certain kind of behavior.), and classes in logical and philosophical sense. For example, five elements in ancient Chinese natural philosophy represent five classes of things with the feature of “family resemblance” in nature: these elements are classes of gold, wood, water, fire and earth in an abstract sense. As we know, there is a controversial question about the meaning of “lei” in ancient Chinese philosophy. Some scholars think that it is not equivalent to the Western concept of “class” and its translation is “kind” rather than “class.” But some other scholars do not fully agree with this view and think that some of the uses of the “lei” are about classes. To answer this question, we think we could make a comparison between the concept of “lei” in the ancient Chinese philosophy of logic and the concept of “class” in modern Western logic. Many scholars have been discussing this question from the 1960s to now, especially some sinologists who are good at studying ancient Chinese logical thought. Of course, this question is open and there are various answers to it. For example, in Janusz Chmielewski’s famous paper named “Notes on Early Chinese logic” (I) which was written in the 1960s, he thought that the word “lei” used by Gongsun Long and the Mohists in pre-Qin times was the same concept of “class” in modern Western logic (Chmielewski 1962: 7–22). But A. C. Graham insisted that these two were not the same in essence and that “class” in modern logical thought was understood in a realistic sense, while “lei” in ancient Chinese philosophy was a nominalist concept which mostly represented the similarity between things (Graham 2003: 336). Christoph Harbsmeier held that “lei” in ancient Chinese logical thought was not an abstract set or class in mathematical logic, but a basic cultural concept of understanding and arguing models in ancient Chinese philosophy (Harbsmeier 1998: 218–229). On the other hand, Thierry Lucas understood “lei” neither as the concept of “class” in modern logical thought nor as “sorts,” but as a fusion concept including these two meanings (Lucas 2005: 361–362). However, “lei” is a very important concept in ancient Chinese logical thought. When we study the philosophical argumentations and debates in the basic classics of Confucianism, Mohism, Daoism, Legalism and the School of Names in pre-Qin times, we will find that ancient Chinese philosophers who claimed that reasonable arguments and debates are required to base on “knowing classes” (zhi lei 知類), “understanding classes” (ming lei 明類) and “examining classes” (cha lei 察類). For example, although Mencius criticized the political and ethical principles advocated by Mozi, both Mencius and Mozi stressed the importance of “knowing classes” and “understanding classes.” So, if people want to clearly know a thing, they must know the class to which the thing belongs, and understand the sameness

210

W. Yang and W. Zhang

among all members of the same class. Furthermore, we can get the unknown from the known through the sameness of their class. However, we think that classes could be made by traditional convention or natural division, and an artificial set could be established by some famous logicians and philosophers. Although “lei” in ancient Chinese logical thought could not be equivalent to the set in mathematical sense, actually it is the most important concept when ancient Chinese people engage in reasoning and demonstration. To understand well the sameness of things in terms of their being in the same class (lei tong 類同) and the difference of things in terms of their being not in the same class (bu lei zhi yi不類之異) is an objective criterion for ancient Chinese analogical reasoning.

4  A Criticism of Chad Hansen’s Mass Noun Hypothesis One of Chad Hansen’s main concerns is about nouns in ancient Chinese language and logical thought. He argued that all nouns in pre-Han Chinese language were mass nouns, which were different from count nouns with singular and plural variations. Mass nouns or uncountable nouns cannot be added with indefinite article and with numerals without classifiers. For example, the terms “water” and “grass” are mass nouns. Mass nouns in ancient Chinese do not refer to individuals of a class, but to stuff as parts of a whole. Because all nouns are mass nouns in classical Chinese, so in ancient Chinese philosophers’ eye, the world was understood in a “part-whole stuff” model, not in an “object-property” model. Almost all ancient Chinese thinkers could not study and think theoretically of individuals and abstract entities. This is the famous Mass Noun Hypothesis proposed by Chad Hansen (Hansen 1998: 37–63). As a result, the theory has had an impact on a wide range of studying language and logic in ancient China, but it also has been criticized by many researchers.2 In our opinion, Hansen’s Mass Noun Hypothesis is clearly unsatisfactory because of the concepts of “class” and “individual” do play important roles in argumentations in the ancient Chinese philosophy of logic. As mentioned above, “names” can be divided into “the extended name” (da ming), “the class name” (lei ming) and “the private name” (si ming) in the Mohist Canons, or “names” can be divided into “the great general name” (da gong ming), “the general name” (gong ming), “the great particular name” (da bie ming), “the particular name” (bie ming) and “the name without further specification” in the Rectification of Names of the Xunzi. It is 2  Among them, the most serious critiques are provided by Christoph Harbsmeier, Chris Fraser and Yiu-ming Fung, etc. Harbsmeier tried to provide some examples in ancient Chinese classics to overthrow the Mass Noun Hypothesis. (Harbsmeier 1998: 312–319) Chris Fraser thought that nouns in ancient Chinese could also refer to individuals, such as the private name in the Mohist Canons (Fraser 2006: 58–107). Yiu-ming Fung criticized that Hansen’s hypothesis was only an epitaxial explanation of nouns while not processing the connotation, and when the hypothesis was used to interpret the chapters “On White Horse”(白馬論) and “On the Hard and White” (堅白論) in the Gongsun Longzi, it would distort the meaning in the text and had a result of treating Gongsun Long’s view as a violation of logic rule (Fung 1993: 161–174).

12  Classes (Lei 類) and Individuals

211

obvious that proper names in classical Chinese, such as “the private name” in the Mohist Canons and “the particular name without further specification” in the Xunzi, are used to refer to individuals. For example, the name “Zang” 臧 is used to refer to a slaver in the Mohist Canons. In regard to the concept of “lei” or “class,” “the class name” in the Mohist Canons and “the general name” and “the particular name” in Xunzi’s naming theory are used to refer to many things of the same class, respectively. Two or more things are of the same class with the same property identified by the same standard and criterion, such as the name “ox” refers to every ox. Obviously, “lei” or “class” is an abstract concept in ancient Chinese logical thought. So we think that Hansen’s Mass Noun Hypothesis is insufficient to grasp the linguistic feature and logical role of these concepts in Chinese philosophy of language and logic in pre-Qin times.

References Chan, Wing-Tsit. 1969. A Source Book in Chinese Philosophy. Princeton: Princeton University Press. Chen, Bo 陳波. 2008. “Xunzi’s Theory of Name: Interpretation and Comparison. (荀子的名稱理 論:詮釋與比較)” Social Science Front (社會科學戰線) 12: 27. Chmielewski, Janusz. 1962. “Notes on Early Chinese logic (I)”. Rocznik Orientalistyczny 26.1: 7–22. Fraser, Chris. 2006. “Mass Noun Hypothesis and Interpretive Methodology.” In Chinese Philosophy and Culture (Series 1): The Reverse Analogical Interpretation and Global Philosophy (中國哲 學與文化 (第1輯):反向格義與全球哲學) 2006: 58–107. Fung, Yiu-ming. 1993. “Philosophy of Language in Chinese Philosophy: Discussing the Mass Nouns Theory” (中國哲學中的語言哲學問題:物質名詞理論的商榷). In Selected Papers of Analytic Philosophy and Philosophy of Language (分析哲學與語言哲學論文集). Hong Kong: New Asia College, Chinese University of Hong Kong: 161–174. Graham, A.  C. 2003. Later Mohist Logic, Ethics and Science. Hong Kong: Chinese University Press. Hansen, Chad 陳漢生. 1998. Language and Logic in Ancient China (中國古代的語言和邏輯). Beijing: Social Sciences Academic Press. Harbsmeier, Christoph. 1998. Science and Civilization in China, Volume 7, Part 1: Language and Logic. Cambridge: Cambridge University Press. Johnston, Ian. 2010. The Mozi: A Complete Translation. Hong Kong: The Chinese University Press. Lucas, Thierry. 2005. “Later Mohist Logic, Lei, Classes and Sorts.” Journal of Chinese Philosophy 32.2: 361–362. Wu, Jianguo 吳建國. 1980. “Production and Development of the Concept of ‘Classes’, and Formation of Logic Science in the History of Ancient Chinese Logic Thought (中國邏輯思 想史上類概念的產生、發展與邏輯科學的形成)”. Social Science in China (中國社會科學) 2: 51–75. Wu, Jianguo 吳建國. 1981. “Production and Development of the Concept of ‘Classes’ in History of Ancient Chinese Logic Thought” (中國邏輯思想史上類概念的產生和發展). In Selected Papers of Chinese Logical Thought (1949–1979) (中國邏輯思想論文選 1949–1979). Beijing: SDX Joint Publishing Company.

212

W. Yang and W. Zhang

Wujin Yang  graduated from the Renmin University of China in 2001 with a PhD. From 2001 to 2003, he conducted postdoctoral research at the Chinese Academy of Social Science’s Institute of Philosophy. He currently serves as a philosophy professor and doctoral tutor at the Renmin University of China. His main areas of research are the history of Chinese logic, Mohism and critical thinking. His publications include the Logical Basis of Dialectics, Reading the Mo Jing, Logic and Critical Thinking, and Capability and Literacy in Logical Thought. Wanqiang Zhang  got his PhD from Renmin University of China in 2016. He is currently a lecture in Philosophy Department of Xidian University, China. His research focus is on the history of Chinese logical thought and philosophy, especially on the Mohism and the School of Names. He has published some papers in philosophical journals in Chinese.

Chapter 13

Sameness (Tong 同) and Difference (Yi 異) Yiu-ming Fung

1  Different Senses of “Tong” and “Yi” in Later Mohism According to Fung Yulan’s 馮友蘭 interpretation, Hui Shi’s 惠施 view on the different and relativist criteria for identifying similarity and difference of things is based on a kind of metaphysical organism which can be named as the thesis of “he tong yi” 合同異 (the unity of similarity and difference), while Gongsun Long’s 公 孫龍 view on the unity and separation of entities in a two-level world is a Platonic-­ like realism which can be labeled as the thesis of “li jian bai” 離堅白 (the separation of hardness and whiteness). (Fung 1948: 86, 90) In comparison, I think, Later Mohists 後期墨家 maintain both “bie tong yi” 別同異 (distinguishing sameness and difference) and “ying jian bai” 盈堅白 (containing both hard and white [properties in one object]) which are based on a kind of common-sense ontology to analyze terms referring to things or objects in the empirical world. In dealing with the topics of tong 同 (sameness, similarity or unity) and yi 異 (difference, dissimilarity or disunity), Later Mohists’ view is much more sophisticated than other ancient Chinese thinkers’ in the sense that they distinguish different senses of tong and yi, including the sameness and difference of classes or the similarity and dissimilarity of kinds. In regard to the conceptions of “tong,” in the Mozi 墨子, Later Mohists have distinguished four different senses of the word in the following:1

1  Most of the Chinese quotations from ancient texts can be found in http://ctext.org of Chinese Text Project, edited by Dr. Donald Sturgeon. Here, most of the English quotations of the Mozi are based on my own translation. Other translations can be found in Ian Johnston 2010 and A. C. Graham 1978.

Y.-m. Fung (*) Department of Philosophy, Tunghai University, Taichung, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_13

213

214

Y.-m. Fung 同:重、體、合、類。(經上:87) (Refer to http://ctext.org/mozi/canon) Tong: Names in duplication, things in one body, things joining together, and things belonging to a class or kind. (Canon A87) (My translation) 同:二名一實, 重同也。不外於兼, 體同也。俱處於室, 合同也。有以同, 類同也。(經 說上:87) (Refer to http://ctext.org/mozi/canon) Tong: There being two names of one thing [object, event or reality] is the sameness of duplication [in terms of identity or equivalence]. Not being outside a composite whole is the sameness in terms of unifying things in one body. Both occupying a room is the sameness of joining together. Having something in common is the sameness in terms of belonging to a class or kind. (Canon Explanation A87) (My translation)

Based on some concrete examples in the Mozi, I think one of the reasonable interpretations of this passage is: (1) Chong-tong 重同 (sameness of duplication in terms of identity or equivalence): two names of one thing (er ming yi shi 二名一實), such as the token-identity between “Confucius” (Kongzi 孔子) and “Zhongli” 仲尼 which are two individual names referring to the same person and the equivalence between “gou” 狗 (dog) and “quan” 犬 (anther name of dog in Chinese) which are two general terms covering the same extension of things. (2) Ti-tong 體同 (sameness of commonality in terms of two names referring to two unifying parts): two things not separated from a composite whole or compound (bu wai yu jian 不外於兼), such as the united relation of a hand (shou 手) and a foot (zu 足) in one body. (3) He-tong 合同 (sameness of commonality in terms of two names referring to two things joining together): two things occupying a common room (ju chu yu shi 俱處於室), such as a table (zhuo 桌) and a chair (yi 椅) in a room (shi 室). (4) Lei-tong 類同 (sameness of commonality in terms of things belonging to a class or kind): two things have something in common (you yi tong 有以同), such as the case that a white horse (bai-ma 白馬) and a white stone (bai-shi 白石) have the common property of being white (bai 白). They can be included into the same class named by “bai-wu” 白物 (white thing). In addition to making distinctions among the four kinds of “tong,” Later Mohists have also distinguished four kinds of “yi” as follows:2 2  In the chapter “Da-qu” 大取 (Choosing the Greater) of the Mozi, there are similar distinctions with minor difference about the conceptions of tong and yi. It says that:

智與意異。重同, 具[俱]同, 連同, 同類之同, 同名之同; 丘[區]同, 鮒[附]同, 是之同, 然 之同, 同根之同。有非之異, 有不然之異。有其異也, 為其同也, 為其同也異。(大 取:22) (Refer to http://ctext.org/mozi/daqu) [Objective] knowing and [subjective] speculating are different. [In regard to tong, there are:] the sameness in terms of duplication [identity or equivalence], the sameness in terms of joining together, the sameness in terms of connecting units of a whole, the sameness in terms of [two things belonging to] the same class, the sameness in terms of [two things with] the same name, the sameness in terms of the same region, the sameness in terms of [two things with] the same attribute, the sameness in terms of [two sentences of] the same assertive value, the sameness in terms

13  Sameness (Tong 同) and Difference (Yi 異)

215

異:二、不體、不合、不類。(經上:88) (Refer to http://ctext.org/mozi/canon) Yi: Two [different entities], not unifying in one body, not joining together, not belonging to a class or kind. (Canon A88) (My translation) 異:二必異, 二也。不連屬, 不體也。不同所, 不合也。不有同, 不類也。(經說上:88) (Refer to http://ctext.org/mozi/canon) Yi: Two distinct things which are necessarily different [in naming] are [called] “two.” Not connected or attached [in a whole] is “not unifying in one body.” Not in the same place is “not joining together.” Not sharing something in common is “not of the same class or kind.” (Canon Explanation A88) (My translation)

This passage can also be interpreted as follows: (1) Er 二 (difference in terms of two things named by two terms): two names refer to two things (er shi 二實) respectively, such as “Confucius” (Kongzi 孔子) and “Mencius” (Mengzi 孟子) referring to two different persons and “ox” (niu 牛) and “goat” (yang 羊) covering different extensions of things. (2) Bu-ti 不體 (difference in terms of two things not in one body): two things are not closely connected (bu lian shu不連屬), such as the relation between a cat and a book which are not parts of a whole. (3) Bu-he 不合 (difference in terms of two things not joining together): two things are not occupying the same place (bu tong suo 不同所), such as the relation between a fish and a bird. (4) Bu-lei 不類 (difference in terms of two things not being of the same class): two things do not have some designated property in common (bu you tong 不有同), such as “attack” (gong 攻) and “execution” (zhu 誅) which are different in the sense that the former belongs to “fei yi zhi lei” 非義之類 (not of the class of righteousness) while the latter is “yi” 義 (righteous). In regard to the first kind of tong, i.e., two names (ming 名) referring to one thing (shi 實), Lao Sze-kwong’s 勞思光 interpretation is that this is about the relation between classes (lei 類) while the other three kinds of tong are about individuals. (Lao 1984: 318) I think Lao’s view is not totally right. In addition to class terms or kind terms, I think the formulae “two names one thing” can apply to other cases. For example: (i) “John” as a private name (i.e., proper name) only refers to this thing/person (shi) (i.e., John). (Canon Explanation A79) But, both names “Kongzi” and “Zhongli” as different proper names can be used to refer to the same person, i.e., Confucius. They are sameness of duplication and can be understood as “two names one thing.” of [two descriptions of] the same case, and the sameness in terms of [two things of] the same root. [In regard to yi,] there is the difference in terms of [different senses of] negation, there is the difference in terms of [different senses of] being not the case, there is difference in terms of [two things] being the same but their sameness is different. [For example, two things can be identified as the same in terms of their being an animal but they are different animals, say, a dog and a cat.]” (My translation and explication)

216

Y.-m. Fung

(ii) A gou 狗 (a name of dog) and a quan 犬 (another name of dog) are different in naming. But, they are same in terms of duplication. (Canon B53) This is also a case of “two names one thing.” (iii) The three names: “zhou” 周, “bian” 遍 and “xian” 咸 are different names or terms of the same thing or meaning, i.e., their meaning is the same. (Zhuangzi I 2: 6) (莊子齊物論 6) These three adverbial terms are used to express universal quantification, meaning all or whole of things. The first case is about the identity of two individual names (a=b) while the second case is about the equivalence of two general terms [(∀x)(Gx↔Qx)]. In ancient Chinese texts, a general term can be used to express a kind of things with similarity or to express a class of things with common characteristic or property. The third case is about the fact that different names or terms can be used with the same meaning. The difference among these three cases is that: (i) is about individual terms [a, b, …], (ii) is about general terms [G, Q, …], and (iii) is about logical terms [(∀x), (∀y), …]. Since the word “shi” can be used to stand for things/objects denoted by a general term distributively and also be used to stand for individual thing/object denoted by an individual term, the formulae “two names one thing” is not restricted to general terms; it can also apply to singular terms such as proper names or demonstrative pronouns. Another example to show that shi as the referent of ming 名 (name) is an individual rather than a collection of things is by descriptive identification or by ostensive demonstration. That is: 舉友「富商」也, 是以名視 [示] 人也。指是臛 [鶴] 也, 是以實視 [示] 人也。(經說 下:53) (Refer to http://ctext.org/mozi/canon) To enumerate a friend with “rich businessman” is to use a [descriptive] name (ming 名) to show to other people. To point this crane is to use a thing (shi 實) to show to other people. (Canon Explanation B53) (My translation)

In these examples, “shi” is used to refer to things/objects distributively (one’s friend as one of the referents of “rich businessman”) or used to refer to an individual thing/ object (the thing pointed out: this crane) rather than to a collection of things/objects. So, I think, “two names one thing” is not exclusively used for the collective use of general terms, either class terms or kind terms; it is more usually used for individual terms or the distributive use of general terms.

2  Sameness and Difference of Lei (Class or Kind) With respect to the problem of sameness and difference of lei, first, let’s go to Canon B65 and Canon Explanation B65: 狂舉不可以知異, 說在有不可。(經下:65) (Refer to http://ctext.org/mozi/canon) To give a wild example cannot help people to know the difference between things. The reason for explanation is given under “not acceptable.” (Canon B65) (My translation)

13  Sameness (Tong 同) and Difference (Yi 異)

217

狂:牛與馬惟異, 以牛有齒、馬有尾, 說牛之非馬也, 不可。是俱有, 不偏有偏無有。曰 之[牛]與馬不類, 用牛[有]角、馬無角, 是類不同也。若[不]舉牛有角、馬無角, 以是 為類之不同也, 是狂舉也, 猶牛有齒, 馬有尾。(經說下:65) (Refer to http://ctext.org/ mozi/canon) Wild: Although ox and horse are different but to take “ox has incisors” and “horse has tail” as an explanation for “ox is not horse” is not acceptable. These are what both have, not what one has and the other one doesn’t have. If you say, “ox and horse do not belong to the same class” with the example that ox has horns and horse doesn’t have horns, this is the case that they do not belong to the same class. But if you do [not] give the example “ox has horns and horse does not have horns” [as evidence], to regard them [i.e., ox and horse] as not belonging to the same class is to give a wild example. This is like the case that “ox has incisors and horse has tails” [which cannot be identified as evidence]. (Canon Explanation B65) (My translation)3

I think the message of Canon 65 is that: wild demonstration or guangju 狂舉 (i.e., using improper or ridiculous examples to make assertion or reasoning) cannot help us to know the difference between things (i.e., objects, events or reality). Canon Explanation B65 says that even though niu 牛 (ox) (used as a predicate, say, F, to modify an individual constant, say, a or a variable, say, x) has a property as described by chi 齒 (incisors) (used as a predicate, say, J) and ma 馬 (horse) (used as another predicate, say, G, to modify another individual constant, say, b or a variable, say, y) has another property as described by wei 尾 (tail) (used as another predicate, say, K), it is not acceptable to use this example to demonstrate that the individual modified by niu is different from the individual modified by ma. It is because both individuals may have both properties. Nevertheless, if we use the example that the individual modified by niu has a property as described by jiao 角 (horns) (used as a predicate, say, H) while the individual modified by ma does not have, with respect to the designated characteristic described by jiao, we can demonstrate that they are different in classes. If we cannot provide any characteristic which is owned by an individual and not owned by another and thus claim that they are different, it would commit a similar fallacy as the example about chi and wei mentioned above. In sum, I think, Later Mohists’ view is that: to identify whether two groups of things belong to the same class or different classes depends on whether there is a common and distinct characteristic shared by the two groups of things. But, in this passage, Later Mohists have not yet explained the ontological status of the characteristic. Is it the nominal essence or real essence of the relevant things? Is it treated as the defining characteristic of the same class or different classes of things? For Later Mohists, the necessary condition for identifying the difference in classes for two things/events or two groups of things/events is that they do not share a relevant or designated property or attribute. They think that “to know lei 類” (class/kind) and “to understand gu 故 (reason/cause)” are necessary for making assertion or inference. If one uses a wrong example to illustrate the sameness or difference between two terms, (Canon B65) i.e., giving a wild example, one cannot

3  Some of the emendations of this passage are based on Wu Feibai 伍非百, 1983: 175-6. See also Ian Johnston 2010: 550-1.

218

Y.-m. Fung

construct an acceptable assertion or parallel inference.4 Because, if one misidentifies different things or events as of the same class, what one can provide as a relevant reason to support the assertion of an event of one class is not the same reason to support the assertion of another event of another class in a parallel inference. If these different things or events seem to be in the same class but actually not, the different referents of the relevant terms cannot be identified as covered by the same concept though the terms look like expressing the same concept as required in the inference. So, in Canon B2, Later Mohists stress “the difficulty of the generalization of class” (tui lei zhi nan 推類之難). In Canon and Explanation A86 and A87, although they give a very sophisticated analysis of the different concepts of “sameness” and “difference” for the identification of various relations including the sameness and difference of classes, they still warn us to strictly abide by the principle of “different classes are not comparable” (yi lei bu bi 異類不吡[比]) (Canon B6).5 Both Early and Later Mohists claim that people’s argumentations should be strictly in accordance with this principle and they should be very cautious in making any kind of parallel inference. For example, Mozi distinguishes the concept of “attack” (gong 攻) from that of “execute” (zhu 誅) though both terms seem to refer to the same kind of action with military force (Against Military Attack III: 5, Johnston 2010: 189). He thinks that the former as an aggression with military force is unjust or not of the same class of righteousness (fei yi zhi lei 非義之類) (Against the Confucians II: 9, Johnston 2010: 360-1) while the latter as a punishment with military force is just or righteous. So, if we accept a statement with the latter concept, it does not follow that we can make a similar acceptable statement with the former concept. Based on this reason, Mozi criticizes those who “do not know the class to which my words belong” and “do not understand the reason on which my words are based” (子未察吾言之類, 未明其故。). In other words, Mozi’s criticism is against those who make their argument on the basis of wild example. In the second part of Tong-bian Lun 通變論 (Essay on Understanding Change) in the Gongsun Longzi 公孫龍子, Gongsun Long extends his view from “two doesn’t have one” (er wu yi 二無一) and “what is changed is not what is unchanged” (bian fei bu bian 變非不變) to the thesis of “two is not one” (er fei yi 二非一) and discusses the problem of the correct or proper demonstration against the wild or improper demonstration. Here, in comparison with Later Mohists, he provides a much more strict condition to identify the sameness and difference of classes. [主] 曰:羊合牛非馬; 牛合羊非雞。 [客] 曰:何哉?

4  Parallel inference is a kind of reasoning with the form “A is B; CA is CB.” In Later Mohism, it is called “mou” 侔, meaning parallelism. Detailed discussion can be found in Chaps. 8, 14 and 16 of this volume. 5  Some concrete examples of this kind are: “Which is longer, a piece of wood or a night? Which do you have more of, knowledge or grain? Which is the most valuable, aristocratic rank, one’s own parents, right conduct, a price? ….” (木與夜孰長?智與粟孰多?爵、親、行、賈四者孰貴?…) (Explanation B6) (Johnston 2010: 474-5 or Graham 1978: 357) Other incomparable examples can be found in Canon and Explanation B2. (Johnston 2010: 476-7 or Graham 1978: 349)

13  Sameness (Tong 同) and Difference (Yi 異)

219

[主] 曰:羊與牛唯異, 羊有齒, 牛無齒, 而牛之非羊也, 羊之非牛也, 未可, 是不俱有 而或類焉。羊有角, 牛有角, 牛之而羊也, 羊之而牛也, 未可, 是俱有而類之不同也。 羊牛有角, 馬無角; 馬有尾, 羊牛無尾。故曰羊合牛非馬也。非馬者, 無馬也。無馬者, 羊不二, 牛不二, 而羊牛二, 是而羊而[與]牛非馬可也。若舉而以是, 猶類之不同。若 左右, 猶是舉。牛羊有毛, 雞有羽。謂雞足一, 數足二, 二而一, 故三。謂牛羊足一, 數 足四, 四而一, 故五。牛羊足五, 雞足三。故曰牛合羊非雞, 非有以非雞也。與馬以雞, 寧馬。材不材, 其無以類, 審矣!舉是謂亂名, 是狂舉。(Refer to (Refer to http://ctext. org/mozi/gongsunlongzi) A: A ram and an ox connected together is not a horse. An ox and a ram connected together is not a chicken. Q: What do you mean? A: Although a ram is different from an ox in that the ram has upper front-teeth while the ox does not, yet one cannot say that an ox is not the same as a ram or a ram is not the same as an ox. They might not both have this characteristic, but still belong to the same class. On the other hand, although a ram has horns while an ox also has horns, yet one cannot say that an ox is the same as a ram, or a ram is the same as an ox. They might both have this characteristic, and yet belong to quite different classes. Rams and oxen both have horns and horses do not, whereas horses have prominent tails, while rams and oxen do not. Therefore, I say, “a ram and an ox connected together is not the same as a horse.” This means that there is no horse. In saying that there is no horse, [the reason is that] a ram is not two and an ox is not two, but ram-and-ox is two. So, this makes it possible for a ram joined with an ox has no horse. When we demonstrate the problem with this example, it is just based on the difference between classes. It is similar to the problem of left and right [i.e., “two doesn’t have left (one) and two doesn’t have right (one)”]; that is also demonstrated with this. Ox and ram have wool and chicken has feathers. The legs of a chicken are one by saying, and two by counting, and therefore three altogether; the legs of an ox or a ram are one by saying, and four by counting, and therefore five altogether. Thus an ox or a ram have five feet each, and a chicken three. Therefore, in saying that “an ox and a ram connected together is not the same as a chicken,” it is not the case that they have something to distinguish them from a chicken. Between [the examples of] a horse and a chicken, it is better to choose a horse [as a proper example for demonstration]. The former [horse] is the stuff suitable for demonstration while the latter [chicken] is not. It is obvious that they are not of the same class. To demonstrate the problem with this [latter] example is called “confusing the name” and thus is a wild demonstration. (My translation with explication)

In the second part of the dialogue, Gongsun Long introduces the concept lei (類 class/kind) to explain how to use a correct example to demonstrate the thesis “two is not one.” Just like the word “mou” (某 some/someone) which is often used as a symbol for some individual, he uses the terms “ram,” “ox,” “horse” and “chicken” as symbols for predicates to modify individual constants or variables, not as names referring to concrete animals as we usually use the terms. Otherwise, we cannot make sense for the dialogue and thus some scholars, such as A.  C. Graham and Chad Hansen, identify it as fake though there is no evidence provided by them. For Gongsun Long, I think, to choose this cluster of names is just like to make a random model for demonstration. He thinks that, from one perspective, in a particular possible world or model of interpretation, although A (ram) and B (ox) are different in terms of their not sharing a particular characteristic (upper front-teeth), only based on this difference, we cannot ascertain that A is not B (ram is not ox) or B is not A (ox is not ram). They may be of the same class. It is because this particular characteristic may not be the defining characteristic which can be used to identify whether

220

Y.-m. Fung

they are of the same class or not. From another perspective, although both B (ox) and A (ram) have a particular characteristic (horns), only based on this similarity, we cannot identify that they are of the same class. It is because what they share may not be the sufficient condition for defining their sameness and thus they may belong to different classes. Here, it is obvious that Gongsun Long uses these animal names as symbols rather than concrete terms; otherwise we cannot make any good sense for the passage.6 As we know, it is a logical fact that ram and ox are either of the same class or not; there is no middle place for them. But Gongsun Long’s example clearly demonstrates that, from two different perspectives, they (A and B) may be of the same class and also not of the same class. In other words, there are four possible combinations: A and B are not of the same class and B and A are of the same class; A and B are not of the same class and B and A are not of the same class; A and B are of the same class and B and A are of the same class; A and B are of the same class and B and A are not of the same class. From two different perspectives, the A or B in “A and B” is not the A or B in “B and A.” To prevent confusion, Gongsun Long is clever to make a distinction between “A and B” (ram and ox) and “B and A” (ox and ram) with the terms “left” and “right.” To use the terminology of “left one” and “right one,” A in “A and B” is the left one while A in “B and A” is the right one. It is randomly mentioned by Gongsun Long. So, I think his message is that: please don’t take the names of animal seriously or factually! The examples with the forms “A and B” and “B and A” are used to explain that only based on a particular characteristic, we cannot set up any criteria of sameness or difference. The criteria must be based on a kind of essential characteristic which can be used as both a sufficient and a necessary condition for identifying sameness or difference of classes. Following this example, Gongsun Long provides another example with the form “AB is not C” (ram joined with ox is not horse) and “BA is not D” (ox joined with ram is not chicken) to explain the correct and incorrect demonstrations of “two is not one.” To demonstrate the type-difference between the entities of the two worlds, Gongsun Long uses the general statement “two is not one” as example; to demonstrate the token-difference, he uses the particular statement “white-horse is not horse” as example. To demonstrate the difference of classes, he uses the symbolic forms such as “AB is not C” and “BA is not D” for explanation. To assert “AB is not C,” we must find out in advance some defining characteristic(s) to identify their difference of classes. For example, how can we assert that A (ram) joined with B (ox) is not C (horse)? Since we find out that AB has a property or characteristic p1 (horns) while C does not have and C has a property or characteristic p2 (prominent tails) while AB does not have, these can be regarded as both a sufficient and a necessary condition to identify that AB and C are different in classes. Based on this criterion, we can assert that “AB is not C” and thus “AB does not have C” ([羊合牛]非馬者 無馬也). In saying that there is no C (horse), it is because A (ram) is not two and B 6  It is similar to the Western logicians in the Medieval Age: some of them use numbers as propositional variables (“the first,” “the second,” …), some others replace numbers with demonstratives (“this” and “that) as a kind of symbols. See Dov M. Gabbay and John Woods 1984: 4.

13  Sameness (Tong 同) and Difference (Yi 異)

221

(ox) is not two, but AB (ram-ox) is two. So, this makes it possible to say that A joined with B (ram joined with ox) has no C (horse). Here Gongsun Long clearly indicates that AB as a combination of two entities is not A or B as a single entity; and it is also true for the case that AB is two but C is one. Based on the fact that AB as two is not A or B as one and is also not C as one, he concludes that two does not have one. In this regard, it is obvious that when A joins with B to form AB, AB does not have A or B. In other words, AB (including A in AB and B in AB) is essentially different from A (not the A in AB) or B (not the B in AB) in terms of their ontological status. This can be explained by the idea of “what is changed is not what is unchanged” (bian fei bu bian 變非不變) in the first part of the dialogue. This can also be explained by the general idea of “wu 物 is not [du-]zhi [獨]指” (“a thing is not an [independent] abstract entity”) and “wu does not have [du-]zhi” (“a thing does not have any [independent] abstract entity”) in Zhi-wu Lun 指物論 (Essay on Zhi-wu) and explained by the specific idea of “white-horse is not horse” and “the fixing white is not the [non-fixing] white” in Bai-ma Lun 白馬論 (Essay on White-­ Horse) in the Gongsun Longzi.7 Although Gongsun Long accepts that “AB is not C” (ram joined with ox is not horse), he does not think that “BA is not D” (ox joined with ram is not chicken) is a proper demonstration of “two is not one.” Most of the scholars think that the reason provided by Gongsun Long is queer and he cannot escape from the charge of sophistry. Some even argue without evidence that the essay is fake. I don’t think these accusation and judgment are reasonable because they do not provide any evidence except that they cannot explain the text in a sensible way. If we treat Gongsun Long’s discussion in a formal or semi-formal way, all the seeming queer things would disappear. However, why “AB is not C” (ram joined with ox is not horse) is a correct demonstration of “two is not one” but “BA is not D” (ox joined with ram is not chicken) is not? The reason is that “AB has p1 (horns) but C does not have and C has p2 (prominent tails) but AB does not have” can be regarded as defining characteristics to identify their difference in classes. On the other hand, we cannot find a defining characteristic to ascertain that BA is different from D in classes. Gongsun Long mentions that, for example, BA (ox joined with ram) has a particular property or characteristic p3 (wool) while D (chicken) has p4 (feathers). Since the difference between p3 (wool) and p4 (feathers) is not an essential one (In regard to this model, it is because in Chinese language, “mao” 毛 (wool) and “yu” (羽 (feathers) can be classified under the same category of “yu-mao” 羽毛 and their difference is in degree, not in essence.), based on this, we cannot distinguish BA from D. In other words, even though BA is different from D with a feature or characteristic in terms of quantity rather than quality, based on the quantitative difference (mao as short and soft hair against yu as long and hard hair), we still cannot make a demarcation between BA and D. In regard to the model of legs, the legs of a chicken are one by saying, and two by counting, and therefore three altogether; the legs of an ox or a ram are one by saying, and four by counting, and therefore five altogether. Thus an

 The explanation of the ideas of “zhi” and “non-fixing white” can be found in Fung 2020.

7

222

Y.-m. Fung

ox or a ram have five feet each, and a chicken three. Even though their difference can be identified either by saying in a subjective mode, by counting in an objective mode, or by both,8 the difference is only in terms of quantity but not in terms of essential characteristic. So, to use this example or model to prove that “two is not one” is an incorrect or wild demonstration. Gongsun Long has a consensus with Later Mohists that “we cannot know the difference between things by wild demonstration” (狂舉不可以知異). Based on the semi-formal explanation mentioned above, he concludes that the “horse” in “ram joined with ox is not horse” is an appropriate example in the model for explaining “AB is not C” while the “chicken” in “ox joined with ram is not chicken” is not a suitable example in the model for explaining “BA is not D.”

3  Arguments About the Sameness of Duplication Sometimes Later Mohists explicitly provide explanation for some real and unreal anomalies of parallelism (mou 侔) or parallel inferences.9 One of the most significant examples is in Canon B39 and B53, respectively: 狗, 犬也; 而殺狗非殺犬也, 不可。說在重。(經下:53) (Refer to http://ctext.org/mozi/ canon) A gou 狗 [a name of dog] is a quan 犬 [another name of dog]; it is not acceptable [to say] that to kill a gou is not to kill a quan. The reason for explanation is given under “duplication.” (Canon B53) (My translation) 狗, 犬也。[殺狗] 謂之殺犬, 可。若兩脾。(經說下:53) (Refer to http://ctext.org/mozi/ canon) A gou is a quan; it is acceptable to say that [to kill a gou is] to kill a quan. Just like [the type-identity of] twin thigh. (Canon Explanation B53) (My translation)

A simplified version of the above example is as follows: (I) 狗, 犬也; 殺狗, 殺犬也。 “A gou is a quan; to kill a gou is to kill a quan.” It seems that (I) shares the same form as that of (II) in the following: (II) 白馬, 馬也; 乘白馬, 乘馬也。(小取:3) (Refer to http://ctext.org/mozi/xiaoqu) “A white horse is a horse; to ride a white horse is to ride a horse.” (Xiao-qu 5) (My translation)

8  Here Gongsun Long’s example of “by saying” and “by counting” is used to illustrate that some features or characteristics are known in a subjective way and some other are known in an objective way. 9  For example, the argument: “A thief is a man; to love a thief is not to love a man.” (Mozi, Xiao-qu (Choosing the Lesser) 6) (盜[人], 人也; 愛盜, 非愛人也。) (墨子:小取) is one of the anomalies of parallelism [Form II] which is discussed in Chap. 18 of this book.

13  Sameness (Tong 同) and Difference (Yi 異)

223

I think (I) does share the same grammatical structure and a similar logical form with (II) except that (II) should be formulated with material implication sign (→) as its main connective while (I) should be formulated with material equivalence sign (↔) as its main connective if both arguments are valid. Here, it is because “gou” and “quan” are two different names of the same thing. The logical form of (I) should be:

 Form I  :

 x   Sx  Px    x   y  Sy & Rxy    y  Py & Rxy  

In comparison, the logical form of (II) should be:

 Form II  :

 x   Sx  Px    x   y  Sy & Rxy    y  Py & Rxy  

The parallel inferences of both (I) and (II) can be proved validly. However, the following example indicates a different criterion for checking the validity or fallacy of inference or reasoning: 知狗而自謂不知犬, 過也。說在重。(經下:39) (Refer to http://ctext.org/mozi/canon) If you know this is a gou, to say of yourself that you do not know this is a quan is a fallacy. The reason for explanation is given under “duplication.” (Canon B39) (My translation) 智狗, 重智犬, 則過; 不重, 則不過。(經說下:39) (Refer to http://ctext.org/mozi/canon) If knowing [that this is] a gou is identical with knowing [that this is] a quan, it is a fallacy. If they are not identical with each other [in knowing], there is no fallacy. (Canon Explanation B39) (My translation)

As we know, in an intensional context,10 although “gou” and “quan” are two different names of the same thing, we cannot infer the conclusion that knowing that something is a quan from the premise that knowing that something is a gou. The author of Canon B39 seems not aware that the sentence is in an intensional context and thus treats the example as a case in an extensional context as that in Canon B53, so the author thinks that to make a negative assertion in the conclusion is a fallacy. It seems that the author of Explanation B39 does not treat this as a real fallacy and clarifies that, in an intensional context, it is fallacious to make the statement that one knows that something is a quan but does not know that something is a gou if one  In philosophy of language, a linguistic context in which a sub-sentential expression e appears is called extensional if and only if e can be replaced by an expression with the same referent and necessarily preserve truth-value. In opposition to extensional contexts are intensional contexts, where synonymous terms cannot be substituted in without potentially compromising the truthvalue. Sentences with “believe-that” or “know-that” clause are typically regarded as in the intensional context. For example, although Clark is the superman, the sentence “Mary believes that the superman is strong” can be true while the sentence “Mary believes that Clark is strong” can be false. Here, “Clark” and “the superman” are co-extensive; but the rule of substitution applied to these two terms cannot preserve truth-value.

10

224

Y.-m. Fung

knows that “gou” and “quan” are two different names of the same thing. If one does not know that the two names are identical, even though they are in fact identical, it is not fallacious to say that to know that something is a gou is not to know that something is a quan. So, to say that: (III) “A gou is a quan; to know this is a gou is not to know this is a quan.” is not to make an anomaly of parallel inference if the speaker does not regard the two knowing events as identical or does not know that the two names are identical in reference; but if one does recognize that the two knowing events are identical or does know that the two names are identical in reference it can be identified as a fallacy.

4  Was Lei Used as a Concept of Class in Classical Chinese? According to Hansen’s view, the term “lei” 類 used by ancient Chinese philosophers cannot be understood as the technical equivalent of “class” or “set” in Western Philosophy. He says: (Hansen 1983: 113) [T]he mathematical notion of class is “born together” with the distinction of subclass and member. Without any such distinction, lei only be regarded as mereological class - using (as we shall see) only the distinction of part-whole. This conclusion, obviously, is reinforced by the masslike grammar of the nouns called lei-names and the tendency to explain their semantics in the same way one explains the semantics of proper names. We will notice that the use of lei makes it clear that it is mainly used to assert similarity between thing-kinds and only derivatively to describe a thing-kind based on similarity of its parts.

He argues that to suggest there is the notion “mathematical class” without the notion of “member” is paradoxical. Follow this line, one may think that: Since classical Chinese does not have these companion concepts, it is impossible for lei to be understood as “class” as used in Western philosophy and logic. So, as stressed by Hansen, “lei” should be translated as “similar-stuff” rather than “class.” (Hansen 1983: 117) A. C. Graham also does not treat lei in the Mozi as class and suggests that “lei” should be translated into “kind” or “of a kind.” (Graham 1989: 169) Nevertheless, if we follow their treatment and thus translate “lei” as “similar-stuff” rather than “class”(Hansen 1983: 117), our question is: what is the translation of “tong-le” 同類? Is it the “same similar-stuff” or “similar similar-stuff”? If “wu” 物 means mass-stuff or thing-kind, how to translate “wu-le” 物類? Does it mean similar-­stuff of mass-stuff or similar-stuff of thing-kind? I don’t think Hansen can translate this cluster of names in a sensible and coherent way. In general, I think it is wrong to treat lei exclusively as a concept without the content of membership. In one sense, instances of a kind or tokens of a type are defined by the similarity of the instances or tokens. The relation between instances and their kind or that between tokens and their type is exemplification in the sense that the instances or tokens share similarity in some respect. Just like the word “cow” in “A cow is bigger than a chicken” and the word “cow” in “A cow is an

13  Sameness (Tong 同) and Difference (Yi 異)

225

animal” are two tokens of the same type of words. These tokens are not members of a class. Similarly, in the sentence “This big stone and that small stone are on the top of the mountain,” we may treat this stone and that stone as the same natural kind, but not members of a class. I think, however, there are still many examples in ancient Chinese texts in which lei is used as a class with membership. For example, in the Xunzi (VIII: 22): 居楚而楚. 居越而越, 居夏而夏, 是非天性也, 積靡使然也。(荀子儒效) (Refer to http:// ctext.org/xunzi) Anyone who lives in Chu state will become a Chu man. Anyone who lives in Yue State will become a Yue man. Anyone who lives in Xia state will become a Xia man. This is not due to their endowed nature from the heaven, but is the result of their accumulated experience of undergoing.” (My translation)

It says that: any person who lives in a particular state will become a member of that class of people in that state. It is not to say that: any person who lives in a particular state will become an instance of that kind of people in that state. Suppose Jack was a Qin 秦 man. Now he moves to Chu 楚 and thus becomes a Chu man. In this case, we cannot treat him as an instance of the kind of Qin man in the past because he as an instance was similar to other Qin man in the past. We also cannot treat him as an instance of the kind of Chu man now because he as an instance is similar to other Chu man now. A more rational explanation is that: he was a member of the class of people in Qin state in the past and, now, becomes a member of the class of people in Chu state. This is his experience of undergoing. One may insist to interpret lei as kind in the sense that: if a lei as a kind is identified by the similarity of things in some aspect, then similarity of things in some aspect is not sameness in terms of things’ sharing a common property/characteristic (or in terms of the predicate of a common property/characteristic being true of them) but similar in the sense of family-resemblance. That is, a as the first instance of a kind is similar to the second instance b in some respect and b is similar to the third instance c in some other respect. But a may not be similar to c in the former respect or the latter respect. In other words, sameness is a relational property of transitivity while similarity is not. So, only based on family-resemblance, it is unlikely to make generalization which is required in inductive reasoning. Nevertheless, one may disagree with me and think that the relation among the instances of a kind is not recognized in this sense. If similarity is not defined in this sense, the relation among the instances of a kind can be treated as having property of transitivity. But, my question is: by what criterion can we determine that the relation among the instances of a kind has the property of transitivity? If similarity is not defined in terms of non-transitive characteristic/property, but in terms of transitive one, the characteristic/property can be understood as a common one which, in turn, can be regarded as a defining characteristic/property of a class. In this sense, I think, there would be no real difference between the concept of similarity and that of sameness and there would be no real difference between the concept of kind and that of class.

226

Y.-m. Fung

Moreover, if a community did not have the concept of a class, its general concept “thing” could be interpreted as a super-kind which has all things in the world as its instances. In this regard, all the instances of a kind must be similar to each other, otherwise they cannot be identified as of the same kind. But what is the similarity between all the things in the world? It seems that they can only share a family resemblance and are thus without any common material characteristic/property. I think the only common characteristic they share is a formal one, that is: they all have a spatial-temporal point in the world. But this cannot be used to identify an instance of a kind in a non-trivial sense, because “thing” as a general concept does not have any material content which can be used to identify it as a kind. Besides, if the general concept “thing” includes events and other abstract entities in the content of its package, then we cannot find any material or non-formal characteristic shared by all the physical objects, events and abstract entities. In this case, the only candidate for us to identify the general concept of “things” in this sense must be a class whose defining characteristic is formal. Since resemblance or similarity is a relation with reflexive, symmetric, and non-­ transitive, it cannot be treated as the basis for defining class. As demonstrated by Jessica Wilson: (Wilson 2006: 243) A long-recognized problem with resemblance nominalism is Goodman’s (1966, pp. 162–4) “imperfect community difficulty,” initially raised against Carnap’s (1927/1967) version. Roughly, Carnap’s suggestion was that membership in a maximal class α of resembling particulars (such that every pair of members of α resemble, and nothing outside the class resembles every member of α) is necessary and sufficient for the members of α to share a property. The imperfect community difficulty cuts against the sufficiency of the analysans. As a simple example: let a be F, G, and H; b be F, P, and Q; and c be R, P, and H. Then every pair of members of a, b, c resembles, but there is no property they all share.

Although Quine is skeptical of the ontological status of intensionalist entities such as properties and attributes,11 he does not give up classes and kinds because our languages and theories presuppose these entities. But what is a kind? Quine proposes to characterize it in terms of similarities because he thinks that: “The notion of kind and the notion of similarity seem to be variants or adaptations of a single notion.” (Quine 1969: 117) But this approach shows to be unsuccessful. As explained by Nathan Stemmer (2007: 77): The similarity notion that we obtain with the help of experiments on conditioning and extinction is the comparative notion of a being more similar to b than to c (for subject s) and it seems to be impossible to define a notion of kind by using this comparative notion. Quine examines various options for solving the problem such as appealing to paradigms (p. 119) or to respects (p. 122), but none is satisfactory. He concludes with the pessimistic conclusion that a “definition of kinds in terms of similarity is unknown” (1969b, p. 121).

 Quine does not accept intensionalist entities and thus does not treat properties or attributes as abstract entities or universals, but he doesn’t refute that, in some cases, the words of properties or attributes in a sentence can be used as being “true of” individuals. See Quine 1953:10.

11

13  Sameness (Tong 同) and Difference (Yi 異)

227

So, Quine’s proposal of treating the terms of natural kinds as projectible predicates12 and the objects of a kind as similar comparatively is not successful. Instead, he later claims that: “Kinds can be seen as sets, determined by their members. It is just that not all sets are kinds.”13 (Quine 1969: 118) If we accept Quine’s view that kinds are sets or classes, sets or classes as abstract entities cannot be understood as concrete objects or mass-stuffs. In general, in one sense (i.e., Hansen’s), the concept of kind is based on similarity while that of class is not and only a common and distinct property or attribute can be treated as the defining characteristic to identify the members of a class. Since a lot of examples in ancient Chinese texts, including those mentioned above in the Mohist Canon and Gongsun Longzi, can be understood as using some common and distinct property or attribute to identify a group of objects, persons or events, it is impossible to treat these particular things as mass-stuffs of a kind which only have the relation of similarity or resemblance. Furthermore, it is obvious from ancient Chinese texts that there are many examples of inductive generalization in ancient Chinese thinkers’ thought. To make generalization, they must make it on the basis of treating things/events as members of a class. That is, under a class, things/events share a common property/characteristic. Without the relation of class-membership, there would be no generalization in ancient China. Hansen adopts Quine’s view on mass nouns to form his own ideas of thing-kind and mass-stuff; but it seems that he doesn’t know that Quine does treat a kind as a set or class. Moreover, Quine has mentioned that not any sort of aggregation of objects can be identified as a class, unless we keep clearly in mind that “the aggregation or collection or combination of say seven given pairs of shoes is not to be identified with the aggregation or collection or combination of those fourteen shoes.” (Quine 1963: 1) So, if we clearly know that the number of the members of the aggregation of seven pairs of shoes is seven, not fourteen, we do regard the aggregation as a class, not merely a mereological sum. In classical Chinese language, “five pairs of kudzu-shoes” (ge-li wu liang 葛屨 五兩) (Shi Jing 詩經齊風南山) and “two pairs of hemp-shoes” (xi-li er liang 枲履 二兩) (Juyan Jian 居延簡 p.406, 260.23) mean “five pairs of shoes” and “two pairs of shoes” respectively. The number of the members of the aggregations described by these phrases is not ten and four, but five and two respectively.

 Goodman has argued that in order to formulate an adequate theory about inductive inferences one must distinguish between projectible predicates like “raven” and “green,” which are well suited for these inferences, and non-projectible predicates like “non-raven”, “raven or fire” and “grue” which do not have this property (Goodman 1946, 1983). This raises the problem of characterizing “generally and rigorously” the difference between these predicates (Goodman 1946: 383). 13  Quine thinks that in order to account for our inductive inferences we need the concept of kind, because we have to distinguish between inductive inferences about classes such as “raven,” which are supposed to be kinds, and inductive inferences about classes such as “non-raven,” which are not. 12

228

Y.-m. Fung

Similarly, “wu jue” (五瑴/珏) (Zuo Chuan, Duke Zhuang 左傳莊十八) is not referring to ten pieces of white jades, but five pairs of white jades. Since “ma yi cheng” (馬一乘) (Analects , Gong Ye Zhang) means “four horses used in a carriage,” “ma shi cheng” (馬十乘) cannot only be interpreted as “forty horses” or “forty horses which are used in carriages.” Instead, it should be understood as “ten carriages of horses” or “ten groups of horses which have four horses used in a carriage respectively.” All these examples indicate that the modes of divided reference between the single object and the ordered pair of single objects are different; and thus cannot be treated only in terms of the part-whole relation of mass-stuff. For human beings, Xunzi makes a hierarchy of classes to arrange their relation of subordination. He says, 水火有氣而無生, 草木有生而無知, 禽獸有知而無義, 人有氣、有生、有知, 亦且有義, 故最為天下貴也。(荀子王制) Fire and water possess vital breath but have no life. Plants and trees possess life, but lack awareness. Birds and beasts have awareness, but lack a sense of morality and justice. Humans possess vital breath, life, and awareness, and add to them a sense of morality and justice. It is for this reason that they are the noblest beings in the world. (Xunzi IX:19) (Knoblock 1999: 236-7)

This classification of hierarchy is closely related to Xunzi’s ideas of lei 類 and shu 屬 whose relation can be understood as that between class and sub-class. Harbsmeier calls this kind of relation “subsumption.” (Harbsmeier 1998: 241-4) I think it is impossible to explain the relation in Hansen’s mass-noun scheme. In the Introduction to his Set Theory and its Logic, Quine presents a promising picture of the identity of a class. He says: (Quine 1963: 1-2) Imagine a sentence about something. Put a blank or variable where the thing is referred to. You have no longer a sentence about that particular thing, but an open sentence, so called, that may hold true of each of various things and be false of others. Now the notion of class is such that there is supposed to be, in addition to the various things of which that sentence is true, also a further thing which is the class having each of those things and no others as member. It is the class determined by the open sentence. [My italic] Much the same characterization would serve to characterize the notion of attribute; for the notion of attribute is such that there is supposed to be, in addition to the various things of which a given open sentence is true, a further thing which is an attribute of each of those things and of no others… But the difference, the only intelligible difference, between class and attribute emerges when to the above characterization of the notion of class we adjoin this needed supplement: classes are identical when their members are identical. This, the law of extensionality, is not considered to extend to attributes.

If we accept Quine’s criterion of identity for the class in terms of open sentence,14 it is clear that there is a concept of class used in classical Chinese. We do have examples in ancient Chinese texts which demonstrate the existence of the concept of class. For example: 14  The distinction between open sentence and closed sentence is that: at least one variable in the former is free, that is, without being constrained by a quantifier while all the variables in the latter are bound, that is, with a quantifier to constrain each variable.

13  Sameness (Tong 同) and Difference (Yi 異)

229

夫章子, 豈不欲有夫妻子母之屬哉?(孟子離婁下) Moreover, did not Zhangzi wish to be one who has the relationships of husband and wife, child and mother in [the family] that he belongs to? (Mencius VIII 58) (My translation) 凡生天地之間者, 有血氣之屬必有知, 有知之屬莫不知愛其類。(禮記三年問) For all living creatures between heaven and earth, if one belongs to the creatures who have blood and breath, it is necessary that he also has a certain amount of knowledge. If one belongs to the creatures who have a certain amount of knowledge, it is not the case that he does not know to love his species. (Liji, San Nian Wen 3) (My translation)

Here, “one who has the relationships of husband and wife, child and mother in [the family] that he belongs to” (有夫妻子母之屬) and “one who has blood and breath” (有血氣之屬) can be formulated into the following forms of open sentence, respectively: (OS1) x has F More detailed form: [(xR1y)&(xR2z)] (R1 stands for the relation between husband and wife and R2 for the relation between son and mother) (OS2) x has G More detailed form: (xRyz)&(By&Qz) [R stands for the relation of having something, B for blood, and Q for qi (vital force)]

Based on (OS1), I think, we can identify the possible members of the family in Zhangzi’s wish or the actual members of the ordinary family; based on (OS2), we can also identify the members of the sentient creature. The family as a group or the sentient creature as a collection is the class whose members are identified by the respective open sentence. As we know, mereology as an instrument of ontology is handicap to address the problem of personal identity. For example, a person, say, Peter preserves his identity from one moment to the next even in spite of the fact that he gains or loses parts of his body. However, in Chinese language, when Huo 獲 was dead, he cannot be identified or called as “ren” 人 (man) and thus, if he had a ghost-soul after his death, his ontological status is changed. We may explain this change as the shift from one kind of stuff to another kind of stuff. But, we cannot explain why he as an elder brother of someone can still be identified or called “xiong” 兄 (brother) after his death.15 The reason is that: “xiong” is not a stuff name and its referent can be identified as the same person from one moment to the next and from one location to another while a stuff name’s referent cannot. When Mencius said that: “Shun is a man, I am also a man” (舜人也, 我亦人 也。) (Mencius VIII 56), he was not talking about Shun as a human-stuff or human-­ mass is similar to Mencius as a human-stuff or human-mass. If, for the sake of argument, it did say that, when Zhuangzi said that: “you are not me” (formulated as “a≠b) during his debate with Hui Shi (Zhuangzi, II 10: 13), it would have led us to assert that Zhuangzi was talking about that Hui Shi as a human-stuff is not similar  A similar example is: “The ghost-soul of a man is not a man; the ghost-soul of your elder brother is your elder brother.” (Mozi, Xiao-qu 9) (人之鬼, 非人也; 兄之鬼, 兄也。) (墨子:小取 9)

15

230

Y.-m. Fung

to Zhuangzi as a human-stuff. If both of them are mass-stuffs, it does make sense to treat them as similar. But, as indicated in the text, they are not similar. Then, in what sense they are not similar or not identical with each other? I think the only option to explain their difference is that they are two different individuals rather than two dissimilar mass-stuffs. If Hui Shi and Zhuangzi are individuals, Shun and Mencius are individuals as well. If Shun and Mencius are individuals, the sentence “Shun is a man, I am also a man” should be understood as the subsumption relation between Shun and Mencius as individuals and man as a class. It is definitely not about stuff-­ kind. Moreover, for the sake of argument again, if “man” or “human” can be treated as a term of stuff-kind, when Mencius said that “One who has no feeling of commiseration is not a man.” (無惻隱之心, 非人也。) (Mencius III 6), it entails that, in addition to treat man as stuff-kind in terms of being physical existence, man can also be treated as non-stuff-kind in terms of being non-physical existence. In this regard, my question is: what is this man which is not a stuff-kind?

References Fung, Yulan. 1948. A short History of Chinese Philosophy. New York/London: The Free Press and London/Collier Macmillan Publishers. Fung, Yiu-ming. 2020. “Reference and Ontology in the Gongsun Longzi.” In The Gongsun Longzi and Other Neglected Texts: Aligning Philosophical and Philological Perspectives, edited by Wolfgang Behr, Lisa Indraccolo and Rafael Suter, in Series: Welten Ostasiens / Worlds of East Asia / Mondes de l’Extrême Orient 28. Gabbay, Dov M., and John Woods. (ed.) 1984. Handbook of the History of Logic, volume 2: Mediaeval and Renaissance Logic. Amsterdam: Elsevier. Goodman, Nelson. 1946. “A Query on Confirmation.” Journal of Philosophy 43: 383–385. Goodman, Nelson. 1983. Fact, Fiction and Forecast (4th ed.). Cambridge, MA/London: Harvard University Press. Graham, A.C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: The Chinese University Press. Graham, A.C. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle: Open Court. Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press. Harbsmeier, Christoph. 1998. Science and Civilisation in China, vol. 7, part 1, Language and Logic in China. Cambridge: Cambridge University Press. Johnston, Ian. 2010. The Mozi: A Complete Translation. Hong Kong: Chinese University Press Knoblock, John (translation). 1999. Xunzi. Hunan People’s Publishing House and Beijing: Foreign Languages Press. Lao, Sze-kwong 勞思光. 1984. A History of Chinese Philosophy (New Edition) (新編中國哲學 史). Taipei: San Min Book Company. Quine, W.V. 1953. From a Logical Point of View. Cambridge, MA: Harvard University Press. Quine, W.V. 1963. Set Theory and Its Logic. Cambridge, MA: Harvard University Press. Quine, W.  V. 1969. Ontological Relativity and other Essays. New  York: Columbia University Press.

13  Sameness (Tong 同) and Difference (Yi 異)

231

Stemmer, Nathan. 2007. “On Universals: an Extensionalist Alternative to Quine’s Resemblance theory.” Journal of General Philosophy of Science 38: 75–90. Wilson, Jessica. 2006. “Review: Resemblance Nominalism: A Solution to the Problem of Universals.” Philosophy and Phenomenological Research 72: 241–246. Wu, Feibai 伍非百. 1983. Language in the Ancient Chinese School of Names (中國古名家言). Beijing: China Social Sciences Press. http://ctext.org of Chinese Text Project, edited by Dr. Donald Sturgeon Yiu-ming Fung  is Emeritus Professor at the Hong Kong University of Science and Technology and currently Visiting Professor in the Department of Philosophy at the Tunghai University, Taiwan. He is an analytic philosopher with special interest in Chinese philosophy and comparative philosophy. Fung is the author of several books, including The Methodological Problems of Chinese Philosophy (1989); Chinese Philosophy in the Ancient Period, 4 volumes (1992); The Gongsun Longzi: A Perspective of Analytic Philosophy (1999); and The Myth of Transcendent Immanence: A Perspective of Analytic Philosophy on Contemporary Neo-­Confucianism (2003). He has also published more than 100 research papers both in Chinese and English. Email: [email protected]

Chapter 14

Definitions in Pre-Qin Texts Thierry Lucas

1  Introduction Definitions are among the most difficult words to define. They have many forms, they serve many purposes and people have distinguished many different types of definition. It would not be very fruitful to review all these classifications and most of the distinctions are decidedly not clear. Moreover, we are dealing here with definitions in early Chinese philosophy, in which there is an extensive use of definitions but no explicit theory of definition. So let us start from the very simple idea that definitions generally answer the question: what is X? The expected answer is in general a definition and, when fully expanded, it takes the form: “X is an object (or a concept or a class or an individual or an adverb or a verb or a word…) which has such and such characteristics”, more formally “X is an entity x such that A(x).” In this standardized form, we have a term to be defined, the so-called “definiendum” X and the words “an entity x such that A(x)” which define it, the “definiens.” In what follows, we concentrate on the definition of predicates. To avoid misunderstandings, let us emphasize that we take the presence of questions like “What is X?” and answers of the kind described above as our criterion for recognizing definitions in pre-Qin texts. This is a very general approach, but it is already sufficient to make some basic observations, which cover all types of definitions. First, a definition is expressed in a certain language and supposes a certain background theory. It is a trivial observation that it is expressed in a certain language, but it is worth noting that we also need a certain background theory. Take the example of the definition of “platypus”: a platypus is an animal which has four legs, a T. Lucas (*) Institut supérieur de philosophie, Université Catholique de Louvain (Louvain-la-Neuve), Louvain-la-Neuve, Belgium © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_14

233

234

T. Lucas

d­ uck-­bill, lays eggs, lives in Australia, etc. We explain “platypus” by appealing to other words such as “animal,” “four,” “leg,” “duck,” “bill,” etc. This means that if we want to understand the meaning of “platypus,” we have to understand some other words and that we have at least some basic understanding of the world in which we are living: that animals often have legs, sometimes two, sometimes four, sometimes six, they may have a bill, etc… In this case, that basic understanding may be said to form the background theory; we are supposed to understand “animal,” “four,” “leg,” etc. and we explain the meaning of “platypus” using what we know. Pondering on this observation, we may say that we have a background language L and a theory T expressed in that language; the theory T gives the relations between the concepts involved in L; a definition consists then in adding to the language L a new term X, thus forming a new language L∗ and expanding the theory T to a theory T∗ by adding the definition “X is an x such that A(x);” formally speaking, that definition is best considered as a new axiom added to T. Even the so-called “ostensive definitions” have a similar structure, albeit much less explicit; when trying to define “red,” I may use a gesture indicating a patch of red color and saying “red is a color like this;” the background theory here is constituted by my understanding of the gesture and of the words “color” and “like this.” This purely formal observation already explains the structure of formal mathematical definitions or definitions of formal logic: methodologically speaking, one gives a meaning to a new term and one is not obliged to conform to a preliminary usage. A good example of such a definition would be the definition of a mathematical group: a group is a structure constituted by a set and a binary operation on that set satisfying the law of associativity, etc. But even in formal definitions, it is a gross simplification to say that we do not conform to a preliminary usage; e.g. words are generally chosen to be evocative and definitions given in set theory are obviously organized so as to closely correspond to previous usages. Most real-life definitions and philosophical definitions are of that type: one already has a certain knowledge of a concept X and one wants to give a definiens of it in such a way that it closely corresponds to what we know of X. When we have a certain knowledge of X, it may mean that we already have a good idea of the extension of X, i.e. the objects to which it applies; in our definition, we aim at giving a compact characterization of that extension. For example, we could have already seen some platypus in zoos and many pictures of the animal and we try to describe its basic features, so that the entities having these basic features coincide as closely as possible with what we know of its extension; our enterprise is successful if we reach an adequate and complete definition. “Adequate” means that all things which have those features are what we know to be platypus; “complete” means that all things which we know for sure to be platypus have those features. Let us be a bit more careful for completeness: there are perhaps animals which look like the platypus which we have seen in zoos or in museums but for which we have no clear idea of what they are (is the fossil Obdurodon tharalkooschild a platypus?): in this case, our notion of platypus is fuzzy and the zoological definition of platypus may help us to decide which borderline cases should be taken as platypus (the “fossil platypus” Obdurodon tharalkooschild is not the same as the “usual” Platypus anatinus); in this case, completeness becomes a prescriptive property and the definition brings in new information.

14  Definitions in Pre-Qin Texts

235

It could also be that we already have a good idea of the intension of X, i.e. the properties which are satisfied by X, but we want to make it more accurate or more fundamental. This is often the case of philosophical concepts or of lay concepts which we want to approach philosophically. Take the example of “virtuous”: we know that it designates a desirable feature of human behavior, that it implies a respect of others, etc., but we would like to get a better apprehension of it. One way to do that would be to exhibit the “essence” of virtue, i.e. its fundamental features, which make that virtue is what it is; this notion of essence is all but clear; however, it is already enough to rule out “superficial” definitions. Take the well-known example of “man.” “Man is a featherless biped”: this is a good definition; it is adequate and complete in the sense defined above: all featherless bipeds are men (adequacy) and all men are featherless bipeds (completeness); but we prefer the definition “man is a rational animal”; the traditional explanation of that preference is that we can imagine a variation of our world in which men would be for example bipeds with feathers, without loosing their basic properties (the property “featherless biped” is contingent); while we cannot imagine a variation of our world in which men would not be rational animals (the property “rational animal” is necessary). This explanation is by no means clear and looks circular: how will we define “necessity,” “basic property,” without using the notion of essence? I prefer a more formal but simpler explanation which goes back to our first observation: “featherless,” “biped” do not belong to the accepted philosophical vocabulary, at least in this context, while “rational” and “animal” do; the first definition is acceptable in the simple theory of an extraterrestrial just looking for a way to identify men and knowing what “featherless” and “biped” mean; the second is preferred by the philosopher, although it is difficult to say exactly why. Definitions need not be unique: a square is a quadrilateral with four equal sides and four right angles; a square is a quadrilateral with four equal sides and one right angle. With Euclidean geometry as a background theory, these are perfectly equivalent definitions and there is no good reason to say that one is preferable to the other. Definitions rarely come isolated. This may already be suspected on the basis of our first presentation: the definiendum is defined using the definiens, which we may question again. In scientific contexts, they are very often part of a system of definitions. This is at the basis of the very natural requirement that there be no “vicious circle.” This means that there should be no chains of length 1: X defined using X. Nor chains of length 2: X defined using Y and Y defined using X. Nor any chain of higher length n: X1 defined using X2, X2 defined using X3, …, Xn defined using X1. With these preliminaries at hand, we will track definitions in a few typical pre-­ Qin texts and judge them by looking at some of the characteristics which we have hinted at. For the ease of the reader, we present them under four headings, which are natural from a logical point of view: 1. syntactic questions: Is the definition clearly announced as such? What is the background language and the background theory? Is the definition clear? Is it compact enough? 2. semantic questions: Is the definition adequate? Is it complete?

236

T. Lucas

3. relation of the definition with other definitions: Is the definition part of a system of definitions? Is that system free of vicious circles? 4. pragmatic questions: What is the purpose of the definition? In relation with the last question, it is good to remind the reader that pre-Qin thinkers were mainly motivated by ethical and political questions. However, language was an important part of their reflections, in a complex sense, depending from author to author, for which the reader can consult (Willman 2018).

2  Definitions in the Lunyu and in the Confucian Tradition We should not expect that a collection of sayings like the Lunyu (Analects) contain formal definitions such as are found in mathematical or contemporary logical contexts. But there is a clear intention and numerous examples of looking for definitions of notions which became the basis of confucianist philosophy: xiao 孝, junzi 君子, li 禮, zhi 知, ren 仁, … The context is generally given by the interaction of a disciple and the Master. The disciple is asking for the meaning of a term or the content of a notion, with the use of the sentence “… [name of a disciple] 問 [notion]…” This sentence has two not exactly identical interpretations which translators have rendered by “disciple … asks what is…” or “disciple… asks what constitutes …” on the one hand and “disciple… asks about …” on the other hand. The first question is awaiting a complete definition, while a partial definition is enough for the second one. In our approach, the difference does not matter, because we judge adequacy and completeness by examining the answer: the Master’s answer is often a short sentence or a metaphor or an example, which do not in general constitute a complete definition, but there are also cases in which the answer shows an attempt at completeness. Here are a few examples of that diversity. All our quotations and translations of the Analects are taken from [Legge 1893]. Chapter 2, Wei Zheng: Zi Gong asked what constituted the superior man (junzi 君子). The Master said, “He acts before he speaks, and afterwards speaks according to his actions.”

This is a rather partial definition, because there are undoubtedly people who obey the description, yet are not superior man. However, in the following dialogue, the double repetition of “and is this all?” shows a preoccupation for completeness; it is also completed by the canonical example of the two exemplary figures of Yao and Shun. Chapter 14, Xian Wen: Zi Lu asked what constituted the superior man. The Master said, “The cultivation of himself in reverential carefulness.” “And is this all?” said Zi Lu. “He cultivates himself so as to give rest to others,” was the reply. “And is this all?” again asked Zi Lu. The Master said, “He cultivates himself so as to give rest to all the people. He cultivates himself so as to give rest to all the people - even Yao and Shun were still solicitous about this.”

14  Definitions in Pre-Qin Texts

237

Here is a celebrated example of an answer which is very short and aphoristic, but an analysis of which, completed by the context of the times shows that it has some features of a reasonably good definition: Chapter 12, Yan Yuan: The duke Jing, of Qi, asked Confucius about government (zheng 政). Confucius replied, “There is government, when the prince is prince, and the minister is minister; when the father is father, and the son is son.” “Good!” said the duke; “if, indeed, the prince be not prince, the minister not minister, the father not father, and the son not son, although I have my revenue, can I enjoy it?”

Confucius’ short answer is traditionally interpreted as meaning that good government obtains when everybody fulfils his role in society according to his “name,” i.e. its rank and function.1 In this case, prince and minister are typical representatives of relations in the “public” sphere, while father and son are typical representatives of the “private” sphere; all in all they represent every function in the organization of society. Another element aiming at completeness is given by what duke Jing is adding and which logically amounts to: if condition “the prince is prince, and the minister is minister, and the father is father, and the son is son” is not satisfied, then there is no good government. Logically speaking, all this leads us to an equivalence between condition “x is a [good] government” and “x is a state of affairs in which everybody is acting according to its function.” Note also that in the example above, I had to interpret the question as a question about “good” government. This is not an isolated phenomenon in the Lunyu and indeed of the vast majority of other pre-Qin texts: definitions are not used to analyze in a Socratic way the “essence” of a notion, but they are intended to answer in a generally concrete way ethical and political problems. For example, when Fan Chi asks about knowledge, Confucius’ answer is that it is to know how to employ people, employing the upright and putting aside all the crooked, so that the crooked will be made upright; the answer is then completed with the two canonical examples of Shun and Tang (see Chapter 12, Yan Yuan).2 Is there a system of definitions in the Lunyu? There are clearly relations between the different notions, but we can hardly speak of an organized system of definitions: key notions such as yi 義, xin 信, zhong 忠, are not defined and the diversity of definitions of the same notion reveals a lack of care for an economy of notions. 1  Referring to Searle’s idea of the construction of the social world, Confucius’ short answer is obviously based on a constitutive rule “making institutional facts through status-function declarations”: “X counts as Y in context C, if such and such conditions are satisfied,” “X counts as a prince (or a minister, or a father, or a son) in the context of the Zhou institutions if and only if such and such conditions are satisfied.” On this account a better translation of his answer would be: “Let the prince be [or act as] a prince, the minister be [or act as] a minister, the father be [or act as] a father and the son be [or act as] a son.” See for example [Searle 2010:10]. I thank the editor of this volume for calling my attention on this important connection and for the suggested translation, see his “Confucius and Xunzi’s Ideas of Naming.” In [Fung 2017] 2  See also the example of Zi Lu’s question hereabove. It is clear that the point of the definition is to give an explanation of how gentlemen behave and implicitly to make the recommendation to behave in the way described.

238

T. Lucas

Despite these limitations, it is an interesting feature of the Lunyu that we have many definitions or attempts at defining notions. A few words about Mengzi’s definitions. A detailed comparison is beyond the scope of this article, but I would like to mention a nice example of an embryonic Socratic approach of xing 性 (nature). (See [Legge 1895], Chapter Gaozi I). The philosopher Gaozi identifies life and nature: sheng zhi wei xing 生之謂性, “Life is what we call nature.” (We need not discuss here the exact interpretation of this famous assertion, let us simply say that it is often understood as an identification of what makes man’s specificity, his mind, his “Nature,” with what is inborn in him, his physical nature, “Life.”) Mengzi does not agree and counterattacks by looking at that identification: “Do you say that by nature you mean life, just as you say that white is white?” “Yes, I do,” was the reply. Mencius [Mengzi] added, “Is the whiteness of a white feather like that of white snow, and the whiteness of white snow like that of white jade?” Gao again said “Yes.”

Then follows a kind of “retorsion” or “reductio ad absurdum,” the essence of which will be that you will have to identify the nature of man with that of an ox and of a dog: “Very well,” pursued Mencius [Mengzi]. “Is the nature of a dog like the nature of an ox, and the nature of an ox like the nature of a man?”

The text does not conclude, but leaves no doubt that we have to reject this last identification, hence that we also have to reject the first identification, which may be understood as a strong kind of definition introduced by using zhi wei 之謂. What interests us here is that this discussion appended to a definition shows a real attempt at making the notion of xing more precise. In the Xunzi, the organization of concepts increases and the concision of definitions do not leave us very far from the system of definitions of the Mohist Canons (or Moist Canons). Here are two typical examples: (1) What does “oneness” refer to (he wei yi 曷謂一)? I say that it refers to steadfastly cleaving to his spiritual nature. What does “spiritual nature” refer to? I say the utterly good and thoroughly ordered are described as “spiritual nature.” [What does “steadfast” refer to? I say] what none of the myriad of things can deflect from its goal is called “steadfast.” A person who is both spiritual and steadfast is described as a sage. (See [Knoblock 1990: 76], Chapter 8.7 vol. 2).

Oneness is defined using “steadfastly” and “spiritual nature,” which are in turn defined and connected with the notion of “sage.” That these are definitions is clearly announced by the expression 曷謂. (2) To be able to employ ritual and moral principles in serving one’s parents is called “filial piety.” To be able to use them in serving one’s elder brother is called “brotherly affection.” To be able to use them in serving one’s superiors is called “obedience.” To be able to use them in commanding one’s subordinates is called “being lordly.” (See [Knoblock 1990: 104], Chapter 7.9.16a vol. 2.)

14  Definitions in Pre-Qin Texts

239

We find here compact definitions of four key notions of confucian philosophy: xiao 孝, ti 弟, shun 順, jun 君, here translated as “filial piety,” “brotherly affection,” “obedience” and “being lordly.” They are clearly announced as definitions by the expression wei zhi 謂之, here translated by “is called.” Another remarkable example is given in Xunzi 22.2, in which we find a long list of fourteen almost successive weizhi“謂之” tying in a rather complex way notions which are at the heart of Xunzi’s philosophy and which are translated in [Knoblock 1994: 128] by “nature”, “emotions”, “thinking”, “conscious exertion”, “business, “virtuous conduct”, “awareness”, “knowledge”, “ability, “being capable”, “illness” and “fate”. These notions are highly interdependent and form a system of definitions which is not so ample as the system of the Mohist Canons, but in some ways is tighter than it. We refer to [Knoblock 1994: 120–124] for a thorough discussion of the economy and strategy of the passage. As a conclusion to the question of definitions in confucian philosophy, we could say that there is a real attempt at defining notions, but that these attempts are realized to various degrees; there is no real system of definitions in the Lunyu and in the Mengzi, but an increased conscience of their importance in the Xunzi. We now turn to the so called “School of Names.”

3  Definitions in the School of Names Remember that the School of Names is mainly represented by Hui Shi and Gongsun Long. Hui Shi is famous for his 10 theses which come to us via Zhuangzi; two of these theses are clearly intended as definitions; they constitute remarkable examples of definitions and will be discussed in detail here below. On the other hand, what we know of Gongsun Long shows him in his Baimalun much more interested with reasoning itself than with the search of good definitions; the mysterious Zhiwulun could perhaps be considered as involving definitions (see e.g. [Shi 2009]), but problems of interpretation are so controversial that we will not tackle the text here; disregarding problems of authenticity of the other texts, we may say that in Mingshilun, we find an embryonic system of definitions which will be briefly presented below, but there are no clear examples of definitions in Tongbianlun and Jianbailun. Hui Shi’s ten theses are probably the remnants of an involved reasoning tending to promote a kind of “universal love.” The first thesis contains twice the expression 謂之 which is a clear signal that we are facing two definitions. The fifth thesis contains twice the expression 之謂, which is also signalling definitions. Theses 2, 3, 4 and 6, 7, 8, 9 contain paradoxical assertions tending to show the relativity of differences. The last thesis was probably intended as a conclusion of the reasoning: “Universally care for the myriad things. Heaven and earth are one body.” We concentrate on theses 1 and 5 and give three translations of thesis 5: Thesis 1: 至大無外, 謂之大一; 至小無內, 謂之小一。

240

T. Lucas

The ultimately great has no outside, call it the Great One. The ultimately small has no inside, call it the Small One. (translation taken from [Fraser 2012]) Thesis 5: 大同而與小同異, 此之謂小同異; 萬物畢同畢異, 此之謂大同異。 The same on a large scale but different from what is the same on a small scale, this is called “same and different on a small scale.” The myriad things all being the same or all being different, this is called “same and different on a large scale.” (translation taken from [Fraser 2012]) Things which have more in common being different from things which have less in common, that is called micro-differentiation. All things having something in common, and all things differing from each other in some respect, that is called macro-differentiation. (translation given in [Harbsmeier 1998: 294].) A great similarity differs from a little similarity. This is called the little similarity-and-­ difference. All things are in one way all similar, in another way all different. This is called the great similarity-and-difference. (Bodde’s translation in [Fung 1952: 198].)

The first definition is probably intended as a characterization of the universe or totality of all things and of the smallest elements of reality, most probably atoms. We need not decide here, because the formal structure, the economy and compactness of the two definitions are already achievements which are remarkable enough to concentrate our comments. Thesis one defines the Great One and the Small One using the two undefined notions of wu wai 無外 (having nothing outside) and wu nei 無內 (having nothing inside). It is not necessary to go deeply into the interpretation of these two notions to remark that the definitions have a purely formal sense in a very simple background theory. A partial ordering relation ≤ on a set S, i.e. a reflexive, anti-­symmetric and transitive relation on S, is enough to give a completely formalized version of the two definitions of thesis one. Think of S as the set of all things and of the relation ≤ as meaning “is a part of” or “is a sub-object of.” The definition of the Great One is exactly that of a greatest element G: G is the greatest element of S if and only if for all x in S, x ≤ G (“everything is less than or equal to G”). The definition of a Small one is that of a minimal element: m is minimal if and only if for all x, x ≤ m implies x = m; equivalently, for no x does one have x ≤ m and x is different from m; or simply said: for no x, does one have x < m (“m has no inside strictly smaller than it”). It is well known that if there is a greatest element, it is unique: were there G and G', then by the definition of G', one would derive G ≤G' and by the definition of G, G'≤ G, from which G=G' would follow. On the other hand, minimal elements need not be unique. This corresponds to the intuition which is most probably behind Hui Shi’s definitions. The dissymmetry behind the greatest and the minimal is hidden by the symmetry of the formulation of Hui Shi’s definitions and it is no inherent feature of the formalized version given here: mathematicians also use smallest and maximal elements, but distinguish greatest and maximal as well as smallest and minimal. The three different translations of thesis 5 exhibit the difficulty of an accurate interpretation. For a fuller discussion of this point, see [Lucas, forthcoming], of which I give here an abbreviated version. Let us begin by what is clear: the key-­ notion is that of tong yi 同異, which I understand as being a relation of similarity: things may be more or less identified according to the scale; on a small scale, mounts

14  Definitions in Pre-Qin Texts

241

and plains have different heights; on a higher scale, they may be considered as being of the same height. The formal version of that relation of similarity is what is called an equivalence relation on a set S. Recall that an equivalence relation E on a set S is a reflexive, symmetric and transitive relation on that set. The E-equivalence class of an element e of S is the set of all elements e' of S such that eEe'. E-equivalence classes are disjoint and the set of E-equivalence classes is called “quotient of S by E” and denoted by S/E. The second part of thesis 5 is obvious in this setting; among the different possible relations on a set S, there are two canonical ones: (1) One equivalence relation, Em, is distinguishing everything, i.e. x Em y iff x = y and the quotient of S by Em remains essentially the same as S, things are identified only with themselves, and this is a minimal identification (“all things differ from each other in some respect”); (2) The other equivalence relation, EM, identifies all elements of S, i.e. x EM y for all x and y of S and the quotient of S by EM is reduced to one element, which represents a maximal identification (“all things have something in common”). This already explains the second part of thesis 5 and most interpretations read only in the first part that there are other identifications, which are intermediate between the extremes Em and EM; to take an informal example in the spirit of the Chinese examples, we may assimilate things having the same shape; horses and oxen will then form different equivalence classes; since horses are assimilated with one another, that relation Es is different from Em ; but since horses are not assimilated to oxen, that relation is different from EM. In my opinion, these explanations are not enough and fall short of explaining the dynamics of identification which Hui Shi wants to induce in his reader: start from the minimal identification Em and go progressively to the maximal identification EM, leading to his thesis 10 “the universe is one”. If you understand “identification” in an active sense, Harbsmeier’s translation is a better representation of that dynamics. In fact it is easy to represent that movement of identification by calling on the notion of comparison of equivalence relations: equivalence E is finer than E' (or E' is coarser than E) if for all x and y of S, x E y implies x E' y. The relation “is finer than” is a partial ordering relation between equivalence relations; Em is the smallest element of that partial ordering, it is the finest relation; EM is the greatest element of that partial ordering, it is the coarsest relation; the relation of “having the same shape” taken above as an example is between these two relations: Em is finer than Es which in turn is finer than EM. I understand the first part of thesis 5 as calling xiao tong-yi the many possible transitions from usual similarity relations to coarser ones and the second part as calling da tong-yi the unique transition from the minimal identification Em to the maximal identification EM. Drawing on the translations given above, I conclude this section with an almost literal translation of thesis 5 displaying that dynamics of identification: A coarse identification and a fine identification are different; the difference between these identifications is called micro-differentiation. All things may be maximally identified or minimally identified; the difference between these two extreme identifications is called macro-differentiation.

242

T. Lucas

Whatever the fine points of interpretation of theses 1 and 5, it is obvious that Hui Shi’s definitions are amazing achievements in the domain of definitions: they are almost ready-made for formalization, they are clear, accurate and compact and they clearly exhibit their background language and theory. On the other hand, the data are not sufficient to judge whether they were part of a more complete system of definitions. For Mingshilun, we mention the first five assertions, which are clearly intended as definitions of wu 物 (things), shi 實 (entity), wei 位 (in position), fei wei 非位 (out of position) and zheng 正 (correct). Heaven and earth, and what they give rise to, are things. When a thing is taken to be the thing which the thing is and there is no “going beyond,” it is an entity. When an entity is taken to be the entity which the entity is and there is no “being empty,” it is “in position.” If it goes away from its position it is “out of position.” If it is positioned in its position it is correct. (see [Johnston 2004])

These definitions have the merit of compactness and economy and they are obviously part of an organized system which explores the relation between name and reality. However, the list of definitions is too short to give a documented appreciation of it as a system.

4  Definitions in the Mohist Canons From the point of view of logic, Mohism constitutes a giant step forward. In the texts of Early Mohism, one can already observe standardized procedures of argumentations, often based on parallelism. But fully logical discoveries occur mainly in the Late Mohist texts, consisting of 6 parts: the Upper Canons (Jingshang 經上), their explanations (Jingshuoshang 經說上), the Lower Canons (Jingxia 經下), their explanations (Jingshuoxia 經說下), the Big pick (Daqu 大取) and the Small pick (Xiaoqu 小取). We concentrate here on the Upper Canons which begin with 75 very short definitions,3 all of them taking the form: X, Y 也. Here is a sample, constituted by the first ten definitions: (A1) 故, 所得而後成也。 (A2) 體, 分於兼也。 (A3) 知, 材也。 (A4) 慮, 求也。 (A5) 智, 接也。4

3  I follow here the ways of referring and translation of [Graham 1978]; the definitions are 76 in number if you count yi 宜, part of A47, as being a lost definition. 4  智 instead of 知, according to Graham’s comment in [Graham 1978: 77].

14  Definitions in Pre-Qin Texts

243

(A6) , 明也。 (A7) 仁, 體愛也。 (A8) 義, 利也。 (A9) 禮, 敬也。 (A10) 行, 為也。 (1) The gu (reason/cause) of something is what it must get before it will come about. (2) A ti (unit/individual/part) is a portion in a jian (total/collection/whole). (3) The zhi (intelligence/consciousness; 1st tone) is the capability. (4) Lü (thinking/forethought) is the seeking. (5) Zhi (knowing; 4th tone) is the connecting. (6) Zhi (understanding/wisdom; 4th tone) is the illumination. (7) To be ren (benevolent/humane/kind) is to love individually. (8) To be yi (righteous/dutiful/moral) is to benefit. (9) Li (manners/courtesy) is respect. (10) Xing (conduct) is doing. This sample is already enough to point at some characteristics of these definitions. 1. The concepts which they define cover many different fields: logic [(1) in our examples], epistemology [(3)–(6) in our examples], geometry [(2) in our examples], ethics [(7)–(9) in our examples], but also space and time, action, etc.; this shows that there is a clear intention of giving a general foundation to scientific and practical knowledge of the times. 2. The definitions are in general extremely compact, defining a one-character word by another one-character word; judging each individual definition, the contemporary reader will often hesitate to qualify them as clear definitions; three cases appear: –– In most cases the definiens is definitely more concrete than the definiendum: “duration is pervasion of different times;” “space is the pervasion of different places;” “li 利 (benefit) is what one is pleased to get;” “the qiong (limit) is wherever at the next advance there is no room for a measured length;” etc. –– In other cases the definiens is less concrete than the definiendum: “wo (sleep) is the intelligence not knowing of anything;” “to shi 始 (commence) is to be plumb with the time.” –– In still other cases, the very short definiens appears to have the same level of abstraction; this is the case of some definitions defining concepts in term of just ONE other concept: “the zhi 知 (intelligence/consciousness) is the capability,” “lü 虑 (thinking/ forethought) is the seeking,” “zhi 智 (knowing) is the connecting,” etc. This shows that the role of these definitions is not so much to give an explanation of the meaning of the terms than to organize a system of reduction of concepts to other concepts which are considered as more basic; in other words, the background theory is given by a certain philosophical and scientific doctrine which constitutes

244

T. Lucas

the basis of Mohist thinking. In still other words, these definitions should not be judged individually, but as constituting a SYSTEM of definitions. 3. Looking at these definitions as an organized system, one is struck by the remarkable qualities of the whole structure (see [Lucas 2012]): –– There is no “vicious circle” in the system (see introduction). –– The system is rather elaborate: concepts are defined using other concepts which are either undefined or themselves defined in terms of other concepts, which in turn are undefined or defined in terms of still other concepts, etc.; this hierarchization may be quite high and reach 6 levels (example of ci 次). –– The set of undefined concepts is not particularly economical, but remains limited (71 concepts according to [Lucas 2012]); a very interesting result is given by the list of the more used concepts: ran 然 and de 得 are basic and each one is used in five definitions; ming 明 and zuo 作 are basic and each one is used in four definitions; although there is some room for discussion, wei 為 in the sense of “to act” is also basic and appears in 3 definitions; these concepts belong definitely to the heart of the background theory: they confirm the opinion that Mohist philosophy is a practical philosophy more interested in action than in contemplation: ran (to be so) and ming (to be clear) presuppose a prevalence of sense experience; de (to obtain), zuo (to initiate) and wei (to act) are intimately tied to practical action: to obtain practical advantages and to transform reality. –– The division of the set of definitions in “clusters” is a matter of debate, but a close study shows that the net of interconnections reveals bunches of concepts which are closely interconnected and almost independent of one another: such is the case of ethical concepts, of epistemological concepts and of geometrical-­spatial-temporal concepts. As a conclusion, we can say that this first part of the Mohist Canons is a remarkable example of an organized system of definitions: no circularity, clear background theory, conciseness, relative complexity of the system.

5  Definitions at the Periphery of Logic This part is perhaps less conventional than the preceding parts. It deals with 3 other types of definitions; strictly speaking, some of them are not usually considered as definitions but they could and they partake with more usual definitions some important features. Their impact on the philosophy of logic is perhaps limited, but they ought to be mentioned, because they play a significant role in the history of Chinese thought: (1) enumerative definitions, (2) honorifics and (3) fu 賦, rhyme-­ prosed poems. (1) Numerical expressions is a favorite way of Chinese expression and rhetorique, be it pre-Qin or contemporary. Even in pre-Qin texts, most of them are best

14  Definitions in Pre-Qin Texts

245

considered as fixed or stereotyped phrases which have taken an enlarged or abstract meaning. In that case, they are generally not defined; take the example of the number five. In the Liji, there is no definition of the following expressions: wugu 五穀 (literally “the five grains” or more generally “the cereals”), wuse 五色 (literally “the five colours” or more generally “the colours”), wuyue 五岳 (“the five mountains” with the connotation of “sacred mountains”), wuxing 五刑 (“the five punishments” with the connotation of “severe punishment”), etc. On the other hand, some of them, perhaps not so traditional, are explicitly defined by enumeration: The five (administrative) officers (wuguan 五官) of the son of Heaven are: the minister of instruction; the minister of war; the minister of works; the minister of offices; and the minister of crime. These preside over the multitude in (each of) their five charges. (see [Legge 1885], Qu Li II)

These definitions are important in that they specify the extension of the definiendum (they are a special kind of “definitions by extension”), but they do not bring out its “essential” properties. However, in some cases, definitions by enumeration appear explicitly in answer to philosophical questions and they bring out some essential characteristics of a concept. A good example is found in the Lunyu, when Zi Zhang asks Confucius what ren 仁 is. The answer describes five fundamental features of ren and goes like this: Zi Zhang asked Confucius about perfect virtue (ren 仁). Confucius said, “To be able to practice five things everywhere under heaven constitutes perfect virtue.” He begged to ask what they were, and was told, “Gravity, generosity of soul, sincerity, earnestness, and kindness. If you are grave, you will not be treated with disrespect. If you are generous, you will win all. If you are sincere, people will repose trust in you. If you are earnest, you will accomplish much. If you are kind, this will enable you to employ the services of others.” (see [Legge 1893], Analects, Chapter 17, Yang Huo)

The value of this type of definition depends on its adequateness and completeness, qualities which are difficult to assess without entering into the background ethical theory. (2) Honorific titles (jingci 敬辭) and self-deprecatory titles (qianci 謙辭) are an important ingredient of Chinese culture and were especially important in Confucianism where they appear as an essential component of li 禮. Books like the Liji contain an extremely elaborate code of behavior dealing with the right conduct. Among other regulations, the Liji deals with the polite way for person X to address person Y; the title to be used depends on the relative position of X and Y, but also on the circumstances and we find for example in the Liji, an extraordinary casuistics of these appellations: X speaking of himself, X speaking of himself when addressing Y or in the presence of Y, Z speaking of X when addressing Y or in the presence of Y:

246

T. Lucas

When they spoke of themselves, the style of the son of Heaven (天子) was, “I, the One man (予一人);” a chief of regions (伯) described himself as “The strong minister of the son of Heaven (天子之力臣);” the relation of a feudal lord (諸侯) expressed itself by “So and So, the guardian of such and such a territory (某土之守臣某).” If the fief were on the borders, he used the style, “So and So, the minister in such and such a screen (某屏之臣某).” Among his equals and those below him (其於敵以下), he called himself “The man of little virtue ( 寡人).” The ruler of a small state (小國之君) called himself “The orphan (孤).” The officer who answered for him (at a higher court) (擯者) also styled him so (孤). (see [Legge 1885], Chapter Yu Zao )

And it goes on in a similar manner for a Great officer of the highest grade 上大夫, a Great officer of the lowest grade 下大夫, the son and heir of a feudal prince 世子, a ruler’s son (by an inferior lady) 公子, etc. These are not definitions in the same sense as our previous examples: they do not present themselves as answering the question “what is X?” or “what is the meaning of X?”, but they could be presented like that and they share with more familiar definitions some important characteristics. Firstly, they could be presented like definitions, e.g. answering the question “what is gua ren 寡人?” “what is the meaning of gua ren 寡人?”, “what is yu yi ren 予一人?” “what is the meaning of yu yi ren 予一人?”, etc. A literal answer would be barely satisfactory and their use as conventional expressions is an important ingredient of their meaning. Compare with a question like “what is a yardstick?” Webster’s dictionary says: “a graduated stick, three feet in length, used in measuring.” (see [McKechnie 1983], definition of “yardstick”) Secondly, they are conventional appellations, but not completely arbitrary: “gua ren” is the conventional auto-designation of X in the presence of Y, but it retains some ingredients of its literal meaning “few” (寡人 = 寡德之人, “man of few virtues”) (See [Li 2001:167].). Compare with mathematical definitions: “the intersection of two sets A and B is the set of elements belonging to A and to B” is conventional (it could have been called otherwise, e.g. “common part of A and B”), but it is not totally arbitrary (the word “intersection” corresponds to a diagrammatic representation). Thirdly, they are prescriptive as are some terminological appellations, for example the Latin nomenclature of plants, which constitutes a norm of designation for scientific purposes. Finally, they form an elaborate system guided by certain rules (the speaker takes a humble position, the addressee is magnified, the vocabulary depends of the socio-­ political position of the speaker and of the addressee) which reminds us of what we saw for the Mohist system of definitions or for modern systems of appellations. (3) Definitions in the rhyme-prosed poems fu 賦. Fu is a literary genre which is very specific of Chinese culture; they are pieces of literature between prose and poetry and their explicit aim is to describe an object or

14  Definitions in Pre-Qin Texts

247

a feeling or a natural phenomenon in many details and from many different points of view. They interest us here, because they may be considered as a very special literary version of a definition. We find five of them in Book 26 of the Xunzi, under the form of riddles: somebody gives a first description, he pretends not to recognize the described thing, asks an authority, who goes on with the description and gives the answer. The five fus of Book 26 describe li 禮 (ritual principles), junzi zhi zhi 君 子之知 (the knowledge of the gentleman), yun 雲 (clouds), can 蠶 (silkworms) and zhen 箴 (needle). Their hidden theme is the qualification of the gentleman, and in particular of Xunzi himself, to hold office. His knowledge of ritual could transform the untutored, animal-like qualities of human nature into composed, elegant forms. His wisdom could improve the individual, perfect the government, and maintain the state. His “clouds” are his agility and comprehensiveness. His “silkworm” is his responsiveness to continual changes. His “needle” is his critical acumen that takes diverse ways of thinking and combines them into a harmonious and useful whole. (see [Knoblock 1994: 189], vol. 3.)

Here are extracts of the first fu: Here there is a great thing: It is not fine silk thread or cords of silk, yet its designs and patterns are perfect, elegant compositions (文理成常). It is not the sun, nor is it the moon, yet it makes the world bright. [...] Your servant stupidly does not recognize it. And presumes to ask Your Majesty about it. The King replied: Is it not something that has cultivated form, yet is not brightly colored? Is it not suddenly and easily understood, yet especially possesses natural order? Is it not what the gentleman reveres and the petty man does not? Is it not something that if inborn nature does not acquire it, one is like a wild beast; [...] I suggest where all qualities come together is ritual principles (禮).

The expression of the definition is of course here extremely far from our canonical form “X is an x such that A(x).” It plays on words and is voluntarily ambiguous in that it has a literal meaning and another more profound meaning which the hearer is supposed to decode. What is interesting here is that these fus may be understood as true definitions of concepts: some basic characteristics of li are indeed described or hinted at here; here is how Knoblock explains the first lines: [...] since the topic of the riddle is li 禮 ritual principles, wen 文 means as well the perfected forms of social conduct embedded in the ritual and culture acquired by those who master its principles. The li 理 lines of the design evoke the principles of order and reason that underpin all nature, as seen in the lines (li) that that distinguish true jade from other stones and in the lines (li) in our hands which vary from individual to individual. Since ritual principles accord with the principles of order in nature, the individual and state that observe them can be but perfected (cheng 成). The overall design, the elegant composition (zhang 章) is the achieved result, the beauty of variation of color, ornamentation and pattern, realized in the fabric, painting, or building as well as the movement of poetry, song or dance. It is made brilliantly manifest (zhang) and apparent as the badge of accomplishment (zhang) of the gentleman who has mastered ritual.

248

T. Lucas (see [Knoblock 1994: 190], vol. 3.)

Another characteristics is also obvious in the king’s answer: li 禮 is not innate, but acquired. From our point of view, this means that we have here genuine definitions, but with a background theory which is a whole set of cultural knowledge and poetical conventions. Needless to say that they are far from being compact, clear or formalized!

6  Conclusion Definitions are everywhere in the pre-Qin texts. Some of them are extremely accurate, concise, almost formal: such are Hui Shi’s definitions. Confucian definitions are trying to find the fundamental characteristics of philosophical concepts such as ren. Mohist definitions constitute a remarkable, almost axiomatic attempt at constructing a philosophical system encompassing logic, ethics, epistemology and cosmology. Definitions by enumeration are frequent, often factual but sometimes philosophically dense. Honorific titles and self-deprecatory titles of the Liji are almost pure conventional definitions. Fus, especially Xunzi’s fus, constitute also a kind of definition sui generis. Our presentation is far from being exhaustive and there is no doubt that it could be extended to other types of definitions. However, our examples have shown that all these types of definition have very different characteristics and at the same time share the basic common feature of definitions: to answer the question “What is X?” in a background language woven by a background theory.

References Fraser, Chris. 2012. “School of Names.” In Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter 2012 Edition). URL = http://plato.stanford.edu/archives/win2012/entries/ school-names/. Fung, Yiu Ming. 2017. “Confucius and Xunzi’s Ideas of Naming.” Ms. Soochow University, Taiwan. (A third chapter of a forthcoming book, named “Language, Truth and Logic in Ancient China.”) Fung, Yu-lan. 1952. A History of Chinese Philosophy, vol. 1. Princeton: Princeton University Press. Graham, A.  C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong/London: Chinese University Press/School of Oriental and African Studies. Harbsmeier, Christoph. 1998. Science and Civilisation in China, vol. 7, Part 1: Language and Logic. Cambridge: Cambridge University Press. Johnston, Ian. 2004. “The Gongsun Longzi: A Translation and an Analysis of Its Relationship to Later Mohist Writings.” Journal of Chinese Philosophy 31.2: 271–295. Knoblock, John. 1988/1990/1994. Xunzi: A Translation and Study of the Complete Works, 3 volumes. Stanford: Stanford University Press.

14  Definitions in Pre-Qin Texts

249

Legge, James, trans. 1885. The Sacred Books of China. The Texts of Confucianism. Part III and IV. The Lî Kî. Volume 27 and 28 of Sacred Books of the East. Oxford: Clarendon Press. Legge, James, trans. 1893. Confucian Analects, the Great Learning, and the Doctrine of the Mean, Revised second edition, Volume 1 of The Chinese Classics. Oxford: Clarendon Press. Legge, James, trans. 1895. The Works of Mencius, Revised second edition, Volume 2 of The Chinese Classics. Oxford: Clarendon Press. Li, Xueqin 李學勤, ed. 2001. Li ji zheng yi “Qu li” 禮記正義 “曲禮”: The Correct Meaning of the Book of Rites, “Summary of the Rules of Propriety,” In volume 21 of Shísān jīng zhùshū: Zhěnglǐ běn 十三經注疏: 整理本. Taipei: Taiwan Gu Ji. Lucas, Thierry. 2012. “Definitions in the Upper Part of the Moist Canons.” Journal of Chinese Philosophy 39.3: 386–403. Lucas, Thierry. Forthcoming. “Mingjia, the School of Names,” submitted for publication. McKechnie, Jean L., ed. 1983. Webster’s New Twentieth Dictionary of the English Language, Unabridged, 2nd edition. New York: Simon and Schuster. Searle, John R. 2010. Making the Social World. The Structure of Human Civilization. Oxford: Oxford University Press. Shi, Ningzhong 史宁中. 2009. Lun dingyi zhongde shuxiang yu gongxiang—Gongsun Longzi “Zhiwulun” pingxi 论定义中的殊相与共相—公孙龙子指物论评析: The Roles of Particularity and Commonality in Definition: A Study of Gongsun Long’s Zhiwulun. Gudai Wenming 古代文明 (The Journal of Ancient Civilizations), 3.1: 22–27. Willman, Marshall. 2018. “Logic and Language in Early Chinese Philosophy”. In Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter 2018 Edition). URL = https://plato. stanford.edu/archives/win2018/entries/chinese-logic-language/ Thierry Lucas  is Emeritus Professor of the Université Catholique de Louvain (Louvain-laNeuve, Belgium). His research interests have always been in the domain of logic and his publications range from algebraic logic to connections of logic with the mathematical theory of categories, modal logic, deontic logic and applications of logic to the understanding of some Chinese philosophical texts dealing with logical problems. He has taught logic at all levels to students of different Faculties and has been a Dean of the Faculté des Sciences Philosophiques of his University.

Part II

Issues and Theories in Chinese Philosophy of Logic

Chapter 15

Logical Thought in Mohism and Later Mohism Thierry Lucas

1  Introduction Mohism flourished during the last part of the fifth, the fourth and the third century BCE during the “Warring states” period (479–221 BCE). It originates in the teachings of Master Mo (Mozi or Mo Di) about whom little is known, but exegesis of the texts suggests that he was a man of modest origin, probably a carpenter (Graham 1978) or perhaps even a slave or a convict (Fraser 2014). One distinguishes two stages in Mozi’s doctrine: the Early Mohism and the later Mohism. Early Mohism is characterized by a number of moral, social and political doctrines, most notably “inclusive care” (a translation of jian ai, now preferred to “universal love”), economy of expenditures and simplicity in funerals, condemnation of military aggression, rejection of fatalism. It is often characterized by the term of consequentialist ethics: in sharp contrast to Confucianism, behavior should always be measured in terms of li(4th tone)/hai, benefit/harm; for example, fatalism has evil consequences for society, people will not strive for being virtuous; consequently, fatalism should be rejected. For the point of view developed here, this doctrine is interesting not because of its moral consequences, but because it is expounded in texts which are full of arguments and implicit logical reasonings, of which we will give examples in Part II. Later Mohism is the doctrine of a branch which evolved from early Mohism and flourished at approximately the same period as the Logicians (Hui Shi and Gongsun Long). Whether Later Mohism predates the Logicians is a matter of scholarly ­dispute and has an impact on considerations of authenticity of the Logicians’ texts. From our logical point of view, Later Mohism is a very important – if not the most

T. Lucas (*) Institut supérieur de philosophie, Université Catholique de Louvain, Louvain-la-Neuve, Belgium © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_15

253

254

T. Lucas

important  – witness of the development of logical ideas in the history of China, which already contains in germ many concepts of contemporary logic and developed along lines which are quite different from the traditional accounts of Aristotle’s syllogistics. Their core doctrine and main concepts will be expounded in Part III. Mohism did not survive Qin Shi Huang Di’s promotion of Legalism and later Han Wu Di’s promotion of Confucianism as a doctrine of State. It was almost completely forgotten by Chinese tradition and it is only at the end of the nineteenth century that it was rediscovered by a number of scholars of the Qing dynasty. The study of Later Mohism was greatly enhanced in the West by Graham’s masterpiece (Graham 1978) and it receives now much attention from Eastern as well as Western scholars. See Defoort and Standaert (2013: 2–3), for a brief but informative summary. Our sources are constituted by the collection of Mohist texts, known as the Mozi (eponym of Master Mo). These texts are of different epochs and different origins, but they show a certain homogeneity of preoccupations. They are constituted by 71 chapters (pian) which are often presented in five Groups: –– Group 1 contains seven chapters on various issues. –– Group 2 contains Chapters 8–39 which expound the main elements of Mohist doctrine and have the particularity of giving three (or two in one case) versions of the same theme; the origin of the different versions is not clear, but could be different stages of redaction. –– Group 3 contains Chapters 40–45: Canons I and II, Exposition of Canons I and II, the Daqu “Major Illustrations” and the Xiaoqu “Minor Illustrations”. It is often referred to as the “Mohist Dialectical Chapters”; a striking feature of these chapters is that logical notions take a very important place; they are most probably of a later origin, which explains why they are referred to as “Later Mohism.” –– Group 4 contains Chapters 46–50, recording Mozi’s conversations. Chapter 51 has been lost. –– Group 5 contains Chapters 52–71, devoted to military theory. As already hinted at, we will begin by some observations on Groups 1 and 2, which are already extremely rich in argumentative procedures and use of some elementary logical reasonings showing the attention of the Mohist School to logical matters. We will then discuss the Mohist dialectical chapters in our Part III and conclude in Part IV.

2  Logic and Argumentation in Early Mohism The texts of Early Mohism do not specifically discuss logical concepts, but they are strikingly different from other texts in that argumentation is explicit and plays a very important role in the exposition of the Mohist theses. Most of these arguments are not cogent (that is why they are under the heading of “argumentation”) but they are systematic and sometimes contain “truly logical” reasonings.

15  Logical Thought in Mohism and Later Mohism

255

Early Mohists use many different types of arguments1 but we restrain our analysis to concentrate on three examples in this section: the first one presents a detailed analysis of a nice piece of reasoning and exhibits its hidden logical structure; the second one illustrates how the frequently used direct parallelism embodies an argument of generalization; and the third one shows how opposite parallelism embodies an argument of equivalence. Example 1  In general, the superficial form of the arguments is not logical, but a slightly deeper analysis shows that some of them are almost ready made for formalization and are logically valid, when supplemented by hidden premises which the context gives us as valid. Here is the analysis of a nice piece of reasoning contained in Book 5, “Condemnation of Offensive War III.”2 I summarize the lengthy argument whose logical substance is this: “If the attitude of the ruler is aggressive (in entering wars), then it will not bless Heaven, it will not bless the Spirits and it will not bless the people”

The intended conclusion is obviously that: “The ruler should not have an aggressive attitude.”

The logical reasoning which is behind this is clearly formalizable when you simplify it a bit and make the hidden premises clear. Let H stand for Heaven, S for the Spirits and P for the people, A(y) for “y is an aggressive attitude of the ruler.” The details of the following reconstruction may be criticized, but they could be refined and they are simply intended to suggest the richness of the background logic. The premise is: “If A(y), then y does not bless H and y does not bless S and y does not bless P.” Standard logical transformations give: “ A(y) →  ∀ x (x = H ∨ x = S ∨ x = P → y does not bless x)” An implicit premise is that: “∀x (x = H ∨ x = S ∨ x = P)” (Everything is either Heaven or the Spirits or the People.) From this, it follows that “A(y)→ ∀x (y does not bless x)” (If y is an aggressive attitude, it does not bless any x.). Another obvious judgment is that “∀x (y does not bless x)” is a highly undesirable consequence U; hence, “(A(y) → U)”. 1  The most frequent arguments are arguments by comparison, appeal to the examples and “counterexamples” (the Sage Kings vs the wicked Kings), arguments of generalization, exhaustion of cases, chains of “implications” or ordering relations, argument “a fortiori,” use of direct parallel structures or opposite parallel structures (see Lucas 2013), positive consequentialist argument (“If you do X, you will have good consequences; hence you should do X.”), negative consequentialist argument (“If you do X, you will have bad consequences; hence you should abstain from doing X.”), reciprocity of relations. 2  References and translations given in this part of the article are from Mei (1929).

256

T. Lucas

This conclusion is presented as a necessity (which will be formalized here by a square ‘□’) and generalized in y, so that it is supposed to prove that “□ ∀y (A(y) → U)” (Necessarily, every aggressive attitude is undesirable.), or equivalently “□ (∃y A(y) →U)” (Necessarily if there is an aggressive attitude, that is undesirable.). We may call here notions of deontic logic and the well-known Anderson-Kanger reduction of deontic logic to modal logic (OA may often be reduced to □ (A → U); see McNamara (2014) for a clear and brief account and for references) to conclude “O ¬∃y A(y)” i.e. “it is obligatory that there be no aggressive attitude of the ruler.” The analysis of numerous other examples allows us to conclude that Early Mohism was especially attentive to disputation by giving careful and transparent argumentation. This paved the way for Later Mohism which began to theorize about bian (disputation) and the notions involved therein. For the sake of later reference here below, we also mention the cardinal importance of parallelism of sentences which is used to support a great variety of arguments. This parallelism does not in general correspond to a logically valid reasoning, but its importance should not be underestimated not only in Mohist writings, but also in all classical writings and even in the foundations of Chinese culture (see Lucas 2013). One may distinguish two kinds of parallelism: direct parallelism and opposite parallelism. Example 2  Here is an example of direct parallelism, given in Chapter 1, “Befriending the Learned” and supporting an argument of generalization: Thus, Bi Gan died of his uprightness; Meng Ben perished by his strength; Xi Shi paid with her life for her beauty; and Wu Qi was torn alive for his achievement. This shows that there are but few who excel other people and do not perish on account of it. Hence the saying: Position of the supreme is hard to keep. To show the underlying generalization, we sketch a first analysis of this example. The premises have parallel forms:

Q  a   P  a  , Q   b   P  b  , Q   c   P  c  , Q   d   P  d 



(where a stands for Bi Gan, b for Meng Ben, etc.; Q stands for “was upright”: Q′ for “was strong”, etc.; and P stands for “perished of a violent death”). A first step generalizes the predicates Q, Q′, Q″, Q‴ to a predicate E (for “excel other people”), yielding

E  a   P  a , E b  P b, E c   P c , E  d   P  d .



15  Logical Thought in Mohism and Later Mohism

257

The second step generalizes a, b, c, d yielding a “nearly universal” sentence:

Most x  E ( x )  D( x )  ,



namely “most persons which excel other people perish of a violent death”. In contrast to our first example in which we discern a logically valid argument, the argument of generalization is not logically valid, but it often has an undeniable persuasive strength. Example 3  Opposite parallelism groups all variations and extensions of the basic scheme: (p → q)/(¬ p → ¬ q) i.e. pass from “p implies q” to “not-p implies not-q”). Of course, the Mohists do not claim that from (p → q), one can deduce (¬ p → ¬ q), but they use this type of parallelism to establish a kind of equivalence between p and q. For example, we find in Chapter 7 “Will of Heaven I”: For, with righteousness the world lives and without it the world dies; with it the world becomes rich and without it the world becomes poor; with it the world becomes orderly and without it the world becomes chaotic. In this case, if we assume that “to die” is the negation of “to live,” “to be poor” is the negation of “to be rich” and “to be chaotic” is the negation of “to be orderly,” logic helps us to exhibit the structure of this paragraph: If the world is with righteousness, then the world lives and if the world is not with righteousness, then the world does not live. If the world is with righteousness, then the world becomes rich and if the world is not with righteousness, then the world does not become rich. If the world is with righteousness, then the world becomes orderly and if the world is not with righteousness, then the world does not become orderly. (1) (p → q) ∧ (¬p → ¬q), (2) (p → r) ∧ (¬p → ¬r), (3) (p → s) ∧ (¬p → ¬s). From (1) one derives the equivalence of p and q: (p⟷q). Similarly, from (2) and (3), one derives (p⟷r) and (p⟷s). In the end, since equivalence is symmetric and transitive, all sentences p, q, r, s will appear to be equivalent. We have exemplified above the logic and argumentation which have been used by the Early Mohists, but one could also set the question of a quasi-explicit theory of argumentation in chapters like Chapters 35–37, in which concepts such as fa 法, biao 表 and bian 辨 are dealt with. It is the opinion of the present author that they are presented there as a general way of evaluating a theory (anti-fatalism in the case), and are not embedded in a complex theory relating them with other logical notions, as is the case in Later Mohism. It seems however without doubt that these concepts prepared the way for the much more theoretical approach which we find in Later Mohism (see Sect. 3.1 of the present text). In any case, it remains interesting

258

T. Lucas

to give a precise account of all the techniques of argumentation of the Mohists; an important step in that direction is given in Liu et al. (2011) which analyzes a typical example with techniques of the theory of games.

3  Logic in Later Mohism Logical considerations appear explicitly in the Later Mohist texts. In modern terms, we could say that there is a considerable reflective work on reasoning, on language, on syntax and semantics, which of course, do not appear as such, but show through a number of key Mohist concepts: bian 辨 (disputation), yin 因 (the criterion), fa 法 (the standard), lei 類 (the kind), ming 名 (names), ju 舉 (to refer), ci 辭 (proposition), dang 當 (fitting the facts), shuo 說 (explanation). Moreover, quantifiers, modal concepts (possibility and necessity), a priori, considerations on time, space and infinity, paradoxes are all present in the Later Mohist texts. The difficulty is that all these concepts are intermingled and it is very difficult to explain one without invoking the other. Hence, we will have to choose some way of exposition of the doctrine, but other ways exist and give complementary points of view on Mohist Logic. Our presentation is organized by starting with what seems to be the backbone of Mohism, from simple to more elaborated, closely following the thread of notions as they appear in the text: Sect. 3.1 deals with names and especially those corresponding to our modern predicates (e.g. “ox”), which serve to build atomic propositions (e.g. “this is an ox”); Sect. 3.2 deals with leis (e.g. “the kind of oxen”), a complementary approach to predicates; Sect. 3.3 deals with compound names (e.g. “white horse”), elementary propositions (e.g. “a white horse is a horse”) and more complex propositions (such as “riding a white horse is riding a horse”); Sect. 3.4 deals with patterns of reasonings, e.g. allowing the transition from “a white horse is a horse” to “riding a white horse is riding a horse,” or forbidding the transition from “a robber is a man” to “killing a robber is killing a man;” Sect. 3.5 deals with argumentation, a central concept of Mohism and of the many schools of philosophers of the time. We will then complete that approach by an enumeration and less organized discussion of logical notions discovered or foreshadowed by the Later Mohists: Sect. 3.6 deals with a priori, necessary and sufficient conditions, propositional connectives, modalities, quantifiers, notions of time, space and infinity, paradoxes. The last section of our exposition, Sect. 3.7 is devoted to the Mohists’ implicit logic; by this, I mean the logical organization and logical qualities of their writings, which was used by them without being the object of an explicit discussion: the organization of the definitions of the Canons, the normalization of their technical language and the systematic comparison of reasonings are the foremost examples of these qualities.

15  Logical Thought in Mohism and Later Mohism

259

3.1  M  ing, fa, yin: Names, Standards, Criterion, Atomic Propositions Ming is of course a fundamental concept in early Chinese philosophy and its meaning varies according to the different schools. To take but one famous example, it is clear that in Confucius’ Analects, zhengming 正名 (rectifying names), is essentially rectifying or recovering the moral behavior which is normally associated with the name; it is not really detached from the linguistic expression, but there is no doubt that its interpretation is intimately tied with a behavior and its possible modifications. It is remarkable that in Later Mohism, names are studied in themselves, detached from immediate moral considerations, with syntactic and semantic considerations. Here are two basic texts of the Mohist Canons, where ming appears3: Canon A78: Ming (name). Unrestricted; classifying; private. (名。達, 類, 私。) Exposition of the Canon: “Thing” 物 is “unrestricted.” 達 –– –– –– –– ––

Any object necessarily requires this name. Naming something “horse” 馬 is “classifying.” 類 For “like the object” (ruo shi 若實) we necessarily use this name. Naming someone “Jack” 臧 is “private.” 私 This name stays confined (zhi 止) in this object. […]

Canon A80: Zhi 知 (know). […]. The name 名, the object 實, how to relate, […]. Exposition of the Canon: […] What something is called by is its “name.” (所以謂, 名也。) What is so called is the “object.” (所謂, 實也。) The mating of “name” and “object” is “relating.” (名實耦, 合也。) [...] And the Xiaoqu confirms the role of names as follows: NO11. One uses names to refer to objects (以名舉實), […] The first text clearly shows that with their classification of names, the Mohist identified three basic syntactic ingredients of language, in which the modern reader will immediately recognize: –– individual variables whose value may be any object of the universe of discourse; –– predicates whose value is extensionally described as a class of objects of the universe; –– individual constants whose value is a fixed object of the universe. Ming has thus the role of “designator,” “term used to designate.”

3  All quotations and translations of the Mohist dialectical chapters are borrowed from Graham (1978) and follow his notations, without mention of the pages where they appear.

260

T. Lucas

For individual constants, the association of name and objects is qualified by the verb zhi, “to stay,” which is defined as “to endure as it was” in Canon A50; an examination of the use of zhi in the Canons shows that its meaning is that of a fixed, definite, lasting association; it is exactly what we expect of individual constants. For predicates, the semantic hint is that they are “classifying” and that a comparison procedure (ruoshi) is at work. We will henceforth concentrate on predicates. The second text clearly spells out the idea of reference of a name: the reference of a name is an object (shi) and the successful association of the name with the object is qualified by the agreement (he) of the name and of the object. The third text qualifies this notion of reference by the term ju, to choose, to select. If we want to fully appreciate these notions, we have to clarify some of the key words which are used. They will give us precious indications on the underlying Mohist theory of reference. The verb ju is defined in Canon A31: Ju (to refer to/pick out by name from others) is to present the analogue for the object. (ju ni shi ye 舉, 擬實也。) Thus, when we want to apply a predicate like “horse” to a concrete entity, we appeal to an “analogue,” which may be conceived of as a prototypical “horse” or a “prototypical way of identifying a horse” and we explain that that concrete entity is like the prototypical horse or obeys our way of identifying horses. This justifies our naming the animal “horse”: this object is like the prototypical horse, and therefore may be named “horse.” In fact the idea of prototype is more liberal than what our word “prototypical” suggests: it may be a concrete example of a horse or a concrete image of a horse or a mental image of a horse and in the Mohist approach, it need not be conceived of as an abstract prototype or as a Platonic form. To be more definite, this notion of prototype is covered by the name of standard (fa) and what this standard covers is exemplified and related with the idea of comparison (suoruo) in:  Canon A70: The fa (standard) is that in being like which something is so. (法, 所若 而然也。) Exposition of the Canon: The idea, the compasses, a circle, all three may serve as standard. (意規員三也, 俱可以為法。) It should also be noted that fa is an important concept of Early Mohism which devotes a chapter to the “necessity of standards” (Mei 1929, Chapter 4). Liu et al. (2011) has also emphasized the notion of criterion yin which gives rise to a notion of comparison: Canon A71: The yin (criterion) is that wherein it is so. (因, 所然也。) Exposition of the Canon: Being “so” is the characteristics being like the standard. ( 「然」也者, 皃 (mao, variant of 貌) 若法也。) The criterion is something like shape or color, and the important point is that each criterion induces a similarity relation: “shape” gives rise to the similarity relation “having the same shape as”, “color” gives rise to the similarity relation “having the same color as”, etc. The importance of this comparison process is confirmed by

15  Logical Thought in Mohism and Later Mohism

261

Canon B70, which uses “color” as a criterion and insists on the derived relation “has the same color as”: Canon B70: When you hear that something you do not know is like something you do know, you know them both. Explained by: informing. Exposition of the Canon: A thing outside you do know, the thing inside the room you do not know. Someone says: “The colour of the thing in the room is like the colour of this.” Then the thing you do not know is like the thing you do know. It is as with “White or black, which does one win? This is like its colour, and what is like a white thing is necessarily white.” In the present case too you know that its colour is like the white, therefore you know that it is white. A name by means of something you are clear about determines something you do not know, it does not bring in the thing you do not know to cast doubt on the thing you are clear about. (Like measuring an unknown length by means of a footrule.) What is outside, you know by experience; what is in the room, you know by explanation. The idea behind this is clear and starts from very concrete examples: when we want to construct or to recognize a circle, we use the notion of circle, or we use a compass, or we are given a concrete example of circle to which we may compare our construction. We can also infer from this that all objects satisfying the norm are similar: all horses are similar in obeying the norm, and this may be seen by applying the criterion (yin) “shape” and its derived similarity relation “has the same shape as” to the prototypical horse. A possible modern approach of this could be given in the spirit of Liu et  al. (2011), by saying that with a predicate H (say “horse”), the Mohist associates a norm faH (“the prototypical horse”) and a similarity relation yinH (“has the same shape as”) in such a way that his interpretation of H would be in Tarskian terms: The object a is in the extension of H iff there are objects b obeying the norm faH and a is related to b by the relation yinH. Another contemporary way of explaining this would be to say that the Mohists consider basic atomic sentences of the form Hx and that they define the truth of Hx with the help of faH and yinH by: Hx is true for the value a of x iff there are objects b obeying the norm faH and a is related to b by the relation yinH. The extension of H results from the grouping together of all objects similar to the prototypical horses; the Mohist call it the lei of horses; this is what we now call the class or the set of horses, but there is much more in the Mohist notion than a pure

262

T. Lucas

set theoretic notion, for leis are embedded in a net of connections to which we devote next section. We have assumed here that we are dealing with rather clear-cut predicates such as “to be a horse,” “to be a circle,” for which the similarity relation may be assumed to be reflexive, symmetric and transitive. For fuzzy predicates such as color predicates, one should only assume that the associated similarity relation is reflexive and symmetric, for transitivity could lead us into contradiction: with imperceptibly different shades of color under the perception threshold, say shades of blue, we could have standard blue objects b and b′ and an object a such that a is similar to b, b is similar to b′, but a is not similar to b′. A detailed study of fuzzy predicates and of general similarity relations is worth being done and there is no doubt that the Mohists as well as the Mingjia were very conscious of limit cases, of infinitesimal differences and the like, but it would lead us far and we do not have enough texts to elaborate their underlying theory. We will therefore make here the simplifying assumptions that any two standard elements satisfying the norm are in the relation yinH and that the relation yinH is an equivalence relation, i.e. a reflexive, symmetric and transitive relation. It follows from that assumption that the definition of the extension of H could be given by: The object a is in the extension of H iff for all objects b obeying the norm faH, a is related to b by the relation yinH. The difference with the preceding formulation is that the quantifier “there are objects b” is replaced by “for all objects b.” We leave to the reader the proof of the equivalence of the two formulations under the simplifying assumption.

3.2  Leis and Similarities, Sets, Sorts or Kinds Lei is the key notion of Mohist logic. It is defined neither in the Mohist Canons nor in the Daqu nor in the Xiaoqu, but it appears in many places, where it is related to many other fundamental notions. It is a notion which the Mohists took for granted and it is precisely for that reason that they are important to study. We consider here a few texts which throw light on some aspects of leis: how are they related to similarity relations, to sets and to characterizing properties and do they presuppose a notion of sort or kind? 1. Leis as equivalence classes of equivalence relations If we return to A78, we see that in the Mohist Canons, there is a complementary approach of predicates, which instead of the standard fa emphasizes the notion of lei, which is a key concept of Mohist logic; as a verb, lei means “to classify;” as a name, it would be the result of that classification and translations with different connotations may be proposed: class, set, kind, type, sort, etc. The reason is that there is considerable difficulty in fixing what a lei is. Besides the association with fa and

15  Logical Thought in Mohism and Later Mohism

263

yin already encountered above, we find a confirmation of the importance of relations of similarity in A86 and A87, which associate leis with the notions “similarity and difference,” tongyi: Canon A86: Tong (same). Identical, as units, as together, of a kind. (同。重, 體, 合, 類。) Exposition of the Canon: There being two names but one object is the sameness of “identity.” (二名一實, 重同也。) Not being outside the total is sameness “as units.” (不外於兼, 體同也。) Both occupying the room is the sameness of being “together.” (俱處於室, 合同 也。) Being the same in some respect is sameness in being “of a kind.” (有以同, 類同 也。) Canon A87: Yi (different). Two, not units, not together, not of a kind. (異。二, 體, 不合, 不類。) Exposition of the Canon: The objects if the names are two necessarily being different is being “two.” (二 必異, 二也。) Not connected or attached is “not units.” (不連屬, 不體也。) Not in the same place is “not together.” 不同所, 不合也。 Not the same in a certain respect is “not of a kind.” (不有同, 不類也。) We find in these Canons an attempt at classifying relations of similarity and difference. The foundation of this classification is not too clear (and it was revised in NO6) but we may confidently describe the four different types by systematically using the expression “the same X”: –– names having the same reference (the stock example of the Mohists “whelp” and “dog” could serve as an illustration); –– objects belonging to the same set or being parts of the same entity (two particular horses are similar in belonging to the set of horses or to this particular group of horses, two parts of this stone are similar in being part of this stone); –– objects having the same location; –– objects being of the same lei. Among the different relations of similarity, the relation of being of the same lei is certainly a fundamental one. There is little doubt that we find again here our similarity relation already hinted at in Sect. 3.1. This confirms that we have to treat the leis as closely associated with relations of similarities and the relation is very easily described in terms of equivalence relations and equivalence classes. On the one hand, when we have an equivalence relation (say “having the same shape as”), it

264

T. Lucas

partitions the universe into disjoint sets, called “equivalence classes” (horses, oxen, squares, circles, … in our example). And conversely, when we have disjoint sets (partitioning the universe), they induce a similarity relation “belonging to the same set”: this animal a is similar to b in being in the same set (e.g. of horses), but is dissimilar from c because c is in another set (say the set of oxen). This gives also a basis for reasonings of the following type: a is in A, a is similar to b, hence b is in A; we deal with this type of reasoning in Sect. 3.4. 2. Leis and sets given by comprehension Another text suggests that the Mohists took very seriously the similarity relation of “being of the same lei” and were not satisfied by simple criteria of dissimilarity; to prove that Horses and Oxen are different leis, it does not suffice to emphasize that they are simply different sets, by exhibiting a property (namely, “having horns”), which Oxen have and Horses have not: it would be as “wild,” as insufficient as using properties which they both have (“having a tail” or “having incisors”). Canon B66. By referring arbitrarily one cannot know differences. Explained: by what they have. Exposition of the Canon. Although oxen are different from horses, it is inadmissible to use oxen having incisors and horses having tails as proof that oxen are not horses; these are things which they both have, not things which one has while the other has none. If you say “Oxen and horses are not of a kind,” and appeal to oxen having horns and horses not, you treat this as the dissimilarity between the kinds, this is referring arbitrarily, as in the case of oxen having incisors and horse having tails. This text naturally rules out the possibility of differentiating Oxen from Horses by properties which they both have. Literally speaking, it does not rule out the possibility of differentiating the two leis by a property which Oxen have and Horses have not or a property which Horses have and Oxen have not, but they rule out the reduction of the comparison procedure (similarity-dissimilarity between kinds) to the existence of that single differentiating property. A possible and often accepted interpretation would be to say that leis carry with them not only the comparison procedure, but also a bunch of “essential” properties which characterize the underlying set: “Oxen are quadrupeds having cloven hooves and horns” could perhaps be such a property identifying Oxen (see 6.1, “The Problem of Distinguishing Kinds,” in Fraser 2013). Fraser (2013) rightly observes that there is a tension between the two conceptions given above: “So we find a tension between Canon A86, which suggests that any group of things similar in some respect can form a kind, and the view implied in B66, that not just anything can form a kind. There is at least one—though nothing in the texts explicitly rules out there being more than one—predetermined, correct scheme of kind distinctions, which we will recognize properly only if we discover and apply the right criteria. This “realist” view is also suggested by canons such as B72, which contends that whether or not something is the kind of thing designated

15  Logical Thought in Mohism and Later Mohism

265

by a term is not determined arbitrarily by our deeming it so, but by whether that is in fact its name (ming).” 3. Leis and typed sets Here is a text which, in my opinion, has not received the attention it deserves. I think that it is important for a good understanding of the notion of lei and of the comparison procedure involved therein. Canon B6. Different kinds are not comparable. Explained by: measuring. Exposition of the Canon. Which is longer, a piece of wood or a night? Which do you have more of, knowledge or grain? Which is the most valuable, aristocratic rank, one’s own parents, right conduct, a price? Which is higher, a deer or a crane? Which is more ∗G’ÅK (sleek), a deer or a ∗G’ÅK (crane)? Which is more ∗ṢIεT (glossy), a ∗ṢIεT (louse) or a ∗ṢIεT (zither)? Technical comment: the symbols following the asterisks represent reconstituted old Chinese pronunciations and exhibit the identical pronunciations of different characters. This text presents paradoxical questions which show that properties or relations having the same name take completely different senses depending on the domain of things to which they are applied: “long” takes “incomparable” meanings for wood and for intervals of time. This suggests that wood and nights belong to very different domains of things which cannot be compared qua length; we should presumably consider two completely different “types of things”: Material-substances and Time-intervals. Similar remarks apply to the other examples given in the Canon and remind us of Ryle’s category mistakes and of Carnap’s infamous example “Caesar is a prime number” given in Carnap (1932). The first example of category mistake given in Ryle (1949) is that of a visitor being shown colleges, libraries, laboratories and similar ingredients of the University of Oxford, and finishing his tour asking “Fine, but where is the university?”; the university is the totality of all these things and of many others and is not reducible to any one of these ingredients; the university and its different ingredients are entities of different categories. Similarly, in Carnap’s example, the oddity is that we apply to a human person a predicate which is exclusively reserved to natural numbers: we should distinguish the type Human-persons and the type Natural-numbers. More or less analogous distinctions are known under various names in the literature: category (in the philosophical and in the mathematical sense), type, kind, sort; we will stick to the word “logical type” in this article to distinguish that notion of type from the lay usage of type. All this suggests that leis should be paired with a certain logical type: if you use the criterion “having the same length,” you will of course be able to group together all pieces of wood of the same length and all intervals of time of the same length; you will also be able to differentiate pieces of wood of 1 meter length and pieces of wood of 2 meters length; you will be able to differentiate intervals of time of 1 second and intervals of time of 1 minute; but you will be unable to differentiate pieces of wood and intervals of time by means of a common measure of length: it

266

T. Lucas

does not make sense. This shows that using the similarity criterion “having the same length as” does not suffice to understand the complexity of this example. We conclude that the two preceding conceptions are not enough to completely master the notion of lei: with a lei, you should mention a logical type. We think that that remark is important to understand the Mohists’ mou reasonings, which will be discussed in Sect. 3.4. 4. So, what is a lei? It seems indeed that the Mohists did feed many things into the notion: standards, similarity relations, essential properties, logical types. Is it necessary to give a unified definition of “the” Mohist notion of lei? We are probably facing a multi-faceted notion which evolved from the Early Mohist notion of fa to the Canons and their explanations and from there to the Daqu and the Xiaoqu. If one insists on giving a unified conception taking into account all these aspects, we would say that a lei requires or determines or is at least intimately connected with the following data: (a) A set S of objects; for example, the lei of Oxen should determine the set of all oxen, the lei of Horses should determine the set of all Horses. (b) A subset stS of S, constituted by the objects which obey the standard; for example, we should be able to give objects which obey the standard for “ox” and the standard for “horse” and they should be subsets of the set of Oxen and of the set of Horses respectively. (c) An equivalence relation yinS such that S is an equivalence class of that relation; for example, Oxen constitute an equivalence class of the relation “having the same shape;” Horses constitute another equivalence class of that relation. (d) A logical type associated with S. The example of wood and time-intervals which is given by the example above suggests that S should be paired with a certain logical type. If you use the criterion “having the same length,” you will have to determine whether you are speaking of length of space extended objects or of length of intervals of time. (e) We should perhaps add that S comes equipped with or determines predicates (or a conjunction of predicates) which give the “essence” of the grouping of objects inside the set S. The extension of the conjunction of predicates should be the set S and from them, we should be able to differentiate a lei S from a lei S′ in a constructive way. But this remains a difficult requirement, since the comparison of different leis may lead us to an ever increasing list of properties. For example, “Oxen are quadrupeds having cloven hooves and horns” suggested above is enough to differentiate Oxen from Birds (by the number of feet), Oxen from Horses (by the presence or absence of horns and cloven hooves) but does not suffice to distinguish them from Sheep; we should perhaps add the nature of fur, etc. It is obvious that these data are not mutually independent, but it is equally clear that the Mohist texts do not give us enough elements to choose the “most basic” one among these. Some of these conditions could be reinforced or on the contrary relaxed. Reinforced: the same set of objects is an equivalence class of many different equivalence relations; instead of one equivalence relation, we could consider all these equivalence relations; is a lei determined by one such equivalence relation or

15  Logical Thought in Mohism and Later Mohism

267

should we include all these equivalence relations in their definition? Relaxed: it has already been mentioned that a similarity relation taking fuzzy cases into account should be reflexive, symmetric, but not necessarily transitive; it is reasonable to drop transitivity, but much of what we have said should be adapted to that more general situation.

3.3  C  ompound Names and Elementary Propositions: Propositions and their Semantics Gongsun Long’s famous Baimalun may be looked at as a problem of understanding the relation between the compound name bai ma 白馬 and the component names bai and ma. Gongsun Long’s provocative answer was conveyed in his Bai ma fei ma 白馬非馬, “White horses are not horses,” to which the Later Mohists echoed Bai ma ma ye 白馬馬也, “White horses are horses.” Similarly, the Jianbai 堅白 problem may be understood as a problem of understanding the relation between the compound name Jianbai and the component names Jian and bai: how do they combine in the concrete stone? Jianbailun’s answer is that they are “separate.” The Later Mohists’ answer was that they “pervade” each other. Here are two texts, whose detailed interpretation is partially uncertain, but which globally shows that the Mohists were concerned with the problem of composition of names. Canon B3. With the same name, of “are two” and “fight” (同名, 「二與鬥」, [....]), “is white” and “is blind” (「白與 (視)眇」), or with others linked to them, of “husband” and “shoes” (「夫與履 lü」), we discard one of the pair, yet inherently the thing is what we called it. Explained by: the criterion. Exposition of the Canon. “They both fight;” “they are not both two” (“Are two” and “fight”). (俱鬥, 不俱二 […]) “Most of a white horse is white,” “Most of a blind horse is not blind” (“is white” and “is blind”). (白馬多白, (視)∗眇馬不多∗眇, 「白與∗眇」也。[...]) For example, someone deemed a fu (husband), when you link yong with it (yong fu ‘brave man’) is not being deemed a husband; (若為「夫」「勇」, 不為 夫。) but something being deemed ju (shoes), when you link mai yi with it (mai yi ju “buy coat and shoes”) is being deemed a pair of shoes (“Husband” and “shoes”). (「屨」, 以「買衣」, 為屨, 「夫與屨」也。) Being two is lost with the one that is lost, does not remain with the one that remains. Whether or not we are dropping “They are two,” only when are these objects is it said of them, without these objects it is not said. It is not like “flower” and “beautiful” (不若敷 fu 與美。). If ‘beautiful’ is said of this, then inherently it is this that is beautiful; […] These examples consider some combinations which are difficult to explain: ju + dou is acceptable; ju + er is not acceptable duo + bai is acceptable; duo + miao is not acceptable yong + fu is not deemed fu

268

T. Lucas

mai yi + ju is deemed ju beautiful+flower is beautiful. Our modern point of view would perhaps too easily see in the last two examples a case of conjunction and read the text as giving conditions for the truth of a conjunction: this object a is a ju and is bought with yi implies that this object a is a ju; this is a beautiful flower implies that this is beautiful. To explain the first two examples, we would probably rely on a study of the number two, and of the quantifiers ju and duo. And we would probably renounce in front of yong + fu, considering this as something which is not analysable from a logical point of view. However, these explanations are probably too far from the Mohist conceptions and we should therefore be facing the question: how did the Mohists justify that “beautiful flowers are beautiful” or for that matter that “a white horse is a horse,” which I call here “elementary propositions;” how did they justify their rejection of other examples such as “a yong fu is a fu”? There is no doubt that they would have appealed to the notion of lei and I see three slightly different styles of answer: 1. They had an implicit notion of conjunction and considered the sentence to be: a thing which is a flower AND is beautiful is obviously beautiful (this is one interpretation of “If ‘beautiful’ is said of this, then inherently it is this that is beautiful.”). Briefly stated, although they did not have an explicit notion of conjunction, they implicitly accepted that ∀x(Fx ∧ Bx → Bx). 2. They compared the leis of beautiful flowers with the lei of beautiful things with the help of characteristic properties: the characteristic properties of the first obviously entail the characteristic properties of the second. This explanation is almost the same as the first one, but without insistence on the notion of conjunction. Briefly said, they accept that ∀x (BFx → Bx). 3. They could have used a special condition relating the faBF, the yinBF and the faB and yinB (this has been proposed by Liu et al. (2011): exists a in faBF, exists b in faB such that for every c, c yinBF a → c yinB b. The second explanation is a bit different from the third one. Here is how. Accepting our definition of satisfaction for atomic formulas, BFc is equivalent to (exists a in faBF such that c yinBF a) and Bc is equivalent to (exists b in faB such that c yinB b). In that manner, ∀x(BFx → Bx) boils down to. For every c (exists a in faBF such that c yinBF a → exists b in faB such that c yinB b), and one sees that the position of the existential quantifiers is not the same.

15  Logical Thought in Mohism and Later Mohism

269

However, the reader can check that for clear-cut predicates associated with equivalence relations, definition (2) is equivalent to definition (3). The differences between (1), (2) and (3) should perhaps not be overstated, because in all the examples of reasoning used by the Mohists, the elementary sentences are considered as obviously true and are given no justification: a white horse is a horse, a robber-man is a man, to read a book is not a book, etc. Another paradigmatic case is that of the combination niu-ma 牛馬 (oxen and horses): Canon B12. Things marked of as a group are one unit. Explained by: both being as one, being this thing specifically. Exposition of the Canon. “Both being as one”: for example, “oxen and horses have four feet.” “Being this thing specifically”: they fit “ox” or “horse.” If you count oxen as an item and horses as an item, oxen and horses are two. If you count oxen and horses as an item, oxen and horses are one. According to the Mohists, we may count oxen and horses as one item, or as two items. In this example, our modern point of view often sees an implicit propositional connective of disjunction: the object a is a niu-ma iff a is a niu OR a is a ma; “Oxen and horses have four feet” is equivalent to “anything which is a horse OR an ox has four feet.” By contrast, the Mohists prefer to explain the set-theoretic union with the help of unit or group, notions undoubtedly akin to their notion of lei: oxen are one lei, horses are one lei and niu-ma are one lei. We could also guess that logically speaking, their notion of pervasion is tantamount to the notion of intersection: the set-intersection of Jian and of Bai is the set of Jianbai things; “A stone is one, hardness and whiteness are two but are in the stone.” (Canon B37); i.e. “Hardness is in the stone” is comprehensive language for the extensional “Hard is a class containing this stone;” “White is in this stone” is comprehensive language for “Bai is a class containing this stone;” “Hardness and Whiteness are in this stone” is comprehensive language for “this stone is in the class-intersection of Hard and White;” Hard and White are two (different classes); Hard intersection White is one (class). This discussion leads us to the notion of ci, proposition, a word which is completely absent in the Mohist Canons and their Explanations, but of which the Mohist become fully conscious in the Daqu and Xiaoqu. What the notion covers is not very far from our notion of proposition, since immediately after the explanation that “one uses names to refer to things,” the Xiaoqu (NO11) explains that “one uses propositions to dredge out ideas.” The notion is described in the Daqu (NO10), with dynamic words describing how it is engendered (sheng), how it grows (zhang), and how it goes (xing): NO10 The proposition is something which is engendered in accordance with the thing as it inherently [故 gu] is, becomes full-grown according to a pattern, and “proceeds” according to the kind (夫辭以故生, 以理長, 以類行者也。). It is irresponsible to set up a proposition without being clear about what is engendered from. Now a man cannot proceed without a road; even if he has

270

T. Lucas

strong thighs and arms, if he is not clear about the road it will not be long before he gets into trouble. The proposition is something which ‘proceeds’ according to the kind; if in setting up a proposition you are not clear about the kind, you are certain to get into trouble. Therefore... Graham (1978) proposes a very precise interpretation of this as a construction in three steps, which is best explained on the paradigmatic examples of the white horses and of the robbers: 白馬馬也, 乘白馬乘馬也 “White horses are horses; riding white horses is riding horses.” 盜人人也, 殺盜人非殺人也 “Robbers are people; to kill a robber is not killing people.” (1) The proposition starts from (sheng) the predicative fragment, which is its inherent reason: ma ye, is a horse; ren ye, is a man (2) It grows (zhang) according to a pattern to become what we called above an elementary proposition: bai ma ma ye; dao ren ren ye (3) It then goes (xing) along transformations. These transformations can make it acceptable: cheng bai ma cheng ma ye. But they could make it inacceptable: sha dao ren sha ren ye. Point (1) is supported by part (C) of NO11 and point (3) by part (D) of NO11, which we have already partially quoted before, but quote now in context: 以名舉實, 以辭抒shu意, 以說出故, 以類取, 以類予。有諸zhu己不非諸人, 無諸己不求 諸人。

One (A) uses names 名 to refer to objects, (B) uses propositions to dredge out ideas, (C) uses explanations 說 shuo to bring out reasons, and (D) accepts according to the kind, proposes according to the kind. What is present in one’s own case is not to be rejected in the other man’s, what is absent from one’s own case is not to be demanded of the other man’s. Our modern point of view has already distinguished the predicates underlying the study of names and which are described in the first step. This section is mainly concerned with the description of elementary propositions as described in the second step. We would now consider the transition from the second step to the third step as an elementary reasoning; this third step will be discussed in the following section. For all we can judge from the material available, the examples of propositions given by the Mohists remain of the elementary type X is Y (or X is not Y) described above and they only theorized complex propositions as can be obtained from those elementary propositions essentially by the addition of a simple syntagm Z: ZX is/is not ZY.

15  Logical Thought in Mohism and Later Mohism

271

This answers some questions about the syntax of propositions, but we have to turn to the explanation of argumentation to guess at what would be the semantics of general propositions. The relevant text is Canon A74: Canon. Bian (disputation) is contending over claims which are the converse of each other. Winning in disputation is fitting the facts (當 dang). Exposition of the Canon. One calling it “ox” and the other “non-ox” is “contending over claims which are the converse of each other.” Such being the case they do not both fit the fact; and if they do not both fit, necessarily one of them does not fit. (Not like fitting “dog.”) This text makes it clear that we have a basically realist semantics: propositions correspond (dang) to the facts or not. This is illustrated by the stock example of the predicate “to be an ox,” to which the opposite predicate “to be a non-ox” is opposed. Exactly one of these predicates should fit the object considered. If we translate this into the language of atomic propositions, the correspondence to the facts gives us exactly what we expect of a classical conception of truth of propositions: the atomic proposition “this is an ox” and its negation “this is not an ox,” equivalent to “this is a non-ox,” cannot be simultaneously maintained (principle of non-contradiction); and of the claims “this is an ox” and “this is not an ox,” one of them must correspond to the facts (principle of excluded middle). The generality of the formulation leaves no doubt that the Mohists subscribe to the two basic principles of non-contradiction and of excluded middle not only for atomic propositions, but also for elementary propositions and arbitrary propositions.

3.4  Reasonings The examples of Mohist reasonings which have been transmitted to us are based on atomic propositions or on elementary propositions and use in both cases the notion of lei. The 4 principal types, “illustrating,” “parallelizing,” “adducing” and “inferring” are consigned in the second half of NO11: (A) “Illustrating” pi 辟 (used for 譬) is referring to other things in order to clarify one’s case. (B) “Parallelising” mou 侔 is comparing propositions and letting all “proceed.” (C) “Adducing” yuan 援 is saying; “If it is so in your case, why may it not be so in mine too?” (D) “Inferring” tui 推 is using what is the same in that which he refuses to accept in order to propose the former. Fraser (2013) gives somewhat different but illuminating translations for these four types: analogy, parallelizing, pulling and pushing. To these four types, one often adds “exemplification” 效 xiao described in NO5:

272

T. Lucas

“Some” is not all [“stone” when broken up]. The loan-named [“Qin horse”] is not now so. An example [“pillar”] is a standard for being deemed such-and-such [“wood”]; the thing exemplified is the standard by which the example is deemed such-and-such. Therefore if something coincides with an example, it is this thing, and if it does not it is not; this is exemplifying. We begin by discussing “exemplification,” give the generally accepted structure of “illustration” and “adduction,” present “inferring” and discuss with more details the technique of “parallelizing.” 1. Exemplification The text of NO5 looks a bit cryptic but seems to be clearly based on the notion of satisfaction of a predicate. If I want to know whether this a is an X, I can pick up a standard object b (the example) obeying the faX and compare the given a to b: if a is similar [here expressed by zhong] to the example b, then a is X (中效, 則是也); we use here the property: if a is similar to an object b obeying the norm faX, then a is in X. Conversely, if a is not similar to the example b, then a is not X (不中效, 則 非也); we use here the property: if a is not similar to the standard object b, then a is certainly not similar to any standard object, hence a is certainly not X. This is valid for clear-cut predicates. The case of fuzzy predicates is more complicated but lends also itself to an analogous discussion. 2. Illustrating, adducing and inferring We group together illustrating, adducing and inferring, because they all seem to be based on analogical argumentation, whose basic structure is of the form: the object a is similar to the object b; b has property X; hence a has property X. So expressed, it is not a logically valid reasoning, e.g.: this white horse is similar to that black horse (in being a horse), this white horse is white; hence this black horse is white. However, its convincing character is often compelling, e.g.: this white horse is similar to that black horse (in being a horse); this black horse has four legs; hence that black horse has four legs. A detailed discussion of analogical reasonings is complicated and largely beyond the scope of this article, but we think that the following observations are relevant: (1) the analogical arguments are extremely abundant in pre-Qin texts and especially in the Early Mohist texts; many Chinese authors consider them as a fundamental characteristic of the so-called “Chinese logic;” from a strictly logical point of view, it has not yet been satisfactorily dealt with, but we have to explain them, if we want to understand the pre-Qin texts; (2) for Mohism, the core of analogical argumentation is clearly based on a logically valid variation of the preceding form: b is in the extension of predicate H;

15  Logical Thought in Mohism and Later Mohism

273

a yinH b (the object a is similar to the object b according to the similarity relation yinH associated with H); hence a is in the extension of predicate H; (3) the differences between illustrating, adducing and inferring do not originate in their basic logical structure, but in their use in argumentation: illustrating is used to explain my position, adducing is used to convince the other of my position and inferring is used to refute the other’s position. 3. Parallelizing This is a typically Mohist type of reasoning. It was undoubtedly inspired by the ever present phenomenon of general parallelism which is common to all writings of the period, and may be seen as an extension of a schema of opposite parallelism of the type (X is Y)/(not X is not Y). The mou procedure in fact is a generalization of that scheme and consists in the study of the transition from sentence (X is/is not Y) to the sentence (ZX is/is not ZY); in this scheme Z generally represents a word which is prefixed to X or to Y. The Mohists classified these transitions into five categories: NO13 夫物或乃是而然, 或是而不然, , 或一∗ 周而一不∗ 周, 或一是而一[不是] [……] 非也。

Of the thing in general, there are cases where (1) something is so if the instanced is this thing, or (2) is not so though the instanced is this thing, or (3) is so though the instanced is not this thing, or (4) applies without exception in one case but not in the other, or (5) the instanced in one case is this and in the other is not. Here are some comments on this classification: (1) The first category deals with the transition from X is Y to ZX is ZY; shi er ran; “the thing is this and remains so;” for example, “White horses are horses; riding white horses is riding horses.” (2) The second category deals with the transition from X is Y to ZX is not ZY; shi er bu ran; “the thing is this and does not remains so;” as an example, they gave the shocking “Robbers are people; to kill a robber is not killing people.” (3) The third category deals with the transition from X is not Y to ZX is ZY; bu shi er ran; the thing is not this but becomes so; as an example, they propose “Reading a book is not a book, but to like reading books is to like books.” Logically minded readers would expect a fourth category, dealing with the transition from X is not Y to ZX is not ZY; that category is missing but the classification is completed by two apparently heterogeneous categories: (4) The fourth category deals in fact with problems of generalization, more exactly with the interpretation of sentences with no explicitly marked quantifiers; an

274

T. Lucas

example is given as this: “ ‘He loves people’ requires him to love all people without exception, only then is he deemed to love people. ‘He does not love people’ does not require that he loves no people at all; he does not love all without exception, and by this criterion is deemed not to love people.” It may be speculated that what underlies these examples is also a transition similar to the preceding one, a transition of the type X is Y to X is ZY (“he loves people” to “he loves all people”) or the non-transition from X is not Y to X is not ZY (“he does not love people” to “he does not love people at all”). (See Lucas 2011 for more details). (5) The fifth category deals with example like “The fruit of the peach is a peach; (however), the fruit of the bramble is not a bramble.” We concentrate on the first two categories, which are the ones which have drawn much attention. First, let us make clear that only the first category seems to be presented as a deduction, and nobody, except Gongsun Long, will object to the transition from “white horses are horses” to “riding white horses is riding horses.” The second category clearly presents transitions from X is Y to ZX is ZY which the Mohists did not admit: from “robbers are people,” we cannot deduce “killing robbers is killing people” (a conclusion which the Mohists wanted to avoid, because they promoted “universal love” – or “impartial care” in modern translations – but at the same time wanted to punish offenders). A considerable part of the literature has dealt with these two transitions. For the first one, it has been observed that its modern formalization nicely explains the validity of the reasoning: from ∀x (WHx → Hx), we may indeed deduce ∀x (WHx ∧ Rx → Hx ∧ Rx) (in this formalization WHx, Rx and Hx stand for “x is a white horse,” “x is ridden” and “x is a horse” respectively). However, the same formalization does NOT explain the non-deducibility of the second example and on the contrary justifies the transition from “robbers are people” to “killing robbers is killing people”: from ∀x (Rx → Mx), one deduces ∀x (Rx ∧ Kx → Mx ∧ Kx) (Kx stands for “x is killed”). Different solutions have been proposed to justify the second example. I describe here three approaches. (1) The first approach considers that “to kill” is somehow taken here as ambiguous; we should distinguish between K1x, “x is killed as a robber” (as a result of a lawful punishment) and K2x, “x is killed as a man” (as a result of homicide). Following this line of explanation, it is easy to see that from ∀x (Rx → Mx) one cannot deduce ∀x (Rx ∧ K1x → Mx ∧ K2x), in symbols:

x  Rx  Mx   x  Rx  K1 x  Mx  K 2 x  .



This solution is basically correct, does not require much addition to the sentence and directly translates the ambiguity. The formalization corresponds to a paraphrased sentence of the type: “To kill a robber by execution is not to kill a robber by murder.” Another argument in favor of that solution is its proximity to some ­analyses of philosophers of the period. For example, Xunzi had a clear conscience of the ambiguity of terms, distinguishing two senses of rong 榮 (honor), which he paraphrased as yirong 義榮 (honor deriving from morality) and shirong 埶榮 (honor

15  Logical Thought in Mohism and Later Mohism

275

deriving from the force of circumstances), and similarly for ru 辱 (disgrace), which he paraphrased as yiru 義辱 and shiru 埶辱.4 An objection is that it does not fit in very well into the scheme of parallelism, which uses a “shocking” identity of words: to be killed as a result of homicide and to be killed as a result of a lawful punishment are expressed by the same words “to be killed.” That is the reason why people have proposed other solutions. (2) A second approach, proposed by Liu and Zhang (2010), involves considerations of intensionality, blocking the monotonic transition from X is Y to ZX is ZY. In the case of this example, one would say that Kx, “x is killed” is an intensional predicate. In a modern semantic approach, this means that the interpretation of K requires the consideration of different possible worlds; to say that “x is killed” varies in meaning may be explained by the consideration of two worlds, say W1 = the moral sphere and W2 = the social sphere; in W1, the extension K1 of K is constituted by all things which are killed as a result of a lawful punishment and in W2, the extension K2 of K is constituted by all things which are killed as a result of homicide. That K varies in meaning is explained as “K is not necessary,” while we can accept that M does not vary in meaning, i.e. “M is necessary.” Among different possible formalizations, we choose the following one, which will make easier the comparison with the third approach:

x  Rx £Mx   x  Rx  Kx £Mx £Kx    .



This is indeed shown by the Kripke model suggested above: take (a) W1 = W2 = a non empty set having at least two elements (say the set of all men); ( b) the accessibility relation is given by: W2 is accessible from W1 (c) interpret R in W1 and W2 by R1 = R2 = a non empty subset of W1 (=W2) (say the set of robbers); (d) interpret M in W1 and W2 by M1 = M2 = W1 = W2; (e) on the other hand, interpret K in W1 in such a way that there is an element a in R1 and K1 and interpret K in W2 in such a way that a is in R2 but not in K2; a clearly represents a robber which “is killed” in world W1 (as a result of lawful punishment) but is not killed in world W2 (as a result of homicide). This model will render the premise of (∗) true and its conclusion false when evaluated at world W1. An element in favor of this approach is that the Mohists were very conscious of intensional notions and would probably have sympathized with such an approach. Note however that the formalization with the necessity operator is not the most adequate, since the proof of non-deducibility requires a bit more considerations than seems to be required, at least for usual modal logics such as S4. It may also be

4  Translations given according to Knoblock (1994: 46). I thank the editor of this book for calling my attention on this text. For an elaboration of this approach, see his Fung (2012).

276

T. Lucas

objected that the formalization is far from the original text and has to supply necessity operators not only for K, but also for M. (3) A third approach has been proposed in Lucas (2011) and is justified by the fact that leis involve the consideration of sorts or kinds . In the example here, one distinguishes two sorts: s and t; say s for a moral sphere containing robbers and other delinquents, and t for the social sphere containing people considered independently of their moral behavior; there is also a function f which to each element of the sort s associates its underlying “social person”: if xs is a robber, fxs is the underlying social person “living” in sort t. Accordingly, to kill will be analysed as a predicate Ks in sort s and a predicate Kt in sort t, with no presupposition that Ksxs automatically implies Ktfxs. With these notions at hand, the reasoning is very neatly analyzed as asserting that (Lucas 2011)

xs  Rs xs  M t fxs   xs  Rs xs  K s xs  M t fxs  K t fxs  .



To give a semantic proof of that non-deducibility in the logic of sorts, one has to give worlds Ws and Wt as well as an interpretation of f as a mapping from Ws to Wt, interpretations of Rs, Ms and Ks as subsets of Ws, and interpretations Rt, Mt and Kt as subsets of Wt. Functionally speaking, this approach may be seen as a syntactically internalized version of the second approach and the technical proof of non-deducibility may be given by exactly the same counter-example, Ws corresponding to W1, Wt corresponding to W2, Rs to R1, Rt to R2, etc.; technically speaking, we also have to define the interpretation of f: simply take f as the identity on Ws, which is equal to Wt. Elements in favor of this approach is that the Mohists seem to have relied on some notion of logical type and moreover, the proof of non-deducibility is “minimal” and does not involve any choice of logic as in the case of modal logic. Another element which the present author thinks important is that this gives a unified approach not only of the examples given here, but of a good deal of the examples of the five categories (see Lucas 2012b). Here is a picture illustrating the proof of non-­deducibility:

15  Logical Thought in Mohism and Later Mohism

277

Of course this approach is also subject to the objection that the formalization is far from the original text and has to supply indices not only for K but also for M. A thorough discussion of this kind of objection is beyond the scope of this article, but should certainly mention that the reconstruction of an argument does not necessarily fit the original text and may be quite involved, as is exemplified by modern analyses of paradoxes, such as the “liar paradox.”

3.5  Argumentation We have already dealt with argumentation here above, but it is fair to repeat that this was an essential notion for the Mohist School and the other schools of the time. What is notable in the Mohist School is that it passed through different stages. The first stage is that of the Early Mohist texts, which contain many types of argumentation which are used in a very competent way, but which remain untheorized. The second stage is that given in the Mohist Canons; disputation is briefly but neatly defined in a dynamic way as the interaction of two agents, one claiming “this is X,” the other one claiming “this is not X;” winning in disputation is given by correspondence to the facts and there is necessarily a winner; if there is no winner, this means that there is no disputation [Canons A74 and B35]. The third stage is given in the Xiaoqu where disputation finds a richer content: it is not only a question of shi/fei 是非 (is this/is not this) but also a question of tong/yi 同異 (distinguishing the same and the different) and a question of li/hai 利害 (settling the beneficial and the harmful). This third role of disputation sends us back to the moral preoccupations which are everywhere present in the texts of the Early Mohists and reminds us that, even if logic was developed with great care by the Mohists, it remained at the service of an

278

T. Lucas

ethics dealing with practical matters and based on the identification of good with beneficial and bad with harmful. A common thread of that evolution remains the idea of a disputation as a dynamic process, which has inspired (Liu et al. 2011) in giving a model based on the theory of games.

3.6  O  ther Logical Notions: A Priori, Necessary and Sufficient Conditions, Etc. The Mohists discussed or evoked many other notions which we now clearly associate with logic. The a priori is present under the word xian 先 in their writings and is illustrated by the image of something which is behind a wall and which I cannot see, but of which I already know something if somebody tells me: “what is behind the wall is similar to this, which you can see.” See Canon A93 and its explanation: “By the things which follow from each other or exclude each other, we may know ‘a priori’ what it is.” See also Canon B38 and its explanation, Canon B57 and its explanation. This is far from being a fully evolved notion of a priori, but the example implicitly contains a logically valid reasoning: –– hypothesis H1: the present object a is similar in color to the object b which is behind the wall (this is a hypothesis given to me) –– hypothesis H2: a is blue (this is a hypothesis confirmed by my observation) –– conclusion C: hence b is blue. The a priori resides in that I do not have to use my experience to look at b to conclude that b is blue: C follows logically from H1 and H2. How H1 and H2 are established does not matter for the validity of the reasoning. Necessary conditions and necessary and sufficient conditions appear explicitly at the very beginning of the Canons. Canon A1 reads: Canon: The gu 故 (reason/cause) of something is what it must get before it will come about. Exposition of Canon: “Minor reason”: having this, it will not necessarily be so; lacking this, necessarily it will not be so. It is the unit . (Like having a starting point.) “Major reason”: having this, it will necessarily ; lacking be so. (Like the appearing bringing about the seeing.) Comment. In (p → q), q is a minor reason for p, i.e. q is a necessary condition of p: having q, one does not necessarily have p, but not having q, one necessarily does

15  Logical Thought in Mohism and Later Mohism

279

not have p. In (p ↔ q), q is a major reason for p (and p is a major reason for q): having q, one necessarily has p and having not-q, one necessarily has not-p. See also the explanation of Canon A83: If without something else something is necessarily absent, the relation is “necessary” [necessary condition]. The judgments of sages, employ but do not treat as necessary. The “necessary” accept and do not doubt. The ones which are the converse of each other, apply on both sides, not on one without the other [necessary and sufficient condition]. On the other hand, they do not explicitly discuss the notion of sufficient condition: in (p → q), p is a sufficient condition of q. However, this does not mean that they lacked means of dealing with implication and logical consequence: ze 則 and erhou 而後, are frequently used in the Canons and in the Daqu and the Xiaoqu. Propositional connectives are present but not clearly thematized as such. For example, negation appears under the words bu 不, fei 非, wei 未, fou 否, wu 無 and others but is always used in context. Disjunction is often expressed by simple juxtaposition of incompatible constituents: “If whether it is limited or limitless is unknowable a priori, then whether it is exhaustible or not, whether men fill it or not, and whether men are exhaustible or not, are likewise unknowable a priori” (有窮無 窮未可智, 則可盡不可盡未可智, 人之盈之否未可智, 人之可盡不可盡亦未可 智。) (Canon B73). “Whether putrid meat will benefit or harm is unknowable in advance” (騷之利害未知也。) (Canon A75). “Whether there was benefit or harm ‘beyond the wall’ was not knowable in advance” (廧外之利害未可知 也) (Canon A75). Note that huo 或 has the meaning of “some,” “someone” and that the construction huo… huo… is better explained as “some …, some…” or “one…, the other…” “One calling it an ‘ox’ and the other ‘non-ox’ is ‘contending over claims which are the converse of each other’” (或謂之「牛」, 或謂之「非牛」, 是爭 (彼)∗仮也。) (Canon A74). Conjunction is almost exclusively expressed by simple juxtaposition. We have argued in Lucas (2018) that the grammatical characteristics of Mozi’s Chinese did not favor the recognition of connectives as a transformation of a proposition (or two propositions or more) into a compound proposition: negation transforms p into the compound proposition not-p, conjunction transforms the two propositions p and q into the compound proposition p-and­q, etc. It would be fair to say that, presumably also for grammatical reasons, the Western world had to wait for the Stoics and more specifically Boole’s and Frege’s modern contributions to recognize that there was a genuine logic of propositions and an interesting logic of the propositional connectives. Modalities are present under different guises. The word bi 必 is extremely frequent in the Canons and expresses necessity; it is defined in terms of time as “the unending” (bu yi 不已) in Canon A51, but it is always used in the strong sense which is given in modal logic. Possibility is expressed by the word ke 可 and appears often negated in the frequent bu ke 不可. There is an inchoative logic of time: they take pains to define qie 且 (will/about to) in Canon A33 and give the relation with yi 已 (already) in the explanation of the Canon:

280

T. Lucas

Canon. Qie (“will/about to”) is declaring something to be so before the event (?). Exposition of Canon. Before the event one says “about to,” after the event one says “already.” The just now so too is about to be. In their considerations on knowledge zhi (4th tone) 智, there is a canon which shows that they had clearly thematized the difference between “not to know” and “to know that one does not know”: Canon B48. Knowing what he does not know. Explained by: picking out by means of the name. Explanation of the Canon. If you mix together what he does know and what he does not know, and ask about them, he is obliged to say “This I do know, this I do not know.” To be capable both of picking out the one and disclaiming the other is to know them both. Quantifiers are present everywhere (see Graham 1978: 127–136 for a list) and the explicit recognition of their interdefinability is perhaps one of the greatest logical achievements of the later Mohists. Canon A43 defines jin 盡 “all” in terms of mo 莫 “none” as “all is none not,” in formulas, “all is not exists not”: Canon. Jin (exhausting/applying to all/all) is none not being so. (盡, 莫不然也。) On the other hand, the Xiaoqu suggests the fact that they understood huo “some” in a strict sense, as excluding “all”: NO5 “‘Some’ is not all.” (「或」也者, 不盡也。) Considerations on time, space and infinity are also very present in the Canons. Preoccupation with paradoxes or puzzles are now considered as a brandmark of true logicians and were explicitly considered in the Mohist Canons. Here are four nice examples. 1. Canon B60 gives a variant of Zeno’s paradox: when you hoe a half of a field, then the half of the remaining half, and so on, you will never exhaust the field. 2. The phenomenon of self-reference, sentences applied to themselves and leading to contradiction may be found in Canons B71, B77, B79, B82. Here is the example of B71: Canon: To claim that all saying contradicts itself is self-contradictory. Explained by: what he says himself. In other words, if the saying “all sayings are self-contradictory” is true, then it applies to itself and is itself self-contradictory. This is not very far from the paradox of the liar. 3. Our third example is the provocative assertions of Canon B59 and its explanation: Canon: One is less than two but more than five. Explained by: establishing the level. Exposition of the Canon. Five has one in it; one has five in it, twelve in it.

15  Logical Thought in Mohism and Later Mohism

281

However, the text is here unfortunately too short to give clear indications on their solution. 4. You can find the problem of infinite regress in Canon B36: Canon: “Defer to the other man in everything” is inadmissible. Explained by: the start. Exposition of the Canon. The time before the man who defers asks the other man to drink is “the start.” It would be impossible for the other to drink first. If A maintains “you drink first” and B maintains the same, A and B will never start to drink!

3.7  Implicit Logic One usually expects of logicians that they write “clearly and distinctly.” This is not a necessity: I could write good logic without being very clear. Neither is it sufficient: I could write very clearly, but with logically uninteresting considerations. However the two often go on a par. By “implicit logic,” I mean here essentially the way the text is organized and written and the methodology used, and which were not the object of an explicit discussion. Among many features, I want to draw the attention to three remarkable examples of the logical quality of the Later Mohists’ writings. The first one is the fact that they had a deep understanding of what a system of definitions is: Canons A1-A75 are clearly organized as a set of definitions, obeying many of the strictest criteria of good definitions; they are systematically presented in a canonical form, without fear of using the same pattern (X-Y-ye/definiendum-­ definiens-­ye); there is no circularity; terms are defined in terms of a current word of pre-Han Chinese or other technical terms which are themselves defined and belong to a chain of definitions; although not particularly economical, the set of undefined basic terms remains limited (see Graham 1978: 262 and Lucas 2012a). The second feature is that the language which is used in the Canons, in the Daqu and in the Xiaoqu is highly normalized, with specialized words or uses for propositional connectives and quantification. They employ again and again vernacular words such as niu 牛, ma 馬, shi 石, Zang 臧, Huo 獲 as stock examples but with the clear intention of conveying an idea of generality: by the systematicity of their repetitive use, niu, ma, shi 石 are genuine variables for predicates (with a distinction between the first two and the third one, which is considered for its parts); Zang and Huo are used as individual constants. The expression yezhe 也者 seems to be ­systematically used as a quotation-unquotation device, marking technical terms introduced for definition.5 5  See Graham (1978:140): “The particle combination 也者 is equivalent to ‘unquote’, implying that what is under discussion is the meaning of the preceding word or phrase. In Names and

282

T. Lucas

The third feature is found in their classification of mou reasonings (NO13-NO18), where we find a rich collection of examples and a systematic comparison procedure between different forms of reasoning; this is not yet formal logic, but it is certainly a natural and useful step to arrive at a satisfactory theory of reasonings. Briefly stated: despite the corruption of some texts, the Mohist writings remain paradigms of clear and systematic logical research.

4  Conclusion Among the different schools of philosophers of the pre-Qin period, the Mohists, along with the Logicians, were the best promotors of logical research. As other schools of the time, they were interested in ethics and in socio-political problems, but their reflection evolved into a more and more precise attention to language. In Early Mohism, promotion of their theses was served by an elaborate argumentation, which remains however an instrument and is not yet the object of their reflections. In Later Mohism, from the Mohist Canons to the Daqu and the Xiaoqu, one may observe a growing attention to language and to what we now call logic, epistemology and science. We have tried to show in contemporary terms that they had a complex, partial but very coherent view of propositions and of their semantics and that they had discovered a wealth of other logical notions. As far as we can judge by the remaining texts and their traditional attributions, their thought was more elaborate and more reflexive than the thought of the Logicians. It is unfortunate that they were forgotten or despised by later thinkers, but their contribution remains a landmark of scientific and philosophical discoveries. Their recent revival testifies now to a fairer appreciation of their achievements.

References6 Carnap, Rudolf. 1932. “Überwindung der Metaphysik durch logische Analyse der Sprache.” Erkenntnis 2: 219–41. Trans. by Pap A. 1959 as “The Elimination of Metaphysics Through the Logical Analysis of Language.” In A.J. Ayer, ed., Logical Positivism. Glencoe: The Free Press.

Objects, it marks technical terms introduced for definition.” For example, NO11 defines pi 「辟」 by the sentence “「辟」也者, 舉也物而以明之” which Graham translates by “‘Illustrating’ is referring to other things in order to clarify one’s case”. Notice the use of 「…」in Chinese and the use of simple quotation marks ‘…’ in English. It is well known that quotation marks are a recent invention and an expression like yezhe was presumably welcome to play a similar role in the original text. 6  There is an abundant list of books and articles written in English or in Chinese on Mozi’s thought and more specially, on the Mohist Canons. The interested reader will find already selected bibliographies in Fraser (2013, 2014) (works in English) and in [Sturgeon] (works in Chinese and in English). We mention here only the works cited in the article.

15  Logical Thought in Mohism and Later Mohism

283

Defoort, Carine, and Nicolas Standaert, eds. 2013. The Mozi as an Evolving Text: Different Voices in Early Chinese Thought. Leiden/Boston: Brill. Fung, Yiu-Ming. 2012. “A Logical Perspective on Parallelism in Later Moism”.Journal of Chinese Philosophy 39.3: 333–350. Fraser, Chris. 2013. “Mohist Canons.” In Edward N.  Zalta, ed., The Stanford Encyclopedia of Philosophy (Fall 2013 edition). URL: http://plato.stanford.edu/archives/fall2013/entries/ mohist-canons/. Fraser, Chris. 2014. “Mohism.” In Edward N. Zalta, ed., The Stanford Encyclopedia of Philosophy (Summer 2014 edition). URL: http://plato.stanford.edu/archives/sum2014/entries/mohism/. Graham A.C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong/London: Chinese University Press/School of Oriental and African Studies. Knoblock, John. 1994. Xunzi: A Translation and Study of the Complete Works. Volume III, Books 17–32. Stanford: Stanford University Press. Liu, Fenrong, and Jialong Zhang. 2010. “New Perspectives on Moist Logic”. Journal of Chinese Philosophy 27.4: 605–621. Liu, Fenrong, Jeremy Seligman, and Johan Van Benthem. 2011. “Models of Reasoning in Ancient China”. Studies in Logic 4.3: 57–81. Lucas, Thierry. 2011. “Basic Concepts of Mohist Logic.” Studies in Logic 4.3: 82–108. Lucas, Thierry. 2012a. “Definitions in the Upper Part of the Moist Canons”. Journal of Chinese Philosophy 39.3: 386–403. Lucas, Thierry. 2012b. “Why White Horses Are Not Horses and Other Chinese Puzzles...” Logique et Analyse 218: 185–203. Lucas, Thierry. 2013. “Parallelism in the Early Moist Texts.” Frontiers of Philosophy in China 8.2: 289–308. Lucas, Thierry. 2018. “Limits of Logic in Moism.” Journal of Chinese Philosophy 45.3–4: 233–251. McNamara, Paul. 2014. “Deontic Logic.” In Edward N.  Zalta, ed., The Stanford Encyclopedia of Philosophy (Winter 2014 edition). URL: http://plato.stanford.edu/archives/win2014/entries/ logic-deontic/. Mei Y. P. 1929. The Ethical and Political Works of Mo-tse. London: Arthur Probsthain; translation quoted according to [Sturgeon]. Ryle, Gilbert. 1949. The Concept of Mind. London and New  York: Hutchinson’s University Library. [Sturgeon] http://ctext.org/mohism/. Thierry Lucas  is Emeritus Professor of the Université Catholique de Louvain (Louvain-laNeuve, Belgium). His research interests have always been in the domain of logic and his publications range from algebraic logic to connections of logic with the mathematical theory of categories, modal logic, deontic logic and applications of logic to the understanding of some Chinese philosophical texts dealing with logical problems. He has taught logic at all levels to students of different Faculties and has been a Dean of the Faculté des Sciences Philosophiques of his University.

Chapter 16

Paradoxes in the School of Names Chris Fraser

1  Introduction Paradoxes are statements that run contrary to common sense yet seem to be supported by reasons and in some cases may turn out to be true. Paradoxes may be, or may entail, explicit contradictions, or they may simply be perplexing statements that run beyond or against what seems obviously correct. They may be proposed for various reasons, such as to overturn purportedly mistaken views, to illustrate problematic logical or conceptual relations, to reveal aspects of reality not reflected by received opinion, or simply to entertain. In the Western philosophical tradition, the earliest recognized paradoxes are attributed to Zeno of Elea (ca. 490–430  B.C.E.) and to Eubulides of Miletus (fl. fourth century B.C.E.). In the Chinese tradition, the earliest and most well-known paradoxes are ascribed to figures associated with the “School of Names” (ming jia 名家), a diverse group of Warring States (479–221 B.C.E.) thinkers who shared an interest in language, logic, and metaphysics. Their investigations led some of these thinkers to propound puzzling, paradoxical statements such as “Today go to Yue but arrive yesterday,” “White horses are not horses,” and “Mountains and gorges are level.” Such paradoxes seem to have been intended to highlight fundamental features of reality or subtleties in semantic relations between words and things. Why were thinkers who advanced paradoxes categorized as a school of “names”? In ancient China, philosophical inquiry concerning language and logic focused on the use of “names” (ming 名, also terms, labels, or reputation) and their semantic relations to “stuff” (shi 實, also objects, features, events, or situations). Hence for classificatory purposes, second-century B.C.E.  Han-dynasty archivists grouped together assorted pre-Han figures whose most prominent ideas seemed to concern C. Fraser (*) Department of Philosophy, University of Hong Kong, Pok Fu Lam, Hong Kong e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_16

285

286

C. Fraser

the relation between names and stuff—or at least strange, unorthodox uses of names—and dubbed them a school or lineage (jia 家) devoted to the study of names. Unfortunately, both the label and the grouping are misleading. Historically, the school was a retrospective, taxonomical fiction. The figures classified under the “School of Names” never formed a distinct circle, movement, tradition, or line of influence devoted to any particular doctrine, theme, method, or way of life. Their intellectual interests overlapped at most only partly, while also overlapping extensively with those of texts associated with other schools or traditions, such as the Mohist “Dialectics” 墨辯, the Zhuangzi 莊子, the Xunzi 荀子, and the Annals of Lü Buwei 呂氏春秋. What perhaps does set the School of Names apart is that some (though not all) of the figures associated with it apparently delighted in propounding paradoxical or preposterous sayings, while the other texts just mentioned generally (though not exclusively) seek to explain and debunk such utterances. Early texts such as the Han History 漢書 associate some eight figures with the School of Names, but records remain of the doctrines of only four of these: Deng Xi 鄧析, Yin Wen 尹文, Hui Shi 惠施, and Gongsun Long 公孫龍. With the exception of a few brief texts attributed to Gongsun Long, there is little or no first-hand evidence of these men’s thought, since none of the writings attributed to them by Han-­ dynasty bibliographers survive.1 Everything we know about them comes from quotations or anecdotes in other texts. These second-hand accounts typically date from long after the lifetime of the figures they describe and may be embellished, biased, or even fictional. Deng Xi (d. 501 B.C.E.) was China’s earliest renowned lawyer and rhetorician and may have served as an official in the state of Zheng. Information about him is limited and entirely second-hand, comprising a one-line entry in the Zuo Commentary 左傳, curt attacks on him in the Xunzi, several anecdotes in the Annals of Lü Buwei, and a few stories in texts of later date. According to the “Bibliographical Record” of the Han History 漢書藝文志, he was the author of two scrolls of writings, neither of which has come down to us. The grounds for associating him with paradoxical statements are several passages in the Xunzi, which may or may not accurately reflect the views of the historical Deng Xi. Little is known of Yin Wen (fl. late fourth century B.C.E.), and it is not clear why the Han History associates him with the School of Names, since no source credits him with views on language or logic. What little information we have about him comes mainly from two sources composed long after his death, the Zhuangzi “Under Heaven” essay (Zhuangzi book 33) and the Annals of Lü Buwei. “Under Heaven” pairs him with a colleague named Song Xing 宋鈃, with whom he shared several

1  Han History 30, “Bibliographical Record,” records two scrolls of writings attributed to Deng Xi, one to Yin Wen, one to Hui Shi, and fourteen to Gongsun Long. None of the writings attributed to Deng Xi, Yin Wen, or Hui Shi survives. Two short texts called Dengxizi and Yinwenzi exist, but the majority scholarly opinion is that these are much later forgeries, not genuine Warring States texts. The extant version of the Gongsun Longzi contains only six scrolls, not fourteen, one of which is a collection of anecdotes cobbled together from other early sources. For further discussion, see (Harbsmeier 1998).

16  Paradoxes in the School of Names

287

salient ethical and psychological doctrines. Xunzi includes two of Song Xing’s doctrines among the paradoxes and sophistries he rebuts. This connection may help explain why Han-dynasty scholars classified Yin Wen as belonging to the School of Names. A variety of early sources mention Hui Shi (fl. 313 B.C.E.), some depicting him as a statesman and political advisor to King Hui of Wei (370–319 B.C.), some as a friend and intellectual sparring partner of the Daoist Zhuangzi, and some as a clever but confused thinker who propounded sophistries and paradoxes. Only two, the Xunzi and “Under Heaven,” give much information about his philosophical views. Both are hostile to him, however, and both merely attribute a series of theses to him without presenting his arguments. Gongsun Long (c. 320–250 B.C.E.) was a retainer to the Lord of Pingyuan (d. 252 B.C.E.) in the state of Zhao. Anecdotes about him are found in the Zhuangzi and the Annals of Lü Buwei, and “Under Heaven” mentions him as a prominent intellectual figure. The Xunzi does not criticize him by name but may cite a version of his white horse sophism in a list of improper uses of names (Xunzi 22/32).2 Similarly, without mentioning him by name, the Zhuangzi “Discourse on Evening Things Out” alludes to his paradoxical claims that white horses are not horses and that pointing is not pointing (Zhuangzi 2/31–32).3 Because at least two pre-Han writings attributable to Gongsun Long survive,4 his are the only paradoxes associated with the School of Names for which we have first-hand presentations of the arguments. In their own day, thinkers such as Hui Shi and Gongsun Long were not associated with the School of Names—since historically no such school existed—but were considered members of a diverse class of intellectuals known as the bian zhe 辯者, or “dialecticians.” The dialecticians engaged in “distinction-drawing” (bian 辯), a form of dialectical, sometimes competitive inquiry, debate, and persuasion aimed at distinguishing the proper semantic relations between names (ming) and stuff (shi), typically on the basis of similarity relations between the things that constitute a kind. For example, individual horses were considered to count as horses by virtue of their similarity to paradigmatic exemplars of horses and in aggregate were regarded as constituting the kind horse. Since the label “dialecticians” refers primarily to participation in an activity, not a doctrinal stance, its scope is broad and its boundaries somewhat vague. Besides figures associated with the School of Names, in some contexts the term could have applied to the later Mohist dialecticians, some of the contributors to the Zhuangzi, and even Xunzi, who engaged in court debates. The dialecticians were typically employed as guest scholars, teachers, or political advisors, counseling rulers throughout the many cities and states that made up the  Citations to the Xunzi give chapter and line numbers in (Xunzi 1966).  Citations to the Zhuangzi give chapter and line numbers in (Zhuangzi 1956). 4  Of the five discourses in the extant Gongsun Longzi, the authenticity of two and a half is disputed. Graham argues that the second half of the third discourse and both of the last two are post-Han forgeries, pieced together partly from misunderstood bits of the Mohist Canons (Graham 1990: 125–215). For an opposing view of the text’s status, see (Fung 2000). For brevity, since the first two discourses present Gongsun’s two most prominent paradoxes, I will discuss only those two. 2 3

288

C. Fraser

world of preimperial China. Some dialecticians may have become interested in paradoxes through their investigations of the basis for kind distinctions, through their exploration of rhetorical techniques, or simply for amusement in witty public debates. Some of their paradoxical sayings seem frivolous, but others were likely the outcome of serious philosophical inquiry or motivated by sincere ethical concerns. Nevertheless, ancient sources that report the paradoxes are overwhelmingly dismissive of their value and hostile to those who propounded them. The early Chinese intellectual mainstream seems to have considered the sort of conceptual explorations reflected by paradoxes to be a fatuous distraction from teaching and practicing the proper dao 道 (way). Any discussion of early Chinese paradoxes must acknowledge a pair of important caveats. First, since the extant sources provide little or no context for most of the paradoxes, interpretation is often partly conjectural and in some cases highly speculative. Accordingly, I will try to make explicit which aspects of the interpretations presented here seem well grounded and which are open to doubt. Second, unsurprisingly, given the obscurity of the texts, for any one paradox a plurality of divergent interpretations can be found in the literature. Indeed, in the case of Gongsun Long, scholars disagree not only about how to interpret the texts but also about which texts to interpret, some researchers accepting all five discourses attributed to Gongsun Long as authentic pre-Han writings, some rejecting half as apocryphal. This brief survey cannot attempt to do justice to every significant interpretation of each paradox. Instead, I will try to present a coherent set of reasonably plausible interpretations while using the notes and bibliography to call readers’ attention to other important readings. Early texts regularly associate the dialecticians with the themes of “same and different” (tong yi 同異), “hard and white” (jian bai 堅白), and the “dimensionless” (wu hou 無厚).5 As we will see, many of the paradoxes associated with the School of Names twist commonsense distinctions of sameness or difference, contending that things normally deemed the same are actually different or that things normally deemed different are in fact the same. Often these paradoxes underscore how judgments of similarity or difference are sensitive to changes in scale or perspective— things considered similar from one standpoint may be significantly different from another and vice versa. In some cases, the paradoxes “separate hard from white,” or treat different, compresent features of things as separate entities, as if we were to treat the hardness and whiteness of a white stone as two distinct objects. Several paradoxes seem to follow from properties of the “dimensionless,” a pre-Han term referring to a geometric point. Although these general motifs are common, overall pre-Han paradoxes and sophisms are thematically diverse, and some are so obscure as to defy easy categorization. Rather than attempting to organize them by topic, then, this chapter groups the paradoxes and sophisms according to the source texts in which they have been

5  See, for example, Zhuangzi 17/66; Xunzi 2/30, 8/34; Annals 17.2. Citations to the Annals of Lü Buwei give section numbers in (Knoblock and Riegel 2000).

16  Paradoxes in the School of Names

289

preserved. I will focus on three sources: the Xunzi, which mentions paradoxes attributed to Deng Xi, Hui Shi, and others; “Under Heaven,” which presents a list of theses ascribed to Hui Shi and a second list of miscellaneous paradoxes and sophisms; and the first two discourses of the Gongsun Longzi 公孫龍子, which nearly all scholars take to be genuine pre-Han texts presenting arguments for two paradoxes ascribed to Gongsun Long. Along the way, I will also draw on material from the Mohist “Dialectics” (Mozi books 40–45) and the Annals of Lü Buwei.

2  Paradoxes in the Xunzi The Xunzi (third century B.C.E.) attacks Hui Shi and Deng Xi in three passages, dismissing them as “fond of managing strange doctrines and playing with bizarre expressions, deeply investigating things without attending to practical matters, drawing distinctions without any use, and having much to do but few accomplishments” (Xunzi 6/9). Their doctrines cannot be applied in government, yet the reasons for their views and the seemingly coherent organization of their statements are “enough to mislead the foolish multitudes” (Xunzi 6/10). Since there is no actual historical connection between Deng Xi and Hui Shi—the two were from different states and lived two centuries apart—the Xunzi is likely using them as figureheads for a sophistical, logic-chopping style of inquiry associated with some dialecticians. The paradoxes ascribed to them are intended to represent this sort of intellectual activity and may or may not actually be their invention. The Xunzi attributes six paradoxes to Hui Shi and Deng Xi (Xunzi 3/2–3): 1 . Mountains and gorges are level. 2. Heaven and earth are alongside each other. 3. Qi and Qin are adjoining. 4. Enter through the ear, exit through the mouth. 5. Elderly women have whiskers. 6. Eggs have feathers. The first two of these are alternate versions of paradoxes also attributed to Hui Shi in “Under Heaven,” where they appear as “mountains are level with marshes” and “Heaven is as low as earth” (Zhuangzi 33/71). Along with the third, they fall among a class of paradoxes concerning spatial relations, which seem to illustrate the relative or perspectival nature of spatial distinctions. By everyday standards, mountains and abysses or the sky and the earth are different in height, but from some sufficiently distant standpoint or by some sufficiently vast standard, the difference between them may be insignificant, such that they count as level or beside each other. Similarly, Qi, on the east coast, and Qin, in the far west, are normally considered far apart, but from some perspective they could be considered to adjoin each other, either because the distance between them is insignificant or because they are both parts of a conjoining landmass.

290

C. Fraser

It is unclear why the fourth statement is included, as it does not seem to be a paradox. It may be an interpolation or a garbled, miscopied version of what were originally two three-word paradoxes, like the other five. The sentence “Enter through the ear, exit through the mouth” also appears in book 1 of the Xunzi, where it refers to a shallow level of learning—merely repeating what one has heard, rather than absorbing it in one’s heart and embodying it in one’s conduct (Xunzi 1/31). The fifth paradox is obscure. Literally, it reads “hooks have whiskers,” but commentators suggest the graph for “hooks” is a loan-word for a homonym referring to elderly women. The sixth also appears in “Under Heaven,” although it is not attributed to Hui Shi there. A speculative explanation is that both the fifth and sixth statements belong to a group of paradoxes based on temporal relations and potentiality. Women can give birth to sons who later grow whiskers, so women “have whiskers.” From eggs hatch chicks, which have feathers, so “eggs have feathers.” Xunzi’s fundamental complaint about Hui Shi and Deng Xi is that their “frivolous investigations” (Xunzi 3/1) are no part of ritual propriety and righteousness (Xunzi 3/3–4), an unsurprising criticism given his Ruist commitment to training in ritual norms, including conventional standards for the correct use of names, along with his general disdain for intellectual curiosity (Xunzi 12/25). Moreover, they have “no regard for the facts about right and wrong, so and not-so” (Xunzi 8/28). Xunzi’s intense hostility toward those who propound paradoxes is motivated partly because they are talented scholars who waste their energy on pointless sophistry, rather than devoting themselves to proper ethical training and political administration, and partly because their theses threaten to disrupt the correct use of names, which is necessary to clarify social ranks, distinguish similar from different things, communicate intentions, and carry out tasks (Xunzi 22/14–15). To Xunzi, effective regulation of the use of names is pivotal to maintaining social and political order. Paradoxical sayings endanger good order by potentially confusing the community about how to draw the distinctions that ground the use of names and thus how to use names properly. Propounding paradoxes that muddle the correct use of names is a “great depravity,” a crime comparable to tampering with tallies and measures (Xunzi 22/8). To help eliminate this depravity, Xunzi develops a sophisticated theory covering the purpose of names, the basis for distinguishing the kinds they refer to, and the key points for regulating their use, three points he claims clarify all of the errors that underlie paradoxical sayings.6 In the course of this discussion, he cites three further sets of paradoxes associated with the School of Names. The first set are “being insulted is not disgraceful,” “sages do not care about themselves,” and “killing robbers is not killing people” (Xunzi 22/29–30). Xunzi claims that these statements must be rejected because they interfere with the purpose of having names—namely, to distinguish different things and social statuses, communicate intentions, and thus complete tasks. The first of these sayings was a doctrine of Song Xing and Yin Wen, who advocated it as part of their campaign

 For a detailed discussion, see (Fraser 2016).

6

16  Paradoxes in the School of Names

291

against war and aggression.7 A widespread cultural expectation seems to have been that gentlemen should defend their honor by answering an insult with aggression or else be disgraced. To eliminate this motive for belligerence, Song Xing and Yin Wen taught that a gentleman could be insulted yet turn the other cheek without disgrace. Honor or disgrace rest in one’s conduct, not in whether one responds to insults with violence. Xunzi seems to hold that this saying muddles the normal use of the names “honor” and “disgrace” and so confuses social statuses. In a separate discussion, he contends that the saying diverges from the sage-kings’ model for the correct use of these names (Xunzi 18/102–112). The sage-kings, he claims, distinguished between moral honor and disgrace, which are qualities of one’s conduct, and social honor and disgrace, which are a matter of social circumstances. A gentleman can accept social disgrace but never moral disgrace. Although Xunzi takes himself to be refuting Song Xing, his discussion is arguably a refinement, not a rebuttal, of Song and Yin’s stance, which is not fundamentally paradoxical. The provenance and significance of the second saying are obscure. One possibility is to attribute it to the Mohists, since they held that sages “care about others,” and thus, one might mistakenly infer, not about themselves. However, the phrase ai ren 愛人 (care about others) can also be read as “care about people,” and in fact the Mohists stipulate that in “caring about people,” one also cares about oneself—the scope of “people” includes not only others but oneself.8 The third saying, “killing robbers is not killing people,” is defended by the Mohist dialecticians. The Mohists prohibited “killing people” (that is, murder) but enforced capital punishment against marauding robbers. Apparently, critics contended that this stance on capital punishment was inconsistent with prohibiting killing and, more important, with the core Mohist ethical doctrine of inclusive moral care for all. The Mohists responded with an analogical argument: just as “disliking there being many robbers is not disliking there being many people,” and “desiring there be no robbers is not desiring there be no people,” so too “caring about robbers is not caring about people, not-caring about robbers is not not-caring about people, and killing robber-people is not killing people” (Mozi 45/16–17). In effect, the Mohists assert that since the extensions of “killing robber-people” (capital punishment) and “killing people” (murder) are different, the two referring to distinct kinds of actions, the term “killing people” should not be predicated of actions of the kind denoted by “killing robbers.” Once we understand the Mohists’ argument, their view is not wildly implausible. Nevertheless, their formulation is needlessly paradoxical and confusing, and Xunzi may well be justified in criticizing it for interfering with the purpose of having names. Arguably, the Mohist “robber paradox” conflates predication with identity in the same way as Gongsun Long’s white horse paradox, discussed below. Xunzi’s second set of examples are “mountains and gorges are level,” “the inherent desires are few,” and “fine meats do not add sweetness, great bells do not add

 For details, see Annals 16.8 and Zhuangzi 33/33–41.  See Mozi 44/17. Citations to Mozi give section and line numbers in (Mozi 1956).

7 8

292

C. Fraser

enjoyment” (Xunzi 22/31). He contends that these sayings are plainly contrary to how our sense organs discriminate similar from different objects. The first saying is another variant of the paradox about the relativity of height and depth attributed to Hui Shi and Deng Xi above. The second Xunzi elsewhere attributes to Song Xing (Xunzi 18/114–115). According to “Under Heaven,” Song Xing and Yin Wen contended that people’s inherent, genuine desires are few, shallow, and easily satisfied, and thus a gentleman should not “put his person in hock for things” or contend with others for goods (Zhuangzi 33/40). Xunzi rejoins that people desire as much sensory pleasure as they can get, and indeed this is why the sage-kings rewarded good conduct with wealth and punished bad conduct with deprivation. Regarding the use of names, his implicit claim seems to be that Song and Yin’s doctrine contradicts the normal perceptual basis for using the terms “few” and “many,” since we plainly observe that people have many desires. Arguably, this trivializes Song and Yin’s position, however. Xunzi interprets their slogan as “people’s constitution is that their desires are few” (Xunzi 18/114–115), a statement that may indeed seem paradoxically contrary to observed facts. According to “Under Heaven,” however, Song and Yin actually advocated that people’s “genuine” or “inherent” desires are few (Zhuangzi 33/41).9 They might agree that people typically have many desires but contend that only a few are indispensable, inherent features of human life. The rest we can live without. The third saying is of uncertain origin and significance. It could be related to Song Xing’s views, insofar as Song and Yin also contended that “five pints of rice are enough” to live on (Zhuangzi 33/38–39). Again, Xunzi’s stance seems to be that normal use of the senses shows that fine foods and musical instruments obviously enhance enjoyment. Xunzi’s final group of paradoxes cannot be interpreted with assurance, as the text appears corrupt. If the text is again presenting three sayings, these seem to be “visiting when it is not the case” [?], “the pillar has oxen” [?], and “horses are not horses” (Xunzi 22/32). Alternatively, perhaps there are only two sayings, of which the first is corrupt and the second is either “having oxen-and-horses is not horses” or “oxen-­ and-­horses are not horses.” Given the textual problems, only the final saying is open to plausible interpretation. If read as “horses are not horses,” it could be an abbreviated reference to the white horse paradox (see below). If read as “oxen-and-horses are not horses,” it could be a variant of a claim discussed in Mohist Canon B67.10 The Mohists consider how compound names such as “oxen-and-horses” should be handled in distinction-drawing debates. They suggest that the same grounds can be offered for deeming the assertion “oxen-and-horses are not oxen” impermissible as can be offered for deeming it permissible. The compound name refers to the aggregate of all oxen and horses. Suppose someone deems it permissible to distinguish this aggregate as “not oxen” on the grounds that some of the aggregated animals are  The two construals hinge on different interpretations of the word qing 情 (facts, conditions, affective states). Xunzi takes it to refer to people’s constitution or actual feelings. Song and Yin seem to take it to refer to what is genuine or inherent. Both construals reflect common uses of qing. 10  References to the Mohist Canons (the first four of the Mohist dialectical books) follow the numbering system in (Graham 1978). 9

16  Paradoxes in the School of Names

293

not oxen. Then, the Mohists point out, analogous grounds support deeming it impermissible to distinguish the aggregate as “not oxen,” since some of the aggregated animals are oxen. The implicit point seems to be that such compound names are not a fit subject for distinction-drawing (bian) (Fraser 2007). However, in the course of their explanation, the Mohists state that, without asserting that oxen are not oxen or horses are not horses, there is no problem with asserting “oxen-and-horses are not oxen and not horses,” since some animals included in the aggregate of oxen and horses are not oxen and some are not horses. Xunzi’s response, whether to this paradoxical-sounding Mohist assertion or to Gongsun Long’s “white horses are not horses,” is that such sayings violate basic conventions for the use of names. Checking against naming conventions, we find that everyone, even proponents of these sayings, conventionally uses the name “horse” of all horses, for example. So the paradoxical claims contradict what even their authors accept in everyday practice.

3  The Paradoxes of Hui Shi “Under Heaven,” the final book in the 33-book edition of the Zhuangzi, is a retrospective survey of major schools of thought of previous eras, probably from the hand of a Qin or Han dynasty writer. Appended to the end of the book is a critical discussion of Hui Shi and his sayings, along with a list of miscellaneous paradoxes and sophisms attributed to unnamed dialecticians. Along with the Gongsun Longzi, these two lists are the most important early records of the pre-Han dialecticians’ paradoxes. The text ascribes ten theses to Hui Shi: 1. The ultimately great has no outside, call it the Great One. The ultimately small has no inside, call it the Small One. 2. The dimensionless cannot be accumulated, its size is a thousand miles. 3. Heaven is as low as earth, mountains are level with marshes. 4. Just as the sun is at noon, it is declining. Just as things are alive, they are dying. 5. The same on a large scale but different from what is the same on a small scale, this is called “same and different on a small scale.” The myriad things all being the same or all being different, this is called “same and different on a large scale.” 6. The south has no limit yet has a limit. 7. Today go to Yue but arrive yesterday. 8. Linked rings can be disconnected. 9. I know the center of the world. It is north of Yan [the northernmost state] and south of Yue [the southernmost]. 10. Universally care for the myriad things. Heaven and earth are one body. “Under Heaven” does not record Hui Shi’s arguments for the theses. Some are reasonably clear or open to educated conjecture, but without the original arguments close interpretation is speculative. With the possible exception of theses 1, 5, and

294

C. Fraser

10, there is insufficient contextual information to offer an authoritative argument for any interpretation, although some readings can be ruled out for failing to pertain to any recognized issues or theories in early Chinese philosophical discourse. As a group, the theses seem to revolve largely around the theme that distinctions are not inherently fixed but relative to a standpoint and thus can be redrawn or collapsed by shifting one’s standpoint. Several turn on negating commonsense distinctions, in particular spatial and temporal ones, partly by appeal to the relativity of comparisons and partly by appeal to indexicality. (A high mountain is not high when seen from space; if I move southward, a spot south of me now will become north of me.) The fifth thesis, on “same” and “different,” offers a key to several of the others. It indicates that on some scale or another, anything can be deemed “the same” or “different.” The theses divide fairly naturally into four groups. The first comprises theses 1, 5, and 10, which state philosophical doctrines about ontology and ethics, are relatively clear, and are not obviously paradoxical. All three deal with the plurality of possible ways to distinguish things, either as “the same” or “different” or as parts of a whole, ranging from the smallest possible part—the infinitesimal—to the largest possible whole, the “Great One,” which includes everything in the cosmos. How we distinguish things is relative to the scale or perspective we adopt. Thesis 5 seems to describe the relative or perspectival nature of relations of similarity and difference. Two things can be the same on a large scale, or in some general respect, while at the same time being different on a smaller scale, or in some more specific respect. Two animals can be the same in being of the kind horse, yet be different in color. If we distinguish finely, every individual horse is different; if we distinguish coarsely, horses are no different from other animals or even from all other things. Because the same/different relation comprises both kind relations and part-whole relations, Thesis 5 can also be taken to include part-whole relations. Deeming all things “the same” is deeming them all parts of the same whole. Deeming them “different” can be understood as separating them off as individual parts of that whole or as something else entirely. To the extent that the other theses are based on relations of sameness versus difference and part versus whole, then, thesis 5 helps to explain the others. Thesis 1 is nearly self-explanatory. How we distinguish things—in this case, how we even count “one”—is relative to some standard of division. The thesis concerns aggregating and dividing. The whole cosmos can be aggregated into a whole to form the Great One, or it can be divided down to the smallest possible unit, the Small One, probably a geometric point. Thesis 10 presents an ethical principle tied together with an ontological one, which presumably is meant to justify it. Since everything can be summed into a whole—the Great One—heaven and earth and the myriad things contained therein can be considered a single “body” or “unit.” (The expression “heaven and earth” refers to the cosmos, including not only the sky and earth but the entire natural world.) Hui Shi’s ethical conclusion, then, is that if everything is one unit, then any care (ai 愛) we have for ourselves should also be directed at all of the other myriad

16  Paradoxes in the School of Names

295

things (wu 物, also “creatures”), since we and they are all parts of the same vast “body.” The second group are paradoxes concerning infinitesimals and part-whole relations. Thesis 2 concerns geometrical points, the “dimensionless” (wu hou 無厚). The sum of two points is a point, and so points cannot be accumulated to form an object with thickness or length. Yet anything with dimensions, such as a length of a thousand miles, is somehow constituted by points and divisible into them. Thesis 8 also may pertain to infinitesmals, although interpretation is speculative, since it is the most obscure of the ten.11 If the linked rings are thought of as circles, formed by points on a plane, then they have no thickness. They appear linked when viewed from above, but on the surface of the plane nothing blocks them from being pulled apart. Alternatively, if, as thesis 2 suggests, two three-dimensional rings are constituted by dimensionless points, then they can be pulled apart: since each point takes up no space, there is nothing preventing the rings from passing through each other. The third group are based on spatial relations, including comparisons of size. Thesis 3, also attributed to Hui Shi in the Xunzi, can be interpreted as illustrating how things deemed different on one scale can be deemed the same on another. By the scale of the infinitely vast Great One, the difference between the height of the sky and the earth or mountains and marshes may be insignificant. The differences between mountains and marshes may be only what thesis 5 calls “differences on a small scale,” while the two count as “the same” on a large scale. Thesis 6 is especially obscure, and interpretation is speculative. One plausible conjecture is that, like theses 7 and 9, it trades on the properties of indexicals (Hansen 1992: 262). Since the referents of indexicals shift with speaker, time, and standpoint, they vividly illustrate how distinctions are relative to perspective. The cardinal directions were thought to have no limit, in that one could continue traveling in any direction without end. However, since the directions are indexical, relative to our point of reference, south always also has a limit, namely the point at which we stand. Thesis 9 also involves indexicals and is equally obscure. One reasonable guess is that if space is infinitely large, then anywhere one stands can be considered the center, whether in the far north or the far south of the known world (Graham 1989: 79). The fourth group deal with temporal relations. Thesis 4 is paradoxical but easily intelligible. Just as, from one perspective, the sun is at its highest, from another perspective, it is beginning to set. Just as things are living and growing, they are also coming closer to death. Again, the availability of different perspectives threatens to collapse the distinction between two apparent opposites, living and dying. Thesis 7 is more difficult. One proposed interpretation is that if I cross the border into Yue at precisely the instant when today turns into tomorrow, then I simultaneously leave one state today and arrive in the other tomorrow.12 Another possibility is that, since the word xi 昔 can refer to either “yesterday” or “the past,” thesis 7 can be read as

11 12

 For discussion, see (Harbsmeier 1998: 296) and (Graham 1989: 79).  See (Harbsmeier 1998: 298) and (Graham 1989: 79).

296

C. Fraser

“today I go to Yue, but I arrive in the past.” The point could then be that by completing any action we take up a standpoint that locates it in the past. Whenever we arrive somewhere, our journey is completed, or “past.” Thesis 7 is quoted—without attribution to Hui Shi—in the Zhuangzi “Discourse on Evening Things Out” as an example of “taking what does not exist to exist,” which “Evening Things Out” dismisses as absurd.13 One common interpretation of Hui Shi’s theses is that thesis 10 articulates their overarching significance.14 The theses attempt to show that the similarities and differences by which we distinguish things can be identified in indefinitely many ways, depending on one’s standpoint. Apart from the standpoints we take up, there are no independent, preexisting standards by which to identify a scheme of privileged, correct distinctions. The world in itself fixes no particular way of drawing distinctions as correct. The neutral or objective standpoint—that of the world in itself—is monistic, drawing no distinctions at all. Hence, as Thesis 10 states, nature in itself forms a unified whole, of which we are parts, and as parts of this whole, we should care for all things. If this is indeed Hui Shi’s position, the move from perspectivalism about distinctions to a form of ontological monism seems questionable. On the basis of Thesis 5, Hui Shi should hold that any scheme of distinctions may be deemed permissible or impermissible, by some standard or other, yet no scheme is privileged, including a scheme that consists in drawing no distinctions at all. Instead, if the monistic interpretation is correct, he mistakenly takes the “Great One” or “one body” view to be an exception to this rule, a privileged or authoritative perspective. The mistake is understandable, since drawing no distinctions at all might seem to be a way of circumventing the perspectival nature of distinctions. It is not, however, since strictly speaking it remains one among other ways of drawing distinctions.15

4  Miscellaneous Paradoxes in the Zhuangzi “Under Heaven” lists twenty-one more paradoxes (Zhuangzi 33/74–78), “with which the dialecticians responded to Hui Shi for their whole lives without end.” All lack explanations, leaving some impenetrably obscure. This section presents interpretations of several of the relatively tractable ones, along with speculations about some of the others.16  Zhuangzi 2/22. This part of the Zhuangzi also cites paradoxes suggestive of theses 3 and 10 (Zhuangzi 2/51–53) and uses phrasing similar to thesis 4 (Zhuangzi 2/28) but without attributing these to Hui Shi. 14  Hu Shi may have been the first to advance this view (Hu 1922). 15  For further discussion, see (Fraser 2015). For alternative interpretations of Hui Shi’s paradoxes, see (Lange 1988), (Stevenson 1991a, b), (Lucas 1993), (Xu 1997), (Solomon 2013), and (Fung 2014). 16  This account is indebted to (Hansen 1992), (Graham 1989), and a range of Chinese commentators cited in the Qing dynasty Zhuangzi Jishi 莊子集釋 of Guo Qingfan 郭慶藩 and the modern 13

16  Paradoxes in the School of Names

297

One of these paradoxes is nearly self-explanatory, a version of Zeno’s racetrack paradox: A one-foot stick, every day take away half of it, in a myriad generations it will not be exhausted.

If we remove half of a stick each day, the stick will never be completely used up, since at each stage half its length remains. Mohist Canon B60 presents an alternate version of this paradox and suggests a solution. (The text of the canon is obscure, so the following interpretation is tentative.) The canon states that, seeking to move a particular distance, if one cannot cut a portion other than half, one cannot move. The explanation is that if one tries to move forward cutting half at a time, one never succeeds even in reaching halfway, as at each stage what was originally the center no longer marks half but instead becomes a new starting point. If cutting must be by halves, then since in every case some portion remains that constitutes half, it is impossible to reach the end. The implied solution is that in moving a measured distance, we do not necessarily first move one fraction of the distance and then another. It is possible to move the whole distance in one stage. Several of the other paradoxes can also be tentatively explained by considering passages in the Mohist Canons. Dogs are not hounds.

In the Mohist Canons, “dog” and “hound” (or “pup” and “dog”) are stock examples of coextensive terms, of which a speaker might know one without knowing the other. A speaker unaware that the two terms are coextensive could know about dogs and yet say without error that he didn’t know about hounds (Canon B40). For related reasons, in some contexts, the Mohists hold, it is permissible to say “killing dogs is not killing hounds” (B54). The paradox may be twisting or extending this point. Fire is not hot.

The point of this paradox could be that when we feel the heat of the fire, the heat is in us, not the fire. If this speculation is correct, Mohist Canon B47 appears to present a rebuttal. It reads: “Fire is hot.” “We call the fire hot, we don’t deem the heat of the fire to belong to us.” The argument is unclear, but it may be that heat is among the distinguishing criteria for the kind fire, so simply to call something “fire” is to deem it “hot.” Alternatively, the paradox could be based on conflating predication with identity, such as in Gongsun Long’s “White Horse Discourse.” Since the extension of “fire” is different from that of “hot,” “fire is not [identical to] hot,” for some hot things are not fire. The eyes do not see.

The Mohists concur with this paradoxical-sounding claim. “The knower sees by means of the eyes and the eyes see by means of the fire but the fire does not see. . . .

editions of Chen (2000) and Wang (1988).

298

C. Fraser

Seeing by means of the eyes is like seeing by means of the fire” (B46). Like the fire, the eyes themselves do not see but are the means by which we see. The shadow of a flying bird has never moved.

Again, the Mohists agree with this seemingly paradoxical statement. Canon B17 states, “The shadow does not move. Explained by: Being made over again.” “When light arrives, the shadow disappears.” The shadow itself does not move across the ground. Instead, as the bird moves, the area it blocks from the light changes as well, causing one shadow to vanish and a new shadow to form over and over again. Interpretation of the remaining paradoxes and puzzling sayings is largely guesswork. Some readings can be excluded as implausible, particularly those that construe them as dealing with concepts or topics which had no role in the context of early Chinese philosophy of language, epistemology, or ontology, such as the dichotomy between appearance and reality or relations between universals and particulars. However, it is difficult to find compelling arguments to justify one or another interpretation as best explaining a particular paradox. The interpretations below are strictly conjectural. Eggs have feathers.

As we saw above, the Xunzi attributes this paradox to Hui Shi and Deng Xi. The point may be that the future, potential feathers of the chick are already possessed by the egg. Chickens have three feet.

A common, conjectural interpretation of the argument for this sophism is that one can assert that chickens have feet and also that they have left feet and right feet. Therefore they have a total of three feet. Another version of the same sophism appears in Annals of Lü Buwei (18.5), where Gongsun Long is reported to have cleverly argued that a man has three ears. In response, Kong Chuan, his opponent, comments to the Lord of Pingyuan, “Asserting that a man has three ears is extremely difficult and in reality is wrong; asserting that a man has two ears is extremely easy and in reality is right. I wonder, will my Lord follow what is easy and right or what is difficult and wrong?” The city of Ying possesses the world.

Perhaps each part of the whole “possesses” the whole (?). Hounds can be deemed sheep.

Mohist Canon B8 addresses cases of “borrowed” names, as when a name for one kind of thing is temporarily used to refer to another kind of thing, and of using a general term for one kind of thing as a proper name for another, as when a person is surnamed “Bird.” Perhaps, for such reasons, hounds can be deemed sheep. Alternatively, perhaps the paradox refers to how the criteria or distinctions by which we name things can be changed arbitrarily. Frogs have tails.

16  Paradoxes in the School of Names

299

A frog was once a tadpole, which has a tail (?). Wheels do not touch the ground.

The point may be that the wheel as a whole does not touch. Only a single point touches at each instant, and the point of contact is dimensionless. Pointing does not reach, reaching does not detach.

Perhaps the claim is that referring to something by a name, or pointing to it physically, is never enough to ensure that one’s audience picks out the correct referent. To actually “reach” something, one must touch it and cannot let go (?). Tortoises are longer than snakes.

Mohist Canon B6 explains that fundamentally different kinds of things cannot properly be compared. For instance, we cannot ask which is longer, a piece of wood or the night. The gist of the paradox might be that a snake’s body has a longer length, but a tortoise has a longer life. The set square is not square, the compass cannot make a circle.

The carpenter’s set square and wheelwright’s compass were standard examples of models (fa 法) used to guide action and check whether it conforms to norms. The paradox implies that the models themselves fail to conform to the relevant norms. Perhaps the point is the obvious, true claim that the set square and compass themselves are not actually square or circular. Perhaps it is that since they are models, rather than the geometric figures themselves, they fall short of being square or circular. The barbed arrow at its swiftest, there is a time when it neither moves nor stops. The time referred to might be the instant the string is released, before the arrow moves. This paradox is distinct from Zeno’s arrow paradox, which is that the flying arrow is at rest in every instant of time and so does not move. Here the paradox is that the arrow is neither in motion nor at rest. A brown horse and a black ox are three.

Possibly the three are the horse, the ox, and color, which both animals have. The white horse paradox (see below) treats the shape of the horse as distinct from its color. White dogs are black.

Mohist Canon A96 discusses how, in determining the application of a compound name such as “black person,” one must fix what part of a person is the criterion for deeming the person “black.” Presumably it is the person’s skin color, not hair color, for example. If one chooses an unorthodox criterion for deeming a dog black, such as the dog’s nose, dogs with white fur might be deemed “black.” An orphan colt has never had a mother.

300

C. Fraser

Since the criterion for being an orphan is that one’s parents have died, the point may be that there has never been a colt deemed an orphan that had a mother at the time it was deemed an orphan. Horses have eggs. (?) Mountains emerge from mouths. (?) The chisel does not surround the handle. (?)

These three puzzling sayings defy educated conjecture.

5  Gongsun Long’s Paradoxes Gongsun Long is famously associated with the paradoxical claim that white horses are not horses. The brief extant dialogue defending this claim has attracted a plethora of interpretations, with no consensus in sight as to the significance and theoretical basis of the text’s arguments.17 Hence the interpretation proposed here must be considered only one of several potentially defensible approaches to the text. The present interpretation stems from two hypotheses. The first is that the likely theme of the “White Horse Discourse” is reflected in an ancient anecdote about Gongsun Long included in the introductory chapter of the Gongsun Longzi.18 Hoping to study with Gongsun Long, Kong Chuan asks him to first abandon the thesis that white horses are not horses, which Kong Chuan cannot accept. In response, Gongsun Long contends that Confucius himself—Kong Chuan’s ancestor—accepted the same view. He cites a well-known story about the King of Chu losing his bow. The King’s attendants asked to search for it. The King said, “Stop. A Chu person lost a bow. A Chu person will find it. Why bother to look for it?” Confucius heard about it and said, “The King of Chu is benevolent and righteous but hasn’t yet followed all the way through. He should simply have said, ‘A person lost a bow, a person will find it,’ that’s all. Why must it be ‘Chu’?” In this way, Confucius took Chu people to be different from what’s called “people.” Now to approve of Confucius’s taking Chu people to be different from what’s called “people” but disapprove of my taking white horses to be different from what’s called “horses” is contradictory.19 (1.1/1/14–18)

 One early, influential interpretation took its theme to be denying the identity of the universals “horse” and “white horse” (Feng 1958; Cheng 1983). Other interpretations have taken it to deal with kind and identity relations (Cikoski 1975; Harbsmeier 1998), part-whole relations (Hansen 1983; Graham 1989), how the extensions of phrases vary from those of their constituent terms (Hansen 1992), and even the use/mention distinction (Thompson 1995). For recent discussions, see (Fung 2007), (Im 2007), (Mou 2007), (Lucas 2012), (Solomon 2013), and (Fraser 2015). For overviews of competing interpretations, see (Hansen 2007), (Cheng 2007), and (Fung 2014). For interpretations which, like that presented here, question the enterprise of interpreting “white horse” as a serious piece of philosophical inquiry, see (Harbsmeier 1998) and (Trauzettel 1999). 18  I adopt this view from (Harbsmeier 1998: 302). 19  Citations to the Gongsun Longzi give index numbers in (Lau et al. 1998). 17

16  Paradoxes in the School of Names

301

The king did not discriminate between recovering the bow himself and letting another person of Chu find it. Confucius suggests the king could reach an even greater degree of benevolent impartiality by ceasing to discriminate between the people of Chu and everyone else. According to Gongsun Long, since Confucius distinguishes Chu people from people in general, he implicitly holds that Chu people are different from what we call “people” and so are not people. In an alternate version of the anecdote found in the Kong Congzi, Kong Chuan rejoins that when Confucius omits the word “Chu,” he is not implying that Chu people are not people, but simply broadening the reference of the noun phrase. Whenever we say “people,” we refer to people in general, just as whenever we say “horses,” we refer to horses in general. “Chu” by itself is the state; “white” by itself is the color. Wishing to broaden the referent of “people,” it’s appropriate to omit the “Chu”; wishing to fix the name of the color, it’s not appropriate to omit the “white.”20

The anecdote suggests that to an ancient audience, the theme of the white horse paradox was how the scope of a general term such as “people” or “horses” changes when modified by an adjective such as “Chu” or “white.” The commonsense understanding is that the adjective restricts the scope and not, as Gongsun Long insists, that it yields a noun phrase referring to an entirely different kind of thing. As Xunzi points out in his discussion of naming, sometimes we refer to things by a single, general name, such as “horses,” and sometimes, to communicate more precisely, we use what he calls a “compound” name, such as “white horses” (Xunzi 22/21–22). Provided one of the two kinds of names is more general, we can use both without their interfering with each other. Gongsun Long puckishly refuses to acknowledge that terms can refer to things at different levels of generality. The second hypothesis concerns Gongsun Long’s intellectual orientation, as implied by the anecdote cited previously about his ingenious argument that a man has three ears. As Kong Chuan puts it, Gongsun Long’s case is extremely difficult to make and plainly does not fit reality. His arguments are exercises in cleverness, a sort of trick performance devoted to defending claims obviously at odds with the facts. If this characterization is accurate, there is little reason to expect the grounds for his assertions to embody convincing reasoning based on a cogent semantic or logical theory, as their original aim may have been to provide whimsical entertainment.21 The “White Horse Discourse” consists of a series of exchanges between a sophist, who presents five arguments that the statement “white horses are not horses” is “permissible” (ke 可), and an objector, who defends the commonsense view that white horses are horses. The first argument runs as follows: “Horse” is that by which we name the shape. “White” is that by which we name the color. Naming the color is not naming the shape. So I say, white horses are not horses. (1.2/3/1–2)

20 21

 From Kong Congzi, Book 12. See (Graham 1989: 84).  This approach to Gongsun Long’s arguments follows (Harbsmeier 1998: 300–301).

302

C. Fraser

The argument seems an obvious non  sequitur. One plausible explanation is that “white horses” names both the color and the shape of white horses, not only the shape. So “white horses” names something different from what “horses” names. Since the two name different things, Gongsun contends, white horses are not horses. He disallows the possibility that the difference between white horses and horses could be that the former are a subset of the latter. The argument can also be understood as “separating hard and white,” in that the shape and color of white horses, two compresent features, are treated as if they were two separate things, such that naming one is not naming the other. Of course, referring to the color is different from referring to the shape. But the object that has the color is the same as the object that has the shape. Hence we should challenge the third premise and insist that naming the object that has the color is also naming the object that has the shape. The objector responds that there being white horses, one cannot claim there are no horses. How does white horses’ being white make them not horses? The sophist replies by presenting his second argument. If someone seeks horses, brown or black horses can comply. If someone seeks white horses, brown or black horses cannot comply. Supposing white horses were indeed horses, in these cases what is sought would be one and the same. What is sought being one and the same is white ones not being different from horses. If what is sought is not different, then how is it that brown or black horses in the one case can comply and in the other cannot? Can and cannot, that they contradict each other is clear. So brown and black horses are one and the same in that they can respond to “having horses” but not to “having white horses.” This confirms that white horses are not horses. (1.2/3/7–11)

The sophist construes “white horses are horses” as “white horses are identical to horses.” In Chinese, as in English, the sentence “white horses are horses” can be interpreted as predicating the term “horses” of white horses, thus making the true assertion that white horses are among the things picked out by “horses,” or it can be interpreted as expressing an identity, thus making the false assertion that the things picked out by “white horse” are identical to those picked out by “horse.” The argument trades on this ambiguity. Because we know that modifying a noun restricts the scope of its extension, when we hear “white horses are horses,” we charitably assume the speaker is not saying something patently false and so interpret the sentence as predicating “horses” of white horses, not stating an identity. By contrast, the sophist insists on interpreting the sentence as an identity. He applies a principle roughly like Leibniz’s law of indiscernibility of identicals, assuming that if two things are identical, they share all their features and are intersubstitutable in any context. In his view, any difference between white horses and horses shows that white horses are not horses. Again, he glibly ignores the possibility that white horses fall within, without being identical to, the extension of “horses.” The objector rejoins that if having a color renders a horse not a horse, then since there are no colorless horses, would the sophist allow that there are no horses? The third argument responds as follows: Horses indeed have color; thus there are white horses. Supposing horses had no color, such that there were simply horses and that’s all, how could we pick out white horses? So white

16  Paradoxes in the School of Names

303

is not horse. White horses are horses combined with white. Are horses combined with white the same as horses?22 So I say, white horses are not horses. (1.2/3/15–16)

White is not horse because the name “horse” alone does not pick out the white horses. Only “white” does. The sophist assumes that “horses” combined with “white” are not simply “horses.” Here he is “separating hard and white,” in that he explicitly treats “white” (the animals’ color) and “horse” (their shape) as two distinct things that are combined to form something different from mere horses. The argument again turns on construing “white horses are horses” as the claim that the kind white horses is identical to the kind horses. In the next exchange, the objector sees the problem behind the previous argument and accuses the sophist of illegitimately naming compresent things—the horses’ shape and color—as if they were separate. Sidestepping this criticism, the sophist’s fourth argument seeks to trick the objector into agreeing that brown horses are not horses, a claim pragmatically inconsistent with maintaining that white horses are horses. “Since you take having white horses to be having horses, can we assert that having horses23 is having brown horses?” “Not permissible.” “Taking having horses to be different from having brown horses, this is taking brown horses to be different from horses. Taking brown horses to be different from horses, this is taking brown horses not to be horses.” (1.2/3/21–25)

Here the sophist explicitly indicates that he construes “brown horses are horses” as an identity claim, since he states that acknowledging a difference between brown horses and horses is denying that brown horses are horses. He also fallaciously treats predication as symmetric: if having brown horses is having horses, he holds, then having horses is having brown horses. The objector now explicitly accuses the sophist of “separating” the color white from the horse, this being the reason he denies that white horses are horses. The sophist’s response is his fifth and final argument. “White” does not fix what is white; we can forget about it. “White horse” speaks of white and fixes what is white. That which fixes what is white is not white. “Horses” excludes no colors, so brown or black horses are all what can respond. “White horses” excludes some colors; brown and black horses are all excluded on the basis of color, and so only white horses alone can respond. Excluding none is not excluding some. So I say, white horses are not horses. (1.2/4/1–3)

The sophist first “separates hard and white.” The color alone does not fix the location that is white; saying “white horse” does. Therefore horse, the shape, is not white, the color. Indeed, “horses” specifies no color at all. “White horses,” on the other hand, does specify a color. So again the sophist claims to have shown that

 The text appears faulty here. I have followed Harbsmeier’s proposed emendation (Harbsmeier 1998: 307, n2). 23  “White horses” has been emended to “horses” here to cohere with the first clause in the sophist’s next sentence. 22

304

C. Fraser

white horses and horses have distinct features and are not identical. Therefore white horses are not horses. To sum up, the arguments of “White Horse” repeatedly equivocate between statements of identity and statements that predicate a more general term of objects denoted by a less general term. The sophist refuses to distinguish the true claim that the extension of “white horses” is not identical to that of “horses” from the false claim that white horses are not part of the extension of “horses.” A second well-known paradox attributed to Gongsun Long is expounded in the brief, enigmatic “Discourse on Pointing at Things.” This text opens with the puzzling, apparently self-contradictory statement that “no things are not pointed out, yet pointing is not pointing.” Although the significance of this statement and the details of the text’s arguments for it have been the subject of intensive interpretive effort, “Pointing at Things” remains profoundly obscure.24 The key term, zhi 指, roughly “pointing,” can be interpreted as “to point,” “to refer,” “what is pointed out,” or “referent,” each of which yields plausible explanations in some but not all contexts. The text slides playfully between verbal and nominal uses of zhi, embedding them in complex strings of quantifiers and negations of vague scope. The following excerpt gives a taste of the discussion: There being no pointing out the world, things cannot be called the pointed out. Not being able to be called the pointed out is not the pointed out. Not the pointed out is no thing is not the pointed out. There being no pointing out the world and things not being able to be called the pointed out is it not being the case that there is not the pointed out. It not being the case that there is not the pointed out is no thing is not the pointed out. No thing is not the pointed out yet pointing is not pointing. (2.1/4/14–16)

One interpretation consistent with key themes and assumptions in pre-Han philosophy of language and ontology is that the discussion concerns a paradoxical feature of unrestricted, maximally general terms such as “things” or “the world.” The extension of such terms includes everything, so “no thing is not pointed to.” Because their scope includes everything, however, they do not distinguish their referents from anything else. So “pointing out is not pointing out,” since by referring to everything, such terms fail to point anything out from anything else. Graham among others develops an interpretation along roughly these lines (Graham 1989: 91ff.). He proposes that the text concerns how the word “world” (tian xia 天下) functions as a name. According to pre-Han theories of language, names designate things by distinguishing them from other things. But since “world” refers to the whole comprising everything, it does not distinguish anything from other things. The gist of the paradox might be that when we refer to the world, we refer to everything: no thing is not pointed to. Yet the world cannot be pointed out from other things, because it is not a thing in itself distinct from the things that constitute it. In referring to the world, we do not point it out from anything, so pointing to it is not pointing out anything.  For a broad sample of interpretations, see (Cheng and Swain 1970), (Kao and Obenchai 1975), (Hearne 1976, Hearne 1985), (Rieman 1980), (Graham 1989), (Stevenson 1991a), (Hansen 1992), (Schleichert 1993), (Lai 1997), (Reding 2002), (Fung 2014), and (Fraser 2015). 24

16  Paradoxes in the School of Names

305

On this interpretation, the text may also include a rebuttal of the paradox (2.1/4/18–21). Although no thing in the world is deemed or named “the world”—no one specific thing or kind of thing is pointed out by the name “the world”—nevertheless, “the world” refers to all things. We aggregate into a whole all things in the world and use the phrase “the world” to refer to them jointly. So there are things referred to by “the world” even though they are not specifically deemed “the world.” Moreover, although “the world” does not point out something distinct from other things, we cannot say there is no pointing out things, since “the world” points out the sum of all things. So the paradox is mistaken to claim that “pointing is not pointing.” Nevertheless, on this interpretation—or most others—the steps in the reasoning remain confusing, the justification for them murky. The text is so rife with apparent contradictions, circularities, and logical gaps that no interpretation inspires much confidence in its superiority over plausible rivals. Indeed, one intriguing conjecture is that the very purpose of “Pointing at Things” was to generate endless perplexity. Given Gongsun Long’s mischievous literary persona, the text could well be an ancient practical joke, one that has successfully bewildered audiences for millennia.

References Chen, Guying 陳鼓應. 2000. Zhuangzi: Contemporary Notes and Paraphrase 莊子今註今譯, rev. ed. Taipei: Commercial Press. Cheng, Chung-Ying. 1983. “Kung-sun Lung: White Horse and Other Issues.” Philosophy East and West 33.4: 341–54 . Cheng, Chung-Ying. 2007. “Reinterpreting Gongsun Longzi and Critical Comments on Other Interpretations.” Journal of Chinese Philosophy 34.4: 537–60. Cheng, Chung-Ying, and Richard H. Swain. 1970. “Logic and Ontology in the Chih Wu Lun of Kung-sun Lung-Tzu.” Philosophy East and West, 20.2: 137–54. Cikoski, John. 1975. “On Standards of Analogical Reasoning in the Late Chou.” Journal of Chinese Philosophy 2.3: 325–57. Feng, Yu-lan. 1958. Derk Bodde, tr. A Short History of Chinese Philosophy. New  York: Macmillan. Fraser, Chris. 2007. “More Mohist Marginalia.” Journal of Chinese Philosophy and Culture 2: 227–59. Fraser, Chris. 2015. “The School of Names.” In Edward Zalta, ed., The Stanford Encyclopedia of Philosophy (plato.stanford.edu/entries/school-names/). Stanford: Metaphysics Research Lab. Fraser, Chris. 2016. “Language and Logic in the Xunzi.” In Eric Hutton, ed., Dao Companion to Xunzi. Dordrecht: Springer. Fung, Yiu-Ming. 2000. Gongsun Longzi 公孫龍子. Taipei: Dong da 東大. Fung, Yiu-Ming. 2007. “A Logical Perspective on ‘Discourse on White-Horse’.” Journal of Chinese Philosophy 34.4: 515–36. Fung, Yiu-Ming. 2014. “Review of Bernard Solomon, On the School of Names in Ancient China.” Journal of Chinese Studies 59: 284–298. Graham, A.  C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: Chinese University Press. Graham, A. C. 1989. Disputers of the Tao. LaSalle: Open Court.

306

C. Fraser

Graham, A. C. 1990. Studies in Chinese Philosophy and Philosophical Literature. Albany: SUNY Press. (Contains Graham’s important textual study of the Gongsun Longzi.) Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought. Oxford: Oxford University Press. Hansen, Chad. 2007. “Prolegomena to Future Solutions to ‘White-Horse Not Horse’.” Journal of Chinese Philosophy 34.4: 473–91. Harbsmeier, Christoph. 1998. Science and Civilisation in China, Vol. 7, Part 1: Language and Logic. Cambridge: Cambridge University Press. Hearne, James. 1976. “A Critical Note on the Cheng-Swain Interpretation of the Chih Wu Lun.” Philosophy East and West 26.2: 225–28. Hearne, James. 1985. “Formal Treatments of the Chih Wu Lun.” Journal of Chinese Philosophy 12.4: 419–29. Hu Shih. 1922. The Development of the Logical Method in Ancient China. Shanghai: Commercial Press. Im, Manyul. 2007. “Horse-Parts, White-Parts, and Naming: Semantics, Ontology and Compound Terms in the White Horse Dialogue.” Dao: A Journal of Comparative Philosophy 6.2: 167–86. Kao, Kung-yi, and D.  B. Obenchai. 1975. “Kung-sun Lung’s Chih Wu Lun and Semantics of Reference and Predication.” Journal of Chinese Philosophy 2.3: 285–324. Knoblock, John, and Jeffrey Riegel. 2000. The Annals of Lü Buwei. Stanford: Stanford University Press. Lai, Whalen. 1997. “Kung-sun Lung on the Point of Pointing: The Moral Rhetoric of Names.” Asian Philosophy 7.1: 47–58. Lange, Marc. 1988. “Hui Shih’s Logical Theory of Descriptions: A Philosophical Reconstruction of Hui Shih’s Ten Enigmatic Arguments.” Monumenta Serica 38: 95–114. Lau, D.  C., Ho Che Wah, and Chen Fong Ching, eds. 1998. Concordances to the Kongcongzi, Dengxizi, Yinwenzi, Gongsun Longzi. Hong Kong: Commercial Press. Lucas, Thierry. 1993. “Hui Shih and Kung Sun Lung: an Approach from Contemporary Logic.” Journal of Chinese Philosophy 20.2: 211–55. Lucas, Thierry. 2012. “Why White horses Are Not Horses and Other Chinese Puzzles.” Logique Et Analyse 56: 185–203. Mou, Bo. 2007. “A Double-Reference Account: Gongsun Long’s ‘White-Horse-Not-Horse’ Thesis.” Journal of Chinese Philosophy 34.4: 493–513. Mozi. 1956. A Concordance to Mozi 莊子引得. Harvard-Yenching Sinological Index Series, supplement no. 21. Cambridge: Harvard University Press. Reding, Jean-Paul. 2002. “Gongsun Long on What is Not: Steps Toward the Deciphering of the Zhiwulun.” Philosophy East and West 52.2: 190–206. Rieman, Fred. 1980. “Kung-sun Lung, Designated Things, and Logic.” Philosophy East and West 30.3: 305–19. Schleichert, Hubert. 1993. “Gong-sun Long on the Semantics of ‘World’.” In Hans Lenk and Gregor Paul, eds., Epistemological Issues in Classical Chinese Philosophy (113–117). Albany: SUNY Press. Solomon, Bernard. 2013. On the School of Names in Ancient China. Monumenta Serica Monograph Series 64. Sankt Augustin: Germany: Steyler Verlag. (An informative monograph covering early Chinese paradoxes.) Stevenson, Frank. 1991a. “Meaning is not Meaning: World, Thing, and Difference in Kung-sun Lung’s Chih Wu Lun.” Tamkang Review 21.3: 297–322. Stevenson, Frank. 1991b. “South Has (No) Limits: Relative and Absolute Meaning in Hui Shi’s Ten Points.” Tamkang Review 21.4: 325–346. Thompson, Kirill Ole. 1995. “When a ‘White Horse’ is not a ‘Horse’.” Philosophy East and West 45.3: 481–99. Trauzettel, Rolf. 1999. “A Sophism by the Ancient Philosopher Gongsun Long: Jest, Satire, Irony—or Is There a Deeper Significance?” Journal of Chinese Philosophy 26.1: 21–36.

16  Paradoxes in the School of Names

307

Wang Shumin 王叔岷. 1988. Collated Interpretations of Zhuangzi 莊子校詮. Taipei: Academia Sinica. Xu Keqian. 1997. “The Unique Features of Hui Shi’s Thought: A Comparative Study Between Hui Shi and Other Pre-Qin Philosophers.” Journal of Chinese Philosophy 24.2: 231–253. Xunzi. 1966. A Concordance to Xunzi 荀子引得. Harvard-Yenching Institute Sinological Index Series, supplement no. 22. Cambridge: Harvard University Press. Zhuangzi. 1956. A Concordance to Zhuangzi 莊子引得. Harvard-Yenching Sinological Index Series, supplement no. 20. Cambridge: Harvard University Press. Chris Fraser  is Associate Professor in the Department of Philosophy at the University of Hong Kong. He is the author of The Philosophy of the Mozi: The First Consequentialists (Columbia University Press, 2016) and numerous scholarly articles and book chapters on Mohism, Confucianism, Daoism, and early Chinese philosophy of language, logic, and epistemology

Chapter 17

Logical Thinking in the Gongsun Longzi Yiu-ming Fung

1  Hu Shih’s Interpretation and Analysis Hu Shih 胡適 is one of the pioneers to use Western logic to deal with logical thinking and logical thought in ancient China. In regard to Gongsun Long’s 公孫龍 logical thinking, he thinks that some of the so-called “paradoxes” of the sophists mentioned in the chapter thirty-three of the Zhuangzi 莊子 were Gongsun Long’s words. He argues that Gongsun Long’s view on these paradoxes was based on his idea of individuation. He also believes that the engagement or comparison between Later Mohists 後期墨家 and Gongsun Long on the problem of individuation reflects the difference between their logical views. In regard to Later Mohists’ view, Hu writes: (Hu 1922: 122–3). The Neo-Mohists had formulated a logic which, as we have seen, is essentially a theory of scientific classification. In classification the individual is explained in the species, and the species in the genus. It is natural that the problem of individuation should soon attract the attention of the logicians as is indicated here and there in Neo-Mohist texts. It has been held that all things which are of one form are similar to one another, as all cubes, be they made of wood or stone, are similar to one another. (Bk. 33: 64; Bk. 35: 64.) If consistently carried out, this view will tend to submerge the individual in the universal. But the individual is not so easily dismissed. “The parents of Hueh [Huo 獲] are human beings, but Hueh’s service to his parents is not the same as his services to mankind. His brother may be a good-looking man. But his love for him is not a love for a good-looking man (or for all good-looking men). ... A thief is a man. But ‘there are many thieves’ is not the same as ‘there are many men.’ ‘No thief’ is not ‘no man.’ For to dislike the numerousness of thieves is not to dislike the populousness of men, and to desire the elimination of thieves is not to desire the extinction of the human race. … Therefore we may say: ‘Thieves are men; but to love thieves is not to love men; and to hate thieves is not to hate men; and to kill thieves is not to kill men.’” (Bk. 37: 6).

Y.-m. Fung (*) Department of Philosophy, Tunghai University, Taichung, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_17

309

310

Y.-m. Fung

It seems that there is a theoretical problem, if not a difficulty, for Later Mohists in making all parallelisms in the positive form (“A is B; CA is CB”) without exception and thus they have to negate the latter part of some parallelisms with the word “not.” So Hu Shih continues: “So long as thieves are classified as men, it is impossible to draw the negative conclusion that to kill thieves is not to kill men.” To explain this problem, Hu compares it with Gongsun Long’s view. He says: “Kung-­ Sun Lung [Gongsun Long] came to rescue with his theory that ‘a white horse is not a horse,’ which is equivalent to saying that a thief is not a man.” Similarly, against the statement that “a puppy is a dog, but to kill a puppy is not killing a dog” in the Neo-Mohist texts (Bk. 34: 53), “Kung-Sun Lung advanced the paradox (q) that ‘a puppy is not a dog’” (Hu 1922: 122–3). I think Hu’s interpretation and explanation mentioned above are inaccurate. First, there is no obvious textual evidence to demonstrate that some of the so-called “paradoxes” of the sophists are Gongsun Long’s statements. As I argued elsewhere, most of the statements are not paradoxes in terms of W. V. Quine’s classification (Fung 2009: 164–188). Second, there is a confusion in Hu’s analysis that he does not distinguish between subsumption or implication, on the one hand, and identity or equivalence, on the other. As indicated in the quotation above, Hu Shih thinks that Later Mohists maintain a theory of classification: “In classification the individual is explained in the species, and the species in the genus.” According to Hu’s interpretation, for example, in the proper parallelism “A is B; CA is CB” (for example, “a white horse is a horse; to ride a white horse is to ride a horse.”), the relation between A and B and that between CA and CB can be understood as “to submerge the individual in the universal,” that is, subsumption. But, in his analysis of the counter-example of parallelism “A is B; CA is not CB” (“A thief is a man; to kill thieves is not to kill men.”), he treats the words “is not” as “is not the same as,” namely, “is not identical with” or “is not equivalent to.” Hu’s explanation is that: based on the subsumption between individual and universal, puppy is an instance or a part of dog and thief is an instance or a part of man, so it is acceptable to say “a puppy is a dog” and “a thief is a man.” But, based on the difference between individual and universal, puppy is small without long hair while dog is big with long hair and thieves are the man of stealing while men are not, so it is acceptable to say “to kill a puppy is not to kill a dog” and “to kill a thief is not to kill a man.” (Hu 1997: 174) Nevertheless, I think, there is no reason to treat the relation between A and B in the first sentence of the proper parallelism in “A is B; CA is CB” as subsumption but the relation between A and B in the first sentence of the counter-parallelism as difference in terms of non-­ duplicated identity or inequivalent. Moreover, although there are different versions of editing for the text it seems that the original text is that “it is not acceptable to say ‘to kill a puppy is not to kill a dog’.” rather than that “it is acceptable to say that ‘to kill a puppy is not to kill a dog’.” It is because the follow sentence of this statement is: “explanation is given under ‘duplicated [identity]’.” (shuo zai zhong 說在重[同]) (Canon B53) It means that it is not acceptable to say CA is not CB because A is identical with B. Besides, the Canon Explanation B53 says that: 狗, 犬也。[殺狗] 謂之殺犬, 可。若兩脾。

17  Logical Thinking in the Gongsun Longzi

311

A puppy is a dog; it is acceptable to say that [to kill a puppy is] killing a dog. Just like [the type-identity of] twin thigh. (My translation)

If the original sentence is “it is acceptable to say that ‘to kill a puppy is not to kill a dog’.” rather than “it is not acceptable to say that ‘to kill a puppy is not to kill a dog’.”, it would lead to a consequence that Canon Explanation B53 cannot match Canon B53. Third, it is a shame that Hu hasn’t given an explanation for the question why there are two kinds of parallelism – the proper parallelism and the count-­parallelism – in the Mohist text. For the Mohists, I think, the two tokens of “kill” in “to kill a thief” and “to kill a man” are not expressed as the same concept. Mozi has made a distinction to explain why killing innocent people or killing people through military attack is unjust or not righteous and killing a thief or executing a dictator is just or righteous in the following dialogue: (Wang and Wang 2006: 159–161 with my minor modification) Those warlike lords attempt to defend their views by reproaching Master Mozi: “You regard offensive warfare as unrighteous and unprofitable, but is it really so? In ancient times, King Yu 禹 attacked Youmiao 有苗, King Tang 湯 attacked Jie, and King Wu 武 attacked Zhou 紂, yet all of them were later honored as sage kings. How can you explain this?” Master Mozi said: “You neither clearly aware the class (lei 類) of what I said to refer to nor understand the reason (gu 故) behind it. What the sage kings did was not ‘to attack’ (gong 攻) but ‘to punish/execute’ (zhu 誅).”

Mozi thinks that no matter how serious the crime made by a thief or warlike lord they should be killed in terms of punishment or execution. But punishment or execution is not the same thing as murder or killing an innocent. So Later Mohists are consistent in claiming that “to kill a thief or dictator [by execution]” is not “to kill a man [by murder].” They are not two events of the same kind or class. So, here, the term “kill” is ambiguous. If there is an ambiguous term in a parallelism, it cannot be a proper instance of the parallel reasoning. Unlike the improper parallelism in the “thief” example, in the proper parallelism such as the “white horse” example, the conclusion “to ride a white horse is to ride a horse” can be validly derived from the premise “a white horse is a horse.” It is because there is no ambiguity in the parallel reasoning.1

1  Why the parallelism in “A thief is a man; to kill a thief is not to kill a man” cannot be subsumed under the same argument form (“A is B; CA is CB”) as that in “A white horse is a horse; to ride a white horse is to ride a horse”? Hansen’s answer is that: the consequent of the former example should be interpreted as conveying different i [yi 意] intentions and, thus, as used in an intensional or opaque context. He thinks this difference demonstrates that Later Mohists reject the inference rule. (Hansen 1983: 132–3 and 1992: 250–1) I don’t think Hansen is right to identify the sentence “A thief is a man” as an equation or identity sentence (Hansen 1983: 187 and 1992: 256) and to treat the former example as a counterexample of the parallel reasoning. I think there is no real problem about the failure of substitution between “to kill a thief” and “to killing a man” in the socalled intensional context and no real problem of referential opacity. Because the relevant sentence is not an equation. Besides, as we know, a sentence puts a predicate or other term in an intensional context if any other predicate or term with the same extension cannot be substituted without the possibility of changing the truth value. However, a sentence used in an (s-) intensional context is

312

Y.-m. Fung

After years of studies in intellectual history and philology about Gongsun Long’s Zhiwu Lun 指物論 (Discourse on Zhiwu), Hu claims that: (Hu 1922: 126) I have concluded that the word “finger” here means “a mark” or “a sign,” that which “signifies.” … It seems that by “mark” or “sign” is here meant the attribute or quality by which a thing is known. “There are no things which are not marks” [wu mo fei zhi 物莫非指] means that things are what their attributes indicate them to be; that is, what they are perceived to be. This subjectivism is immediately qualified by the realistic statement, “But marks are no marks” [er zhi fei zhi 而指非指]; that is, marks are not entities in themselves but marks of things. “For without things, can there be marks?” [tianxia wu wu, ke wei zhi fu? 天下無物/ 可謂指乎]

Taking the word “zhi” 指 as “fingers” to mean “marks” or “attributes of things,” Hu believes that some of the so-called paradoxes assigned to the Gonsun Longzi can be explained in a sensible way. For example, he says that: “We know a horse by its horse-ness [not interpreted as a Platonic universal but as a perceived attribute], a white horse by its white-horse-ness [also not interpreted as a Platonic Idea], and a white stone by its white-ness and solidity.” Based on this general thesis, the statement “white horse is not horse” (baima fei ma 白馬非馬) can be interpreted by Hu as follows: Since the term “white horse” signifies the attribute of white horse, the term “horse” signifies the attribute of horse, and the former attribute is not identical with the latter attribute, therefore we can assert that “white horse is not horse.” (Hu 1922: 126). Hu’s conceptions of “horse-ness” and “white-ness” are not Platonic Ideas. According to his conception of “biaode” 表德 (attribute or external property), the relation between name and reality can be understood as that between description and the property described. Put in the words of contemporary philosophy of language, it is an indirect relation of reference in the sense that a ming 名 (name) used as a description can fix or identify its referent through its descriptional property, attribute, or condition. So it cannot directly refer to the object which has the property or attribute. For example, Hu mentions that: (Hu 1997: 178). 有了「物指」, 然後有「名」。一物的名乃是代表這物一切物指的符號。如「火」 代表火的一切性質, 「梅蘭芳」代表梅蘭芳的一切狀態性質。 After obtaining ‘wuzhi’ [物指 attribute of thing], we can have ‘ming.’ The name of an object is the symbol which represents all the attributes of the object, such as ‘fire’ representing all the attributes of fire, ‘Mei Lanfong’ [梅蘭芳] representing all the states and attributes of Mei Lanfong. (My translation)

This seems to be an application of the descriptional theory of reference or the indirect theory of reference originated from Gottlob Frege, Bertrand Russell, and Leonard Linski. According to this theory, a term’s meaning can be divided into two parts: its reference (or referent) and its sense (or the Fregean Sinn), and the sense of a term can be used as a criterion to identify its reference. For Russell and also for not necessarily related to (t-) intention. As indicated by John Searle, “These are often confused, even by professional philosophers.” (Searle 2004: 174–5) So the above “thief” example is not a real counterexample of the parallel reasoning. My argument against Hansen’s view can be found in (Fung 2012: § II: 334–41).

17  Logical Thinking in the Gongsun Longzi

313

Hu, a proper name is nothing but an abbreviation of a series of descriptions (i.e. a conjunction or disjunction of descriptions). In this sense, each referring term can be used to refer to their referent only indirectly through its descriptional sense; it is quite different from John Mill’s idea that a proper name has no sense but refers to its referent directly if its referent is not empty. For example, the proper name “Plato” can be understood by a theorist of indirect reference as an abbreviation of a series of descriptions such as “Socrates’s student,” “Aristotle’s teacher,” and so on. If an object or person satisfies the attributes described by a series of descriptions, based on this criterion we can identify the object or person as the referent of the name.2 Let’s put aside the problem whether the descriptional theory is right or not. Even though the theory is right, Hu Shih’s (unconscious) application of the theory to the interpretation of the text still has serious problems. With respect to the issue in Baima Lun 白馬論 (Discourse on White Horse), according to Hu’s interpretation, “white horse” represents two attributes while “horse” represents only one attribute, we can arrive at the conclusion as claimed by Gongsun Long that “white horse is not horse.” Although an object with two attributes is not identical with an object with only one attribute, it is obvious that having two attributes can be said having one of the two attributes. However, we know from the text that Gongsun Long does assert that “having white horse cannot be called having horse” (you bai ma bu ke wei wu ma 有白馬不可謂有馬), because, as indicated by the text, horse can be separated from white as an independent entity. It is what Gongsun Long says about the mere horse without selecting or excluding any color (wu qu qu yu se zhi ma 無去取於色 之馬) and the mere white without fixing into any things (bu ding suo bai 不定所 白). We know that Jianbai Lun 堅白論 (Discourse on Hard and White) also has the same idea of “non-fixing” for “white” and for “hard;” it says that these non-fixing entities can be separated and self-hided (li er zi cang 離而自藏) from the sensible world (tianxia 天下). Moreover, in Tongbian Lun 通變論 (Discourse on Understanding Change), if we follow Hu’s idea to treat the word “one” as designating only one attribute owned by an object and the word “two” as designating two attributes owned by the same object or another object, it would be impossible for us to assert that “two does not have one” (er wu yi 二無一). However, if we treat “one” as a non-sensible simple and “two” as a sensible compound which is combined from two separated one and thus changed into a fixing whole in the sense that “the right [one] with a compound partner (the left one) can be called ‘changed’ ” (you you yu ke wei bian 右有與可謂 變), then the above sentence would be intelligible. The simple (one) is separated from the compound (two) in the sense that it is without any partner to constitute a compound, and is thus not yet to be changed into an object in the sensible world. In other words, this non-fixing simple is unchanged. When it is combined with another one in the sense that it has a partner to form a compound object in the sensible world, it is changed. Since “the changed is not the unchanged” (bian fei bu bian 變

2  The theoretical characteristics of the descriptional theory of reference and those of the direct theory of reference are discussed in detail in (Salmon 1982: 9–22).

314

Y.-m. Fung

非不變), we can follow Gongsun Long to say that “the left [one] combined with the right [one] can be called two” (zuo yu you ke wei er 左與右可謂二) and at the same time to say that “two does not have one” (er wu yi 二無一), “two does not have left” (er wu zuo 二無左), and “two does not have right” (er wu you 二無右). It is obvious that this coherent picture inside or outside Baima Lun cannot be drawn by Hu Shih. In other words, he cannot give an interpretation for the text in a sensible way.

2  Fung Yulan’s Interpretation and Analysis In contrast to Hu Shih’s descriptional interpretation of names in the Gongsun Longzi, Fung Yulan’s 馮友蘭 interpretation is Platonic. He treats terms such as “horse” and “white horse” as names referring to different (Platonic) universals. With respect to the seeming paradox “white horse is not horse,” he interprets the statement as a rejection of the identity of two different universals or the identity of their names. According to his explanation, Gongsun Long provides three kinds of argument to demonstrate the truth of the statement: (Fung 1948: 87–8) (I) Difference in the Intension of Names: The word “horse” denotes a shape and the word “white” denotes a color. That which denotes a shape is not that which denotes a color. Therefore, a white horse is not a horse. Since the intension of each of the three terms, namely, a shape, a color and a color cum shape, are different, it can concludes that a white horse is not a horse. (II) Difference in the Extension of Names: In searching a horse, a yellow horse or a black horse may be brought forward; but in searching a white horse, a yellow horse or a black horse cannot be brought forward. So, if a yellow horse and a black horse can only respond to a call for a horse but not for a white horse, it implies that a white horse is not a horse. In terms of Western Logic, the argument can be understood as about the difference between the extension of the name “white horse” and the extension of the name “horse.” (III) Difference between Names of Universals: A horse without color can be understood as the horse as such, that is a universal or Idea in the Platonic sense. A white horse is horse together with white and thus is not a horse as such. So, the distinction is about the difference about two kinds of universal: one is “horseness” and the other is “white-horse-ness.” Although Fung’s explanation is much more sophisticated than those provided by his contemporaries, he didn’t aware that the last argument and the first two arguments presuppose different ontological commitments which cannot be coherently maintained. As we know, a Platonic universal is an abstract entity or non-physical individual. Platonic entity in the intellectual or noumenal world is transcendent of a particular thing/object in the physical or phenomenal world. So, physical t­hings/ objects as particulars are distinct from Ideas as universals in the sense that the former is about concrete individuals while the latter is about abstract individuals. Whiteness as an entity of universal referred by the individual term “white” is with-

17  Logical Thinking in the Gongsun Longzi

315

out color; but white as a color referred by the general term “white” includes all white objects in its extension, including white horses and white stone. It seems that Fung has never noticed that there is a serious problem in his explanation. In addition to Baima Lun, Jianbai Lun is also explained by Fung in terms of Platonic Ideas. He thinks that there are mainly two kinds of arguments in Gongsun Long’s view: one is epistemological, the other is metaphysical. He writes: (Fung 1948: 89) This dialogue uses epistemological proof to show that hardness and whiteness are separated from each other. Here we have a hard and white stone. If we use our eye to see it, we only get what is white, i.e., a white stone. But if we use our hands to touch it, we only get what is hard, i.e., a hard stone. While we are sensing that the stone is white, we cannot sense that it is hard, and while we are sensing that it is hard, we cannot sense that it is white. Epistemologically speaking, therefore, there is only a white stone or a hard stone here, but not a hard and white stone.

That is an epistemological argument. The metaphysical one is: (Fung 1948: 89) Its general idea is that both hardness and whiteness, as universals, are unspecified in regard to what particular object it is that is hard or that is white. They can be manifested in any or all white or hard objects. Indeed, even if in the physical world there were no hard or white objects at all, none the less, the universal, hardness, would of necessity remain hardness, and the universal, whiteness, would remain whiteness. Such hardness and whiteness are quite independent of the existence of physical stones or other objects that are hard and white. The fact that they are independent universals is shown by the fact that in the physical world there are some objects that are hard but not white, and other objects that are white but not hard. Thus it is evident that hardness and whiteness are separate from each other.

I think Fung’s explanation of these two arguments are also incoherent. The metaphysical argument presupposes that whiteness and hardness as universals cannot be known by perception. But the epistemological argument presupposes that both white and hard can be sensed though the former can only be seen by eyes while the latter perceived by hands. Let alone the problem whether the existence of external objects can only be claimed in terms of being perceived by a single sensory organ and thus all physical objects can only be defined in terms of phenomenalism. It is obvious that we cannot use “white” and “hard” as terms referring to both universals in terms of Platonism and to objects in terms of phenomenalism. Unlike Hu Shih’s thesis of “attribute” which is consistent with nominalism, Fung Yulan’s thesis of “universal” has to presuppose two levels of existence or two different but related worlds. Similar to Plato’s two-world theory, he also makes a distinction between the world of particulars and the world of universals. He thinks that: (Fung 1961: 257) On the one hand, the referent of a name is an individual, this is the so-called ‘a name is denoting reality.’ On the other hand, the referent of a name is a universal. For example, in addition to this [particular] horse and that [particular] horse, there is the horse[ness] in the sense of ‘having horse as such;’ in addition to this [particular] white thing and that [particular] white thing, there is the white[ness] in the sense of ‘the white as non-fixing white.’ These ‘horse[ness]’ and ‘white[ness]’ are the so-called ‘universals’ or ‘elements’ in modern [Western] philosophy. These are also the referents of names. Gongsun Long makes a distinction of zhi 指 and wu 物, his zhi can be recognized as the universal referred by the name.”

316

Y.-m. Fung

It is obvious that Fung’s interpretation is based on Plato’s realism. According to Fung’s view, Gongsun Long’s “wu” 物 designates all the concrete individual things in the sensible world, his “zhi” means all the abstract universals which are ontologically independent of the objects in the sensible world, and his “wuzhi” 物指 is used to characterize all the concrete properties of sensible objects which are emerged from the zhi’s embodying or fixing into particular things. Based on this triple distinction, Fung is able to explain what Hu cannot explain. For example, Fung can explain why Gongsun Long says that “having white horse cannot be called having horse” in the sense that having a particular white horse (or having the universal white-horse-ness) is not having the universal horseness. In other words, Gongsun Long’s specific thesis “white horse is not horse” can be understood as an example of the general thesis “wu is not zhi,” “the fixing white (which is fixing into a concrete object such as white horse or white stone) is not the non-fixing white” or as an example of “[wu-]zhi is not [du-]zhi” 獨指 (single, simple, and separate universal). Unfortunately, Fung does not use this triple distinction to interpret the whole book of Gongsun Longzi in a consistent way. He considers all the referring terms in the book, no matter it is a simple or a compound term, can be used to refer to universals. For example, he asserts that, “the referents of ‘horse,’ ‘white,’ and ‘white horse’ are what are called by ‘zhi’ in Zhiwu Lun of the Gongsun Longzi.” (Fung 1961: 258) They are all considered by Fung as independently separate universals. However, if both white horse and horse are universals, we would be unable to use the distinction between wu and zhi to interpret the difference between these two universals. Furthermore, if Fung accepts that “white horse” represents something of the characteristic of “an object of fixing into white,” it is definitely not something of the characteristic of non-fixing. Or, if he accepts that “white horse” refers to something which selects some color (i.e. white), it is clearly not something which does not select any color. Put the words more generally, if white horse is a fixing object, and fixing is not non-fixing, then white horse is not non-fixing. Since non-fixing is universal, we can say that white horse in this sense is not universal. Based on Gongsun Long’s acceptance of his opponent’s assertion that “there is a white horse in tianxia 天下 (i.e., the sensible world)” and his own assertion that “zhi is not something in tianxia; but wu is something in tianxia” (zhi ye zhe tianxia zhi suo wu ye, wu ye zhe tian xia zhi suo you ye. 指也者天下之所無也, 物也者天下之所有 也。), it is obvious that his “white horse” is used as an individual name like “this [white horse]” or “that [white horse]” which refers to something in the sensible world, while horse[ness] as zhi is not a member of things in the sensible world but something separated and self-hided from the sensible world. Besides, Fung’s interpretation is also handicap in dealing with Tongbian Lun. He interprets “two” as referring to a universal which is different from the universal referred by “one.” But this cannot help in explaining both “two has no left/right [one]” and “two is combined from left [one] and right [one]” (er zhe zuo yu you 二 者左與右) consistently. Because what is referred by “two” is “changed,” and what is named by “one” is “unchanged.” Gongsun Long’s thesis of “the changed is not the unchanged” (bian fei bu bian 變非不變) is inevitably committed to two ontological levels for the different residences of the changed and the unchanged. If I am

17  Logical Thinking in the Gongsun Longzi

317

right at this point, to interpret “two” as universal is inaccurate. The accurate interpretation is to understand the term “two” as referring to something combined from fixing two independent and separated zhi (one) into a compound wu. According to this interpretation, the expressions “two has no left/right [one],” and “two is combined from left [one] and right [one]” can be interpreted as that “the compound wu has no simple and unchanged zhi in it, though it is emerged from two unchanged zhis in terms of their fixing or combining into something whose ontological status is changeable.”3 In comparison with this interpretation, I conclude that the shortcomings of Fung’s interpretation are coming from his peculiar realism which does not consistently use the distinction between particulars and universals in his interpretation.

3  Janusz Chmielewski’s Interpretation and Analysis Chmielewski may be the first sinologist who uses symbolic logic to analyze the logical thinking in ancient China, especially that in the Gongsun Longzi. He is not satisfied with Zhang Binglin’s 章炳麟 syllogistic analysis and H. Greniewski and O. Wojtasiewicz’s rational construction in terms of the algebra of sets. Instead, in dealing with the logical thought in ancient China, his proposed approach is sinological and his analysis is logical. He says: (Chmielewski 2009: 176) The task I have set myself may be briefly stated in the following terms: Without losing sight of philological and historical background (which I believe, is always the necessary prerequistite in sinological research) I propose to single out some more or less typical froms of reasoning (whether already interpreted by others, or not) occurring in early Chinese philosophers; to define them from the standpoint and in terms of elementary symbolic logic; to find out general logical laws and notions underlining them; and, as far as possible, to compare them with ancient logical theory of the West.

Joseph Needham follows Kou Pao-koh’s view and states that syllogistic reasoning is not infrequently implicit in ancient Chinese texts and its complete form can be found in the Gongsun Longzi. (Needham 1956: 200) However, Chmielewski thinks that both writers have been mistaken. For example, he does not agree with Kou’s elaboration of the following argument in Baima Lun: 1. 馬者無去取於色 (Horse has not rejecting-selecting colour.) 2. 故黃黑[馬]皆所以應 (Hence either yellow or black [horse] can be used to respond the request.) 3. 白馬者有去取於色 3  Here, “one” and “two” are used as symbols to signify a name of universal or single (non-fixing) entity and a name of particular or compound (fixing) entity, respectively. For example, “white” and “horse” are used by Gongsun Long as single names for different single (non-fixing) entities while “white-horse” and “white-stone” are used as compound names for different compound (fixing) entities.

318

Y.-m. Fung

(White horse has rejecting-selecting colour.) 4. 黃黑馬皆所以色去 (Both yellow horse and black horse are excluded on the basis of colour.) 5. 故唯白馬獨可以應耳 (Therefore only white horse can be used to respond the request.) 6. 無去[取於色]者非有去[取於色]也 ([The property of] what has not rejecting-selecting colour is not [the property of] what has rejecting-selecting colour.) 7. 故曰白馬非馬 (Therefore: White horse is not horse.) He thinks that Kou leaves out the sentences (2) (“hence either yellow or black [horse] can be used to respond the request.”), (4) (“Both yellow horse and black horse are excluded on the basis of colour.”), (5) (“Therefore only white horse can be used to respond the request.”) and (6) (“[The property of] what has not rejecting-­ selecting colour is not [the property of] what has rejecting-selecting colour.”), which he considers as “explanatory elements,” picks up the sentences (1), (3) and (7), and obtains thus a whole composed of three members to form an Aristotelian syllogism. He argues that it is not the Aristotelian syllogism in its right form. One of the reasons is that the relation of connection in each sentence of a syllogism is implication (⊃/→), not identity (=). (Chmielewski 2009: 177) So, he thinks that, instead of forcibly pressing Gongsun Long’s reasoning into a syllogistic from, we had better analyze it in a way having nothing to do with the syllogism, either Aristotelian or traditional. He says: (Chmielewski 2009: 178) Leaving out the explanatory sentences (2), (4) and (5), we obtain a complex composed of four propositions, namely (the translation deliberately deviated from the normal English usage): (1) Horse has not rejecting-selecting colour; (3) White horse has rejecting-selecting colour; (6) [The property of] what has not rejecting-selecting colour is not [the property of] what has rejecting-selecting colour; (7) … White horse is not horse.

For Chmielewski, this can be best conceived as a reasoning on classes or sets as follows: (Chmielewski 2009: 178–9)4 (1) Φ′ A (2) Φ B (6) (X̂ )Φ′ X . (X̂ )Φ X = 0 (7) A ≠ B Here, A and B stand for “class ‘horse’” and “class ‘white horse’,” respectively; “′” is the sign of negation – in the present case referring to the function Φ; Φ′ means non-Φ, i.e., “has not rejecting-selecting colour.” The formula X . Y = 0 means that 4  To avoid confusion, I suggest to formulate (6) as (X̂ )Φ′ X . (Ŷ)Φ Y = 0 rather than (X̂ )Φ′ X . (X̂ )Φ X = 0.

17  Logical Thinking in the Gongsun Longzi

319

the product of classes X and Y is an empty class, i.e., it states the exclusion of classes X and Y; “≠” is the sign of non-identity. Chmielewski claims that there is sufficient philological evidence to support that Gongsun Long firmly believed in the validity of the statement “white horse is not horse” and did not consider it as a mere paradox. This means that (7) can be conceived only as stating the non-identity of the classes “horse” and “white horse” – just as is indicated by the formula A ≠ B. (Chmielewski 2009: 179). Another example of reasoning in Baima Lun analyzed by Chmielewski is: (1) (2) (3) (4)

馬者所以命形也 白者所以命色也 命色者非命形也 故曰白馬非馬

Its translation is: ( 1) Horse is what commands shape [and only shape]; (2) White is what commands colour [and only colour]; (3) What commands colour [and only colour] is not what commands shape [and only shape]; (4) White horse is not horse. In contrast to A. C. Graham, who regards this argument as faulty, Chmielewski argues that it is a valid argument. Its argument form is: (Chmielewski 2009: 181).5 (1) (2) (3) (4)

ΦA ΨB (X̂ )Ψ X . (X̂ )Φ X = 0 B . A ≠ B

Chmielewski interprets Gongsun Long’s “horse” and “white,” not as terms either referring to abstract entities (i.e. horseness and whiteness) or referring to individual objects within the scope that the concept characterizes, but as names of class or set. He thinks that, from the premises “horse is what commands shape [and only shape]” (1), “white is what commands colour [and only colour]” (2), and “what commands colour [and only colour] is not what commands shape [and only shape]” (3), the conclusion “white horse is not horse” (4) cannot be drawn as commonly thought. However, he continues, if we consider Gongsun Long’s argument as based on a peculiar kind of theory of classes, “narrow and incomplete” as not admitting class inclusion (⊂) and its negation (⊄), we can infer the conclusion “the class of white objects [including white horses] is not identical with the class of horses.” Since (2) “what commands colour [and only colour] is not what commands shape [and only shape]” indicates that the intersection (∩) is null or empty, and white horse is an instance of this kind of intersection, we can validly derive the conclusion “white

 Again, to avoid confusion, I suggest to formulate (3) as (X̂ )Ψ X . (Ŷ)Φ Y = 0 rather than (X̂ )Ψ X . (X̂ )Φ X = 0.

5

320

Y.-m. Fung

horse is not horse.” (From (B ≠ A) we can have (B . A) ≠ (A . A) and (B . A) ≠ A) if not admitting (A  ⊂  B).) From this and from another Gongsun Long’s argument, Chmielewski argues that, in the conclusion “white horse is not horse,” Gongsun Long uses the negative copula fei 非 (not/non) solely to deny identity, and is misunderstood by the attacker who supposes him to be denying class inclusion. I think one of the reasons why Chmielewski’s “Gongsun Long” regards “white horse” as a name of empty class is that: if the class is not empty, Gongsun Long would have to accept that the class of white horse is included in the class of horse, i.e. the former is a sub-set of the latter. According to Chmielewski’s view, however, Gongsun Long accepts only the relation of non-identity but not that of class inclusion between them. (Chmielewski 2009: 183–5) If we agree that white horse is an empty class and horse is not, the conclusion “white horse is not horse” would be naturally obtained. Although Chmielewski’s interpretation seems to be coherent in some sense, it cannot escape from the awkwardness or oddity in interpreting Gongsun Long’s text. According to the naïve set theory, for example, an empty set can be regarded as a sub-set of any set; it is no evidence to assert that Gongsun Long does not accept the class inclusion between the white horse and the horse because the former is null or empty. To understand Gongsun Long’s thought as rejecting class inclusion and as asserting the empty status of the class of white horse is not in accordance with the text. It is obvious that one of the plausible interpretations of Gongsun Long’s expression, say, “someone who searches horse will be just as satisfied with yellow horse or black horse” (求馬黃黑馬皆可致), is that the class of yellow horses or the class of black horses is included in the class of horses. “White horse” is not a name of null or empty class, because Gongsun Long agrees with his opponent that “there is white horse” (you baima 有白馬). He says: “Certainly horse has color, which is why one has white horse.” If white horse is an empty class, so is yellow horse and black horse, how can Gongsun Long assert the difference between white horse and yellow horse through the question “If you deem having white horse having horse, is it admissible to say that having white horse is to be deemed having yellow horse?” In other words, if both white horse and yellow horse are empty classes, they cannot be different from each other. It is clear from the text that white horse is not an empty class and Gongsun Long does not reject the possibility of having the relation of class inclusion between white horse and horse if the relevant terms are really used as class names. Furthermore, most importantly, it is evident from the text that Gongsun Long makes a distinction between the fixing white and the non-fixing white, and their names can both be simplified as “white.” If we use the term “white” as a class name, their difference could not be identified unless we provide two different class names for these two kinds of white. In regard to the second argument of Gongsun Long mentioned above, Chmielewski’s translation deviates from the common one. According to J.  W. Hearne’s observation, Chmielewski’s interpretation is quite different from Graham’s. I think at least three points deserve our attention: (1) In the first round debate, he translates the term “ming” 命 in “ming xing” 命形 and “ming se” 命色 not as “ming ming” 命名 (to name), but as “ming ling” 命令 (to command). (2) He

17  Logical Thinking in the Gongsun Longzi

321

does not treat “bai” 白 (white) and “ma” 馬 (horse) as names mentioned or as names used to refer to itself, but as names used to refer to the class of white objects and to refer to the class of horses respectively. (3) He considers Gongsun Long’s reasoning as based on a class theory in a special narrow sense: a theory which stresses the relation of non-identity and excludes the relation of class inclusion. (Hearne 1980: 36) In this regard, to treat “ming” not as “to name,” but as “to command,” seems to be able for Chmielewski to explain away the complicate problem of interpretation in meta-language, but his translation still makes no sense though “ming” used as “to command” could be understood as a metaphor. Furthermore, to add the extra words “and only shape” and “and only colour” into the text is not an acceptable reading; and to treat the original sentence or the common reading of the sentence “to name the color is not to name the shape” (ming se zhe fei ming xing ye 命色者非命形也) as an empty intersection is also not a reasonable interpretation. This interpretation for (3) cannot put the word “fei” (not or non-) into a readable place in classical Chinese as meaning the same as the word “fei” in (4). I think to interpret the so-called special set-theory of Gongsun Long as not allowing both inclusion (A ⊂ B) and its negation (A ⊄ B) is unintelligible. For example, if there are no class inclusion and its negation in Baima Lun, it is not only that the intersection of the class of white and the class of horse is null but also that the intersection of the class of white and the class of white is empty. In standard set-theory, we can say that the class of white is a (non-proper) sub-class of itself. They are overlapping in the sense that they share the same members. Return to the two arguments discussed above. Let’s see (6) in the first argument and (3) in the second one: (6) (X̂ )Φ′ X . (X̂ )Φ X = 0 (in the first argument) (3) (X̂ )Ψ X . (X̂ )Φ X = 0 (in the second argument) I think both sentences can be expressed in predicate logic with the following general form6:

 x   y  ~  Fx  Gy    x   y  ~  Gy  Fx 

or

 x   y   Fx  ~ Gy    x   y   Gy  ~ Fx 



If we use the predicates “F” and “G” to define two sets respectively, the above formula means that these two sets do not share the same members. It also means that the set defined by “Fx” is not included in the set defined by “Gy” and vice versa. So, at least implicitly, Gongsun Long has to admit the negation of inclusion in his thought.

 It can be simplified as:

6

 x   y  ~  Fx  Gy   ~  Gy  Fx  or  x   y   Fx  ~ Gy    Gy  ~ Fx  .

322

Y.-m. Fung

The only evidence provided by Chmielewski to support his set-theoretical analysis for Baima Lun is his interpretation of the first two sentences in Zhiwu Lun. He claims that the word “zhi” 指 means “class” and the original parallel sentences “wu mo fei zhi, er zhi fei zhi.” (物莫非指, 而指非指。) can be translated as “No thing is without class, but classes themselves are without classes.” (Chmielewski 2009: 187) I think the cost of accepting this translation is to sacrifice a coherent interpretation of the whole essay of Zhiwu Lun. It is also not consistent with the ideas expressed in Tongbian Lun and Jianbai Lun. In Jianbai Lun, Gongsun Long says that we have two by seeing a white stone or by touching a hard stone, respectively. That is: “white as a colour and stone as a shape can be seen” or “hard as a texture and stone as a shape can be touched.” In other words, the compound of white stone and that of hard stone are not concepts without extension or not null classes. In Tongbian Lun, why “two does not have (left or right) one,” but “left (one) together with right (one) can be called ‘two’?” Here, “two” means entity referred by compound term such as “white horse” and “hard stone.” “One” means entity referred by single term such as “white” and “horse.” If we adopt a set-theoretical interpretation, I think we cannot make any coherent interpretation for this ontological statement. Moreover, in both Baima Lun and Jianbai Lun, there is a concept of “ding suo bai” 定所白 (fixing white) in contrast with “bu ding suo bai” 不定所白 (not fixing white). The former, for example, is the white fixing in a [white] horse, while the latter the white without joining with another entity. Both can be used with the same type of word “white.” But how can we distinguish these two tokens of the same word in terms of class or set? So, I think that the set-theoretical interpretation is not an option for understanding all the essays in the Gongsun Longzi.

4  A New Logico-Philosophical Interpretation Chmielewski is right to say that the word “fei” in “baima fei ma” (“white-horse not horse” or “a white horse is not a horse”) means not-identity. But I think the two arguments mentioned by him can be elaborated in first-order predicate logic with a much more coherent interpretation and reasonable analysis for the text. The first argument can be formulated as follows: (Let’s use “a” for “horse,” “b” for “white-horse,” and “S” for “color-selection.”) (1) ~Sa (2) Sb (6)  x   y   ~ Sx  Sy   ~  x  y   (7) ∴ a ≠ b This is a full-fledged argument that is very easy to proof its validity: (7) can be derived from (1), (2) and (6). To elaborate the second argument, let’s use “a1,” “a2” and “a3” to stand for the name “horse” mentioned, for the name “white” mentioned, and for the name “white-­ horse” mentioned, respectively, use “b1,” “b2” and “b3” to represent “the entity

17  Logical Thinking in the Gongsun Longzi

323

shape” (or “the horse-shape”), “the entity color” (or “the white-color”), and “the entity of both shape and color,” respectively, and use “R” to represent the relation “to name.” The argument form can be elaborated as below: (i) (a1Rb1) (ii) (a2Rb2) (iii)

 x   y   xRb2    yRb1   ~  x  y 

∴ (iv) ~(a3 = a1) According to this formulation, (iii) says that, for all names used to refer to color and for all the other names used to refer to shape, they are not identical. Based on these three premises, we cannot validly derive the conclusion “a white horse is not a horse.” However, if we accept (i) and (ii), it is quite natural to think that Gongsun Long implicitly takes for granted (in background knowledge) that “the name ‘white-­ horse’ is used to name the entity of both color and shape (or the entity of both the white-color and the horse-shape)” (baima ji ming se yi mingxing 白馬既命色亦命 形). For the reason of economy of writing and printing in the ancient time of China, if something is understood, it is no need to say it explicitly. To make it explicit, we can derive the conclusion from the premise (i) together with the following implicit premises which are consistent with and shown by the three explicit premises. These implicit sentences are: (iia) (a3Rb3) and





(iiia)  x   y   xRb3  •  yRb1    ~  x  y 

The conclusion “a white horse is not a horse” can be validly deduced from (i), (iia), and (iiia). The function of (ii) is to reflect the implicit premise (iia) through its comparison with (i), and the explicit sentence (iii) is significant in suggesting the implicit sentence (iiia). If the above analysis is right, I think it is not necessary to interpret the argument as based on a class theory in a special narrow and peculiar sense. One may be skeptical of the fact that there is any logical reasoning in the Gongsun Longzi. One of the skeptical examples provided by Chad Hansen is an over-­ simplified interpretation of the original argument as follows7: ( 1) All yellow horses are horses. [A is B] (2) No yellow horse is a white horse. [A is ~C] (3) Therefore, a white horse is not a horse. [B is ~C] This argument is clearly invalid according to Western logic. Hansen thinks, “No culture that routinely accepted such inferences could have had sagacity enough to rule that empire for two thousand years!” (Hansen 1983: 161) So, he claims, this 7  According to the text, if we follow Hansen’s formulation, (3) should be formulated as “C is ~B,” not “B is ~C” as made by Hansen.

324

Y.-m. Fung

paragraph should be understood as discriminating the boundaries of the substances or stuffs referred by names, instead of representing the individual objects in the world. In this regard, he thinks, it is not appropriate to use Western logic to interpret the argument’s form. However, his accusation of inappropriateness is unfortunately based on his inappropriate treatment of the logical form of the argument. Let’s put aside his inappropriate treatment of the original statements which include quantification as statements without quantifiers, Hansen’s translation together with his formulation is still not acceptable.8 If we look more carefully into the inner structure of Gongsun Long’s sentences, we can know that the argumentative structure is an application of Modus Tollens including an application of the Leibniz’s Law which is not the ridiculous form as elaborated above by Hansen. The argument embedded in the paragraph can be naturally elaborated as follows: (1) If white-horse is horse, it would be no difference between yellow-horse (or black-horse) in response to searching white-horse and searching horse. (2) There is a difference. (3) Therefore, white-horse is not horse. Its argument form is: (1’) [(a = b) → (Fa↔Fb)]

8  Hansen treats all sentences of the form “XY yeh [ye 也]” as equations. He says, “I use equation as a grammatical form. But it could be argued that sentences of this type are indeed equations and not classic subject-predicate assertions. The second term would be regarded as implicitly quantified existentially so that the logical form would be ‘horse is identical with some animal.’ Or, to capture the part-whole structure of the mass-stuff ontology, ‘horse is identical with part of animal.’ I suspect that, contrary to Frege, English nominal predications are also equations.” (Hansen 1983: 187). In dealing with Gongsun Long’s argument of “searching horse” (qiu ma 求馬), Hansen also treats the sentences of the argument as equations. He regards the formal structure of the argument (∗ A is B; A is not-C; therefore B is not-C) (∗ Yellow horse is [identical with part of] horse; yellow horse is not [identical with part of] white horse; therefore white horse is not [identical with part of] horse.) as “a simple and common deductive fallacy.” (Hansen 1983: 161) I think this treatment is not right. Let’s put aside his inaccurate translation, Hansen’s treatment is certainly inconsistent with other places in the text. For example, Gongsun Long asserts that, “if white horse is horse, the search for both should be the same.” If we follow Hansen to treat the antecedent as “∗ white horse = (part of) horse,” the consequent cannot be obtained. But if we treat the antecedent as “white horse is identical with horse,” the consequent can be maintained with a good sense. In other words, if we treat the sentence “baima nai ma ye” (白馬乃馬也) as an identity statement (not as “∗ white horse = (part of) horse,” but as “white horse is identical with horse”) and treat the sentence about “qiu ma” as predicating expressions, we can interpret it as a sound argument with an application of Leibniz’s law. (求馬, 黃、黑馬皆可致。求白馬, 黃、黑馬不可致。使白馬乃馬也, 是所求一 也, 所求一者, 白者不異馬也。) Besides, if we follow Hansen to treat the sentence “someone who searches horse will be just as satisfied with yellow-horse (or black-horse)” as the equation of terms – “yellow horse (or black horse) is [identical with part of] horse,” it would have the following troubles: (1) If Gongsun Long uses the so-called “one-name-one-thing” principle as assigned to him by Hansen, he should not assume this equation. (2) If Gongsun Long treats yellow horse, white horse, black horse, etc., like ox-horse, as mass-sum rather than mass product, as interpreted by Hansen, he has no point to conclude that “therefore white horse is not [identical with part of] horse.”

17  Logical Thinking in the Gongsun Longzi

325

(2’) ~(Fa↔Fb) ∴(3’) ~(a = b) This argument is not only valid, but also sound if the two sides of the debate agree that (2) is empirically true. This is a pretty good argument which presupposes Leibniz’s law in the reasoning.9 With respect to the problem of reference in the Gongsun Longzi, I do not treat all terms (simple and compound) as referring to abstract entities though I agree that there are abstract entities in Gongsun Long’s philosophy. My arguments and philosophical explanations are different from Fung Yulan’s. I think only single or simple terms, such as “white” (bai 白), “horse” (ma 馬), “hard” (jian 堅), and “stone” (shi 石), are used by Gongsun Long to refer to abstract entities or concrete universals in a broad sense. This is quite different from the view of Fung Yulan who treats all the terms in Baima Lun as names referring to Platonic universals. I regard the compound (referring) terms, such as “white horse”(baima 白馬), “yellow horse” (huangma 黃馬), and “hard stone” (jianshi 堅石), used by Gongsun Long as implicit form of expressions which have the same function as the phrases explicitly expressed in English denoting definite or indefinite individual objects. In ancient Chinese language, for example, the term “baima” (white horse) can be used as having the same function as “[a particular] white horse,” the term “huangma” (yellow horse) as “[this/that] yellow horse,” and the term “jianshi” (hard stone) as “[some] hard stone” in different contexts. These compound terms function like terms with demonstratives which directly refer to concrete objects in the world. If I am right, both single terms and compound terms used as referring terms by Gongsun Long can be recognized as Saul Kripke’s rigid designators, (Kripke 1980: 3–15) they should not be understood as descriptions and their logical forms should be formulated as individual terms rather than predicates. As I argue elsewhere,10 if all the arguments in Baima Lun formulated in descriptional form with predicate construction for the single and compound terms used by Gongsun Long, it cannot provide a coherent interpretation for the whole dialogue. But, if we interpret the debate between Gongsun Long and his opponent as the conflict between the use of language based on a direct theory of reference and that based on a descriptional theory of reference, the arguments provided by both sides will be understandable and can be judged as reasonable. According to my interpretation, the opponent is also reasonable to say that, “white horse is horse” because it is an analytic truth of the form as follows:

 x   Wx & Hx   Hx 



In other words, “white” and “horse” in the opponent’s statement should be represented by the predicates “W” and “H” respectively. However, if we formulate Gongsun Long’s key words in his statement “white-horse is not horse” by the same

 Here, premise (1’) is an instance of Leibniz’s Law: (∀x)(∀y)[(x = y) → (Fx↔Fy)].  Yiu-ming Fung, Language, Truth and Logic in Ancient China, chapter seven (forthcoming).

9

10

326

Y.-m. Fung

predicates, his statement would become wrong, i.e., analytically false. Based on the direct theory of reference, I think, Gongsun Long basically uses a referring term to refer to the same referent in all possible worlds if it does exist in each world, that is, to use a referring term as a rigid designator. A simple term “white” which refers to a non-fixing (bu ding 不定) entity, for example, can be understood as referring to the same simple and unchanged zhi 指 in all possible worlds and cannot be regarded as referring to a class because a class of some entities in this world could be different from a class of the same kind of entities in other possible world. Based on this consideration, I think his key words should be represented by individual terms “a,” “b,” “c,” etc. or understood as the value of individual variables “x,” “y,” “z,” etc. So, instead of formulating Gongsun Long’s statement “white-horse is not horse” as:

 x   Wx & Hx  ~ Hx 



I think we should elaborate its form as: ~(a = b). This can be proved validly from his premises. Of course, Gongsun Long does not notice very well the distinction between use and mention and thus sometimes uses these words such as “horse,” “white” and “white-horse” not to refer to entity but to refer to itself, i.e., to mention the name itself. However, his thesis can be proved to be valid based on his premises whether “white-horse is not horse” means “white-horse is not the same [entity] as horse” or “The name ‘white-horse’ is not [referring to] the same [thing] as the name ‘horse’.” The consequence of this arrangement is that the arguments provided by both sides can be formulated in an intelligible way and nothing has to be explained away. For the whole piece of Baima Lun, in another article, I have used the first-­ order predicate logic to formulate and to prove all the arguments provided by both sides; but I have to stop here.11

References Chmielewski, Janusz. 2009. Language and Logic in Ancient China. Warszawa: Komitet Nauk Orientalistycznych PAN. Fung, Yiu-ming. 2009. “School of Names.” In Bo Mou, ed., History of Chinese Philosophy (164– 188). London and New York: Routledge. Fung, Yiu-ming. 2012. “A Logical Perspective on the Parallelism in Later Moism.” Journal of Chinese Philosophy 39.3: 333–50. Fung, Yiu-ming. forthcoming. Language, Truth and Logic in Ancient China, chapter seven. Fung, Yulan. 1948. A Short History of Chinese Philosophy. London: Collier-Macmillan Publisher. Fung, Yulan. 1961. A History of Chinese Philosophy (中國哲學史). Hong Kong: Pacific Book Company. Hansen, Chad. 1983. Language and Logic in Ancient China. Ann Arbor: University of Michigan Press.

11

 Yiu-ming Fung, Language, Truth and Logic in Ancient China, chapter seven (forthcoming).

17  Logical Thinking in the Gongsun Longzi

327

Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. New York: Oxford University Press. Hearne, J. W. 1980. Classical Chinese as an Instrument of Deduction (Dissertation). University of California Riverside. Hu, Shih. 1922. The Development of the Logical Thought in Ancient China. Shanghai: The Oriental Book Company. Hu, Shih. 1997. An Outline History of Chinese Philosophy, vol. I (中國哲學史大綱卷). Shanghai: Ancient Books Publishing House. Kripke, Saul. 1980. Naming and Necessity. Cambridge: Harvard University Press. Needham, Joseph. 1956. Science and Civilisation in China, vol. II.  Cambridge: Cambridge University Press. Salmon, Nathan. 1982. Reference and Essence. Basil Blackwell: Oxford University Press. Searle, John. 2004. Mind: A Brief Introduction. Oxford: Oxford University Press. Wang, Rongpei and Hong, Wang (translation). 2006. Mozi. Hunan People’s Publishing House. Fung, Yiu-ming  is Emeritus Professor at the Hong Kong University of Science and Technology and currently Visiting Professor in the Department of Philosophy at the Tunghai University, Taiwan. He is an analytic philosopher with special interest in Chinese philosophy and comparative philosophy. Fung is the author of several books, including The Methodological Problems of Chinese Philosophy (1989); Chinese Philosophy in the Ancient Period, 4 volumes (1992); The Gongsun Longzi: A Perspective of Analytic Philosophy (1999); and The Myth of Transcendent Immanence: A Perspective of Analytic Philosophy on Contemporary Neo-­Confucianism (2003). He has also published more than 100 research papers both in Chinese and English.

Chapter 18

Correcting Names in Early Confucianism Hui Chieh Loy

1  Introduction According to a widely held conventional wisdom, the doctrine of zhèngmíng 正名, “rectification of names”, or perhaps “correcting names”, appears to be central to the tradition that springs from the thought of Confucius (Kǒngzǐ 孔子). But whether zhèngmíng really is central to early Confucianism, what the doctrine itself amounts to remains the subject of continuing controversy.1 This paper aims to present an account of the zhèngmíng doctrine in Early (i.e., Pre-Imperial) Confucianism, especially as it is presented in the Analects (Lúnyǔ 論語), and secondarily, the “zhèngmíng” essay of the Xúnzǐ (荀子). Bracketing the issue as to whether zhèngmíng is “central” to Early Confucianism, the contention here is that an interesting set of meditations on the ethical and political implications of the use of language can be found in the selected passages, meditations which are properly identified under the traditional heading of “zhèngmíng”. These meditations point to a view regarding how the correct use of words (or language more generally) when referring to things can be ethically and political significant. The correct use of words is integral to people’s ability to carry out tasks, follow 1  A useful collation of traditional opinions on zhèngmíng in the Analects can be found in Liu Zhenghao 劉正浩 1978. See also Makeham 2003: 333–340. Hu Shih 胡適 is likely the source of the modern conventional wisdom—for him, Confucius saw zhèngmíng as “the heart of the problem of social and political reformation,” while “the reform of society” as itself, the “central problem of Confucius”, thus implying that zhèngmíng is central to the thought of Confucius (Hu 1963: 22). This notion, however, probably has much to do with Hu’s idiosyncratic concerns rather than completely based upon an accurate assessment of the tradition.

H. C. Loy (*) Department of Philosophy, Faculty of Arts and Social Sciences, National University of Singapore, Singapore, Singapore e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_18

329

330

H. C. Loy

norms, and contribute to flourishing cultural activities. Conversely, the incorrect use of language—at least by the political elite—can lead to social and political disorder, so much so that guarding against the corrupt use of language is a political imperative. The sources, however, are not as detailed regarding exactly how names are to be corrected or preserved, especially by the ruling powers, or exactly how the incorrect use of language causes disorder. Nevertheless, clues can be found if the zhèngmíng doctrine is set in the context of the wider Analects teaching regarding how people’s speech relates to their ethical virtue, or lack thereof. Finally, key elements of the zhèngmíng doctrine as it is implied by the Analects are found in the Xunzi as well, even though the latter also contains additional ideas that reflect changes in the intellectual milieu. Before proceeding, it might be useful to say something about the locus classicus of the doctrine: Analects 13.3. That the opening section from the thirteenth book of the Analects should be a starting point for discussing the zhèngmíng doctrine is uncontroversial: the passage preserves the earliest occurrence of the phrase zhèngmíng in Confucian philosophical literature. In fact, it is the only place where the phrase occurs in the Analects at all. But there are several difficulties relating to this text that should be addressed. First, Analects 13.3 does not present Confucius as expounding a systematically formulated doctrine. Confucius says that he would first “correct names” should the prince of Wey employ him in government. Looking to the passage to furnish the contents of a zhèngmíng doctrine may involve reading into it what is not explicitly expressed there. Nonetheless, this is not an insurmountable difficulty. There is simply no systematic doctrine of anything to be found in the Analects. Not even when he speaks of such famously Confucian concepts as rén 仁, lǐ 禮, jūnzǐ 君子, was Confucius presented as invoking a systematic doctrine. There is even reason to think that the writers or compliers of the Analects and early Confucians in general would not have thought well of the idea that Confucius’ Way (dào 道) can be reduced to a systematic doctrine given their particularist tendencies.2 But despite all that, it might still make sense to speak of the Analects’ doctrine of rén, lǐ, jūnzǐ, i.e., what the text has to say about such things as rén, lǐ, jūnzǐ, and beyond that, what it has to say about ethical excellence or the morally exemplary agent. In the same spirit, it makes sense to consider what the text has to teach when it presents Confucius as saying that he would “correct names”. A different difficulty, however, is posed by the potentially problematic status of the passage within the Analects text. Modern critical scholars tend to see the passage as a later interpolation. While there have been various forceful rejoinders, the balance of evidence tends to favor the critics.3 For the purposes of the present study, I will take for granted that the passage belongs to a later rather than earlier layer of  See e.g., Van Norden 2007: 36, 58–59.  See Waley 1939: 21–22, 171–172; Creel 1951: 321–322; and Brooks and Brooks 1997: 190 for the text critical case that the passage is a later interpolation. For the responses, see e.g., Knoblock 1988:114–116, Reding 1985: 252–255, and Makeham 1994: 163–165. For a more evaluation of the evidence, see Van Norden 2007: 86–90.

2 3

18  Correcting Names in Early Confucianism

331

the Analects. Consequently, I assume it unwise to take the passage as recording a conversation that took place between a historical Confucius and his disciple the historical Zǐlù 子路. This also meant that any doctrine we managed to draw from the passage should not be thought of as something taught by the historical Confucius—we simply don’t know. In any case, the working procedure here is to take the Analects not as a record of the historical Confucius’ teachings, but as a source for a broad tradition of thinking that had its beginning in, but is not restricted to, the expressed thought of the historical Confucius—Analects Confucianism. With this in mind, it would be more accurate to call any zhèngmíng doctrine derived from the text the Analects’ zhèngmíng doctrine, as it was expressed through the passage’s presentation of the Confucius character, rather than Confucius’ zhèngmíng doctrine. The paper will proceed as follows. Section §2 will argue, on the basis of an analysis of the passage, that Analects 13.3 contains a concern with the correct use of words rather than directly with stations or roles. In addition, three different aspects of the doctrine—to do with a diagnosis of disorder, the preservation of order, and the remedy against disorder—can be distinguished. Section §3 picks up on that result to consider how other ideas and themes in the Analects can motivate the concern with language so identified. Finally, Sect. §4 shows how Xúnzǐ’s discussion both connects with the zhèngmíng doctrine of the Analects but also diverges from it.

2  The Doctrine of Analects 13.3 As indicated in the introduction, the first order of business is to outline the zhèngmíng doctrine as it is presented in Analects 13.3. In this section, I will argue that the doctrine is ultimately about the importance of the correct use of language (or the correct application of names to things and affairs) for social and political order. Such a conclusion needs to be argued for because it is distinct from, and sometimes confused with, a different interpretation. On the alternative interpretation, Confucius was concerned with the importance of people living up to or be properly treated in accordance with their stations or roles. While this other notion is attested to in the Analects, the argument here is that zhèngmíng is a different thing. In the last part of the section, I will distinguish between three aspects of the doctrine—one to do with the diagnosis of disorder, another with how the correct use of names preserves order and a third to do with the remedy against disorder. But first, the passage: Zǐlù said, “The prince of Wey is waiting to leave the government of his state to you, what are you going to put first?” The Master said, “That must be to correct names (zhèngmíng).” Zǐlù said, “Is that right? Surely you are far [off the mark]. Why consider correcting [names] at all?” The Master said, “You! How boorish! The gentleman, when faced with something he does not know, is wont to keep his peace. When names are not correct then speech will not accord; when speech does not accord then affairs will not be effected; when affairs are not effected then ritual and music will not flourish; when ritual and music do not flourish then punishment will not hit the mark; when punishment does not hit the mark then the people will not know where to place hand or foot. Hence what the gentleman names surely

332

H. C. Loy

can be put into speech, what he puts into speech surely can be done. The gentleman with regard to his speech, is careless about nothing.” (13.3)4

One first conclusion can be easily drawn: Confucius’ reply to the disciple Zǐlù’s query tells us that zhèngmíng is part of what a governing minister might need to do; in other words, it is an aspect of proper government. The reason why zhèngmíng is an aspect of proper government is further hinted at by the main part of what Confucius said to Zǐlù: grave social and political consequences follow when names are not correct—ritual, music, criminal punishment will not be properly ordered and the common people will lack proper guidance for their conduct. Presumably, it is part of the remit of proper government to insure against such outcomes and to find remedies for them when they arise. Note that even though more general considerations applicable to a variety of circumstances might be inferred from what he said, Confucius’ reply is framed in response to the putative situation in Wey. The passage thus implies that, from the point of view of Confucius, names were likely not correct in the state of Wey, and perhaps worse things have already followed (more about this later). So what did Confucius mean to do by his saying that he will correct names if he really were given the reins of the government in Wey? Relatedly, what does it mean for a situation of “names are not correct” to obtain? Here, the crucial differences between the proposed answers can be explained in terms of how interpreters answer two questions. The first question concerns the denotation of the “names” (míng 名) in question, and the major divide is between interpretations that take it to refer to words, i.e., an element of language, though perhaps restricted to a specific class of words (e.g., names, titles), on the one hand, and those that take it to refer to stations or roles (míngfèn 名份 in Modern Chinese), on the other hand. Interestingly, the term often take the sense of “reputation” (míngshēng 名聲 in Modern Chinese) in other passages of the Analects (4.5, 9.2, 15.20). Now, míng as “names” and the same as “reputation” are related to each other—if John has a good reputation (“a good name”), that meant, among other things, that people have been saying good things about him, giving him a good report, as we might say. The sharper distinction is between míng as an element of language and the same as stations or roles. John may be called a “ruler”, whether in writing or in speech; and he may possess a certain status, i.e., occupy the station or perform the role of a ruler. The latter obtains when he stands in a series of relationships with others, relationships that entail his having certain duties towards and rights that he can claim against them, relationships instantiated in existing and expected patterns of interactions between him and others. Normally, someone who has the station or role of a ruler would be called “ruler”, and vice versa. But the two—being called a certain name or title by others, on the 4  All translations of the Analects are my own; and the text will be cited using the division of the text in Lau 1992. The following editions were consulted: Legge 2006, Waley 1939, Yang 1984 (modern Chinese translation), Lau 1992, Leys 1997, Ames and Rosemont 1998, and Slingerland 2003; and also, the collation of the traditional commentaries on the Analects in Cheng Shude 1990. For a more detailed analysis of Analects 13.3, see Loy 2008.

18  Correcting Names in Early Confucianism

333

one hand, and possessing the status that goes with having a certain station or role, on the other hand—are analytically distinct and should not be confused. In principle, one and not the other could be a subject of Confucius’ attention, even as he might also be concerned about both. The second question concerns the intended object of the correcting. Suppose there is a case of míngbùzhèng (i.e., “name is not correct”), does the doctrine say that the míng is to be corrected? Or does it say that something else is to be correct, perhaps in light of the míng? Suppose the míng in zhèngmíng denotes names (i.e., words), which implies that the situation of “names are not correct” occurs when people use those names in some incorrect fashion, e.g., when they apply an incorrect name to something. The first possibility now says that to zhèngmíng is to correct people’s use of those names and insist upon their using them in the right way. An example of such an interpretation would be that offered by Joseph Needham, who explains Confucius’ concern as “the determination to call a spade a spade, no matter what powerful influences might be desirous of having it called something else” (Needham 1954: 9–10).5 The second possibility says that it’s the thing named that needs to be corrected so that it matches the name. This is probably what Fung Yu-lan 馮友蘭 had in mind when he takes Confucius to be teaching in Analects 13.3 that “the actual must in each case be made to correspond to the name” (Fung 1952–53: 59).6 Some scholars combine the two options. Zhang Dai’nian 張岱年, for instance, explains zhèngmíng as “making name and reality correspond to each other, causing the name to match the reality, the reality to match the name” (Zhang 1982: 560).7 Now, the answers to both questions are reasonably clear if the whole passage is kept in mind in one’s interpretation. Consider the concluding part of what Confucius said to Zǐlù: “the gentleman, when it comes to his speech (yán 言), is careless about nothing.” If Confucius did not change the topic along the way of his reply to Zǐlù, the whole matter of zhèngmíng is ultimately a matter of people using language, in some proper, non-careless manner. But if that is so, then it seems much more plausible, firstly, to see the míng in zhèngmíng as denoting words, perhaps a restricted class of words, rather than stations or roles; and secondly, to take Confucius as calling for correcting the use of these míng rather than correcting something else with reference to them. In addition, the last part of the passage also suggests that it is mainly the gentlemen’s (jūnzǐ) language that is of concern. This might mean that it is the political elite’s use of language that is at issue, or it might mean that Confucius saw the correct use of names as something expected of exemplary agents regardless of whether they are actual members of the political elite. More likely, the concern is  Another example is Creel 1951: 321–22.  See also Legge 2006: 263–64, Hsiao 1979: 98–100, Reding 1985: 251. 7  Cheng Zhongying and Cai Mingtian spell out zhengming in terms of the injunction that speech and deed must match (yán xíng yī zhì 言行一致); Cheng Zhongying (1974) 1981: 66, and Cai 1984: 6–8. See also Xu 1966: 5, and Hu 1963: 24–27. Examples of scholars who see the zhèngmíng if 13.3 as being concerned about station or role include Lao Sze-kwang 勞思光 and Lin Yuanqi 林遠琪; see Lao 1984: 122–27 and Lin 1979: 9–18. 5 6

334

H. C. Loy

with the political elite’s use of names because, ideally, members of the political elite ought also to be exemplary agents. Interpreters who take the míng in Analects 13.3’s zhèngmíng to denote stations or roles may well have another passage of the text in mind, though the reasoning behind citing it is seldom expressed clearly.8 This passage is 12.11, which begins with a ruler asking Confucius regarding the proper way to govern. Confucius replied with an extremely concise “君君臣臣,父父子子 jūn jūn chéng chéng fù fù zǐ zǐ”. On the majority reading of the line, Confucius is enjoining people to properly fulfill the roles they occupy: “rulers (should) be proper rulers. Ministers (should) be proper ministers. Fathers (should) be proper fathers. Sons (should) be proper sons.”9 On another syntactically possible reading, he is enjoining people to accord those who occupy certain roles with consideration that ought to be due given those roles: “(People should) treat rulers in ways as befitting rulers. Treat ministers in ways as befitting ministers. Treat fathers in ways as befitting fathers. Treat sons in ways as befitting sons.” The duke’s subsequent response can also be understood accordingly as saying either that trouble comes when people fail to properly fulfill the stations and roles they occupy, or when people fail to accord those who occupy those stations and roles the consideration due given the relationships they stand with them. But either way, the passage does not mention anything specifically about the use of language. That is, while, both Confucius and the Duke used the terms “jūn”, “chéng”, “fù”, and “zǐ” to make their point, neither of them focused on those terms as, themselves, subject matters of discussion. Rather, the speakers used those names to make a point about the social and political importance of certain stations or roles, e.g., that of ruler, minister, father and son. Consequently, one might think that there is support for the stations or roles reading of míng in Analects 13.3 given the traditional association between that passage and 12.11. But unfortunately, this is not a sound conclusion to draw. Quite apart from the fact that 12.11 is not associated with 13.3 in the earliest Analects commentaries (See Defoort 2013: 82), the forgoing conclusion simply does not follow. I take for granted that in 12.11, Confucius was talking about the social and political significance of people living up to their stations or roles or treating others as befitting their stations or roles. But this doesn’t imply that he was talking about the same thing in 13.3, or that he wasn’t concerned about the social and political significance of the use of language in 13.3. In fact, I will take it for granted that Analects Confucianism was concerned about both issues; but that’s not a reason to confuse them. While there may yet be some less direct connection between 13.3 and 12.11, on the face of it, 12.11 does not provide evidence that Confucius was talking about social roles and stations themselves in 13.3. Another reason why someone might prefer the stations or roles reading of Analects 13.3 relates to the passage’s putative historical context. Sima Qian 司馬遷 8  In fact, the passage is often cited in discussions of zhèngmíng without explanation as to why it belongs in such discussions. For a recent example, see McLeod 2015: 48–52. 9  See e.g., Legge 2006: 256, Waley 1939: 166, Yang 1984: 128, Lau 1992: 113, Leys 1997: 57, Ames and Rosemont 1998: 156, and Slingerland 2003: 130–131.

18  Correcting Names in Early Confucianism

335

implies that the conversation between Confucius and Zǐlù took place against the backdrop of an ongoing palace struggle in the state of Wey 衛, an episode also mentioned in the Zuozhuan 左傳.10 Briefly, the previous crown prince, Kuǎi Kuì 蒯聵, went into exile after a failed attempt to murder his father Duke Líng’s (衛靈公) wife Nánzǐ 南子. After Kuǎi Kuì’s own son Zhě 轍 became the ruler upon the death of Duke Ling, Kuǎi Kuì attempted to return with the help of a foreign power. He was promptly repulsed by the new Duke, i.e., his own son, who went on to besiege the father in a small town. If the above is the context of the conversation in Analects 13.3, then the “ruler of Wey” that Zǐlù refers to in his initial question to Confucius would be none other than Prince Zhě, now Duke Chù of Wey (衛出公). The Zuozhuan tells us that Zǐlù was an employee of Prince Zhě and will later die fighting to protect him against Kuǎi Kuì when the later made yet another attempt to return to power. John Makeham believes that taking this imputed historical context of the passage seriously “leaves little doubt that the term [míng] refers to primarily to two roles: father and son” (Makeham 1994: 334). So in calling for names to be corrected in Wey, Confucius was basically referencing the conflict between erstwhile father and son and the resulting chaos. Note that the point above does not depend upon the accuracy of the passage as an authentic report of a historical event. The only assumption required is that the backstory provides the intended context of the conversation for early readers of the text. Nonetheless, the conclusion that míng in Analects 13.3 refers to stations or roles, even the specific roles of father and son, does not quite follow from backstory taken as the intended context of the passage. What the backstory might suggest is at best the following modification: the míng in zhèngmíng denotes not words in general, but a more restricted class of words, i.e., those names that denote the various stations or roles that are so important to the Confucian conception of social and political order, “father” and “son”, “ruler” and “minister”, for instance. If this is granted, then perhaps we can at least rule out some interpretative proposals. Ma Rong 馬融, for instance, thought that the zhèngmíng of Analects 13.3 refers to “rectifying the names of the hundred things” (正百事之名 zhèng bǎi shì zhī míng) and Zheng Xuan 鄭 玄, on the other hand, thought that it refers to any erroneous words in writing (wénzì 文字) (Cheng 1990: 890); both proposals might be considered too broad if the intended backstory were to be taken seriously. But even granting that Confucius was thinking specifically about people’s use of “father” and “son” (and perhaps “ruler” and “minister”) in relation to Zhě and Kuǎi Kuì, it doesn’t follow that zhèngmíng is only concerned with such a restricted class of míng. The concern with how the specific names “father” and “son” are used could just as well follow from more general considerations to do with the proper use of language. Having established that Confucius was ultimately talking about the importance of the correct use of language to social and political order in Analects 13.3, we can

 See Shiji (Sima Qian 1982 37:1599–1601, 47:1933–1934); and Zuozhuan, Duke Ai 15.5 (see Durrant et al. 2016: 1938–1941).

10

336

H. C. Loy

now unpack his doctrine a bit more. As earlier mentioned, three aspects of the doctrine can be distinguished. A first aspect concerns a diagnosis of disorder. This point is encapsulated in the long chain of conditionals that form the main bulk of Confucius’ subsequent reply to Zǐlù. That is, ritual and political disorder arise when people (probably the social and political elite) use names incorrectly. There is no evidence in the text, however, that Confucius thought that using names incorrectly is the only cause of disorder, or that all other causes can ultimately be traced to this one cause. A second aspect concerns the preservation of order against disorder. This idea is also hinted at in the last part of Confucius’ response to Zǐlù, where he expressed his concern that the gentleman (i.e., the social and political elite as they ought to be) be careful in his speech, i.e., including the way he uses individual “names”, and possibly even his keeping his peace when confronting something he does not know. In a sense, this point follows from the previous. If ritual and political disorder arise when certain people use names incorrectly, it follows that a way to prevent the baleful outcomes from happening is that the antecedent cause not happen, i.e., those same people not use names incorrectly. And finally, there is also a third aspect to do with a remedy to ritual and political disorder, at the very least, such disorder as caused by the situation of names not being correct. This third part of the doctrine is captured in the Master’s initial statement that he will “correct names” if he were to be appointed to government by the ruler of Wey. The passage here highlights the role of a ‘corrector of names’—the Master should he take up office in Wey—whose task it is to see to it that names are used correctly. But the text does not say or imply that if “gentlemen” used names correctly, or if a ‘corrector of names’ should successfully do his thing, then ritual and political order will automatically follow. From the fact that you can’t have ritual and political order if people use names incorrectly, it doesn’t follow that if people use names correctly, you will get ritual and political order. Let’s say that the above is, in outline, the kernel of Analects 13.3’s zhèngmíng doctrine. But how can we best understand the doctrine, especially within the terms of Analects Confucianism? What exactly does it mean for someone to use names incorrectly? And how exactly does that lead to the social and political problems listed in 13.3 such that correcting the way names are used can become a pre-­ condition of political order? Could it be that, as Bao Zhiming speculates, Confucius was working from a “belief that language possesses a magical power which has unfailing influence on affairs both human and natural?” (Bao 1990: 197–199)

3  Language and Ethics in the Analects Even though outside of 13.3, the Analects does not speak of zhèngmíng, various passages do say or imply things about how the use of language might be related to ethics and politics. Collectively, these passages point to at least two distinct though related cluster of ideas regarding how an improper use of language can be ­associated

18  Correcting Names in Early Confucianism

337

with ethical or political problems, and both cluster of ideas can be taken as motivating what the zhèngmíng doctrine asserts regarding a diagnosis of disorder and what can be done to preserve order, i.e., the social and political elite being careful with their language. In contrast, there is much less certainty regarding the political remedy of disorder, i.e., the aspect of the doctrine to do with the putative ‘corrector of names’. Rather, the text at best portrays Confucius as a teacher without an official position seeking to ensure that names are used correctly among his disciples. Suppose we take for granted that for Confucius, what ultimately matters most is that people conduct themselves well and so enable the social and political world to be well-ordered, rather than that they merely possess the right notions or can mouth the right pieties. Nonetheless, it may yet be the case that for someone to conduct himself well, he needs to have the right notions or at least to avoid mistaken ones, and furthermore, that there is such a thing as a proper expression of the relevant notions in words. Consider here the following four passages: Mèng Yízǐ asked about filial piety. The Master said: “Never disobey.” Fán Chí was driving. The Master told him about the exchange, saying: “Mèng Yízǐ asked me about filial piety. I answered, “Never disobey.” Fán Chí asked: “What does that mean?” The Master said: “When your parents are alive, serve them in accordance with the rites; when they die, bury them in accordance with the rites and sacrifice to them in accordance with the rites.” (2.5) Mèng Wǔbó asked about filial piety. The Master said: “Give your father and mother no cause for worry other than illness.” (2.6) Zǐyóu asked about filial piety. The Master said: “Nowadays, by a ‘filial person’, it is meant that he is able to feed his parents. Even dogs and horses are provided with food; without reverence, by what shall we tell them apart?” (2.7) Zixia asked about filial piety. The Master said: “Showing the right countenance on one’s face is the difficult thing. When there is work to be done the younger ones take on the burden, when wine and food are served the elder ones are given precedence—can such be considered filial?” (2.8)

All four passages feature a disciple asking Confucius regarding filial piety, with Confucius obliging with an answer. The presumption throughout is that the students signal a desire to conduct themselves in a filial manner—that being the reason why they have approached the master for instruction on the issue. But there are some interesting differences between the passages. In 2.5 and 2.6, Confucius’ point is that something should be done or that some attitude ought to be adopted, if one is to be filial. In both cases, the suspicion is that Confucius’ injunctions relate to aspects of the questioner’s conduct or character that is still wanting from the point of view of filial piety. 2.8 contains an additional layer of complexity. The bulk of Confucius’ answer is a rhetorical question which implies that filial piety goes beyond the young shouldering burdensome work on behalf of the elders, or giving them good food and drinks when these are available. The reply hints that, in Confucius’ estimation, Zixia harbors a mistaken notion about what behavior is good enough to count as filial piety, a mistake requiring correction. Like 2.8, 2.7 also suggests that the interlocutor—Zǐyóu in this case—harbors a mistaken notion about filial piety: perhaps he thought that making sure to feed one’s parents is good enough to count one as filial. But it adds yet another layer to the mix by asking whether what people nowadays meant by “a filial person” is enough to

338

H. C. Loy

qualify that person as filial. In other words, the problem is not just sub-filial behavior, or even a mistaken notion that might have led to it, but something about the way people use language—misapplying the name “filial” to what is not good enough to count as filial behavior (in Confucius’ estimation). When seen in this light, 2.7 hints at the zhèngmíng doctrine’s diagnosis of disorder.11 But what if the ‘wrong notions’ expressed by the improper use of language springs not so much from a passive ignorance of truth, but something more active? After all, our words don’t just passively reveal what we are like to others; they are often very much our own attempt to reveal ourselves to them, to actively present a certain picture of ourselves to them. Consider this passage—which has been associated with zhèngmíng from early on (Cheng 1990: 913–916)—in which Confucius literally corrected a disciple’s use of a term (i.e., a míng): Ranzi [i.e., Rán Yǒu] returned from court. The Master said, “Why so late? He replied, “There were affairs of government (zhèng政).” The Master said, “They could only be (private) business (shì 事). If there should be affairs of government, even though I am not in office, I would get to hear of them.” (13.14)

The “affairs of government” in question were likely related to partisan activities of the Jì 季clan (cf. 3.1, 11.17 and 16.1) in the court of Lu 魯, hence Confucius’ retort (see Cheng 1990: 913–916; Slingerland 2003: 145). To see what’s going on, consider the passages below where Confucius brings into contrast two terms which are close in meaning but with a difference in moral overtones: The Master said, “Gentlemen help others (zhōu 周) without colluding (bǐ 比); petty men collude without helping others.” (2.14) The Master said, “Gentleman agree (hé 和) without echoing (tóng 同); the petty men echo without agreeing.” (13.23) The Master said, “The gentleman is dignified (tài 泰) without being haughty (jiāo 驕); the petty man is haughty without being dignified.” (13.26) The Master said, “The gentleman is proud (jīn 矜) without being contentious (zhēng 爭), he forms alliances (qún 群) but not cliques (dǎng 黨).” (15.22)

Notice how the first three passages each makes use of a pair of adjectives, a positively meant one to describe the jūnzǐ 君子, and a negatively meant one to describe the xiǎorén 小人. And even though the fourth passage does not make the explicit

11  There is one additional passage that is traditionally associated with discussions about zhèngmíng, namely, the cryptic 6.25: “The Master said, ‘A gū (觚) that is not truly a gū. A gū indeed! A gū indeed!” According to Ma Rong, the term gū, apart from naming a kind of ritual drinking vessel, also names a certain measurement of liquid (two shēng 升 or approximately …). The ritual drinking vessel was apparently so named because it is supposed to hold that amount of wine. The Qing Dynasty scholar Mao Qiling 毛奇龄 (1623–1716) thought that “gū means ‘small,’ that is, ‘the drinking [of wine] should be in small quantity,’. He proposes that in ancient days, a measure of three shēng of wine was considered appropriate, five as excessive, and two as moderate [lit. “small”], and drinking vessels were named according to their capacity. … (in the Master’s day) though the name of the vessel is gū, yet it was frequently used for excessive drinking… hence his exclamation: “A gū indeed! A gū indeed!” One can almost imagine the Master, upon the pattern of 2.7 saying, “by ‘drinking in moderation’ nowadays, it is really meant …” (in Cheng 1990: 412– 413) But all this is very speculative.

18  Correcting Names in Early Confucianism

339

comparison between jūnzǐ and xiǎorén, the parallelism suggests a similar contrast (cf. 3.7). What’s interesting is that in each case, the terms are practically synonymous within suitable contexts. Take 15.22: “forming an alliance” (qún 群) and “forming a clique” (dǎng 黨), for example. The latter is commonly associated with “partisanship” or “being partial” (cf. 7.31), which might also explain the contentiousness of petty men in 15.22 (cf. also 3.7).12 Still, from a certain perspective, “forming an alliance” is not very different from “forming a clique”— both involve people coming together and co-­operating for some common purpose.13 So likewise, it may well be that in some contexts, there isn’t a lot to choose from between calling a particular activity the public affairs of state or the private affairs of a clique. But for Rán Yǒu to say that he was attending to the former having come back from court is, in the eyes of Confucius, tantamount to his saying that the activities of his paymasters the Ji Clan—and not the legitimate ruling house—constitutes the affairs of government (cf. 11.17, 16.1). One can imagine similar examples constructed from the terms contrasted in the other cited passages. At this point, two slightly different possibilities present themselves depending on how sincere we think Rán Yǒu was.14 First, suppose that Rán Yǒu was a true believer in his paymasters’ cause. (This is a plausible supposition given what other passages of the Analects say about Rán Yǒu; see 3.1-2, 3.6, 11.17, 16.1.) On this supposition, Rán Yǒu was sincerely applying the term “affairs of government” to his business and the business of his employers—and so signaling to Confucius where his true loyalties lie. On the other hand, suppose Rán Yǒu was insincere. That is, in calling their business “affairs of government”, he was trying to represent his conduct and that of the Jì Clan as being better than they are even according to his own estimation. Perhaps he did this to avoid the Master’s censure, or to raise his own status and that of his employers in the Master’s eyes. On either interpretation, Rán Yǒu’s choice of words is not a passive revelation of a mistaken notion, but an aspect of an active attempt to present a picture of himself and those he served. The above discussion should also be seen against the backdrop of a larger theme in the Analects—the idea that virtue is not only a matter of proper conduct, but proper speech as well (Loy 2014). One motif that stands out within this larger theme is the idea that being cautious and slow in speech is “close to rén” (13.27; see also 12.3). In contrast, eloquence is problematic as it often the way whereby a person “goes beyond” his actual conduct and thus puts himself in a better light than warranted (1.3, 5.5, 5.25, 17.17). On this score, keep in mind the courtly political context where, as one might say, a man’s worth is a matter of what he says and what others say about him. Consequently, self-interested courtiers might gain preferment

 Yang Bojun 楊伯峻 links up the contrast with 2.14 and 13.23 (Yang 1984: 165).  To borrow some words from Kenneth Burke, they are “same motions but different acts” (Burke 1969: 108). 14  For a more extensive discussion of the distinction between the two interpretations, see Loy 2014: 150–151. 12 13

340

H. C. Loy

because of their eloquence in ways that are at odds with the common good. Given such a context, it is not surprising that Confucius’ suspicion against eloquence relates not merely to the clever speakers’ own character but also to the potential harm that they can bring to family and state (15.11, 17.18). Given the above discussion, we now have at least two ways to motivate what I’ve called the dimension of the zhèngmíng to do with a diagnosis of disorder—and by the same token, the idea that the social and political elite preserves order in the polity when they are careful with their language—without having to appeal to magic. First, the writers of the Analects noticed that the widespread misapplication of terms (how “people nowadays” used the term “filial” in 2.7), especially among the elite, is symptomatic of a more general moral decadence (people considering sub-filial attitudes and conduct as good enough). Second, they also noticed that individual misapplications of terms are often the tool in trade of a self-serving hypocrisy on the part of agents who stand to gain in status and conventional goods through misrepresentation of themselves. Either way, given the very Confucian assumption that social and political order depends upon the rule of exemplary agents, we can see how the writers of the Analects could have thought that when names are incorrectly used by the political elite, there is an increased likelihood for the baleful consequences described in Analects 13.3 to follow. This diagnosis of disorder offered in the zhèngmíng doctrine is all the more plausible if we focus mainly on terms referring to ideas that are meant to be action-guiding import, i.e., not so much “tables” and “chairs”, but “filial piety”, (legitimate) “rulership”, and so on. After all, to the extent that he was interested in the subject, Confucius—as with other early Chinese thinkers in general— was probably thinking mainly of the role that language plays in guiding conduct, e.g., as the means for us to attribute action guiding status to various affairs and things in the world, rather than, for instance, in describing the world. (As we shall see later, this point becomes more explicit in Xúnzǐ; see §4 below.) But what of the aspect of doctrine to do with the work of a ‘corrector of names’— what Confucius would be doing should he had taken up office in Wey—as a remedy against disorder? Recall also that Confucius presents “correcting names” as an aspect of what a minister should do—in other words, an aspect of proper government—in his initial remarks to Zǐlù Analects 13.3. Presumably, this is not merely the act of pointing out individual or widespread misapplications of terms, but the positive inculcation of the right forms—given that “correcting” (zhèng 正) and “teaching” (jiào 教) are two sides to the same coin, and both being functions and prerogatives of government (zhèng 政) conceived along Confucian lines (2.20, 12.17, 13.9). But one looks in vain in the Analects for a discussion of how zhèngmíng should or could be conducted from a position of political authority. A moment’s reflection, however, suggests a reason why such a discussion might not be forthcoming in the Analects. There wouldn’t be an explicit discussion of how the ancient sage kings—the likes of Yáo and Shùn, or the founders of the Zhōu 周 Dynasty—conducted zhèngmíng when they were in power for the obvious reason that there wasn’t supposed to be a need for such a thing when the ruler’s charisma was such that all he had to do was to sit facing south and order prevailed without his

18  Correcting Names in Early Confucianism

341

deliberate action (15.5). In fact, as Confucius also reminds us, “When the Way ­prevails in the world, commoners do not discuss affairs of state” (16.2). Consequently, a discussion of how an official ‘corrector of names’ would proceed is more plausibly set in a world in which the Way does not prevail. (Xúnzǐ, ever the realist, seemed to dissent from this line of thinking; see later.) But now we have the other horn of our dilemma: In such a world, it becomes extremely unlikely for someone capable of zhèngmíng—perhaps someone like Confucius—to assume a position of power and so have opportunity to put his scheme into action. Nonetheless, a way out is suggested by a clue already found in 13.3 itself. Recall, again, the supposed backstory about Kuǎi Kuì and Zhě as the imputed context of Analects 13.3. One additional conclusion can be draw with its help: notice that when Confucius says that gentlemen, i.e., the political and ethical elite as they ought to be, are to be careless about nothing when it comes to his speech, he was concluding his rejoinder to Zǐlù’s dismissal of his earlier claim regarding what is politically important or urgent in Wey. But if Zǐlù was—as indicated in the backstory—an official serving Zhě and thus part of the relevant political elite in Wey, then it would seem that he was himself part of the group of people whose use of language was the subject of Confucius’ concern. And if that is so, the passage hints that Confucius’ own rebuke to Zǐlù is itself plausibly a covert specimen of zhèngmíng, even though he wasn’t occupying a position of political authority. This need not be surprising given Confucius’ comment at 2.21 that “by being a good son and friendly to his brothers” a man can already exert influence upon those in government. So likewise, by being a good teacher to his disciples—some of whom have actual political connections—he can already make a difference to the situation regarding names. It is thus in this spirit that we return to the text to consider a phenomenon that shows up in some passages, namely, how Confucius is portrayed as explicitly teaching how certain terms should be used. Here, we find Confucius speaking of conditions under which someone “can be (properly) called” (kěwèi 可謂) called “filial”: The Master said: Observe what a man sets his heart on when the father is alive, and what his actions are when his father is dead. If for three years he makes no changes to his father’s ways, he can be (properly) called filial. (1.11; see also 4.20)

In other passages, the Master uses the same formulation to talk about how someone “can be (properly) called rén” (6.22), or how something “can be (properly) called the method for rén” (6.30) in response to disciples’ queries. At other times, it is the disciples who ask him regarding when things “can be (properly) called” by this term or that, as Zi-Zhang asked concerning “being successful” (dá 達) in 12.20, or Zi-Gong and Zǐlù asked (separately) concerning what it means to be proper “man of service” (shì 士) in 13.20 and 13.28. And then, we have this: Jì zǐrán asked, “Zhōng Yóu [i.e., Zǐlù] and Rán Qiú [i.e., Rán Yǒu]—can they be (properly) called great ministers?” The Master said, ‘I expected a different question, rather than about Yóu and Qiú. As for what is meant by “great minister”, it refers to those who serve their ruler with the Way and who, when that is not possible, set aside office. Now as for Yóu and Qiú, they can be (properly) called adequate ministers.” “If that’s the case, are they the sort who will always do as they are told?” “They will not do that if it comes to patricide or regicide.” (11.24)

342

H. C. Loy

The above don’t exactly form a large body of passages from across most of the books of the Analects—I listed a mere eight passages from six books. Their presence imply that some strands of the Analects tradition thought Confucius was in the habit of giving what looks to be definitions, even though none of them seem to amount to a set of properly necessary and sufficient conditions for the application of a term. For most of them, it’s easy to imagine a pedagogical context in which Confucius was attempting to impress upon someone a proper view of things by way of a comment about the proper use of a term with some ethical import. In other words, they are moments of ethical instruction (rather than lessons in logic or lexicography) where the Master impresses upon the student the norms or ideals of conduct implied in the referent of an ethical term. The normative guidance comes from the student trying to “live up to” a míng so defined. But collectively, what these passages suggest is that even to the extent that the Analects tradition had a view regarding the remedy involved in zhèngmíng, the doctrine is only implicit and less definite. That this should be the situation probably has something to do with the fact that Analects Confucianism tends to think of ethical instruction—Confucius’ teaching the proper way to use those terms included—in a highly particularistic manner where a Master and a disciple where the former tailors his instructor to the specific situation, strengths and weaknesses of the latter.

4  Early Confucian Zhèngmíng from the Analects to the Xúnzǐ Within the corpus of Early Confucian philosophical literature, the Xúnzǐ offers the second explicit mention and the first sustained discussion of zhèngmíng. This is found in the Chap. 22 of the received text, titled “zhèngmíng”, though in this case, the sense of the title is probably better captured in the noun phrase “Correct Names”. But before we turn to Xúnzǐ, let me make a couple of observations regarding the surrounding literature to explain why the discussion here moved from the Analects to the Xúnzǐ. While the phrase “zhèngmíng” does occur in various texts of the early period normally taken to post-date the time of Confucius,15 it does not occur in the Mencius (Mèngzǐ 孟子), the other major text in the corpus of Early Confucianism. Nonetheless, at least one passage from the text has sometimes been associated with zhèngmíng:16 King Xuān of Qí asked, “Is it true that Tāng banished Jié and King Wǔ marched against Zhòu?” “It is so recorded,” answered Mencius. “Is regicide permissible?” “He who mutilates benevolence is a mutilator; he who cripples rightness is a crippler; and a man who is  To give some examples, the phrase shows up in the logical chapters of the Mozi (Graham (1978) 2003: 440), several places in the Guánzǐ 管子, and an entire chapter of the Lǚshìchūnqiū 呂氏春 秋 has it as the title. 16  More detailed discussions of the passage with reference to Confucian zhèngmíng can be found in Defoort 1998: 111–118 and Defoort 2000: 85–109. 15

18  Correcting Names in Early Confucianism

343

both a mutilator and a crippler is an ‘outcast’. I have indeed heard of the punishment of the ‘outcast Zhòu’, but I have not heard of any regicide.” (Mencius 1B8; D.  C. Lau trans. Updated to pinyin) (Lau 1984: 39)

The reason for the association is not hard to perceive—the passage presents Mencius correcting King Xuān’s use of a term with moral and political significance. It is one thing for people to apply the term “rulers” to Jié and Zhòu, the implication being that killing them would be properly called “regicide”. It is something else to call them “outcasts”—because they are properly considered “mutilator” and “crippler”—the implication being that their punishment was entirely justifiable. Mencius’ argument even proceeds by way of (partial) definitions, much the way Confucius was presented as doing (e.g., in 11.24 discussed previously). The important point remains that the names we apply to various people or states of affairs matter, even morally and politically. Still, without the assumption that there is a zhèngmíng doctrine in the background—a doctrine of the sort described in the previous sections— there will be much less reason to read Mencius as being motivated by specifically zhèngmíng concerns. Xúnzǐ’s contribution to Early Confucian thinking about zhèngmíng thus far overshadows anything from Mencius. (Nevertheless, as we shall see below, a certain development recorded in the Mencius may well explain something in Xúnzǐ’s zhèngmíng discussions.) Now, since Xúnzǐ’s concerns in his essay are wide ranging, the discussion here will be constrained.17 The aim will be to link up the discussions in Xúnzǐ’s essay with the zhèngmíng doctrine of the Analects, and in so doing, highlight the key points of continuity between the two and touch upon some of the way in which Xúnzǐ’s discussion makes explicit what is merely implicit in the Analects, or presents an innovation over the earlier text. As mentioned earlier, the three aspects of the zhèngmíng doctrine identified earlier are readily found in Xúnzǐ’s essay. In fact, they come out fairly early on in Xúnzǐ’s essay, though now expressed in terms of a contrast between the regimes of the former sage kings and the situation in Xúnzǐ’s own time: …when the kings established names, the names were fixed, and the corresponding objects were thus distinguished. This way was followed, and the kings’ intentions were thus made understood. They then carefully led the people to adhere to these things single-mindedly. Thus, they called it great vileness to mince words and recklessly create names such as to disorder the correct names and thereby confuse the people and cause them to engage in much disputation and litigation. This wrongdoing was considered to be just like the crime of forging tallies and measures. Hence, none of their people dared rely on making up strange names so as to disorder the correct names, and so the people were honest. Since they were honest, they were easy to employ, and since they were easy to employ, tasks were accomplished. Because none of the people dared rely on making up strange names so as to disorder the correct names, they were unified in following the proper model of the Way and were diligent in following commands. Because they were like this, the legacy of the kings was long-lasting. To have such a long-lasting legacy and to achieve such accomplishments

 For a recent discussion of the philosophy presented in the essay as a whole, see Fraser’s contribution to the Dao Companion to the Philosophy of Xúnzǐ (Fraser 2016: 291–321).

17

344

H. C. Loy

is the height of good order. Such is what can be accomplished by diligently preserving the agreed names. Nowadays, the sage kings have passed away, and the preservation of these names has become lax. Strange words have arisen, the names and their corresponding objects are disordered, and the forms of right and wrong are unclear. As a result, even officers who assiduously preserve the proper models and scholars who assiduously recite the proper order are also all thrown into chaos. If there arose a true king, he would surely follow the old names in some cases and create new names in other cases. (Hutton 2016: 236–237)18

While we get a general sense that for Confucius, the model of proper conduct (and presumably, naming as well) have their sources in the practices of the ancient sages and a corresponding sense that the “moderns” have fallen far from that model, Xúnzǐ is much more explicit. Even though the beginning of the quoted passage refers only to “kings”, the sequel makes clear that these are the “sage kings” of the past. The correct names (i.e., the correct ways to use various terms) were once established by sage kings, who also sought to enforce their proper use against deviancy. In fact, the legacy of sages was long-lasting because, through their enforcement of correct names, they ensured an honest population “unified in following the proper model of the Way” and “diligent in following commands”. The unstated implication is that when names are not correct—or when their correct use is not enforced—there is a tendency for dishonesty and divisions to arise. On that last point, Xúnzǐ seems less sanguine than the Analects that the quiescent charisma of a sage ruler is sufficient for orderly rule—it is part of the remit of a sage king to take remedial action against the incorrect use of names to the extent of making it a crime to “mince words and recklessly create names such as to disorder the correct names” on par with that of “forging tallies and measures”. This goes far beyond the ideal of having officers and scholars “preserve the proper models” by being exemplary users of correct names themselves. In any case, the diagnostic, preservative and remedial aspects of the zhèngmíng doctrine as they might be inferred from the Analects are all clearly attested in Xúnzǐ’s remarks. There is an implied diagnosis of disorder as being caused by the incorrect use of names. There is the thought that the social and political elite as they ought to be (here identified as the officials and scholars) diligently preserved the proper models forms by being careful in their own use of language. And there is also the idea that the ruler and his ministers can ensure the correct application of names from a position of political power—Xúnzǐ both talked about how the earlier sage kings’ enforced the correct names and also called for any future “true king” to do the same. If anything, Xúnzǐ’s remarks on the activities of the kings make explicit what was merely implicit in Analects 13.3. Something else that Xúnzǐ made explicit was a notion I could only assume when discussing the Analects, an assumption offered to make the diagnosis of disorder offered in the Analects’ zhèngmíng doctrine more plausible without having to appeal

 The alternative reading is that he would “follow along with the old names and change back the new names”; see Hutton 2016: 237n2 and Wang 1988: 2:414. All references to the Xúnzǐ will refer to Hutton 2016.

18

18  Correcting Names in Early Confucianism

345

to a belief in magic. The assumption was that in talking about zhèngmíng, Confucius of the Analects was thinking mainly of the role that language plays in guiding conduct, rather than, for instance, in bloodlessly describing the world (see §3 above). In the immediate sequel to the previous quote, Xúnzǐ has this to say: Thus, one must examine the reason for having names, the proper means for distinguishing like and unlike, and the essential points in establishing names. When different forms make contact with the heart, they make each other understood as different things. If the names and their corresponding objects are tied together in a confused fashion, then the distinction between noble and base will not be clear, and the like and the unlike will not be differentiated. If this is so, then the problem of intentions not being understood will surely happen, and the disaster of affairs being thereby impeded and abandoned will surely occur. Thus the wise person draws differences and establishes names in order to point out their corresponding objects. Most importantly, he makes clear the distinction between noble and base, and more generally, he distinguishes the like and the unlike. (Hutton 2016: 237)

Taken in the abstract, language, for Xúnzǐ, exists so that we can mark distinctions between things and affairs in the world and communicate such distinctions to others.19 But both from the essay and what he says elsewhere in the corpus, the paradigmatic “distinctions” Xúnzǐ has in mind are meant to be action-guiding and constitutive of dào.20 In other words, they are more like the distinction between noble and base, virtue and vice, wisdom and folly, rather than that between apples and oranges, tables and chairs, especially if we were to think of the latter pairs in a purely descriptive manner without any action-guiding import. Given this background, when someone uses a name correctly, what’s ultimately going on is that she is properly drawing an action guiding distinction; and conversely when someone misapplies a name, she is ultimately improperly drawing an action guiding distinction and so accords an incorrect normative status to something. The caveat, however, is that this language is conventional in nature: Names have no pre-determined appropriateness. One forms agreement in order to name things. Once the agreement is set and has become custom, then they are called appropriate, and what differs from the agreed usage is called inappropriate. Names have no pre-­ determined objects. One forms agreement in order to name objects. Once the agreement is set and has become custom, then they are called names of objects. Names do have a pre-­ determined goodness. If they are straightforward, simple, and do not conflict, then they are called good names. (Hutton 2016: 239)

In other words, the correct names are correct not because there is a natural fit between the names (it is unclear if Xúnzǐ was thinking of the written or spoken form) and the distinctions they mark. Rather, the relevant connections hold as a matter of human convention. While the conventionality of language is not directly mentioned in the Analects, the notion is compatible with what the earlier text says about language and its relation with ethics and politics. A more striking innovation from Xúnzǐ lies in his

19 20

 See Fraser 2016: 292.  See especially Chapter of the Xúnzǐ (Hutton 2016: 35).

346

H. C. Loy

understanding of what is happening when names are not correct: he highlights the rise of “disputation and litigation” amongst people. In fact, much of the rest of the essay will target specific philosophical slogans that, according Xúnzǐ, “are cases of confusion about the use of names leading to disordering names”. These slogans include: “To be insulted is not disgraceful,” “The sage does not love himself,” “To kill a robber is not to kill a man,” “Mountains and gorges are level,” “The desires of one’s natural dispositions are few,” “Fine meats are not any more flavorful,” “Great bells are not any more entertaining” and “Oxen and horses are not horses” (Hutton 2016: 239–240). So unlike the Analects, which seemed to have the occurrent misapplication of individual terms to things and affairs (especially by the political elite) in mind when talking about zhèngmíng, Xúnzǐ identifies general philosophical claims as examples of names incorrectly used, or at the very least, as being ultimately due to names incorrectly used. If anything, this aspect of Xúnzǐ bears a closer connection with something that can be traced in the Mencius, namely, the idea that some moral and political problems have their causes in the propagation of pernicious doctrines (yán 言)—often expressed or identified by verbal maxims. The parallels with Mencius are striking. Consider what Xúnzǐ says about the need for persuasion: The people can easily be unified by means of the Way, but one should not try to share one’s reasons with them. Hence, the enlightened lord controls them with his power, guides them with the Way, moves them with his orders, arrays them with his judgments, and restrains them with his punishments. Thus, his people’s transformation by the Way is spirit-like. What need has he for demonstrations and persuasions (biànshuō辨說)? Nowadays the sage kings have all passed away, the whole world is in chaos, and depraved teachings are arising. The gentleman has no power to control people, no punishments to restrain them, and so he engages in demonstrations and persuasions. (Hutton 2016: 240)

And so in the very passage (3B9) in which he explains to his disciple Gongduzi why he seemed “fond of disputation” (hàobiàn 好辯), Mencius also pointed out that given the absence of sage kings, “The feudal lords do as they please; people with no official position are uninhibited in the expression of their views, and the doctrines of Yáng Zhū and Mò Dí fill the world”. And all that is left for men like him who desire to “safeguard the way of the former sages” is to forcefully dispute with the purveyors of “excessive views”, i.e., the Yang and Mo of the world. The crucial point is that the above cluster of ideas is missing from the Analects: there is no clear sense in the Analects that bad thinkers and their pernicious doctrines—as distinguished from bad rulers, self-serving ministers, or more generally, a social-political elite that conducts itself badly, even if because of mistaken notions—are part of the problem facing the world. What Xúnzǐ did is to subsume the issue of bad doctrines, previously identified by Mencius (and Mòzǐ 墨子), under the rubric of zhèngmíng. In so doing, Xúnzǐ goes beyond Mencius: for unlike Mencius (and Confucius), Xúnzǐ is much more explicit in being optimistic that words—a discourse made up on names—can in some sense be a fit vehicle for conveying the Way: Names are the means by which one arranges and accumulates objects. Sentences combine the names of different objects in order to discuss a single idea. Persuasion and demonstration use fixed names of objects in order to make clear the proper ways for acting and

18  Correcting Names in Early Confucianism

347

remaining still. Procuring agreement and naming are the functions of demonstration and persuasion. Demonstration and persuasion are the heart’s way of representing the Way. The heart is the craftsman and overseer of the Way. The Way is the warp and pattern of good order. When the heart fits with the Way, when one’s persuasions fit with one’s heart, when one’s words fit one’s persuasions, then one will name things correctly and procure agreement, will base oneself on the true disposition of things and make them understood, will discriminate among things without going to excess, and will extend by analogy the categories of things without violating them. When listening to cases, one will accord with good form. When engaging in demonstration, one will cover thoroughly all the reasons. One will use the true Way to discriminate what is vile just like drawing out the carpenter’s line in order to grasp what is curved and what is straight. Thus, deviant sayings will not be able to cause disorder, and the hundred schools will have nowhere to hide. (Hutton 2016: 241)

That is, the crucial task, as Xúnzǐ conceived it, was not merely to confute specific bad doctrines (what Mencius might be thought of as doing when he debated various objectors and doctrinal rivals), but to attempt a comprehensive and argued presentation of the Confucian Way in words—exactly what he did in his writings. At this point, the way Xúnzǐ’s conception of zhèngmíng has gone beyond Confucius parallels the way the form of Confucian philosophical literature has evolved between the Analects, the Mencius, and the Xúnzǐ. In particular, Confucian zhèngmíng has now become intertwined with the concept of biàn (disputation). For Xunzi, the point of a disputation is, in effect, to “correct” or settle the application of names. This means that, in practice, the activity of zhèngmíng now overlaps with that of bian: The latter highlights the idea that reasons or explanations are given to support an act of distinguishing things given their kinds (lèi), which culminates in the use of a name (míng) one way rather than another. The former, in contrast, emphasizes the idea that the use of a name (míng) one way rather than another—an act to be supported by reasons and—is itself embedded within a larger concern with living well both individually and collectively. Reflections on the significance of zhèngmíng are thus also informative about the significance of biàn more generally, and insofar as biànshuō is (partly or roughly) an early Chinese analogue to logic, the discussion of zhèngmíng tells us something about how the purpose and function of logic were understood by early Chinese thinkers.

References Ames, Roger T., and Rosemont Jr, Henry. trans. 1998. The Analects of Confucius: A philosophical translation. New York: Ballantine. Bao, Zhiming. Apr 1990. “Language and World View in Ancient China.” Philosophy East and West 40.2: 197–99. Brooks, E.  Bruce and Brooks, A.  Taeko eds., trans. 1997. The Original Analects: Sayings of Confucius and His Successors. New York: Columbia University Press. (Translation and study of the Analects, and major attempt to discern the different lays of composition in the text.) Burke, Kenneth. 1969. A Grammar of Motives. University of California Press. Cai Mingtian 蔡明田. 1984 “On the Thought of Correct Terms in the studies of Benevolence of Confucius 論孔子仁學中的正名思想.” Journal of the Kongzi-Mengzi Society 孔孟學報 48: 1–26.

348

H. C. Loy

Cheng, Shude 程樹德. 1990. Collected Explanations of the Analects 論語集釋. 4 vols. Beijing: Zhonghua shuju. (Very useful compilation of the major traditional commentaries on the Analects.) Cheng, Zhongying 成中英. 1981. “A Discussion of Confucius’ ‘zhèngmíng’ 論孔子的正名思 想.” In Chinese Philosophy and Chinese Culture中國哲學與中國文化. Taibei: Sanmin shuju. Creel, Herrlee G. 1951. Confucius, the Man and the Myth. New York: Routledge & Keegan Paul. (Older but still useful study of the Analects.) Defoort, Carine. 1998. “The Rhetorical Power of Naming: The Case of Regicide”. Asian Philosophy 8.2: 111–118. Defoort, Carine. 2000. “Can Words Produce Order? Regicide in the Confucian Tradition”. Cultural Dynamics 12.1: 85–109. Defoort, Carine. 2013. “Modern Interpretation of “zhèngmíng” in the West: How Much of an Influence Did Hu Shih Have 当代西方人对 “正名” 的解释 :胡适的影响有多大.” In Fudan Lecture Series on Literature and History vol. 5: Open the Window to Enlighten the House 复旦 文史讲堂之5: 牖启户明. Shanghai: Zhonghua shuju. Durrant, Stephen, Wai-Yee Li, and David Schaberg trans. 2016. Zuo Tradition/Zuozhuan: Commentary on the “Spring and Autumn Annals.” 3 Vols. Seattle: University of Washington Press. Fraser, Chris. 2016. “Language and Logic in the Xunzi”. In Eric Hutton, ed., Dao Companion to the Philosophy of Xunzi. New York: Springer. Fung, Yu-lan 馮友蘭. 1952. A History of Chinese Philosophy, vol. 1. Trans. Derk Bodde. Princeton: Princeton University Press. (Major survey of Chinese Philosophy.) Graham, A. C. 1978. Later Mohist Logic, Ethics, and Science. Hong Kong: The Chinese University Press. Hsiao, Kung-chuan. 1979. A History of Chinese Political Thought Vol. 1: From the Beginnings to the Sixth Century A. D. F. W. Mote trans. Princeton: Princeton University Press. Hu, Shih胡適. 1963. The Development of the Logical Method in Ancient China. 2nd ed. New York: Paragon Book Reprint Corp. (Study of Early Chinese Philosophy the author of which is a major 801 twentieth century Chinese thinker in his own right.) Hutton, Eric L. trans. 2016. Xunzi: The Complete Text. Princeton: Princeton University Press. Lau, D.C. trans. 1992. Confucius: The Analects. 2nd ed. Bilingual edition. Hong Kong: Chinese University Press. Lau, D.C. trans. 1984. Mencius. Volume One. Bilingual edition. Hong Kong: Chinese University Press. Lao, Sze-Kwang 劳思光. 1984. History of Chinese Philosophy 中国哲学史. Taibei: Sanmin Shuju. Lin, Yuanqi 林远琪. 1979. “Kongzi’s zhèngmíng Ideology and Political Thought.” Cong-Meng Monthly 孔孟月刊. 18.2: 9–18. Legge, James. 2006. The Chinese Classics. Vol. I: Confucian Analects, The Great Learning, The Doctrine of the Mean. Taipei: SMC Publishing. (Older but still very useful translation of the Analects with copious notes.) Leys, Simon trans. 1997. The Analects of Confucius. New York: W. W. Norton. Liu, Zhenghao 劉正浩. 1978. “An Examination of Kongzi’s zhèngmíng.” Journal of the Kongzi-­ Mengzi Society 孔孟學報. 36: 157–168. (Useful survey of the traditional opinions on the doctrine of “Correcting Names”.) Loy, Hui-Chieh. 2014. “Language and Ethics in the Analects.” In Amy Olberding ed., Dao Companion to the Analects. New York: Springer. Loy, Hui-Chieh. 2008. “Analects 13.3 and the Doctrine of ‘Correcting Names’.” In David Jones ed., Confucius Now: Contemporary Encounters with the Analects. Chicago: Open Court. (One in a collection of recent studies of the Analects.) Knoblock, Johned. and trans. 1988–1994. Xunzi: A Translation and Study of the Complete Works. Stanford: Stanford University Press.

18  Correcting Names in Early Confucianism

349

Makeham, John. 1994. Name and Actuality in Early Chinese Thought. Albany: State University of New York Press. (A study of XU Gan’s theory of naming, with a chapter on the Analects’ doctrine of correcting names.) Makeham, John. 2003. Transmitters and Creators: Chinese Commentators and Commentaries on the Analects. Cambridge, MA/London: Harvard University Asia Center. McLeod, Alexus. 2015. Theories of Truth in Chinese Philosophy: A Comparative Approach. London: Rowman & Littlefield International. Needham, Joseph ed. 1954. Science and Civilization in China, Vol 2. Cambridge: Cambridge University Press. Reding, J.P. 1985. Les fondements philosophiques de la rhétorique chez les sophistes grecs et les sophistes chinois. New York: Peter Lang. Slingerland, Edward T. trans. 2003. Analects: With Selections from the Traditional Commentaries. Indianapolis: Hackett. Sima, Qian 司马迁. 1982. Records of the Grand Historian 史记. Vols. 4 and 5. Beijing: Zhonghua shuju. Van Norden, Bryan W. 2007. Virtue Ethics and Consequentialism in Early Chinese Philosophy. Cambridge: Cambridge University Press. (Important recent study of early Chinese Philosophy presenting Confucianism as a form of virtue ethics.) Waley, Arthur. 1939. The Analects of Confucius. London: George Allen and Unwin. Wang, Xianqian 王先谦. 1988. Collected Explanations of the Xunzi 荀子集解. Beijing: Zhonghua Shuju. Xu, Fuguan 徐复观. 1966. Sparse Explanations on the Gongsunlongzi 公孙龙子讲疏. Taizhong: Sili Donghai Daxue. Yang, Bojun 楊伯峻 trans. 1984. Annotated Translation of the Analects 論語譯注. Beijing: Zhonghua shuju. (Translation of the Analects in Modern Chinese.) Zhang, Dai’nian 张岱年. 1982. An Overview of Chinese Philosophy 中國哲學大綱. Beijing: Zhongguo chehui kexue chubanshe. Hui-chieh Loy,  is Associate Professor in the Department of Philosophy at the National University of Singapore. His articles can be found in International Philosophical Quarterly, Journal of Chinese Philosophy, Monumenta Serica, Philosophy East and West, Oriens Extremus, and Dao. He has also contributed to the Dao Companion to the Analects, the Dao Companion to the Philosophy of Xunzi, and other volumes on Chinese philosophy.

Chapter 19

Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi Kim-chong Chong

When I was an undergraduate, one of my Philosophy teachers remarked that “There are no arguments in Chinese Philosophy.” This seemed to be confirmed when I took a course on Confucius’s Analects. The person who taught the course was a traditionalist who emphasized moralistic teachings in the text. It was not until many years later when I seriously began to read the Chinese texts that I realized the Chinese philosophers do argue, in the logical sense. In previous publications, I have demonstrated the logical structure of arguments in the debate between Mencius and Gaozi about human nature, and in Xunzi’s criticism of Mencius’s theory of human nature (Chong 2007). In this essay, I shall mention some of these arguments again, but more in the context of describing and illustrating the nature of analogical arguments and how they function as such. In the same regard, I shall also bring together work that I have done on the Zhuangzi (Chong 2006, 2011, 2016). One reason for the impression that “there are no arguments in Chinese Philosophy” may have to do with how analogies and arguments by analogy are ordinarily viewed and described. This is reflected in Irving Copi’s well-known Introduction to Logic, for instance. According to Copi, argument by analogy is inductive, where the conclusion can only be said to be probably true. For example, the inference that the new pair of shoes I have bought will wear well because my old pair, purchased from the same store, has worn well. Another example is that I will enjoy a new book by an author because I have enjoyed reading other books by the same author. In terms of these examples, Copi says that “Analogy is at the basis of most of our ordinary reasonings from past experience to what the future will hold” (Copi 1986: 404). Also, “analogy is not difficult to define. To draw an analogy

K.-c. Chong (*) Division of Humanities (Philosophy), Hong Kong University of Science &Technology, Kowloon, Hong Kong e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_19

351

352

K.-c. Chong

between two or more entities is to indicate one or more respects in which they are similar” (Copi 1986: 405). Contra Copi, these examples should not be labeled as arguments by analogy. In analyzing the first example above, he says, “There are three points of analogy involved: the respects in which the two entities are said to resemble each other are, first, in being shoes; second, in being purchased from the same store; and third, in wearing well” (Copi 1986: 404). In the second example, the supposed analogies are the comparison of two books by the same author and that they make for enjoyable reading. But the comparison of these similar things should not be called “analogies” in the first place. More exactly, the form of these examples is inductive. They state probabilities in the way mentioned above, namely, reasoning “from past experience to what the future will hold.” However, this states the problem of induction (whether we are justified in thinking that future events will be like what has happened in the past) instead of saying anything much about analogical argument. And as we shall see, analogies not only compare similar things in the very dull ways in the examples mentioned. Instead, they may also compare different things in a striking manner to bring about an effect, both argumentative and non-argumentative. Copi describes two analogies that he thinks are non-argumentative. One serves the purpose of “lively description”: “The literary uses of analogy in metaphor and simile are tremendously helpful to the writer who strives to create a vivid picture in the reader’s mind.” He cites a description of the encyclopedic nature of certain history books which make no connection between different facts: “Events stand isolated below headings of black type, like islands in some archipelago where no one has yet invented the canoe.” Another serves the purpose of explanation, “where something unfamiliar is made intelligible through being compared to something else, presumably more familiar, to which it has certain similarities.” For example, the explanation “Science is built up with facts, as a house is with stone. But a collection of facts is no more a science than a heap of stones is a house” (Copi 1986: 405). Note that the analogies just mentioned compare different things. And it is correct to note that analogies sometimes serve non-argumentative purposes. However, they may be more than just “lively description” or “explanation.” In other words, we should not miss the critical nature of the analogies mentioned. In the first example, history books which make no connection between events are compared to isolated islands lying throughout an archipelago. This makes the critical point that a book which does not explain connections between different events lacks coherence and is unworthy as a book of history. The analogy makes the point more effectively than if it is plainly stated. To understand the criticism though, one needs some shared background in the study of history. Someone who has no conception of history and standards of historical explanation may not be able to appreciate the point of the analogy. Similar things may be said about the analogy between the facts of science and the stones of a house, namely, that a collection of facts does not constitute science just as a collection of stones do not make a house. Copi says that its purpose is to explain what science is. But there is a context for the analogy, namely, the critique of a

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

353

c­ onception of science as being inductive in nature. This criticism was most notably made by Karl Popper in his book The Logic of Scientific Discovery (Popper 2002). Thus, the analogy between the facts of science and the stones of a house is not just an explanation of what science is. Instead, it has a critical purpose in the context of a debate about the nature of science and scientific discovery. The above discussion allows us to see the following features of analogies. First, analogies do not just compare similar things in an inductive way. Instead, they may compare different things in critical and analytical ways as we shall see in discussing the arguments of Mencius (Mengzi 孟子), Zhuangzi (莊子) and Xunzi (荀子). These comparisons point to both similarities and differences and lead to awareness and questioning of certain assumptions in a way of thinking or a line of argument. Second, even when an analogy is seemingly non-argumentative, it may still have a critical function. An awareness of this function will be available, however, only given knowledge of contexts of argument, such as debates about certain ideas between different schools of thought. Someone without knowledge of these contexts or who is insensitive to argumentative forms will be apt instead to think of the analogies as just “lively descriptions.” Third, an analogy may be so deeply entrenched in a school of thought as to serve as a metaphorical model for a particular concept and a way of thinking. In order to bring out and to question this entrenched model, a rival thinker may make use of alternative analogies and metaphors. We shall elaborate on this below in terms of how Zhuangzi questions the Confucian model of the xin (心) or the heart-mind as a “ruler” in charge of the other organs and senses of the body.

1  Mencius and Analogical Parallel Despite his definition of analogies and analogical thinking as being inductive, Irving Copi separately describes “refutation by logical analogy” as follows: There is a special kind of argument that uses an analogy to prove that another argument is wrong, or mistaken. It is intended to refute that other argument….by showing that the other argument’s premises do not really support the conclusion that was claimed to follow from them. We have here a basic method of appraising an argument as unsatisfactory from a logical point of view. Underlying this method of criticizing arguments is the fact that from the point of view of logic, the form of an argument is its most important aspect (Copi 1986: 421).

This application of analogies to show the unsatisfactory form of an argument is exactly what Mencius does when he questions Gaozi’s (告子) claim that ren (仁 here understood as “love”) is internal, while yi (義 rightness or righteousness) is external. This is in Mencius 6A:4 (Lau 2003). Gaozi has just claimed that “Appetite for food and sex is nature. Ren is internal, not external; rightness is external, not internal.” The mention of desires for food and sex together with ren indicates that Gaozi thinks that ren is “internal” in the same sense as these desires. Later on, he gives as an example of ren (仁) the love for one’s own brother and not someone else’s brother. The word yue 悅 is used here, indicating that in loving my brother, I

354

K.-c. Chong

am doing what “pleases” me. Again, this shows that by “internal,” Gaozi has in mind the desires. By contrast, he thinks that yi is external, citing respect for an elderly person. The respect is due to his social position as an elder. Thus, one does what is yi in circumstances under which it is appropriate. Here, Gaozi makes an analogy with the color white—elderliness and respect for the elderly are external qualities, just as something is said to be white in color. Mencius replies that (admittedly) there is no difference between the “whiteness” of a white horse and of a white person, but is there no difference between the two “zhang” in zhang ma 長馬 (old horse) and zhang ren 長人 (old or elderly person)? The ambiguity of zhang either as adjectival (attributing oldness) or as verbal (to treat with respect) is brought out in his next question: Whether we say that it is the object of respect—that is, the elderly person (zhang zhe 長者)—or the person evincing respect (zhang zhi zhe 長之者), who is yi? Here, Mencius is drawing attention to the fact that yi comes from the person who shows respect, not the object of respect. Gaozi replies that since there is no difference between respecting an elder from the state of Chu and my own elder, this is due to the quality of being elderly and hence it is called external. Presumably, in the case of one’s own elder, there is love as well as respect. But because the respect shown is identical in both cases, this is yi and it is an external quality that identifies whether respect is due. Replying to this, Mencius asks: Although there is no difference between relishing the roast of another person and my own, does it follow that relishing is external? In the next passage (6A:5) Mengjizi (a follower of Gaozi’s) speaks with Gongduzi (Mencius’s disciple). Mengjizi claims that since jing 敬 or respect varies with the circumstances, this shows that respect (as an expression of yi) is external. Thus, normally, respect is accorded my eldest brother over an elderly villager. In a village gathering, however, priority would be given to the villager. Gongduzi consults Mencius, who tells him to cite respect for a younger brother over an uncle in a ritual where the former is impersonating an ancestor. Responding to this example, Mengjizi replies that “It is the same respect whether I am respecting my uncle or my younger brother. It is, as I have said, external and does not come from within.” To which Gongduzi replies: “In winter, one drinks hot water, in summer cold. Does that mean that even eating and drinking can be a matter of what is external?” I refer to the analogies in the last paragraph as the roast and drink analogies. The following remarks of D. C. Lau’s (in an essay entitled “On Mencius’ Use of the Method of Analogy in Argument”) with reference to 6A:4 and 6A:5 helps us to see their overall point. The arguments are obviously not conclusive, but this is in part due to Mencius’ limited purpose. All he set out to do, in both cases, was to show that his opponents failed to establish the externality of yi. He did not attempt to go beyond this and to establish positively that yi was internal (Lau 2003: 384).

The analogical arguments are “obviously not conclusive” means that they are not deductive arguments. Nevertheless, they question how the claim that yi is “external” is established. Mencius is seeking a clarification of both “internal” and “external,” extending the question he had posed earlier in 6A:4, “Why do you say that ren is internal and yi is external?” He is pointing to a problem with how Gaozi uses inter-

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

355

nal/external because the evincing of respect is not just an external quality like whiteness or being elderly. The roast and drink analogies bring out the problem in terms of formal parallels with Gaozi’s and his follower’s analogies. The roast analogy parallels Gaozi’s argument that since respect for both a family and non-family member is the same, it must be external. Mencius points out that by the same token, there is no difference between relishing my own roast and another’s. Following Gaozi’s form of argument, the relishing must be external. The formal parallel can be shown schematically. Taking X to be “my” and Y to be “another,” Gaozi argues in 6A:4 that:1 (1) But Not: Therefore: (2) Similarly: Therefore:

愛X之弟 愛Y 之 弟

Love my brother Love another’s brother Ren is internal

長X之長 長Y 之 長

Respect (literally, “to elder”) my elder Respect another’s elder Respect is external

With reference to (1) and (2) Mencius provides the parallel: (3) Similarly: Therefore:

嗜X之炙 嗜Y 之 炙

Relishing my own roast Relishing another’s roast Relishing is external

It can be seen (3) has the same form as (2). However, the conclusion of (3) is absurd. This is in fact a reductio ad absurdum (or a reductio, in short) where an argument is shown to be unsatisfactory by displaying the absurd consequences of another argument which parallels it in form. The same formal parallel occurs in the drink analogy of 6A:5. Here, we take “X” for “uncle” and “Y” for “younger brother impersonating an ancestor in a ritual.” Mengjizi responds to the analogy of the younger brother impersonating an ancestor as follows: (4) Therefore:

敬 X 則敬 敬 Y 則敬 果在外

Respect for X then (makes up) respect Respect for Y then (makes up) respect (Respect is) external

Gongduzi provides the parallel to (4): (5) Therefore:

1

冬則飲湯 夏則飲水 然則飲食在外也?

 The following modifies Chong 2007: 43.

Winter then drink hot water Summer then drink cold water Drinking and eating is external?

356

K.-c. Chong

In both parallels, an absurd consequence follows: (1) Since there is no difference between relishing my own roast and another’s, it follows that relishing a roast is external. (2) Since different seasons call for different drink temperatures, it follows that drinking/eating is external. The formal parallel in each case questions the sense in which yi is held to be “external.” This means that Gaozi has failed to properly establish that yi is external. At the same time, we are made aware that this does not mean that yi is internal, as Gaozi understood it—in the same sense as the desires for food and sex. The same holds for ren.2 Most commentators have assumed that since Gaozi holds that ren is internal Mencius does not question this, instead questioning only the assertion that yi is external. However, Mencius is not counter-asserting that yi is internal. Remember D. C. Lau’s statement above, that “All (Mencius) set out to do, in both cases, was to show that his opponents failed to establish the externality of yi. He did not attempt to go beyond this and to establish positively that yi was internal.” Lau cites the following statements from the Xiaoqu of the Mohist Canons: Analogy is to put forth another thing in order to illuminate this thing. Parallel is to set [two] propositions side by side and show that they will both do (Lau 2003: 388).3

Here, Lau mentions the statement by Gaozi in 6A:3, “That which is inborn is meant by ‘nature’.” In order to clarify or “illuminate” what Gaozi means, Mencius offers the statement, “White is what is meant by ‘white’.” Gaozi accepts this tautology. Mencius then asks whether this means that the nature of a hound is the same as that of an ox, and the nature of an ox is the same as a man’s. Mencius means to show that it is absurd to hold that human nature is the same as the nature of an animal. Again, this refutation is not conclusive. Gaozi can insist, for instance, that there is no difference between the nature of humans and animals. Nevertheless, the very question whether Gaozi means to assert that the nature of humans and animals are the same raises doubts about the thesis. Lau also cites the following statement from the Mohist Canons: Things may have similarities, but it does not follow that therefore they are completely similar. When propositions are parallel, there is a limit beyond which this cannot be pushed (Lau 2003: 388).

Commenting on this, he says: This is a very good description of the method as it was used by Mencius. In his hands the method of analogy was used to throw light on things which were otherwise obscure. It is by proposing analogies and showing in what way they broke down that this was achieved (Lau 2003: 389).

2  Mencius will argue that yi and ren are internal in a certain moral sense but not in Gaozi’s sense of “internal” such as the desires for food and sex. 3  The clause “they will both do” (俱行也) seems only to say that there is a parallel, not that the parallel is acceptable. Compare A.C. Graham’s “‘Parallelising’ is comparing propositions and letting all ‘proceed’” (Graham 2003: 483).

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

357

From the above, we see the following functions of analogy in argument: drawing distinctions, clarifying terms, pointing to ambiguities, bringing out assumptions, and drawing formal parallels to show invalidity. In his concluding remarks about Mencius’s use of the method of analogy in argument, Lau notes the indispensable and wide use of the method for philosophical problems in the fourth and third centuries BCE in China, “so much so that the only surviving treatise on the methods of argument deals with it in some detail. Seen in the light of the Mohist treatise, Mencius was, indeed, a very skillful user of this method, who never failed to throw light on philosophical issues that were discussed. This is an impression somewhat different from the ineffective debater that he is sometimes made out to be” (Lau 2003: 389–390).

2  An Analogical Model of the Heart-Mind A further function of analogies is mentioned by Lau in the penultimate paragraph of his discussion of Mencius: It is perhaps worth pointing out that the use of analogy is often the only helpful method in elucidating something which is, in its nature, obscure. Two examples come readily to mind. Theories about the mind are often presented through the medium of models, and so are physical theories of the atom. In either case the models are not only helpful in enabling us to see something of the “structure” of the mind or the atom which is not open to inspection by the senses, but also instructive in the way they break down (Lau 2003: 389).

It is unclear whether Lau is thinking of Chinese theories of the xin or the “heart-­ mind” when he mentions “mind” in the above passage. However, what he says is applicable to how the heart-mind was widely thought of by Mencius, Xunzi and others. This is the model of the heart-mind being a “ruler” over the other organs and senses of the body. In other words, the heart-mind is thought of as being superior to the other organs and senses of the body and has a supreme role in guiding action. This is in fact an instance of what George Lakoff and Mark Johnson has described as “conceptual metaphor”: deep structures of thought where abstract ideas are modeled upon everyday things and interactions in our environment (Lakoff and Johnson 2003). In the present case, the heart-mind is modeled upon the role of being a ruler or master in relation to subordinates, including associations such as giving commands, leading them back to the right path, and so on. Lau’s remark about how a model may “break down” points to the limitation of a particular model of thought once its structure is revealed. The model of the heart-mind as ruler was deeply entrenched in the thinking of the Confucians and others.4 One indication of this is that although Xunzi disagreed with Mencius about the innate moral nature of the heart-mind, he too thinks of it in

4  In the Guanzi (管子), for instance, it is stated that “In the body the mind/heart is in the place of the prince” (Zhang 2002: 393).

358

K.-c. Chong

terms of the same model. Mencius and Zhuangzi lived during the same period. Sima Qian tells us that Zhuangzi was a contemporary of King Hui of Liang (梁惠王 370–319 BCE) and King Xuan of Qi (齊宣王319–301 BCE) (Sima Qian 1982: 2143–2144). The Mencius records conversations between Mencius and both kings. Although there is no record of Mencius and Zhuangzi interacting with or referring to each other, they lived amidst the same cross-current of ideas. Zhuangzi was aware of the claims about the heart-mind as the ruler over other organs and senses of the body. Before showing this, let us first describe both Mencius’s and Xunzi’s views of the heart-mind. For Mencius, the heart-mind possesses moral sprouts that allow for an integrally moral and reflective capacity which he calls the “greater” part (da ti 大體) of the person (6A:15). For instance, the sprout of compassion would arise in anyone who witnesses a child about to fall into a well (2A:6). Presumably, the person who witnesses this would rush forward spontaneously to save the child. Put in contemporary terms, this means that there is no cognitive “reason” here that stands independently of the sprout of compassion and there is no need for reflection either. Reflection comes in, however, when for some reason the agent has gone astray and in this respect the moral sprouts are always there to be recalled. As such, the heart-mind is different from the other organs of the body. Mencius says: 耳目之官不思, 而蔽於物。物交物, 則引之而已矣。心之官則思, 思則得之, 不思則不 得也。此天之所與我者。 The organs of hearing and sight are unable to (reflect) and can be misled by external things. When one thing acts on another, all it does is to attract it. The organ of the (heart-­ mind) can (reflect). But it will find the answer only if it does (reflect); otherwise, it will not find the answer. This is what Heaven has given me (6A:15).

As the “greater part,” the heart-mind has a superior guiding role in relation to the other organs of the body which make up the “smaller” part (xiao ti 小體). In other words, the desires and other affective states are subordinate to the heart-mind. We may control our desires and affective states by reflecting upon what is within our heart-minds. Thus, a higher position is given to the heart-mind. It has the capacity to think in specifically moral terms so as to guide and control action especially when the desires have gone astray. As mentioned, Xunzi disagrees with Mencius about the heart-mind’s having an innate moral capacity. Nevertheless, he too stresses its supreme nature. In “Dispelling Blindness (解蔽篇),” Xunzi uses the word bi (蔽) to describe what he thinks is a common fault, namely, to be “blinded” or “blinkered” by a single aspect of a situation and thus fail to see it comprehensively (Knoblock 1994: 21.1 onwards). This might involve holding to an extreme position or insisting on a particular doctrine. Alternatively it could involve being partial toward one’s own views or indulging one’s desires. In all these cases, it is held that the cognitive capacity of the heart-­ mind fails to function properly. Nevertheless, it is described as authoritatively independent of the body in judgment and command of action:

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

359

心者, 形之君也, 而神明之主也, 出令而無所受令。自禁也, 自使也, 自奪也, 自取也, 自 行也, 自止也。故口可劫而使墨云, 形可劫而使詘申, 心不可劫而使易意, 是之則受, 非之則辭。故曰:心容──其擇也無禁, 必自現, 其物也雜博, 其情之至也不貳。 The heart-mind is the lord of the body and master of the spiritual intelligence. It issues commands but does not receive commands. On its own authority it forbids or orders, renounces or selects, initiates or stops. Thus, the mouth can be forced to be silent or to speak. The body can be forced to crouch down or stretch out. But the heart-mind cannot be forced to change its ideas. If the heart-mind thinks something right, it will accept it; but if it thinks something wrong, then it will reject it. Therefore, it is said that the state of the heart-mind is such that of necessity it perceives on its own. No prohibitions can be placed on what it selects. Its objects are diverse and extensive. When it has perfect concentration, it is not divided in purpose. (Knoblock 21.6a modified, Li 1994: 488).5

Xunzi further elaborates this supreme position of the heart-mind. At one level, it is described as authoritative in commanding bodily movement and making judgments according to rational principles. Xunzi is impressed by its infinite capacity for fresh input, its ability to make distinctions and to concentrate on specific purposes. At another level, however, it can be “unsettled (bu ding不定).” An analogy is made with a pan of water (Knoblock 21.7b). If the pan is upright and still, sediment will sink to the bottom and the water will be clear. It can then reflect objects clearly. Similarly, the heart-mind’s ability to make distinctions can be affected if it is unsettled. From the examples provided it is clear that this occurs when the senses are affected by external conditions or impaired in some way. In other words, the heart-­ mind can function clearly only if the senses are in good working order under normal circumstances. Thus, poor lighting conditions affect vision and may lead one to identify objects wrongly. Alternatively, the senses may not be in good working order when a person is drunk or when a sense organ is interfered with in some way (Knoblock 21.8). Thus, it can be seen that both Mencius and Xunzi assume the supremacy of the heart-mind in different but related ways. For Mencius, the moral sprouts and the capacity to reflect upon them are integrated in a way that expresses moral superiority of the heart-mind. For Xunzi, superiority is maintained by the proper cognitive functioning of the heart-mind. And moral order is achieved by the discovery of principles governing human action in general. These principles are external to the heart-mind in that they may only be discerned through investigating the natural order, including understanding the inclinations and propensities of human beings. Contra Mencius, they are not described in terms of innate moral dispositions. Despite their disagreement, however, they share a picture or model of the ­heart-­mind as being in a superior position to the other senses and organs of the body, acting independently and being in charge of them.

5  See also Knoblock 17.3a where Xunzi says that the heart-mind can control and is lord over the other faculties, and 22.5a in which the heart-mind is said to be able to override the desires.

360

K.-c. Chong

3  Zhuangzi’s Questioning of the Model Zhuangzi both reveals and questions this model of the heart-mind in the “Qiwulun” (齊物論). In this chapter, the character Nanguo Ziqi contrasts the sounds of the hollows in the forest with the affective states of those—like the Confucians and the Mohists—who debate over the right and wrong of ethical, social and political issues. The sounds are what they are of themselves. That is, there is no motivational source behind these sounds which arise and cease naturally with the wind. The debaters, on the other hand, are driven by the states of their heart-mind: 日以心鬥。縵者, 窖者, 密者。小恐惴惴, 大恐縵縵。其發若機栝, 其司是非之謂也; 其留如詛盟, 其守勝之謂也; 其殺若秋冬, 以言其日消也; 其溺之所為之, 不可使復之 也; 其厭也如緘, 以言其老洫也; 近死之心, 莫使復陽也。喜怒哀樂, 慮歎變慹, 姚佚啟 態; 樂出虛, 蒸成菌。日夜相代乎前, 而莫知其所萌。 Daily, their heart-minds are in strife. (There are the) deliberately hesitant ones, sly ones, or secretive ones. Small worries sadden them, great worries dispirit them. They shoot off like an arrow, aiming at the rights and wrongs of others. They hold on to their views as if having made a covenant while guardedly awaiting victory. Like things that fade in autumn and winter, they get worn out by the day. They are drowned by what they do, and there is no making them turn back. As if sealed up and bound by cords, as if declining with age—their heart-minds are near death and there is no way to make them return to brightness. Joy and anger, sorrow and pleasure, anxiety and regret, fickleness and determination, vehemence and indolence, indulgence and extravagance—music emerging out of the emptiness, mushrooms springing up in warm moisture, alternating day and night before us—we know not from where they sprout. (My translation, Chen 1999: 46).

This description of the affective states enables us to make sense of what it means for Nanguo Ziqi’s heart-mind to become like “dead ashes” and why he has “lost” his self. The analogy of “dead ashes” suggests that his heart-mind is no longer “smoldering”—in other words he is not afflicted by the states just described. Zhuangzi equates these with wo or the self when, after the description, he says: “Without them there is no wo and without wo there is nothing for them to take hold of (非彼無我, 非我無所取).”6 This implies that the heart-mind does not have any capacity that stands over and above the affective states, independent of them and exercising control over them. As Zhuangzi says, “It seems as if there is a true lord but we just do not obtain any sign of it (若有真宰, 而特不得其眹).” This is taken further by the following analogy of the relation between the different parts and organs of the body: 百骸、九竅、六藏, 賅而存焉, 吾誰與為親?汝皆說之乎? 其有私焉?如是皆有為臣妾 乎? 其臣妾不足以相治乎?其遞相為君臣乎?其有真君存焉? 如求得其情與不得, 無益 損乎其真。 The hundred parts of the body, the nine openings and six inner organs, they all come together to exist as my body—which of them would I take as being intimately close? Would you be pleased with all of them? Or would there be a partiality here? Or would I just treat

 Some translators take bi 彼 to refer to an “Other” against which the wo 我 or the “I” or “Self” is contrasted. But it is more natural to take bi to refer to a list of emotions or feelings mentioned just before this passage in the context of the interminable debates between the Mohists and the Confucians. See for instance Chan 1963: 181.

6

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

361

them all as my servants? Would these servants not adequately govern one another? Or would they take turns to be ruler and servant? Would there be a true ruler existing among them? Whether we succeed in getting at the fact of the matter, there is no loss or gain from reality. (My translation, Chen 1999: 52)

The last sentence in this passage has been variously translated. For instance: “Whether or not we succeed in specifying His attributes has neither positive nor negative effect upon the truth of the Lord” (Mair 1994: 14); and “If we seek without success to grasp what its identity might be, that never either adds to nor detracts from its genuineness” (Graham 1981: 51). It might seem natural to take the last two words其真 qi zhen to refer to the the 真君 zhen jun or “true ruler.” Thus, qi zhen would mean “its truth” or “its reality,” to assert the existence of the “true ruler.” But the series of questions before this last sentence throw doubt on this. After posing these questions, however, it would seem that Zhuangzi is careful not to make any categorical denial or assertion and I take qi zhen as referring to a state of affairs that we would have no knowledge of. Hence my translation, “Whether we succeed in getting at the fact of the matter, there is no loss or gain from reality.” The last passage cited above suggests that the heart-mind is just one organ among the others and doubt is thrown on whether there is a true ruler. The organs may be said to represent different motivational drives that alternate between being ruler and servant. Thus, the passage indicates that the heart-mind cannot be taken as governing the affective states in any cognitively independent sense. As illustrated by the debaters, deliberations of right and wrong are inseparable from these states. In other words, deliberations about right and wrong are enmeshed with the debaters’ affective states of mind. Thus, the resulting judgments of right and wrong presuppose what Zhuangzi shortly after calls the cheng xin: 夫隨其成心而師之, 誰獨且無師乎? 奚必知代而心自取者有之?  愚者與有焉。未成乎 心而有是非, 是今日適越而昔至也。 If a man follows his cheng xin and makes it his teacher, then who is without such a teacher? What need is there for someone who comprehends the processes of change on the basis of his heart-mind? Even a stupid person has his teacher. Not to have established something in your heart-mind before (judging) right and wrong is like setting off for Yue today and arriving yesterday. (My translation, Chen 1999: 56).

The term cheng means that something is “settled,” “formed,” or “established.” Hence, to have a cheng xin is to have a “pre-established heart-mind.” In other words, something is already established in one’s heart-mind prior to judgments about right and wrong. As such, if we allow the cheng xin to be our “teacher”—to be authoritative—then even the stupid person would have such an authority. The analogy “setting off for Yue today and arriving yesterday” means that an inversion takes place in the process of judgment. Instead of being formed on the basis of some prior e­ vidence (such as through “comprehending the processes of change”) it is the judgment which comes before the evidence. This seems to merely say that people are prejudiced. However, the concept of the cheng xin has much wider implications. It makes a claim about the nature of the xin which I have all along translated as “heart-mind” instead of “heart” or “mind.” The advantage of “heart-mind” is that it preserves an ambiguous relation between its

362

K.-c. Chong

cognitive and affective (that is, emotional and motivational) functions. This ambiguity is exploited by Zhuangzi through the cheng xin which, when fully described, offers the view that the cognitive function of the heart-mind is not over and above the affective states and in charge of them in developing and controlling virtue as assumed by the Confucians. Zhuangzi’s characterizations of the cheng xin also question its association with epistemic moral certainty. For the Confucians, knowledge is closely associated with moral knowledge. This knowledge is objective in that what is right does not depend on an individual perspective. For Mencius, in any given situation (such as the paradigmatic example of the child about to fall into a well) the right action is determined by knowledge and expression of what is in the heart-mind and this is something which is universally given by Heaven (tian 天). Even where decisions seemingly need to be made, reflection upon the heart-mind will bring about the right answer. For Xunzi, moral knowledge is gained by the cumulative experience and hard work of sages who discerned the principles governing a moral and social order (Knoblock 23.5a). These principles are objectively out there in the world, so to speak, and action based upon them will constitute right action. Despite their differences, both Mencius and Xunzi reflect the confidence that the Confucians had about moral knowledge and right or righteous action, and also about the possession of necessary virtues that correspond to such action. Appropriate virtuous behavior is directed by the heart-mind which is able to control the influence of desires, motives, feelings, and other so-called “external” factors. Put against the Confucian model of the heart-mind, it can be seen that Zhuangzi’s ideal conception of the heart-mind as “like a mirror—going after nothing, welcoming nothing, responding but not storing” (Watson 1968: 97) is non-epistemic. In other words, obviating the influence of the cheng xin does not involve certainty of knowledge such as in upholding an ethical conception of self. When Zhuangzi speaks of the “clarity (ming 明)” of the heart-mind in the Qiwulun, he uses the metaphor of the mirror to clear the heart-mind of being attached to distinctions. Thus, “clarity” in this context does not refer to an epistemic project of critical inquiry in determining the “truth” about things or moral knowledge. Instead, it is a stilling of the heart-mind. Similarly, the “fasting” of the heart-mind and the related notions of “losing” or “forgetting” the self are congruent with the metaphor of the mirror and also, elsewhere, “goblet words (zhi yan 卮言)” (Watson 1968: 304). These are all different aspects of a central metaphorical motif in the Zhuangzi. The metaphorical structure of goblet words has been described by Shuen-fu Lin as follows: [T]he zhi—a wine vessel used as a metaphor for the mind [xin]—is originally empty and gets temporarily filled with liquid—a metaphor for words—which comes from a larger wine container only when the occasion requires one to do so….Since the mind [xin] is like the zhi vessel without any fixed or constant rules or values of its own stored in there, and takes ideas always from outside when the occasion for speech arises, it will never impose artificial distinctions and discriminations upon things. This is what [is meant by] ‘mindless’—the ‘mind’ to be done away with here is, of course, the cheng xin or ‘fully formed mind-heart’…. Zhi yan, then, is speech that is natural, unpremeditated, free from preconceived values, always responding to the changing situations in the flow of discourse, and always returning the mind to its original state of emptiness as soon as a speech act is completed (Lin 1994: 65).

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

363

In several cases, the structure of Zhuangzi’s words resembles the operation of the goblet. The strategy here is to take a particular distinction (good/bad, right/wrong, this/that, allowable/not allowable, self/other, being/nonbeing, beginning/no beginning, true/false, etc.) and through an outpouring of paradoxes and infinite regresses, “empty” the heart-mind of the distinction. The following example comes directly after Zhuangzi says that he does not know whether his words fit into the category of other people’s, but it obviously fits into some category, and in that respect, is no different from the words of others. However, let me try making my statement (yan 言). There is a beginning (you shi 有始). There is a not yet beginning to be a beginning (未始有始). There is a not yet beginning to be a not yet beginning to be a beginning. There is being (you 有). There is nonbeing (wu無). There is a not yet beginning to be nonbeing (未始有無). There is a not yet beginning to be a not yet beginning to be nonbeing. Suddenly there is nonbeing. But I do not know, when it comes to nonbeing, which is really being and which is nonbeing. Now I have just said something (you wei 有謂). But I don’t know whether what I have said has really said something or whether it hasn’t said something (Watson 1968: 43, Chen 1999: 79).

These infinite regresses have the effect of taking away what was initially a clear distinction, in this case, between beginning/no beginning on the one hand, and between being/nonbeing on the other. Although Zhuangzi has said something, it could also be that he has not said anything. Either he has said something or he has said nothing. But in saying that he has said nothing, he has said something. In contemporary terms, this is a semantic paradox. A well-known example is the liar’s paradox where the cognitive content of “This sentence is false” cannot be determined. The metaphorical structure of goblet words serves to deflect a particular criticism of the status of Zhuangzi’s own words. This arises because Zhuangzi seems critical of the Confucians and the Mohists for their never-ending disputes about what is right and what is wrong. Zhuangzi alleges that they have pre-established heart-­ minds. But what is the status of Zhuangzi’s own words when he says this? The use of goblet words allows Zhuangzi not to be pinned down to any particular position. We may say that Zhuangzi has a creative use of metaphor, instead of using metaphors unconsciously as described by Lakoff and Johnson  (Lakoff and Johnson 1999: 13). There is a conscious and active use of metaphor in the Zhuangzi, with particular reference to “goblet words” and also other forms of metaphor.

4  Xunzi’s Capacity/Ability Distinction and “Potter Analogy” We described above how Xunzi, like Mencius, holds the Confucian model of the heart-mind as ruler. However, Xunzi has a more nuanced position regarding human agency. This can be seen in terms of the distinction that he makes between capacity and ability (Knoblock 23.5b) as compared with Mencius, who regards them as equivalent. For Mencius, moral action is a matter of calling upon resources entirely within the individual agent’s control. In conversation with King Xuan of Qi, Mencius

364

K.-c. Chong

makes the analogy of not needing strength (bu yong li 不 用 力) to lift a feather (1A:7). This highlights the fact that King Xuan’s subjects are not protected because the king does not practice benevolence (bu yong en 不 用 恩). Mencius concludes that King Xuan is not being a true king because he does not act (bu wei 不為), not because he is unable (bu neng 不能). The point also is made when Mencius is asked whether everyone is capable of becoming a Yao or a Shun (sages). His reply is that there is no difficulty: all one has to do is act. The problem is not one of strength but simply not acting (fu wei er 弗 為 耳) (6B:2). When King Xuan asks about the difference between not acting and inability, Mencius cites “Striding over the North Sea with Mount Tai under your arm” as an example of inability. Someone who claims that he is unable to make obeisance to elders is not acting. The king’s not being a true king is like the latter. No one possesses the strength to “Stride over the North Sea...” Obeisance to elders and being a true king, on the other hand, do not require strength at all. In using analogies such as “lifting a feather” and “striding over the North Sea,” Mencius is persuading King Xuan that he has both the capacity and ability to act with compassion. Referring to an incident in which King Xuan had saved an ox from being sacrificed, Mencius asks, “Why is it then that your bounty is sufficient to reach animals yet the benefits of your government fail to reach the people?” All that is needed is for the king to “take this very heart and apply it to what is over there.” Mencius is trying to make the king see that if he can feel for the ox, he can also feel for his people. At the same time, however, Mencius’s way of talking belies an assumption about how a moral agent is able to act so as to fully realize his or her moral capacity. Behind this is an organic model of the heart-mind according to which it will grow and develop to have the ability to function as it naturally should. This organic analogy is evident in how Mencius talks of the “sprouts” of the heart-­ mind and how it is like plants which, properly nurtured, will grow well (6A:7). When reminding King Xuan that he has both the capacity and the ability to be compassionate, Mencius likens this to the ability of the eyes to see—all one has to do is to look. Xunzi brings out this organic model with the following analogy. Those who say man’s inborn nature is good admire what does not depart from his original simplicity and think beneficial what is not separated from his childhood naiveté. They treat these admirable qualities and the good that is in man’s heart and thoughts as though they were inseparably linked to his inborn nature, just as seeing clearly is to the eye and hearing acutely is to the ear. Thus, inborn nature they say is “like the clear sight of the eye and the acute hearing of the ear” (Knoblock 23.1d).

Barring any interference, the capacity and the ability to see are the same. In this sense, the eye (with its capacity to see) is inseparable from eyesight (the actual ability to see). The structure that allows for the capacity to see at the same time constitutes the ability to see. Mencius talks of moral capacity and ability in the same organic terms. Thus, this assumes that the alleged innate capacity to be sensitive to the suffering of others has the same organic structure as being fully compassionate.

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

365

Xunzi, however, states that capacity and ability “are far from being the same, and it is clear that they are not interchangeable.”7 One way of interpreting this is to say that although someone may have a certain capacity she may not succeed in manifesting this as ability to do something. However, this can be accommodated by Mencius in terms of a lack of the appropriate circumstances that would allow for this manifestation. What prevents the capacity from being manifested as ability is the interference of inappropriate circumstances (or the absence of appropriate circumstances). But Xunzi pre-empts this move by denying a structural similarity between an innate capacity and its manifestation as ability. He does this in terms of the following “potter analogy”: The potter molds clay to make an earthenware dish, but how could the dish be regarded as part of the potter’s inborn nature? The artisan carves wood to make a vessel, but how could the wooden vessel be regarded as part of the artisan’s nature? The sage’s relation to ritual principles is just like that of the potter molding clay. This being so, how could ritual principles, morality, accumulated effort, and acquired abilities be part of man’s original nature? (Knoblock 23.4a).

Xunzi is questioning the assumption that before an object can be produced, its structure must be innately present in the producer. Similarly, it should not be assumed that the processes, rules, and structures of ritual and morality must already have been present (innately) before being instituted and practiced. As we have mentioned, Mencius thinks of moral development in organic terms of plant growth, say, from seed to plant. Under this metaphor, the capacity for compassion is not ability until it is developed, “nourished” or “cultivated.” At the same time, however, it is implicitly thought that compassion is already in the heart-mind as a sprout. There is certainly a legitimate use for a metaphor such as the development of seed to plant in the context of moral development and education. Mencius is talking in a morally earnest way of the need for displaying compassion. This is put both in terms of the idea that one possesses the seed of compassion and the idea that one is already fully compassionate. What enables this kind of ambiguity, however, is the belief that there is an incipient organic structural similarity between the seed and the plant, as an analogue to the sprouts of the heart-mind and the fully developed heart-mind. The potter analogy made by Xunzi is aimed precisely at this belief. It argues that ritual/morality is a constituted product with a different structure from the original capacities that may have given rise to it. Insofar as it has a different structure, it cannot be said to have been inborn. We can construe Xunzi to be pointing to a variant of the genetic fallacy: one should not assume that morality or goodness has the same form as the capacities that give rise to them. These capacities are not necessarily “moral” capacities.

7  My translation differs from Knoblock 23.5b: “That something is possible or impossible is entirely dissimilar from having or not having the ability to do so.”

366

K.-c. Chong

It should be noted that Mencius builds upon similar remarks made by Confucius. In Analects 6.12  (Lau 1992), Ran Qiu says that he lacks the strength to practice Confucius’s teaching, namely, about moral and ritual self-cultivation. In reply, Confucius makes an analogy between someone who collapses in a race and Ran Qiu who has “set the limits” beforehand. In other words, he simply does not make any effort. This analogy spells out a difference between the requirement of strength on the one hand, and the voluntariness of self-cultivation. Mencius emphasizes this difference with the analogy of “Striding over the North Sea with Mount Tai under your arm.” Against this, Xunzi takes a more nuanced position, providing us with the following analogy: The feet have the capacity to walk across the width of the world, but no (ordinary) person has been able to do so.8 In other words, everyone has the capacity to do so, but only certain people (sages) are able to do so.

5  Conclusion We began with an analysis of Irving Copi’s view of analogies and analogical argument as being inductive in nature. We showed, however, that analogies are not inductive comparisons of similar things. The examples of analogical arguments from Mencius, Zhuangzi and Xunzi serve critical functions in the course of philosophical argument and debate. These analogical arguments are not deductive in nature and therefore cannot conclusively prove a point of view. But this is a misleading comparison, making it seem as if philosophical argument must be conclusive. By pointing to both similarities and differences between things, analogies serve to make distinctions, force the clarification of terms, question assumptions, and reveal entrenched metaphorical models of thought. These are analytical, argumentative functions and not lively descriptions, as sometimes thought. We mentioned D. C. Lau’s remarks about the indispensable and wide use of the method of argument by analogy for philosophical problems in the fourth and third centuries BCE in China, to the extent that the Mohist Canons deals with it in some detail. There is therefore a history and a context for debates about these problems. We have focused in this essay on debates about human nature and the nature of human agency. Someone unfamiliar with the contexts of debate may not be alert to the arguments, thinking perhaps that a particular writer is only giving a lively description with his analogies. For instance, someone who reads Zhuangzi’s fanciful description of the heart-mind in the Qiwulun without an understanding of the Confucian model of the heart-mind as ruler will miss the critical nature of his description. Similarly, there is a critical context for Xunzi’s distinction between capacity and ability as seen in the light of Mencius’s statements about their equivalence in terms of an organic model of the heart-mind. 8  Compare Knoblock 23.5b: “It is possible for a man to travel by foot across the width of the whole world, yet there has never been a case where anyone was able to travel across the world by foot.” I interpret Xunzi to be saying that certain people, namely the sages, have the ability which the ordinary person does not have. For details, see Chong 2007: 70–71.

19  Analogical and Metaphorical Thinking in the Mencius, Xunzi and Zhuangzi

367

References Chan, Wing-tsit. 1963. A Source Book in Chinese Philosophy. Princeton: Princeton University Press. (Contains translations and comments on the central sources of Chinese Philosophy). Chen, Gu-ying 陳鼓應 (ed.). 1999. Zhuangzi Jin Zhu Jin Yi  莊子今註今譯. Taipei: Taiwan Shangwu. (A contemporary commentary on the Zhuangzi). Chong, Kim-chong. 2006. Zhuangzi and the Nature of Metaphor. Philosophy East and West 56.3: 370–391. (Discusses the nature and modes of metaphor in the Zhuangzi). Chong, Kim-chong. 2007. Early Confucian Ethics. Chicago: Open Court. (Discusses the ethics of Confucius, Mencius and Xunzi). Chong, Kim-chong. 2011. “Zhuangzi’s Cheng Xin (成心) and its Implications for Virtue and Perspectives”. Dao 10.4: 427–443. (Discusses Zhuangzi’s critique of the Confucian assumption of moral autonomy). Chong, Kim-chong. 2016. Zhuangzi’s Critique of the Confucians: Blinded by the Human. Albany: State University of New York Press. Copi, Irving M. 1986. Introduction to Logic (Seventh Edition). New York: MacMillan Publishing Company. (An introductory textbook on both formal and informal logic). Graham, A. C. (trans.). 1981. Chuang-tzŭ—The Seven Inner Chapters and other writings from the book Chuang-tzŭ. London: George Allen & Unwin. Graham, A. C. 2003. Later Mohist Logic, Ethics and Science. Hong Kong: The Chinese University Press. (An analysis and translation of the Mohist Canons). Knoblock, John. (trans.). 1994. Xunzi—A Translation and Study of the Complete Works. Vol. 3. Stanford: Stanford University Press. Lakoff, George, and Johnson, Mark. 1999. Philosophy in the Flesh—The Embodied Mind and its Challenge to Western Thought. New York: Basic Books. Lakoff, George, and Johnson, Mark. 2003. Metaphors We Live By. Chicago: The University of Chicago Press. (Presents the theory of conceptual metaphor). Lau, D. C. (trans.). 2003. Mencius. Hong Kong: The Chinese University Press. Lau, D. C. (trans.). 1992. The Analects. Hong Kong: The Chinese University Press. Li, Disheng李滌生 (ed.). 1994. Xunzi ji shi 荀子集釋. Taipei: Xuesheng shuju. (Contemporary commentary on the Xunzi). Lin, Shuen-Fu. 1994. “The Language of the ‘Inner Chapters’ of the Chuang Tzu.” In W. J. Peterson, A. H. Plaks, and Y. S. Yü, eds., The Power of Culture—Studies in Chinese Cultural History. Hong Kong: The Chinese University Press. (Describes the modes of metaphor in the Zhuangzi). Mair, Victor (trans.) 1994. Wandering on the Way—Early Taoist Tales and Parables of Chuang Tzu. Honolulu: University of Hawai’i Press. Popper, Karl. 2002. The Logic of Scientific Discovery. London: Routledge. (Describes the logic and nature of scientific discovery). Sima Qian 司馬遷. 1982. Shi Ji 史記 (Vol. 7, juan 63). Beijing: Zhonghua Shuju. (The first history of ancient China). Watson, Burton (trans.) 1968. The Complete Works of Chuang Tzu. New York: Columbia University Press. Zhang, Dainian. Translated and edited by Edmund Ryden. 2002. Key Concepts in Chinese Philosophy. New Haven: Yale University Press. (Describes central concepts in Chinese Philosophy). Kim-chong Chong  is Emeritus Professor at the Hong Kong University of Science and Technology. He taught at the National University of Singapore from 1980 to 2003 and is the author of Early Confucian Ethics: Concepts and Arguments (Open Court, 2007), and Zhuangzi’s Critique of the Confucians: Blinded by the Human (SUNY Press, 2016)

Chapter 20

Problems of Language and Logic in Daoism Eske J. Møllgaard

1  Language and Logic The title of this chapter, “Problems of Language and Logic in Daoism,” implicitly raises the philosophical question of the relation between language and logic. Is logic the essence of language? Or is the essence of language prior to and more profound than logic? This question lies at the heart of Daoist philosophy, and leads Daoists to penetrating meditations on the essence of language. Daoists question logic in a broad sense encompassing the use of language to make valid inferences, to establish semantic coherence in naming things, and to form a coherent normative pattern of behavior. In questioning such logic Daoists bring to view language itself, not the knowledge language communicates, but language as pure communicability, which, as we will see, has its own logic. Daoists do not define the idea or a concept of language, they attain rather to an experience of the temporal unfolding of language, and since Daoists are concerned with conveying an experience rather than a concept of language, their discourse on language often takes a form that is literary rather than philosophical. The early Daoist texts collected in the thirty-three chapters of the Zhuangzi 莊子 present us with the most sustained reflection on the question of the relationship between language and logic not only in Daoism but in the Chinese tradition at large. In the following the term “Daoism” refers to these texts. Daoists experience language as such and not the special features of Chinese language. To be sure, Daoists speak and write in Chinese, but this fact does not determine their experience of language. If we approach Daoist thought with the assumption that different languages enclose us in different worldviews, then we will be limited to making philosophical remarks on Chinese language and will never encounter the Daoist experience of language. Daoists express themselves d­ ifferently E. J. Møllgaard (*) Department of Philosophy, University of Rhode Island, South Kingstown, RI, USA e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_20

369

370

E. J. Møllgaard

than Greek philosophers, and modern philosophers of language express themselves differently than both Daoists and ancient Greeks, but they all address language as such. The science of language (linguistics) is based on Saussure’s distinction between the language itself (la langue) and speech (parole). Daoists do not make this distinction. They do not propose a science of language, nor do they put forward a philosophy of language. Daoists present us with an experience of language. But the Daoist experience of language is not subjective play of the imagination. As in science and philosophy of language, although in an entirely different manner, it is an encounter with language itself. Classical Chinese does not have a word that corresponds to the English word “logic,” but, as the present volumes shows, there is plenty of logic in ancient China. Lack of a particular word does not deny a culture access to the thing itself. The English word “language” was borrowed from French around 1280 and first meant “a way of speaking.” Only later did “language” come to mean “national language.” (Cassin 2014: 541) The English, of course, knew about language before they borrowed this word from the French, just like the Chinese knew about logic before they coined words for it in their encounter with Buddhists and Western philosophers. Different languages have different words for “language,” but once philosophers in these different languages begin to ponder their chosen word for what we in English call language, then they will all in one way or another be drawn towards the essence of language. There are many words for language in English (tongue, speech, discourse, words, locutions, and so on). Classical Chinese also has many words for what we call language. Shuo 說 is rendered into English with “talk about,” “explain,” “discuss,” “persuade,” and can also mean “theory” or “doctrine.” Yu 語 is rendered “speak of,” “speech,” “language,” “aphorism” or “saying” (as in the famous “sayings” of Confucius). Ci 辭 is rendered “words,” “phrase,” “utterance.” Ming 名is used as a noun meaning “name,” and as a verb it means “to designate” and “to call.” Wei 謂 means “speak to,” “speak about,” “to call,” “to refer to as.” Yan 言 is rendered “speak,” “words,” “language,” and also “theory,” “teaching,” or “doctrine.” (See Kroll: 2015). We notice that with these various words for language a difference is made between language itself (to speak, to say), and language that conforms to a certain logic, be it the logic of naming and designating, or the logic of language that forms behavior. For Confucians it is important that when we name a ruler “ruler” then the name is used normatively correct, if not it must be rectified (zheng 正). For Mohists it is important that when we name an ox “ox,” then the name is used semantically correct so it “fits” (dang 當). When Daoists think about the essence of language it is mostly the word yan 言 that guides them, and for Daoists the central connotations of the word yan are “speech,” “words,” or “language,” and not “theory,” “teaching,” or “doctrine.” In other words, when Daoists consider the essence of language, they are not thinking of language used to make valid inferences or establish semantic and normative coherence, the kind of language that is formed into theories and doctrines, but of language itself before it has undergone this formation. This view of language is opposed to that of Confucians and Mohists, who consider language only

20  Problems of Language and Logic in Daoism

371

as it is formed in the process of normatively correct naming and the semantically proper fitting of names to objects. Daoists experience language before this process has begun. This distinction is crucial in order to understand the Daoist experience of language. According to Daoists, the experience of language before it is formed by logic is the same as the experience of the dao 道, or the Way. The word dao has a range of meanings. It can mean a particular “way” or “route” to something, “a way of doing something,” “method” and “proper procedure,” “principle,” “guidelines,” and “doctrine.” It can mean “skill or art of a particular kind, specialization.” As a philosophical word dao points to “the fundamental reality” of things, and it can also mean “say, speak; express, communicate orally.” (Kroll 2015: 79–80) We see that there is in the word dao (way) a split in meanings similar to the split in meanings of the word yan (language). On the one hand, we have dao itself as “fundamental reality,” or, as I would prefer to say, as what brings forth reality, the world of things. On the other hand, we have dao as method, procedure, and guidelines, which are imposed on the things brought forth by the dao in the first sense of the word. This distinction correlates with the distinction between yan as language itself (to speak, to say) and yan as logical discourse that forms language into theories and doctrines. For Daoists the essential dimension of dao and yan are the same. Both language (yan) and Way (dao) bring forth the ceaseless emergence of things. The Way “things things” (wuwu 物物), it brings forth all things but is itself “not a thing” (feiwu 非物) and is therefore nothing. (Zhuangzi 2000: 63) Language has precisely the same structure. In the Daoist experience of language things do not first appear and then they are named. Things appear in and through language. But language itself does not appear as a thing, just as the Way does not appear as a thing. The famous opening lines of the received version of the Daodejing 道德經 make this point. Here it is stated that the dao that can be walked (that is to say, “way” as procedure or method) is not the same as the constant dao (that is to say, dao as fundamental reality, or, better, the dao that brings forth all things, including all daos as ways and methods). Correspondingly, says the Daodejing, the name (ming) that can be named (that is to say, the name that is designated and so can be rectified and disputed) is not the eternal name (that is to say, the name that is not yet designated and therefore cannot be rectified and disputed). This latter “name” is language itself, which cannot be named. The dao itself cannot be walked (in a step by step logical procedure), but it can be experienced. Language itself cannot be named (and so become an object for theorizing and disputation), but it can be experienced. This close relation between language (yan) and the Way (dao), and the distance of both from logic, is made clear by Zhuangzi 莊子, when he asks: “How is the Way (dao) concealed (yin 隱) so there is true (zhen 真) and false (wei 偽)? How is language (yan) concealed so there is right (shi 是) and wrong (fei 非)?” (Zhuangzi 2000: 4) Assertions in terms of true and false, right and wrong, are the basis for logical procedure. The Way and language are concealed together with the rise of logical procedure. And they are revealed together in the Daoist experience of language. The Daoist meditation on language centers on the philosophical question of the relationship between language and logic. This question also lies at the heart of

372

E. J. Møllgaard

Western philosophy. The range of meanings of the Greek word λóγος is similar to those of the Chinese word dao 道 and split in the same way: on the one hand, ultimate reality and language itself (to speak), on the other hand, method and logical procedure. In the West normative discourse was early on identified with the logical dimension of language in the form of the logos of science, or “speech that exhibits the fact of the matter in a logical sequence.” (Gadamer 1991: 20) There is science in ancient China, but a discourse like the logos of science developed in ancient Greece did not become dominant in China. Confucians developed a logic of correctness, where they “correct” (zheng 正) names and behavior according to what they take to be the normative order of the world. This logic became dominant in the Chinese tradition. The later Mohists, for their part, developed a logic of naming, where they lay down the rules for which names “fit the facts” (dang 當) according to a particular way of “proceeding” (xing 行). The early Daoists opposed both the Confucian logic of correctness, and the Mohist logic of naming. It is in contrast to these two logics, and not Greek logic, that we must try to understand the Daoist experience of language. In the following I first show how the Daoist experience of language has been misunderstood from the point of view of the Confucian logic of correctness that became dominant in the Chinese tradition. Second, I explain the Daoist experience of language on the basis of the second chapter in the Zhuangzi, which Christoph Harbsmeier characterizes as “a central document in the history of Chinese science and epistemology.” (Harbsmeier 1998: 260) I see this chapter rather as a central document in the Daoist meditation on language. Chad Hansen rightly reads this important text in terms of philosophy of language, but his analysis takes its point of departure in the special characteristics of the classical Chinese language, which in my view are largely irrelevant to understanding the Daoist experience of language, and he ascribes a theory of language to the Daoists, which in my view is a mistake. (Hansen 1992) Third, I show that Daoists do not simply negate the logic we impose on language. Daoists accept the law of non-contradiction, but do not immediately brush aside language that falls into contradiction. They tarry with the contradiction and show its strangeness in order to go beyond but not negate the logic we impose on language and experience the logic of language itself. Finally, I show how this logic of language itself, a logic beyond logic, is exhibited in Daoist stories.

2  Language and Correctness The Daoist experience of language that is one with the appearance of things and prior to logical discourse is difficult to grasp. Luckily, the first section of the Yuyan 寓言 chapter in the Zhuangzi explicates this kind of language. (Zhuangzi 2000: 79–80) In this passage three kinds of language (yan 言) are described that are supposed to overcome antagonistic disputation (bian 辯) where interlocutors simply affirm by saying “right” (shi 是) what agrees with their preconceived ideas and

20  Problems of Language and Logic in Daoism

373

reject with a “wrong” (fei 非) what differs from their particular set of prejudices. As pointed out above, according to Zhuangzi this fall into discourse in terms of right and wrong is a fall away from the Way (dao). The first kind of language, yuyan 寓言, avoids disputes by letting someone else promote one’s position, like a father will have a go-between recommend his son for marriage, so that if some dispute arises about the arrangement he himself will not be directly blamed. The second kind of language, zhongyan 重言, avoids disputes by ascribing one’s position to worthies of old, such as Confucius 孔子 and Laozi 老子, persons who people generally will not attack. These two kinds of language are strategies for avoiding disputes, and like all strategies they only work some of the time. The third kind of language, the zhiyan 卮言, is not a strategic way of saying but words that simply “appear” (chu 出). “To appear” (chu) is a philosophical term in Zhuangzi that indicates the coming-into-being of all things in the movement of the Way (dao). This coming-into-being is prior to any strategic comportment to things, for if nothing comes into being there is, obviously, nothing to relate to at all. To be sure, the zhiyan “harmonizes within the bounds of Heaven” (he yi tianni 和以天倪) and allows us “to live out our years” (qiongnian 窮年), but this does not imply mastery of the real. For in the moment of enunciation the saying that simply “appears” or “comes forth” (chu) is one with the moment of pure appearance, the moment when “there has not yet begun to be things” (weishi youwu 未始有物), that is to say, the moment prior to there being objects to master and manipulate. (Zhuangzi 2000: 5) The Yuyan chapter goes on to describe in more detail this marvelous language that is one with the appearance of things. In saying something we make distinctions between things, but zhiyan equalizes (qi 齊) things. When we are silent, of course, we make no distinctions and things are equalized. But silence is not true equalization, for in silence the difference between saying and equalizing remains in place. To truly equalize we must perform the paradoxical activity of saying non-saying (yan wuyan 言无無言). This is what zhiyan does, and in doing so opens up to an enchanted world where “there is no thing that is not so, no thing that is not allowable” (wuwu buran, wuwu buke 无無物不然, 无無物不可). Zhiyan eludes logical discourse altogether and is neither permissible nor non-permissible according to the rules for making valid inferences. How do Chinese commentators understand this language? Cui Dahua 崔大華 identifies four ways of interpreting the phrase zhiyan. Most traditional commentaries follow, and all are aware of the early commentary of Guo Xiang 郭象 (252–312), who edited the Zhuangzi in the form we have it today. Guo says that the character zhi 卮 in the phrase zhiyan means a wine vessel that tips and pours out the wine when full and straightens up when empty. In the same way Zhuangzi’s zhiyan is goblet words that pour out without any purpose or intentionality of their own but only in response to input from the outside. The goblet words, says Guo, “follow things and changes” (yinwu suibian 因物隨變). The one who speaks them has a complete grasp of the real and is therefore able “to exhaust his or her natural allotment” (jin qi ziran zhifen 盡其自然之分). For Guo goblet words

374

E. J. Møllgaard

implies a kind of super-correctness. As Chen Jingyuan 陳景元 (1024–1094), following Guo, says, goblet words are words that “clarify by always being spot on and correct” (qu qi zhong zheng er ming 取其中正而明). Here the “appearing” (chu) word of the Yuyan chapter is obscured by the correctness (zheng 正) of the word as measured against things that have already appeared (objects). The second line of interpretation of the phrase zhiyan also reads the character zhi as “wine vessel.” But now the wine vessel indicates a joyous celebration where guests greet each other and engage in what was later called “pure conversation” (qingtan 清談), a relaxed informal exchange where one can speak one’s mind freely. The third line of interpretation takes its departure from the fact that the wine vessel is round (yuan 圓). The vessel indicates then satisfactory and tactful speech that does not offend anyone (yuanman zhiyan 圓滿之言). The fourth line of interpretation departs from the other three in giving a different gloss for the character zhi. Now zhi is not a wine vessel but is read as zhi 支 “branch,” “fragmented,” or “broken,” and zhiyan then means incoherent saying that does not hit the mark. Sima Biao 司馬彪 (246?–306?), one of the earliest commentators on the Zhuangzi, says that zhiyan means “fragmented speech without head or tail” (zhili wu shouwei yan 支离 无無首尾言). (Cui 2012: 734–736) This is also the view of the last chapter of the Zhuangzi itself, the Tianxia 天下 chapter, which gives an assessment of early Chinese thinkers and says that Zhuangzi’s language is “absurd,” “extravagant,” and “bizarre.” (Zhuangzi 2000: 100) As can be seen, the traditional commentaries diverge on a fundamental question: Is Zhuangzi’s language a super effective form of language that is always correct and hits the mark? Or is it a kind of scattered, incoherent language with no relation to correctness at all? Surely, it is not reasonable to regard Zhuangzi’s language as being complete nonsense, even if this is at times the view of the one who knew Zhuangzi most intimately, namely the sophist Hui Shi 惠施. Should we then follow the commentators who explain that Zhuangzi’s saying embodies a kind of super-­ correctness? Much depends on how we understand what it means to say non-saying (yan wuyan). Traditional commentators suggest that this means, first, that in this kind of saying the self-other dichotomy is transcended, and, second, that language is freed from its function of making assertions that can be deemed right (shi) or wrong (fei). That is correct. But then the commentators go on to assimilate Zhuangzi’s saying to what they know best, namely Confucian discourse. Confucius wished he did not have to speak but, like Heaven (tian 天), could abide in silence while everything worked correctly, and Mencius says that he is not fond of disputation, but is forced to argue in order to oppose perverse doctrines. (Lunyu 17.19; Mengzi 3B.9) Confucians do not like to argue, but they want to speak the transformative word that once and for all cuts through all the chatter and wrangling and regulates and controls the world. The transformative Confucian word is beyond opposing assertions of right (shi) and wrong (fei), and from this superior position it regulates the entire field of rights and wrongs so all is in the proper order. According to traditional commentators on the Yuyan chapter, with his words Zhuangzi does something similar. Wang Pang 王雱 (1044–1076), for instance, says that, just like

20  Problems of Language and Logic in Daoism

375

the Confucians, Zhuangzi wants to correct the world, and he does so by “returning to the correctness of nature and destiny (fan xingming zhi zheng 反性命之正).” (Fang 2012: 492–500.) In this way Zhuangzi’s saying is assimilated to the Confucian logic of correctness. The commentators see, correctly, that Zhuangzi, like the Confucians, wants to overcome antagonistic disputation (bian), but they also assume that Zhuangzi, again like the Confucians, imagines a language that can regulate and control the world because it embodies a super-correctness that is beyond dispute and therefore is super-effective. This Confucian interpretation of Zhuangzi’s language became dominant in the tradition. Even later Daoist and Buddhist commentators tend to fall back on this ruling interpretation, not only because Zhuangzi’s language is so hard to understand, but also because they consider the essence of “the three teachings” (sanjiao 三教) (Confucianism, Daoism, and Buddhism) to be one and the same, and therefore they could not see that Zhuangzi opposes Confucian language as much as he opposes the language of disputation. Furthermore, Guo Xiang edited the Zhuangzi centuries after Zhuangzi, and many of the traditional commentaries are from a thousand years or more after Guo Xiang, and we cannot assume that any of these interpretations of zhiyan corresponds to Zhuangzi’s experience of language. So, what is really Zhuangzi’s experience of language? To answer that question we should turn to the Qiwulun 齊物論 chapter, which was most likely written by Zhuangzi himself and is the chapter the author of the Yuyan chapter borrows from when he explains zhiyan.

3  The Daoist Experience of Language A unique experience with language is at the center of the Qiwulun chapter, the second chapter of the Zhuangzi. The chapter begins with a description of a meditative trance in which the self-other dichotomy is transcended. To explain this state, three kinds of musical pipes are introduced. First, the pipes of man (renlai 人籟), or the sounds of contending assertions put forth in the human realm. This is the realm of difference. Second, the pipes of the Earth (dilai 地籟), where the contending human voices are silenced from the point of view of the oneness of Nature. This is the realm of the same. Third, the pipes of Heaven (tianlai 天籟), where difference and sameness are transcended and every thing articulates itself in its unique coming-into-­ being. (Zhuangzi 2000: 3) The aim of the essay that follows is to transform language so it neither produces antagonistic differences in the manner of disputation nor obliterates differences by merging them into Nature or the One but says something that is beyond difference and sameness. After the intimation of pipes of Heaven follows, as a counterpoint, a chilling description of a human world where language is reduced to assertions. We make assertions about what is right (shi) and wrong (fei) in order to control our world and dominate each other. Our assertions fly “like arrows from the crossbow trigger,” and we hold on to our victory in argumentation as if we had sworn a blood oath. Caught

376

E. J. Møllgaard

up in this violent discourse we fail to see that self and other are interdependent, and therefore we “cut together with things.” We do not realize that right and wrong are not inherent in things themselves but are products of our own prejudice. And so we enter the world of disputation “where each asserts to be right what the other asserts to be wrong, and asserts to be wrong what the other asserts to be right.” According to Zhuangzi, these antagonistic disputes are without hope of reconciliation and only clarity (ming 明) will get us out of this self-imposed predicament. (Zhuangzi 2000: 3–4) In the following extremely difficult passage Zhuangzi pushes language to its limit in an attempt to attain in saying the clarity (ming) that lies beyond assertion. Zhuangzi turns the language of disputation (bian) into a joke and destroys the opposition between self and other. The breakthrough comes when one realizes that A way is complete when we walk it. Things are so when we designate them. What makes them so? Being so rests with being so. What makes them not-so? Not-so rests with not-so. [When we see this then it becomes clear that] things inherently (gu 固) have what is so, and things inherently have what is affirmed. No thing is not so, no thing is not affirmed. (Zhuangzi 2000: 4)

Parts of this passage are repeated in the Yuyan chapter to explain the zhiyan. Zhuangzi goes beyond the language that designates (wei 謂) things as objects and experiences language that coincides with the emergence (chu) of things before they become objects, the elusive moment when “there has not yet begun to be things.” Since there is no thing as object to speak “about,” language is entirely free of the assertive mode, and each “thing” is affirmed as being “so” (ran 然) by itself, in other words, by being “self-so” (ziran 自然). When each thing appears as self-so, can we distinguish between one thing and another? When language is freed from the differentiation imposed by propositions and assertions, then like the Way (dao) language is beyond the dichotomy between same and different, and like the piping of Heaven (tianlai) it articulates the unique. Language does not propose that a thing is this or that but exposes the thing as thing. This exposure of the thing in language was misunderstood either as a form of super-­ correctness or as pure nonsense. Zhuangzi’s friend Hui Shi considered Zhuangzi’s language useless (wuyong 無用) because it does not conform to the logic of disputation where only assertions that “hit the mark” (dang 當) and correctly names the object are useful. This is the only kind of logic that Hui knows, and therefore he misses the logic of Zhuangzi’s language, which lies beyond using language as a regulative tool. Zhuangzi opposes any regulation of the world, and Zhuangzi’s language is just as useless from the point of view of the Confucian logic of rectifying names (zhengming 正名) as it is from the point of view of the later Mohist logic of fitting names (ming 名) to objects (shi 實). But just because Zhuangzi rejects Confucian and Mohist logic it does not follow that he rejects logic as such.

20  Problems of Language and Logic in Daoism

377

4  Language and Non-contradiction Are Daoists free to contradict themselves? If so, then Daoism is outside the field of philosophy, which takes as its basic logical laws the law of non-contradiction and the law of the excluded middle. Aristotle stated, “The most indisputable of all beliefs is that contradictory statements are not at the same time true” and “it is necessary for the affirmation and the negation to be true or false.” (Aristotle 1984: 1597, 28) In early China the Mohist logicians also affirmed the law of non-­ contradiction and the law of the excluded middle as fundamental logical laws. The Mohist says “The other is not admissible; two are not admissible,” and explains, “The other: Everything is either ‘ox’ or ‘not-ox.’ It is like a hinge. There are just the two possibilities—there is no way to deny (this).” (Johnston 2014: 225) Janusz Chmielewski remarks that this is “the most general and the most explicit statement (in metalogical terms) of the conjunction of the laws of non-contradiction and of excluded middle which the Chinese ever made until modern times.” (Johnston 2014: 225) The Mohists may well have formulated these logical laws in reaction to Zhuangzi’s language. When the Mohist says “To deem all saying inconsistent with itself is inconsistent with itself. Explained by: what he says himself,” then he identifies the performative contradiction, which is a variant of the law of non-­contradiction, and Graham suggests that this is done in response to Zhuangzi. (Graham 1989: 185) Apparently the Mohist logicians believed that Zhuangzi rejected the most basic law of logic. Does Zhuangzi reject the law of non-contradiction and the law of the excluded middle? And is he unaware that his language falls into a performative contradiction? An exchange between two of Zhuangzi’s fictive characters throws light on this question. (Zhuangzi 2000: 6–7) Ju Quezi 瞿鵲子 says that he has heard Confucius remark on the following description of the sage: The sage is not engaged in any business. He does not go for benefit, nor does he avoid harm. He does not enjoy seeking, and he does not follow the Way. When he says nothing, he says something; when he says something, he says nothing. And he wanders beyond the dust and grime.

This is not a description of any old sage. These words are trying to capture Zhuangzi’s philosophy and in particular Zhuangzi’s language. The sentences “when he says nothing, he says something; when he says something, he says nothing,” finds echoes throughout the Zhuangzi. The Yuyan chapter says, “if you speak all life, you have never once spoken; if you do not speak all life, you have never once not spoken.” (Zhuangzi 2000: 79) Zhuangzi’s Confucius thinks that these are crude words (menglang zhi yan 孟浪 之言). Ju Quezi, for his part, thinks that they are “the working of the mysterious Way” (miaodao zhi xing 妙道之行). As we know, some find Zhuangzi’s saying to be vacuous nonsense; others think it is a sublime form of expressing the ineffable. According to Chang Wuzi 長梧子, however, both are wrong. If you say it is nonsense, you are wrong, if you say it makes sense, you are also wrong.

378

E. J. Møllgaard

It is not surprising that Confucians think Zhuangzi’s language is nonsense. After all this language does not conform to the logic of correctness. But why is Ju Quezi wrong, when he says that Zhuangzi’s language not only makes sense but is “the working of the mysterious Way”? Chang Wuzi explains that in saying this Ju Quezi “calculates too quickly” (dazao ji 大早計). In praising Zhuangzi’s language he jumps ahead of himself and misses the point. He does not get the experience with language this saying, which eludes the opposition between sense and nonsense, is supposed to give you. Therefore Chang Wuzi gives Ju Quezi a second chance to experience Zhuangzi’s language. He says: “Let me try to say some nonsense to you, and you listen with nonsense [that is to say, without making sense of what you hear]” (予嘗為女妄言 之, 女以妄聽之奚). The nonsense (wangyan 妄言) in question here is the kind of nonsense that Zhuangzi speaks, which, more precisely, is language beyond the dichotomy of sense and nonsense. It is suggested that in order to understand this kind of nonsense, we should not immediately try to make sense out of it. We must experience Zhuangzi’s language without jumping to conclusions. Chang Wuzi then proceeds to give his own description of the sage. This description is more obscure than the first, but it is clear that it also implicitly refers to Zhuangzi and his language. Chang Wuzi says, in Watson’s translation, that for the sage “all the ten thousand things are what they are, and thus they enfold each other” (萬物盡然, 而以是相蘊 ). (Watson 1968: 47) Or, as Brook Ziporyn translates more elaborately, for the sage “each thing is just so, each thing is right, and so he enfolds them all within himself by affirming the rightness of each.” (Ziporyn 2009: 19) This is an implicit reference to Zhuangzi’s saying “so comes from so … not-so comes from not-so … no thing is not so” (然 於 然 … 不然於不然 … 无無物不然 ), a passage found in both the Qiwulun and the Yuyan chapters. In short, in describing the sage Chang Wuzi speaks more of Zhuangzi’s nonsense. Then Chang Wuzi remarks on dreaming. He says that humans are so immersed in language and, more broadly, in the symbolic order of their communities that they believe it to reflect the order of the real. They are sure that “that’s a ruler!” and “that’s a shepherd!” But it is all a dream. If we are asleep in language, then everything we say is like a dream. It makes no difference if you say that Zhuangzi’s language is nonsense, or if you say it is sublime. And when you say that “all is a dream,” then this saying is also like a dream. It does not correspond to “reality” (shi). What then are we saying when we say, “everything is a dream”? Do we not fall into a performative contradiction? To be sure, but that objection is too easy to make. As Chang Wuzi would say, we “calculate too quickly.” Let us wait a while so we get a sense that we are saying something strange and become aware that the situation here is more subtle than a logician may suspect. Let us look closer at Chang Wuzi’s argument. It has three steps: 1 . Confucius and you are both dreaming. 2. When I say you are dreaming I too am dreaming. 3. These words [about the sage] may be called strange.

20  Problems of Language and Logic in Daoism

379

Step three is the crucial step, and it requires some interpretation. First, what is the reference of the phrase “these words” (shi qiyan 是其言)? It refers to the statement about the sage quoted at the beginning of the dialogue. But since Chang Wuzi substitutes his own description of the sage for the first one, the phrase also refers to his description of the sage. Furthermore, since both descriptions of the sage refer to Zhuangzi’s language, the phrase ultimately refers to Zhuangzi’s language. Finally, and most importantly, Chang Wuzi himself speaks this kind of language. His statement, “when I say you are dreaming, I too am dreaming,” is just the kind of nonsense Zhuangzi would say. It is the kind of saying that Confucius considers crude. We can now formalize the three steps in the argument further: 1 . Everybody is dreaming. 2. When I say, “everybody is dreaming,” I am also dreaming. 3. When I say, “everybody is dreaming,” I am saying something strange. What does Chang Wuzi mean when he says that he is saying something “strange”? The phrase diaogui 弔詭 is translated into English variously as “Supreme Swindle” (Watson), “a flight into the extraordinary” (Graham), and “enigmatic” (Mair). Brook Ziporyn has an interesting interpretation. He takes the phrase shi qiyan 是其言 to mean “if you were to ‘agree’ with these words as right,” and he renders diaogui 弔 詭 with “offering condolences for the demise of the strangeness.” The entire passage then reads: “Confucius and you are both dreaming! And when I say you’re dreaming, I’m dreaming too. So if you were to ‘agree’ with these words as right, I would name that nothing more than a way of offering condolences for the demise of their strangeness.” (Ziporyn 2009: 19) This interpretation is, as Ziporyn himself says, “adventurous,” but it does get to the heart of the matter. For, as Ziporyn explains, “to merely judge these paradoxical words as ‘true’ or ‘right’ … is a way of killing off their salutary strangeness and then eulogizing the corpse with these laudatory titles.” (Ziporyn 2009: 19, note 29) This is precisely what Ju Quezi did. We still wonder, however, what is the “salutary strangeness” of these words? How do they help us? The exchange between Chang Wuzi and Ju Quzi shows that Zhuangzi is aware of and avoids what Western philosophy calls the performative contradiction, which is a variant of the law of non-contradiction. If you say, “there is no truth,” then your interlocutor will ask you, “What is the truth value of that statement?” If it is true, then it is false, for then there is at least one truth. This means you have contradicted yourself in the very utterance of your proposition. One way to elude this performative contradiction is to say that the proposition, “There is no truth,” is not in the register of propositions that can be deemed true or false but indicates the existential situation within which the proposition is made. That is the way Zhuangzi’s argument takes. When Zhuangzi says, “Everybody is dreaming,” he is not making a proposition about an objective state of affairs, but indicates our existential situation—an indication that often strike us as strange, since it unsettles our everyday relation to things. The saying “everybody is dreaming,” and Zhuangzi’s language in general, does not propose something but exposes us to the experience of language itself. This

380

E. J. Møllgaard

experience awakes us from the dream of language. Zhuangzi’s argument goes like this: All language, and more broadly the entire symbolic order, is a dream. It does not name the real, what it says is not fixed; it is hardly different from the senseless chirping of birds. When we say this, then what we say is also a dream. It does not name the real, and one may wonder if it is different from the chirping of birds. Since what we say is not in the register of right and wrong and the logic we impose on language, we do not know if we have said something, but we call attention to the existence of language as language and the dream as dream. This move does not negate the law of non-contradiction, but it does open up to the more profound logic of language itself. I will now turn to this other logic.

5  Logic Beyond Logic Daoist language does not conform to the Confucian logic of correctness, nor does it conform to the later Mohist logic of naming. But this does not mean that Daoist language is arbitrary and without logic. Daoist language follows the logic of language itself, as opposed to the logic imposed on language. This logic emerges when we are awake in language. Confucians are asleep in language. They believe that the distinctions they make with language correspond to the normative order of the world. When the ruler rules (as a true ruler does), the father fathers (as a true father), the son sons (as a real son does), and so on, society is harmonious. For Zhuangzi this harmony is sleep, and to be asleep in language has political consequences. Somebody tells Mencius 孟子about the egalitarian teachings of the Divine Agriculturalist (Sheng Nong 神農) he has learned from a teacher from the southern state of Chu 楚. In response, Mencius explains that the community based on the words of the sages is necessarily unequal. It is constituted by the normative divisions between ruler and ruled, noble and vulgar, intellectuals and manual laborers, Chinese and barbarians. Mencius then accuses his interlocutor of following the teachings of “a southern barbarian with a bird tongue” (nanman juehua zhi ren 南蠻鴃舌之人). (Mengzi 3A4) For Mencius, the language of the barbarians is as senseless as the sound of birds, not because he does not understand what they say, but because what they say contradicts the normative order that is revealed only in his own language. It is precisely because he regards his own language as normative that he regards the language of the other as barbarian. The words of the sages institute the split between the human and the non-human, and, as we know from history, Chinese history as well as the histories of other peoples, this split has led to much suffering. Mencius does not realize that it is not only the language of the barbarians but also his own language that is hard to distinguish from the twitter of birds. Zhuangzi draws attention to this fact when he says: Saying (yan) is not just the blowing of air, saying says something. It is only that what it says is not fixed. Is there really saying then? Or has there never been saying? If saying is ­considered to be different from the sound of baby birds, is there really a distinction or is there no distinction? (Zhuangzi 2000: 4)

20  Problems of Language and Logic in Daoism

381

With his double question—Is there saying? Or is there not saying?—Zhuangzi suspends judgment in order for the phenomenon of language itself to appear. This experience of language itself is beyond the Confucians and their logic of correctness. In rectifying names the act of naming itself is obscured. That is to say, language itself is hidden. Mencius is asleep in language; he does not see language as such but only the distinctions it establishes, and among them first of all the distinction between human and non-human. Mohist logicians too are asleep in language. They view things in terms of what is typical of them, namely in terms of their category (lei 類); and they understand knowing (zhi 知) as passing beyond the appearance of the thing. The later Mohists explain knowing (zhi) with “having passed the thing one is able to describe it.” (Graham 1978: 267) That is to say, knowing is not directed at the immediate coming into presence of the thing but at the representation of the thing that remains after the thing has passed from view. The Mohists pass over precisely that self-­emergence of the thing that Zhuangzi experiences and indicates with the word chu (to appear). According to Zhuangzi, to pass over the experience of the self-­emergence of the thing is to miss the experience of language, and to miss the experience of language is to be asleep in language. Confucians and Mohists are asleep in language and not awake to the logic of language itself. What then is the logic of language itself? Zhuangzi shows that there is a logic of language beyond the logic imposed on language by Confucians and Mohists. He shows it not by developing a theory of language but by pointing to the experience of language in every significant story he tells. All of these stories awake us to language, and they show us the logic of language itself. They point to the wonderful experience of language before the quest for valid inferences begins. Look I can speak! And I can also not speak! This experience corresponds to the experience of the wonder at the world—look there is something!—that is said to be the beginning of philosophy (for if there is something then, immediately, there is also nothing, and so reflection begins). Let me give a few examples of how the experience of language itself, and its logic beyond logic, is exhibited in stories from the Zhuangzi. These stories are often understood to address epistemological questions, but that is to read them within the field of knowledge that Zhuangzi explicitly leaves behind. For, according to Zhuangzi, knowledge (zhi) misses the self-emergence of the thing and the experience of language itself. First, take the famous story where Zhuangzi dreams he is a butterfly and upon waking up is not sure if he is Zhuangzi dreaming he is a butterfly or a butterfly dreaming it is Zhuangzi. This story is often taken to raise an epistemological issue, but that is not Zhuangzi’s concern. To be sure, Zhuangzi problematizes knowing (zhi), and shows that knowing (zhi) is not as good as not-knowing (buzhi 不知), but his observations are not, as Harbsmeier suggests, “the product of an advanced piece of scientific theorizing” about the lack of foundation of our knowledge. (Harbsmeier 1993: 25) For Zhuangzi not-knowing is simply that, not knowing. It is outside the field of knowing altogether and contributes nothing to that field. Zhuangzi is not interested in discussing the foundation or lack of foundation of knowing, he simply brushes knowing aside as of no importance. What is impor-

382

E. J. Møllgaard

tant for Zhuangzi is language, and the butterfly dream story is about language. Specifically it is about the proper name 周 Zhou (Zhuangzi’s given name). When you utter a proper name then the name constitutes the named in the act of naming. Zhou is Zhou by being called Zhou. One may logically dispute that a horse is a horse by arguing, as did the sophist Gongsun Long 公孫龍, that a white horse is not a horse. (Zhuangzi thinks that is silly, if the horse is standing there right in front of you.) But no one can logically dispute that Zhou is Zhou. And yet, says Zhuangzi, since language is a dream, we do not know if Zhou is Zhou or a butterfly. To be sure, between the two there must be a difference. Zhuangzi does not negate the logic imposed on language. He does not pretend that the law of non-­contradiction does not hold, and that by some sophistry one can maintain that Zhou is both Zhou and not Zhou and still remain within the field of imposed logic. A. C. Graham is mistaken when he suggests that Zhuangzi “sees it as the lesson of disputation that one is entitled to affirm or deny anything of anything.” (Graham 1981: 53) Zhuangzi does not negate the imposed logic of disputation but suspends it by suggesting that language is a dream. He accepts the differences we make with language, and within that system of differences there is a difference between Zhou and the butterfly. But if we regard these differences as the normative order of the world, then we are asleep in language. To wake up we must become indifferent to differences, that is to say, indifferent to the logic imposed on language. In such indifference we do not wake up to another normative order but simply to language itself. The Mohists lodge their certainty in the law of non-contradiction: it is either an ox or not an ox. But to be Zhou, although the reference cannot be disputed, is existentially uncertain. To be Zhou is also to potentially be not-Zhou. The remarkable acceptance of death in Zhuangzi is the acceptance of this potentiality to not-be. The potential to not-be is the logic of our existential situation and beyond any logic we may impose on life. For Zhou to be unsure if Zhou is Zhou is for Zhou to wake up to being Zhou. Similarly, in speaking to be uncertain if one says anything is to wake up to language itself. The butterfly cannot name itself. Therefore it cannot dream. And so it cannot wake up to language. That is the real difference between Zhou and the butterfly. As a second example of the Daoist experience of language itself and its logic beyond logic, consider the story of Mengsun Cai 孟孫才. (Zhuangzi 2000: 18–19) Mengsun Cai gained a name as a great mourner in the state of Lu, Confucius’ home state, even as he “observed the mourning without sorrow” (jusang buai 居喪不哀). Yan Hui asks Confucius if it is possible to gain a name (ming名) for being good at something, in this case to be good at mourning—and Mengsun Cai is truly a good mourner, he is not a fraud or deceptive—without having the corresponding substance (shi實), in this case sincere feelings of sorrow that makes the name “good mourner” correct. In other words, is it possible to live beyond the Confucian logic of correctness, where names must correspond to what is understood to be the reality or substance of what is being named? Zhuangzi’s Confucius explains that Mengsun Cai has gone beyond knowledge (jin yuzhi 進於知) and reached a state of perfect simplicity. It is not that Mengsun Cai himself has simplified things, for to simplify things requires knowledge, he

20  Problems of Language and Logic in Daoism

383

simply does not impose his logic on things and so leaves all knowledge behind. He is now Mengsun Cai the famous mourner, but he is also potentially not Mengsun Cai. Such transformations (hua 化) have no cause or reason, they cannot be predicted and known. It is like shifting from one dream to the next. And yet Zhuangzi’s Confucius says that Mengsun Cai is awake, whereas he himself as well as Yan Hui have not yet begun to awaken from the dream. And why is Mengsun Cai awake? Simply because “others cry, he cries too” (ren ku yi ku 人哭亦哭). This is the way things are simplified for him. Similarly, language is already simplified for us. Others speak, I speak too. That is all. That is the experience of language in all its simplicity before any kind of logic is imposed on language. We complicate this simplicity by “bestowing a sense of self on it” (wuzhi 吾之), and then the self appropriates language for its own ends. But how do we know that what we call a self (wu 吾) is really a self? We cannot know, for with the self the dream begins. And even saying this is also part of the dream. “You dream you are a bird and reach the heavens, you dream you are a fish and dive into the depths. I do not even know if the one who is saying this is dreaming” (不識今之言者, 其覺者 乎). This is the same argument as in the passage discussed above where Chang Wuzi says that when you say that all is a dream, then this saying is also like a dream. As explained, this argument does not negate the principle of non-contradiction but calls attention to the strangeness of saying what one says and so begins to awaken us to language. Tarrying with the performative contradiction—if you say “all is a dream” then this saying is also a dream and so not a reliable source of valid inference—we begin to shift from the logic imposed on language, the logic of valid inferences, to the logic of language itself, the uncertainty of saying anything that corresponds to the uncertainty of being anything—a “self” for instance. After the logic imposed on language has been suspended by regarding it as a dream, and things have been simplified for us, all we have to do is “be at ease in the sequence [of whatever appears] and go ahead and transform, then we will enter the unity of vast Heaven” (安排而去化, 乃入於寥天一 ). Mengsun Cai does not know that he is now a famous mourner, he just cries when other cries, that is to say, he is at ease (an 安) in whatever he is. Similarly, when we say something, we cannot know if we have said anything, we just speak when others speak, and if we are awake to language itself, then we are at ease in the sequence of whatever we say. We are at ease with whatever appears (chu) in language. The sequence (pai 排) of what appears has its own logic, namely the logic of whatever takes place. Others cry. I cry. Others speak. I speak. This is not the logic of the immanent order of the world but the logic of whatever happens and of whatever is said. When others speak, I speak. The experience of this simplicity transforms language. For we now take part in the performance of language (communication), without being captured by a supposed substance (shi) of language. We use language without supposing that it corresponds to the normative order of the world, just like Mengsun Cai takes part in the mourning ritual without being captured by the supposed substance of that ritual (sincere sorrow). Language has no hidden substance it is just like the chirping of birds. It is pure communication

384

E. J. Møllgaard

and empty of a meaning that reflects some normative or logical order. It is Confucians and Mohists who impute this deeper meaning on language and so obscure the experience of language itself. To awaken to the experience of language itself is not to exist in a state of clarity (ming) outside language. It is simply to notice the fact that language exists. I speak. And there is no why. You may say “I” speak because there is language, or, conversely, there is language because “I” speak. But that is already to complicate things, when things have already been simplified for us. It is better simply to rest in the unfolding of language itself. Like the great awakening (dajue), “to enter the unity of vast Heaven” is not to be transported to a state beyond language. It is to be awake to and at ease in the logic of the sequence of things brought forth by language, or the Way (dao), for the two are one. The sequence cannot be known and predicted, but it is not random. It is, in a sense, fated. Zhuangzi says, “nothing is as good as fulfilling your destiny, but this is the most difficult of all.” (Zhuangzi 2000: 11) This refers to the difficulty of being at ease in the logic of the sequence of whatever takes place. Finally, consider the famous dialogue between Zhuangzi and Hui Shi about the happiness of fish. (Zhuangzi 2000: 47) The two friends are out walking by the Hao river, when Zhuangzi says, “the fish come out and swim around so at ease, that is the joy of fish.” Hui objects that since Zhuangzi is not a fish he cannot know if the fish are happy. Zhuangzi counters, you are not me, so how do you know that I do not know that the fish are happy. Hui wins the argument by pointing out that if we cannot know what others know, then surely Zhuangzi cannot know the joy of fish. Zhuangzi realizes that Hui has logic on his side, but he maintains that even in questioning it Hui did know that Zhuangzi knew the joy of fish. And he adds that “I knew it from being here above the Hao River” (wo zhizhi haoshang ye 我知之濠上也). This dialogue is often understood to be about the epistemological question of knowing other minds. It is suggested that Zhuangzi claims to have intuitive knowledge of the feelings of other beings, or that he has an implicit knowledge of the environment around the Hao river that allows him to know that the fish are happy. Again it should be pointed out that knowledge of any kind is not the issue for Zhuangzi. One should read the last sentence of the dialogue, “I knew it from being here above the Hao River,” not to affirm some special knowledge, but rather to brush away the topic knowledge altogether as being of no importance. For this last sentence of the dialogue makes no sense from the point of view of logical discourse, and, at any rate, Hui Shi has already shown that Zhuangzi is incoherent in the field of logic. And yet Zhuangzi maintains that what he says, “that is the joy of fish,” has been communicated properly to Hui. Not because he knows something, which would submit his saying to the rules of logic, but simply because he says something. It is language not knowledge that is at stake in the dialogue. In Zhuangzi’s view of language it is of secondary importance that one is able to assert something that may be true or false. What is of primary importance is the simple fact that one is able to speak and so communicate with others. The wonderful event of the story about the happiness of fish is not that Zhuangzi knows that the fish are happy, but that he is able to communicate with his friend. Zhuangzi is indicating a dimension of language similar to what Giorgio Agamben, building on Walter

20  Problems of Language and Logic in Daoism

385

Benjamin’s reflections on language, calls “communicability.” This is a dimension of language before we say something about something and before meaningful discourse is formed. It is language as pure medium of communication, which cannot become an object of reference in language because it is the very medium in which such reference takes place. Communicability implies the trust and wonder that we can communicate, which we have every time we speak. It is not how and what we communicate, no matter how beautiful and potent it may be. Since in the experience of language itself one is indifferent to the efficacy and beauty of language, language seems to be no different from the twitter of birds. Benjamin wrote that language itself is “understood by all humans just as the language of birds is understood by those born on Sunday.” (Agamben 2000: 49, 53) And Agamben suggests that: It is perhaps time to call into question the prestige that language has enjoyed and continues to enjoy in our culture, as a tool of incomparable potency, efficacy, and beauty. And yet, considered in itself, it is no more beautiful than birdsong, no more efficacious than the signals insects exchange, no more powerful than the roar with which the lion asserts its dominion. (Agamben 2011: 71)

This line of thought moves us close to the Daoist meditation on the essence of language. But it is the distinctive achievement of the stories collected in the Zhuangzi to show the logic of pure communicability, or language itself. Like the Way (dao), language is there before what we say is deemed right or wrong, true or false. Language has tremendous power, for it constitutes the subject that speaks (“I am here by the river”) as well as the world spoken about (“the fish are happy”). Before Zhuangzi spoke about the joy of fish there was no Zhuangzi wandering above the Hao river and there were no happy fish. Both appear (chu) together in language. There is a logic, a sequence (pai), to this appearance, but it is not the logic of Hui’s discourse. The logic of Zhuangzi’s saying is beyond Hui’s logic, but logical nevertheless. It is the logic of language itself, language that is not forced to conform to logic, but requires that we stay awake to the appearance of things. Things appear when they are prompted to become things by the Way (dao), and when they are announced in language. The two are one and the same movement. To be attuned to this movement is to be awake to language. But language itself cannot be known. Language itself, not the question if one can know the happiness of fish, is the hidden mystery of the story about the joy of fish. Compared to the meditation on the mystery of the essence of language found in this story, the dispute over the question if one can know the happiness of fish is rather silly. Zhuangzi calls his logic beyond logic the logic of yin 因, the logic of “following.” The term yin is a technical term in later Mohist logic, where it means “criterion.” Going by a criterion is the very basis for the logic imposed on language. According to Mohist logicians, once we have “fixed the criterion” (zhiyin 止因), then we can proceed (xing 行) logically. Zhuangzi shifts the meaning of yin from the register of the logic imposed on language to the register of the logic of language itself. In Zhuangzi yin means the same as the phrase anpai discussed above. It means to follow and be at ease in the sequence of things that self-emerge in language. If you do not impose any logic on language, it does not follow that you can say whatever you like and reject the law of non-contradiction. It means that you now

386

E. J. Møllgaard

say what you say. That you speak. And this is something that is of greater import than whatever you may say. Others speak, I speak. Language exists as pure communicability before any logic is imposed. As pure communicability language itself establishes a level of equality and indifference before all hierarchies and differences are imposed. Therefore the experience of language itself has profound ethical and political implications. The discourse of imposed logic may, in the best of cases (such as democratic discourse ethics in the style of Jürgen Habermas), work towards this equality and indifference, but it will never reach it. Confucian logic explicitly works for inequality, and Mohist logic misses the experience of language itself and so the very foundation for equality. Early Daoists indicate a logic beyond the logic of correctness of the Confucians and the logic of naming of the later Mohists. In the West there have also been those who searched for this logic beyond logic (here beyond the logic of science). Ancient Greek philosophers engaged in “the task of lived logic,” the constant awareness of what he or she is doing and thinking. (Hadot 2004: 138) This “lived logic” is similar to Zhuangzi’s attention to language itself. Christian philosophers broke with logic altogether when they made faith the center of their discourse. But does faith not have its own logic? Trying to combine the Christian life experience with Aristotelian philosophy, the early Heidegger proposed a “logic of the heart” and “the logic of ‘pre-theoretical and practical’ thought.” (Heidegger 2009: 53) Radical artists have dreamt of another logic. Jean Arp wrote that “Dada wished to destroy the hoaxes of reason and to discover a new, ‘unreasoned order’ (ordre déraisonnable).” Philosophers like Benjamin and Agamben are looking for the logic of language itself beyond the logic of valid assertions. In early Daoist stories, however, this logic beyond logic is laid out for all to see. This is a major achievement in world philosophy and literature.

References Agamben, Giorgio. 2000. “Language and History: Linguistic and Historical Categories in Benjamin’s Thought.” In Daniel Heller-Roazen ed. and trans., Potentialities: Collected Essays in Philosophy. Palo Alto: Stanford University Press. Agamben, Giorgio. 2011. The Sacrament of Language: An Archaeology of the Oath. Translated by Adam Kotsko. Stanford: Stanford University Press. Aristotle. 1984. The Complete Works of Aristotle, edited by Jonathan Barnes. Princeton: Princeton University Press. Cassin, Barbara ed. 2014. Dictionary of Untranslatables: A Philosophical Lexicon. Princeton: Princeton University Press. Cui, Dahua 崔大華. 2012. Zhuangzi qijie 庄子歧解 (Divergent interpretations of the Zhuangzi). Beijing: Zhonghua. Fang, Yong 方勇. 2012. Zhuangzi zuanyao 莊子纂要 (Collected material for the Zhuangzi), 8 volumes. Beijing: Xueyuan. Gadamer, H. G. 1991. Plato’s Dialectical Ethics: Phenomenological Interpretations Relating to the Philebus, translated by R. M. Wallace. New Haven: Yale University Press.

20  Problems of Language and Logic in Daoism

387

Graham, A. C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: The Chinese University Press. Graham, A. C. 1981. Chuang-tzu: The Inner Chapters. London: Allan and Unwin. Graham, A. C. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle: Open Court. Hadot, Pierre. 2004. What is Ancient Philosophy. translated by Michael Chase. Cambridge, MA: The Belknap Press of Harvard University Press. Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. Oxford: Oxford University Press. Harbsmeier, Christoph. 1993. “Conceptions of Knowledge in Ancient China,” In Hans Lenk and Gregor Paul, eds., Epistemological Issues in Classical Chinese Philosophy. Albany: State University of New York Press. Harbsmeier, Christoph. 1998. Science and Civilisation in China, vol. 7, part 1, Language and Logic in China, edited by Joseph Needham. Cambridge: Cambridge University Press. Heidegger, Martin. 2009. “Indication of the Hermeneutical Situation.” In The Heidegger Reader, edited by Günter Figal, translated by Jerome Veith. Bloomington: Indiana University Press. Johnston, Ian trans. 2014. Mo Zi: The Book of Master Mo. London: Penguin Books. Kroll, Paul W. 2015. A Student’s Dictionary of Classical and Medieval Chinese. Leiden: Brill. Watson, Burton. 1968. The Complete Works of Chuang Tzu. New York: Columbia University Press. Zhuangzi. 2000. Zhuangzi zhuzi suoyin 莊子逐字索引 (A Concordance to the Zhuangzi, the ICS Ancient Chinese Texts Concordance Series). Hong Kong: The Commercial Press. Ziporyn, Brook. 2009. Zhuangzi: The Essential Writings. Indianapolis: Hackett. Møllgaard, Eske J.  is Professor in the Department of Philosophy at the University of Rhode Island, USA. He has published An Introduction to Daoist Thought: Action, language, and ethics in Zhuangzi (Routledge, 2007), The Confucian Political Imagination (Palgrave Macmillan, 2018), and articles on Chinese and comparative philosophy.

Chapter 21

Paradoxical Language in Chan Buddhism Chien-hsing Ho

1  Introduction Chinese Chan or Zen Buddhism is renowned for its improvisational, atypical, and perplexing use of words. In particular, the tradition’s encounter dialogues, which took place between Chan masters and their interlocutors and are recorded in its koan (“public case”) literature, abound in puzzling, astonishing, and paradoxical ways of speaking. In this chapter, we are concerned with Chan’s use of paradoxical language. In philosophical parlance, a linguistic paradox comprises the confluence of opposite or incongruent concepts in a way that runs counter to our common sense and ordinary rational thinking.1 One naturally wonders about Chan masters’ rationales for their use of paradox. There are also concerns about whether the use violates the logical law of noncontradiction to the effect that nothing can be both P and not-P all over in the same way at the same time. Recently, Yasuo Deguchi, Jay Garfield, and Graham Priest have argued that certain Mahayana Buddhists are committed to the view that some contradictions are true, and that modern developments in paraconsistent logics show that such a view can be rational and acceptable.2 For example, some Buddhists describe certain  There are at least two kinds of paradox: logical paradox and linguistic paradox. To see the contrast, Russell’s paradox regarding Cantor’s set theory is a logical paradox: the set of all sets that are not members of themselves contains itself as a member. In this chapter, we are solely concerned with linguistic paradox; also, in my usage, the paradoxical is not to be identified with the contradictory. 2  These authors themselves endorse the philosophical position known as dialetheism, which holds that the law of noncontradiction fails, that some contradictions are true. For discussions on the law and the position, see Priest et al. 2004. 1

C.-h. Ho (*) Institute of Chinese Literature and Philosophy, Academia Sinica, Taipei, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_21

389

390

C.-h. Ho

things about an indescribable reality such that the indescribable is described. The reality is both describable and not describable  – a contradiction. On the view of these authors, the Buddhists take such contradictory statements to be true, which they say is rationally coherent and intelligible.3 However, Deguchi, Garfield, and Priest focus mainly on Indian Madhyamaka and Japanese Zen. It is doubtful that Chinese Chan masters and thinkers would affirm the truth of contradictory statements. In the Platform Sūtra of the Sixth Patriarch, Huineng 惠能 (638−713 CE), the putative sixth patriarch of Chan Buddhism, is reported as saying that some people “slander [scriptural] teachings and say they make no use of words. As they say so, they should not even speak, for speech is a form of words.”4 It is implied that the claim that one makes no use of words is self-contradictory and should not have been made in the first place. Nevertheless, this sutra’s teaching of no-thought (wunian 無念) instructs one, paradoxically, to have no thought while engaged in thought. One would be both thinking and not thinking. We need to clarify this issue. In any case, I intend in this chapter to investigate Chan masters’ rationales for their use of paradox. My aim is also to show that Chan’s paradoxical expressions are only seemingly contradictory and do not really violate the law of noncontradiction. Chan became a viable Chinese Buddhist tradition during the Tang dynasty (618−907) and continued to develop for several centuries. The tradition produced a huge literature; consequently, our investigation of its use of paradox cannot but be limited and selective. It is now widely recognized among scholars of Chan studies that baffling encounter dialogues that appear in the sayings of Tang Chan masters are generally interpolations made in the post-Tang era, whereas the masters’ own sermons are more reasonable and bear more affinity with Mahayana scriptural teachings than the dialogues do. Thus, we need to distinguish the sermons from encounter dialogues. In the next (second) section, I first sketch key ideas of Chan that are pertinent to our investigation and then examine the use of paradox in the sermons associated with certain Tang masters of the Southern Chan.5 In the third section, I analyze the presence of paradoxical language in post-Tang encounter dialogues. The fourth section concludes.

 Deguchi et al. 2008: 399–401. For further related discussions, see Priest 2002, 2014. For Priest, the nature of reality is contradictory in the sense that it is such as to render certain contradictory statements true. 4  Yang 2001: 64. For this sutra, I rely chiefly on Yang 2001, which contains a well-collated text of a Dunhuang version that may date about 780 CE. In this chapter, all translations from traditional Chinese texts are mine. 5  A number of contemporary scholars have questioned the traditional opposition between “Northern School’s gradual awakening” and “Southern School’s sudden awakening.” Yet, I shall not concern myself with this issue here. By “Southern Chan” I mean the lineages that were traditionally seen as belonging to the Southern School of Chan. 3

21  Paradoxical Language in Chan Buddhism

391

2  P  aradoxical Language in the Sermons of Tang Dynasty Chan For Chan, the purpose of religious training is to realize or awaken to one’s original mind (benxin 本心) or buddha-nature in order to free oneself from the samsaric cycle of life and death. I understand the notion of original mind as referring to our mind-heart in its limpid, nonabiding (wuzhu 無住), and unfettered state. In a sense we already live in original mind, though, being clouded by mental afflictions and habitual sedimentations of dualistic thinking, it is not consciously manifest to us. From the Chan perspective, original mind is all-inclusive and all-pervasive, just like empty space (xukong 虛空). As the pivotal experience of awakening is that of the mind being realized and becoming manifest, one would here experience a dynamic coalescing of one’s mind and the myriad objects, which in the Platform Sūtra is compared to a great sea that merges the waters of various streams into a whole. The subject-object duality is transcended and the awakened Chan practitioner may declare that the myriad things are present in their (original) mind. Since original mind is nonabiding, to realize it, an unawakened mind should learn to flow freely with the changes of thoughts and things without abiding in them or attaching itself thereto. Nonabidingness and nonattachment are valued highly in Chan, but care is needed here. One must not abide in, or attach oneself to, nonabidingness or nonattachment either. Significantly, for Chan, the myriad things are ever changing and devoid of any independent and substantial nature. They are then said to be empty. This, together with the above hinted nondual perspective, indicates that things are devoid of determinate natures and forms (xiang 相), intimately interrelated, and are not really distinct from each other. Thus, for instance, we must not see mountains definitively as mountains nor waters definitively as waters, and so forth.6 To cognize X definitively as X is to impose on it a fixed, determinate identity, which differentiates intrinsically X from non-X things and easily results in attachment to it. Meanwhile, language provides us with an all-important access to the world around us and to other people, and Chan’s attitude toward language is not as negative as scholars used to think. However, the use of nominal words tends to seduce one into reifying their referents, taking the latter to be substantial and distinctly demarcated entities, which may induce an attitude of attachment toward the referents. To note a way of countering this tendency, we may attend to the famous paradoxical formula in the Diamond Sūtra to the effect that what is said to be X is not X, and so is called X. This sutra is highly influential in Chan and the formula presumably leads the Chan master Baizhang Huaihai百丈懷海 (749−814) to claim that a teaching speech in Buddhism consists of three conjoined phrases such as “a

6  I am alluding to the celebrated Chan koan of seeing mountains as mountains; refer to footnote 28 for a related remark. See also the Essentials of the Transmission of Mind, in Taishō Shinshū Daizōkyō (abbreviated as T) 48: 381a15–18 (volume 48, page 381, column a, line 15–18), and the Discourse Records of Dazhu and Visiting Students from All Quarters, in Shinsan Dainihon Zokuzōkyō (abbreviated as X) 63: 25a23–b3.

392

C.-h. Ho

bodhisattva, is not a bodhisattva, so is called a bodhisattva” (henceforth, T1) and “dharma, not dharma, not not-dharma” (T2).7 The formula of the Diamond Sūtra appears to state that something is both X and not-X at the same time.8 Likewise, if we treat the term “bodhisattva” in T1 as a predicate, T1 would seem to involve the following logical form, which is a contradiction: (1a) (∀y)[(y is bodhisattva)→(y is not bodhisattva)] which is equivalent to: (1b) (∀y)(By→~By) Yet, in light of Baizhang’s related explication and Chan teaching in general, the formula can be unpacked as follows: that which is designated by the word “X” (which can be any nominal word) is not a determinate and substantial X, not definitively an X, and ought not be an object of attachment; if this is recognized, one can use “X” provisionally to refer to the designated item as X. Thus construed, the formula, far from implying any contradiction, nicely helps to eschew attachment to the word’s referent while indicating the provisionality of the use of the word. Similarly, Baizhang’s second phrase in T1 (“is not a bodhisattva”) helps to eschew attachment to the referent of the first phrase (“a bodhisattva”) as well as to empty the referent of determinacy and substantiality. Although other Chan masters may not employ the formula in its complete form, many negative expressions in Chan texts serve functions similar to that of the second phrase. Baizhang’s third phrase in T2 (“not not-dharma”) can be construed as signifying the negation of a negation. The point is that one must not attach oneself to the negation expressed by the second phrase (“not dharma”) nor take the latter to signify complete absence of the item negated by it. Now, the second and third phrases in T2 together represent a double negation of the form “neither X nor not-X,” which negates both attachment to X and attachment to the negation of X. Yet, it may be converted to the seemingly contradictory form “non-X and not non-X” or “X and not X.” Consider, for instance, the sentence “The Buddha is neither sacred nor profane.” Here, “the Buddha” denotes an individual (say, “b”). If we view “neither sacred” as equivalent to “non-sacred,” and then to “profane” (designated by “P”), we seem to be able to derive the following logical form, which is a contradiction: (2a) [(b is P)&(b is not P)] which is equivalent to: (2b) [Pb&~Pb)]  Comprehensive Recorded Sayings of Master Baizhang, X 68: 8a10–12. Baizhang’s explication, in X 68: 8a13–15, of the claim indicates that what is negated by the negative particle not is attachment to the referent of the negated phrase and any understanding that takes the referent to be definitively so and so. 8  In the beginning of their aforementioned paper, Deguchi et al. (2008: 395–396) give a list of nine passages from Buddhist texts that they think contain contradictions. Three of the passages are cited from the Diamond Sūtra and concern the application of the formula. 7

21  Paradoxical Language in Chan Buddhism

393

However, for Baizhang, to speak of the Buddha as neither sacred nor profane is not to affirm that the Buddha is non-sacred or non-profane. That is to say, the negation “not” in the sentence S, “x is not-P,” negates any substantial relation between x and P without predicating of x any property (say, non-P); S does not commit its utterer to accepting that x has any property.9 Consequently, the double negation “neither X nor not-X,” as implied in Baizhang’s conjoined phrases, would not for him entail any contradiction. Likewise, contrary to appearances, we cannot derive (2a) or (2b) from the sentence “The Buddha is neither sacred nor profane.” Indeed, even the sentence “x is P and not-P” would not be contradictory if the word “P” is ambiguous, or is used provisionally such that “x is P” connotes that x is not non-P.10 Chan expressions are typically provisional in the sense that they do not predicate of their referents determinate properties, but are used expediently for immediate purposes and are to be negated especially when one becomes attached to their referents or takes the latter to be definitively such and such. In particular, the parity between T1 and T2 suggests that, for Baizhang, even when the third phrase is stated in an affirmative way (as in T1), the affirmation expressed may actually be a provisional affirmation that should be construed as the negation of a negation (as in T2). In this vein, “is P” in the sentence “x is P and not-P” would not predicate the property P of x, but rather express the negation of the negation of P, namely, not not-P (or, not non-P). As a result, the sentence would not imply the contradiction that P belongs and does not belong at the same time to x. Moreover, for Chan, Buddhist teachings are, like medicines, remedial in nature. Language is used not so much for its cognitive function of expressing factual descriptions of the world as for its noncognitive function of fulfilling therapeutic, heuristic and evocative purposes. For example, as Baizhang comments, a Buddhist teacher may sometimes teach that “There is Buddha,” some other times that “There is no Buddha,” depending on the audience’s intellectual illness.11 The teacher may  Comprehensive Recorded Sayings of Master Baizhang, X 68: 8a15–16: “… the Buddha is just neither sacred nor profane. Do not wrongly say he is non-sacred and non-profane.” We here allude to the distinction between two kinds of negation drawn by Indian and Chinese Mādhyamika thinkers. The negation in S is nonimplicative negation. By contrast, if the negation in S is implicative negation, then, while denying P of x, S also implies the affirmation of some other property of x such that S commits its utterer to acceptance of that property in x. 10  The Chinese Mādhyamika thinker Sengzhao (374?−414 CE) is quite influential in Chan. On his view, the provisional use of the word “existent” with respect to the thing x is intended to show that x is not nonexistent, whereas that of the word “nonexistent” is to show that x is not existent. The words function by negatively differentiating x from what x is not, rather than positively characterizing x as definitively existent or nonexistent. See the Treatise of Sengzhao, T 45: 152c12–14, 159b11. Although Chan masters may not have stated similar ideas explicitly, their recognition of the provisionality of language may well point in the same direction. See also the following remark on Baizhang. 11  For the provisionality of language and its remedial character, see the Comprehensive Recorded Sayings of Master Baizhang, X 68: 7c21–24, 12c14–19, and the Essentials of the Transmission of Mind, T 48: 382c27–383a02. Wang (2003: 179) writes: “For the Chan masters and students, if the therapeutic functions of these paradoxical expressions are effective, there are no contradictory meanings at all within the pragmatic context.” 9

394

C.-h. Ho

teach “There is Buddha” in order to remove the audience’s doubt about the existence of buddha-nature, but may teach “There is no Buddha” to quell one’s attachment to buddha-nature. What matters is the effectiveness of the teachings in achieving their purposes, but not their truth, concerning which alone the issue of contradiction occurs. Given the above discussion, we can now look at a few instances of paradox in Tang dynasty Chan. Let us attend to the Platform Sūtra teaching of no-thought first. Clearly, this sutra opposes the extinction of thought. It states that the purpose of the negation implied in the term “no-thought” is to negate one’s attachment to the objects of thought and free oneself from afflictions that arise from dualistic ideas. It is suggested that one whose thought arises directly from original mind, presumably one who has awakened to the mind, knows best how to think without attachment to the objects.12 Then, the word “thought” in the aforementioned “to have no thought while engaged in thought” is ambiguous: its first occurrence signifies a thought of attachment while the second occurrence does not. Since the word is used in two different senses, there is no contradiction to the effect that one is thinking and not thinking in the same respect at the same time. Some of the paradoxical expressions in Tang Chan are not unusual in Chinese Buddhism. For example, the Chan master Shenhui 神會 (684−758) claims that wisdom (Skt. prajñā 般若) does not know and yet knows everything, while Huangbo Xiyun 黃檗希運 (751?−850?) states that the Buddha speaks all day without having spoken.13 Both sayings may come from Sengzhao’s Treatise of Sengzhao and could be said to be rooted in certain Mahayana scriptures.14 For Huangbo, original mind is all-inclusive, capable of pervading the myriad things, which are held to be empty. This results in a nondualistic ontology, in which the mind is nondual with the myriad things and there is no genuine hindrance or demarcation between the things. In consequence, the mind is nondual with language, which is among the things pervaded by it. Given that silence can be taken to symbolize the ineffable original mind, speech and silence are nondual as well.15 Thus, the Buddha’s speaking is not really distinct from his silence or not speaking. Herein, the paradoxicality of Huangbo’s statement is due not so much to the ambiguity of the words used as to Chan’s nondualistic ontology, in which are  Yang 2001: 19–20, 37–38. See Ho 2016 for relevant discussion.  Yang 1996: 26 (Cf. Treatise on the Essential Gate of Entering the Way through Sudden Awakening, X 63: 18b24–c4); Wanling Record of Chan Master Huangbo Duanji, X 68: 21c17–18. Shenhui does not give reasons for his claim. Probably, wisdom does not know because it does not know its object through a mind of attachment and dualistic discrimination; wisdom knows everything because it knows original mind, which comprises all things within itself. 14  Treatise of Sengzhao, T 45: 153a27–29, 153c26. For a recent exposition of Sengzhao’s use of paradoxical expression, see Ho 2013. 15  Wanling Record of Chan Master Huangbo Duanji, X 68: 21c13–14: “Speech is silence, silence is speech, so speech and silence are nondual.” Since, in Chan Buddhism, original mind is generally taken to be ineffable, the question arises as to whether Chan thinkers fall into the contradiction of describing the indescribable. I have, in Ho 2015, attempted to show how one can express the ineffable without contradiction. Then, this issue may not pose a serious problem to Chan. 12 13

21  Paradoxical Language in Chan Buddhism

395

embedded the notions of original mind and emptiness. Although this may imply the nonduality of P (speaking) and not-P (not-speaking), this nonduality is not a strict or logical identity and so we cannot derive a contradiction from it. As a metaphorical illustration, consider the case of foam on the sea surface. The foam is not strictly identical with the sea water around, yet is existentially part of, and not distinct from, the water. The relation between speech and original mind (or the fusional unity of myriad objects and the mind) may be understood by analogy with the foam and sea water. In any case, Huangbo’s statement that the Buddha speaks all day without having spoken is not contradictory. Similarly, appealing to the all-inclusiveness of original mind and the emptiness of all things as a basis, Huangbo claims, paradoxically, that life-and-death and nirvana are not distinct but form a unity.16 Indeed, because he appears to identify nirvana with original mind, Huangbo can readily take nirvana to be nondual with life-and-death while being aware of their differences in certain respects. Incidentally, within this conceptual framework, the best practice for a Chan practitioner would paradoxically be to practice nothing. Or, as Huangbo makes clear, the practitioner simply needs to learn how not to seek for, and attach oneself to, anything. As noted above, the goal of Chan training is to realize or awaken to one’s original mind. Since we already live in the mind, to seek for, and so objectify, it would only counteract its realization; similarly, to attach oneself to the myriad objects would only cloud its manifestation. When one truly refrains from seeking for, and attaching oneself to, original mind and the myriad objects, then and only then may one become awakened to the mind. Often, Chan’s paradoxical expression resembles the combination of Baizhang’s first two phrases (as “A Bodhisattva is not a Bodhisattva”) and is intended to quell the audience’s attachment to the linguistic referent and indicate the latter’s emptiness, indeterminacy, or illusoriness. For example, Huangbo teaches that “mind is not mind,” “form is not form,” “existence is not existence,” “dharma is non-dharma, non-dharma is dharma,” and so forth.17 These expressions are uttered to indicate that the referents of such words as “mind,” “form,” “existence,” and “dharma” are empty, illusory, and unobtainable. The audience is thereby advised that one must not be attached to any of them. Many paradoxical expressions in Chan sermons of Tang dynasty concern the notion of mind. Dazhu Huihai 大珠慧海, a pupil of the illustrious Chan master Mazu Daoyi 馬祖道一 (709−788), affirms, quoting an unnamed patriarch, that only by realizing that mind is not mind can one truly understand mind and its various functions. Further, Huangbo instructs that “this mind is the mind of no-mind,” that

16  Essentials of the Transmission of Mind, T 48: 381a2–29; Wanling Record of Chan Master Huangbo Duanji, X 68: 20b1–8. Recall that a linguistic paradox comprises the confluence of incongruent concepts. Since “life-and-death” and “nirvana” are incongruent concepts, we have a paradoxical expression here. Incidentally, the equation of life-and-death with nirvana is a recurrent theme in Chan and Chinese Buddhism. 17  Wanling Record of Chan Master Huangbo Duanji, X 68: 20a4–24.

396

C.-h. Ho

“mind is itself mindless, and there is no mindless one either.”18 It would seem that one both has and does not have mind. The thrust here, I think, is that, for Chan, our mind is really void of any cognizable form and is ungraspable and unobtainable. In its true nature, the mind is none other than the all-inclusive original mind. Yet, we fail to cognize its true nature and tend to conceive of it as, say, an inner mental entity confined within the body. Such a conception is to be negated. Then, what is negated by Huangbo’s “no-mind” and “mindless” is this internally confined and seemingly identifiable mind, a mind that is constructed while in a state of delusion. As there should not be any attachment to the idea of no-mind, it is added that there is no mindless one either. Meanwhile, this emphasis on no-mind has a practical significance. As one empties one’s mind, not treating it as something inwardly locatable, one may rather expand the mind outwards. This may help one to realize eventually what one’s mind really is, namely, the all-inclusive original mind. The above discussions suggest that paradoxical expressions in Chan texts of Tang dynasty are not truly contradictory. Obviously, there are reasons for using such expressions. In the Chan context, their use may help to indicate the nonsubstantial and indeterminate nature of the referent of the word used and help one to refrain from attaching oneself to the referent. It also helps to show the provisional and remedial characters of language. In addition, their use flows naturally from Chan’s nondualistic ontology, in which are embedded the notions of original mind and emptiness. On the other hand, paradoxical language may serve the functions of dislodging the interlocutor’s habitual way of thinking and of evoking in them an experience of awakening. This is most likely the case in Chan’s encounter dialogues, to which we now turn.

3  Paradoxical Language in Post-Tang Encounter Dialogues An encounter dialogue in Chan Buddhism takes place between a Chan master and an interlocutor (a pupil, a monk, or a layperson) or interlocutors.19 Such dialogues take many linguistic forms and may involve shouting, silence, gesture, and such physical acts as hitting and kicking. They are similar in that the master’s words (and gestures) typically transgress upon our commonly shared sense of things and dis18  Discourse Records of Dazhu and Visiting Students from All Quarters, X 63: 29b18; Essentials of the Transmission of Mind, T 48: 380b1–2, b12–13. The notion of no-mind is a common theme in Chan Buddhism. The Treatise on No Mind (無心論), a Chan text probably composed in the Tang dynasty, states paradoxically that though one has no mind, one can very well see, hear, feel, and know, that genuine mind is none other than no-mind. The text makes it clear that we are said to have no mind because our mind is really formless and so unobtainable, and that what is negated by “no-mind” is the mind of delusive thought. See App 1995: 35–64. 19  An encounter dialogue may occur between two Chan masters, especially when one master is assessing the other’s attainment of awakening, but I shall generally disregard this type of encounter dialogue.

21  Paradoxical Language in Chan Buddhism

397

lodge the interlocutor’s expectation and habitual way of thinking. In most cases, the words are not strictly paradoxical. They are simply perplexing or, in many cases, seemingly irrelevant to the interlocutor’s question. For instance, a question most often asked in an encounter dialogue is, “What was the first patriarch’s intention in coming from the West?” To this question, Chan masters give, among others, these answers: “The cypress tree in the yard;” “To wash this old monk’s feet;” “To walk in the day and lay down at night;” “There is no intention in coming from the West;” “I’ll tell you once the stone turtle understands human speech.” The answers are puzzling, if not nonsensical. Meanwhile, to the question “Two dragons vie for a pearl, which one gains it?” Chan masters’ answers include: “Where is the pearl?;” “The gainer loses it;” “Gain is loss;” “The one with claws and teeth does not gain it;” “The pearl is everywhere on the ground but looks like mud to the eyes.” Some of the answers are indeed paradoxical, while some others are at most at variance with the background assumptions. Although our focus should be on paradoxical expressions, much of the following discussion applies to encounter dialogue in general. We begin with an illuminating encounter dialogue that was traditionally believed to take place between Bodhidharma (470?−543?) and Huike 慧可 (487−593), the first and second patriarchs of Chan Buddhism:20 Huike: “I have no peace of my mind. Teacher, please pacify it for me.” Bodhidharma: “Bring forth your mind, and I will pacify it for you.” Huike: “I seek for it but the mind is unobtainable.” Bodhidharma: “There! I have pacified your mind.” Notably, Huike’s reply to the effect that his mind is unobtainable does not show that he has realized the true nature of his mind and Bodhidharma would be pleased to give him his sanction. In the present context, it does not require a Chan-like awakening realization for one to take one’s mind to be unobtainable: after all, we are unable to bring forth our mind (or physical heart where the mind may be thought to be located) to someone else. What is really perplexing is that Bodhidharma then claims to have pacified Huike’s mind when he seems to have done nothing at all. He claims, somewhat paradoxically, to have pacified Huike’s unobtainable and so seemingly unpacifiable mind. In order to bring into relief Huike’s and Bodhidharma’s different understandings of the unobtainability of mind, let us consider these two statements:21 ( A) This mind is unobtainable. (B) This unobtainable is the mind.  Jingde Record of the Transmission of the Lamp, T 51: 219b21–23. Given the popularity of this dialogue in post-Tang Chan Buddhism, even though it plausibly did not occur between the two patriarchs, we can reasonably regard it as an encounter dialogue, which may well have occurred between some Chan practitioners. For convenience sake, I write as if it took place between the patriarchs. 21  The following elucidation draws some ideas from Abe 1985: 11–14, though Abe’s interpretation of the dialogue differs from what is proposed here. 20

398

C.-h. Ho

In his reply, Huike basically expresses statement (A), which can be simply a description of the empirical fact that one cannot bring forward one’s mind to others. However, it may be the case that Huike actually has a deeper understanding of his mind than merely grasping the fact; that is, he rightly cognizes that, in its true nature, the mind is formless and unobtainable. Even so, Huike does not truly realize this unobtainable mind as such. With (A), indeed, one may tend to take the subject term “this mind” to designate something inside within one’s body. One may tend to delimit and reify the referent of the term and conceive of it as an inner object or substance, a property of which is revealed by the predicate “unobtainable.” Even though the mind is disclosed to be unobtainable or bear the property of unobtainability, it remains an inner entity that is distinct from things outside. Plainly, no Chan master will sanction this understanding of the mind.22 On the other hand, in claiming to have pacified Huike’s mind, Bodhidharma induces him to realize that what he is taking to be unobtainable is none other than his genuine, originally limpid and peaceful mind. Bodhidharma’s response is tantamount to expressing statement (B). Since the subject of (B) is primarily a negative adjective, one is less likely to project something out there or in here and delimit its existence. The word “unobtainable” may generate a sense of oddity and puzzlement. Yet, it may draw an earnest Chan practitioner’s mind to a state of emptiness wherein conceptual delimitation and grasping is quiescent, indeed a state of the dynamic fusion of the mind and the myriad objects. Such a state can present itself in an infinite number of ways. In the present context, it is expressed by the predicate “the mind.” As “the mind” is merely a predicate, instead of delimiting and reifying its referent, one may see it as indicating a way in which the state presents itself. Now, if one truly realizes the unobtainable state of emptiness, which presents itself as one’s genuine mind, one surely will experience peace of mind. Upon hearing his teacher’s claim, we are told, Huike immediately attains awakening while realizing that his mind resides neither inside nor outside, nor in the middle, and yet is everywhere.23 The mind to which Huike is awakened is a mind of no-mind. Although statement (B) is somewhat bizarre, it serves the purposes of evading conceptual delimitation and reification of one’s mind and of drawing the mind to an unobtainable state of emptiness. In a sense, (B) is a miniature of Chan encounter dialogues, which more or less serve the two functions: (i) to dislodge the interlocutor’s habitual way of thinking, and (ii), given appropriate conditions, to transform and open up the interlocutor’s mind so as to evoke in it an experience of awakening. Let us discuss the two functions in some detail. Chan tradition distinguishes live words (huoju 活句) from dead words (siju 死 句) and recommends the use of the former. According to the Chan master Dongshan Shouchu 洞山守初 (910−990), if there are words within words, the (latter) words 22  It is not meant that this understanding is always present whenever one expresses (A). The tendency may be blocked if the language user recognizes the provisionality of the words used. In normal cases, a Chan master would prefer (A) to (B) simply because (A) accords better with people’s conventional way of using language. 23  Recorded Sayings of Master Fenyang Wude, T 47: 604a19–20.

21  Paradoxical Language in Chan Buddhism

399

are dead words; if there are no words within words, the words are live words.24 Basically, dead words are those that induce the listener to stick to their literal meaning and evaluate and react to the words in light of the commonly held sense of things, which may generate further explanatory words. By contrast, live words are words that are seemingly nonsensical and run counter to our habitual way of thinking, which induces the listener to disregard their literal meaning and semantic function of representing things in the world. (Further effects of live words are explained below.) Clearly, paradox is a good candidate for the use of live words. When used in an encounter dialogue, such paradoxical expressions as “The gainer loses it,” “A rootless tree grows on a rock,” and “One rides a water buffalo while walking on foot” may render the interlocutor at a loss to make sense of their representational content. Since the expressions are not meant to be representational in function, none of them is contradictory in the sense of representing mutually contradicting states of affairs. In this context, consider for comparison an oft-used Chan method for dislodging one’s habitual way of thinking: the master holds up, or points at, something, say, a stick, and says to the interlocutor, “Don’t call it a stick, but tell me what it is. Quick! Quick!” Hearing such words, the interlocutor may be at a loss, deprived of their customary way of using words to designate things. Here the order “Don’t call it a stick” precludes one from following one’s (most) habitual way of objectifying and identifying the stick.25 Similarly, in Chan’s use of paradoxical expression, the paradoxical sense conveyed by the expression prevents one from following one’s habitual way of taking the expression to represent the way things are. While paradoxical expressions in encounter dialogue are not representational, some of them may be intended to lead the interlocutor beyond their literal meaning to approximate the import that the master intends to convey but may have difficulty putting into words. For instance, the expression “The gainer loses it” may, apart from its dislodging function, also be meant to indicate that one’s original mind (the pearl!) is not to be realized by conceptual delimitation and grasping, which import may as well be indicated by the expression “The one with claws and teeth does not gain it.” Here we may understand indication as an indirect mode of expression: an indication indirectly expresses and makes known its indicated import or referent, which, for some reasons, cannot be properly described or represented by words. (The notion of indication draws on the famous Buddhist simile of a finger pointing to the moon: the finger points toward, or indicates, the moon but in no way represents it.) Indeed, many expressions, paradoxical or otherwise, in encounter dialogue are partially indicative and what they indicate concern chiefly the ineffable original

 Recorded Sayings of Ancient Worthies, X 68: 248a10. Cf. Wright 1992: 133. The linguistic form, paradoxical or otherwise, of the words used may not itself determine whether the words count as live words. We might also need to take into account the context of using the words as well as the understanding of the listener. 25  Of course, one can think for another word to identify the object. Yet, such thinking may instantly arouse the master’s shouting or hitting. 24

400

C.-h. Ho

mind or the aforesaid state of emptiness.26 In any case, the key point for us is that, when construed indicatively, such expressions as “The gainer loses it” do not involve any contradiction. The dislodging function of an encounter dialogue is only preliminary; it paves the way for the other function, which is to transform and open up the interlocutor’s mind so as to evoke in it an experience of awakening in which one realizes the dynamic fusion between one’s mind and the myriad objects. Here it must be said beforehand that, for those of us, myself included, who have never had any experience of what it is like to attain Chan awakening, it is difficult to understand how the dialogue or its live words can have such a transformative or evocative force. Consequently, my explanation cannot but be brief and conjectural. Common people tend to take their mind to be an inner entity that is distinct from external objects. They also take a stick to be an external object distinct from the mind. The customary use of such distinct words as “mind” and “stick” to designate their referents reinforces the latter’s delimitation. Now, a dialogue that embodies statement (B) may help to prevent one from internalizing the mind and demarcating it from its objects. The above Chan method may help to prevent one from externalizing the stick and demarcating it from the mind. In particular, Chan’s emphasis on no-mind, the emptiness of all things, and nonattachment, as well as an encounter dialogue’s dislodging function, may play a role here. All these, together with proper meditation practice, may help a practitioner, on hearing the master’s live words, to overturn the cognitive chasm between the mind and its objects such that they come to experience a dynamic subject-object fusion. One is then awakened to one’s original mind, which pervades all things yet abides in none. Of course, some live words are paradoxical in form. When a Chan master’s utterance functions to transform the interlocutor’s mind and evoke therein an experience of awakening, we may say that the master performs a perlocutionary speech act that produces in the interlocutor’s mind a specific mental effect.27 That is, apart from conveying its literal meaning and, possibly, indicated import, the utterance also brings about the effect of transforming the interlocutor’s mind such that they come to experience that which previously transcends their experience. Here, the expression uttered may be said to be evocative in function as it is intended to call forth the awakening experience of realizing one’s original mind and seeing the world in a refreshingly nondual way. Evocation is a noncognitive, nonreferential mode of expression, so it is inadequate to speak of truth, falsity, and contradictoriness in respect of evocative sentences. One can only speak of efficacy and inefficacy in respect of them.  An indication is broadly cognitive as it is meant to refer indirectly to, and intimate something about, a reality while denying any correspondence between it and the reality. By “partially” is here meant that a Chan expression can be both indicative and remedial or evocative. 27  See Fung 2008: 236–242. Fung’s ideas draw on Davidson 1978. I do not find convincing Davidson’s view that a metaphor carries no message other than its literal meaning. However, I agree with Fung that a Chan master’s utterance can produce a perlocutionary effect on the interlocutor’s mind. 26

21  Paradoxical Language in Chan Buddhism

401

The arising of an awakening experience requires that one abandon one’s deep-­ rooted cognitive habit of demarcating one’s mind from its objects. When the experience arises, the fundamental subject-object boundary collapses, and the awakened practitioner finds it hard to fit the new experience into our ordinary conceptual grids. As the boundary collapses, all things become paradoxically both one and many, same and different.28 It is then natural for the awakened to resort to paradoxical expressions to express their awakening experience or whatever is considered ultimate in Chan. The expressions may also indicate figuratively their indirect referents’ linguistic indeterminability. Consider, for example, the following dialogues: Question: “What is the truth of reality?” Answer: “A rootless tree grows on a rock.” Question: “The myriad things return to one. Where does the one return to?” Answer: “One faces south to see the Northern Dipper.” Question: “What is it like for one to come to life from within death?” Answer: “One is not allowed to walk in night but must arrive at dawn.” Significantly, here we cannot conclude that Chan masters take certain contradictory expressions to be true, and that, for them, the nature of reality is contradictory in the sense that it is such as to render those expressions true. The paradoxicality of the masters’ words is due largely to the content of the awakening experience, which is impossible for language to adequately categorize and determine. Insofar as one, in undergoing the experience, feels a strong sense of oneness with an object, it makes sense to say oneself and the object are the same. Insofar as one’s body and the object are different in experienced form and location, it is also correct to say oneself and the object are not the same. Since the two sayings draw on different perspectives, they do not together violate the law of noncontradiction to the effect that nothing can be both P and not-P in the same way at the same time. In addition, for Chan, the linguistic indeterminability concerned betokens the deficiency of language and thought in capturing the way things really are. By contrast, the view that certain contradictory expressions of reality are true appears to assign to reality a determinate, albeit contradictory, nature that can be precisely captured by language. Before ending this section, we may briefly mention yet another dimension of Chan’s use of language. In numerous cases, a Chan master’s utterance arises instantly and spontaneously from their experience of the immediate situation around them. For instance, a master was once asked about the first patriarch’s intention in coming from the West, as he was washing his feet, he then answered, “To wash this

 Remarking on the renowned Chan discourse about seeing mountains as mountains and seeing mountains not as mountains, Abe (1985: 18) writes: “Thus in the Zen Awakening attained by Weihsin, on the one hand, mountains are really mountains in themselves, waters are really waters in themselves – that is, everything in the world is real in itself; and yet, on the other hand, there is no hindrance between any one thing and any other thing – everything is equal, interchangeable, and interfusing. Thus we may say: ‘Mountains are waters, waters are mountains.’ It is here in this Awakening … that Zen says, ‘A bridge flows, whereas water does not flow’ …”

28

402

C.-h. Ho

old monk’s feet.” In cases like this, it is futile to look for any hidden message. This helps to explain why in many cases the master’s words in encounter dialogue seem irrelevant to the interlocutor’s question.29 In this section, we have discussed several key linguistic functions and dimensions of Chan encounter dialogues. The master’s words as live words are not representational in function: they are not meant to represent the way things are. Then, even if a master says in an encounter dialogue that “x is P and not-P,” the master is very unlikely to affirm the expression as true in the sense of truly representing reality. In some cases, the expression uttered, paradoxical or otherwise, is evocative, yet it is inadequate to speak of contradictoriness in respect of evocative expressions. In some other cases, the expression is indicative in function, and this recognition, as we saw above, may conduce to resolving literal contradiction.

4  Concluding Remarks In this chapter, we have investigated the presence of paradoxical language in Chinese Chan Buddhism, explored the rationales for the use of paradox, and discussed whether Chan masters would affirm the truth of contradictory statements. As Chan is a long tradition that consists of various sects, lineages, texts and individuals, there remains much work to be done in this regard. However, our study has suggested that, in Tang Chan masters’ sermons and post-Tang encounter dialogues in general, the use of paradox does not violate the logical law of noncontradiction to the effect that nothing can be both P and not-P in the same way at the same time. As mentioned at the beginning of the chapter, Deguchi, Garfield, and Priest focus mainly on Indian Madhyamaka and Japanese Zen in their argument for the thesis that certain Mahayana Buddhists are committed to the view that some contradictions are true. Given that Japanese Zen developed out of Chinese Chan, let us consider a passage from the Zen master Dōgen (1200−1253) that they cite in support of their thesis:30 Just understand that birth-and-death is itself nirvana. There is nothing such as birth and death to be avoided. There is nothing such as nirvana to be sought. Only when you realise this are you free from birth and death.

The last sentence of this passage hints at the general Buddhist view that to attain nirvana is to be free from life and death (birth-and-death), which in turn entails that nirvana and life and death are different. Yet, the first sentence states to the effect that nirvana and life and death are not different. In addition, the passage seems to imply contradictorily that there both is and is not nirvana.  Cheng (1973) proposed two principles to explain the significance and logical dissolvability of Chan puzzles and paradoxes: the principle of ontic non-commitment and the principle of contextual demonstration. Some of my ideas in this section may resemble the two principles. 30  Deguchi et al. 2008: 396. I shall only discuss this passage from the Chinese Chan perspective. 29

21  Paradoxical Language in Chan Buddhism

403

Now, we saw above that Huangbo, on account of the all-inclusiveness of original mind and the mind’s sameness with nirvana, claims that life-and-death and nirvana are the same. He can readily claim them to be different on the ground that nirvana alone, but not life and death, entails genuine realization of original mind as well as freedom from life and death. These two claims are made from different perspectives. Meanwhile, the emptiness of all things indicates that there is no nirvana to be sought, though this does not mean the latter’s nonexistence. In addition, to seek for nirvana qua original mind is to conceptually objectify the mind, which is least effective for the realization of nirvana. Thus, there is no nirvana to be sought yet there is nirvana to be realized. It should be evident from this analysis that, for Chan, the above passage does not involve any real contradiction. Many factors, we have seen, contribute to Chan’s characteristic paradoxical way of speaking, and we must not be led by the surface grammar of expressions into supposing the presence of contradiction therein. As we consider language’s abundant wealth of meaning and the various ways in which words can be used, we may refrain from making the supposition and learn to appreciate Chan’s atypical but skillful use of language.31

References Abe, Masao. 1985. Zen and Western Thought, edited by William R. LaFleur. Honolulu: University of Hawai‘i Press. (Its first chapter brilliantly and thoughtfully elucidates the well-known Chan koan of seeing mountains as mountains.) App, Urs. 1995. “Mushinron—Tonkō shutsudo no ichi tekisuto” 無心論:敦煌出土の一テキス ト. Zenbunka Kenkyūjo kiyō 21 (May): 1–69. (Containing an English translation of the Treatise on No Mind, a Chan text probably composed in the Tang dynasty.) Cheng, Chung-ying. 1973. “On Zen (Ch’an) Language and Zen Paradoxes.” In Journal of Chinese Philosophy 1.1: 77–102. (An early but significant essay on Chan’s use of paradoxical language.) Compendium of Five Lamps 五燈會元, edited by Puji 普濟, X, vol. 80, no. 1565. Comprehensive Recorded Sayings of Master Baizhang 百丈廣錄. Compiled in the Recorded Sayings of Ancient Worthies, X 68: 6a11–14a7. (A record of the sermons of the Chan master Baizhang Huaihai.) Davidson, Donald. 1978. “What Metaphors Mean.” In Sheldon Sacks, ed., On Metaphor. Chicago: University of Chicago Press. Deguchi, Yasuo, Jay L.  Garfield, and Graham Priest. 2008. “The Way of the Dialetheist: Contradictions in Buddhism.” In Philosophy East and West 58.3: 395–402. (Arguing that ­certain Mahayana Buddhists, including the Zen master Dōgen, are committed to the view that some contradictions are true.) Discourse Records of Dazhu and Visiting Students from All Quarters 諸方門人參問語錄. By Dazhu Huihai. X, vol. 63, no. 1224. (A record of the sayings of the Chan master Dazhu Huihai.) Essentials of the Transmission of Mind 黃檗山斷際禪師傳心法要, edited by Pei Xiu 裴休, T, vol. 48, no. 2012A. (A well-known record of the sayings of the Chan master Huangbo Duanji.) Fung, Yiu-ming. 2008. “How to Do Zen (Chan) with Words? An Approach of Speech Act Theory.” In Mou Bo, ed., Searle’s Philosophy and Chinese Philosophy: Constructive Engagement

 The author thanks Professor Fung Yiu-ming and an anonymous reviewer for their very valuable comments and suggestions on earlier drafts of this chapter.

31

404

C.-h. Ho

(229–242). Leiden: Brill. (A philosophical analysis of the functions of Chan koan from the perspective of speech act theory.) Ho, Chien-hsing. 2013. “Ontic Indeterminacy and Paradoxical Language: A Philosophical Analysis of Sengzhao’s Linguistic Thought.” In Dao: A Journal of Comparative Philosophy 12.4: 505–522. (An analysis of Sengzhao’s use and understanding of paradoxical language.) Ho, Chien-hsing. 2015. “Resolving the Ineffability Paradox.” In Arindam Chakrabarti and Ralph Weber, eds., Comparative Philosophy without Borders. London: Bloomsbury Academic. Ho, Chien-hsing. 2016. “Interdependence and Nonduality: On the Linguistic Strategy of the Platform Sūtra.” In Philosophy East and West 66.4: 1231–1250. (A rational analysis and reconstruction of the main linguistic strategy taught in the Platform Sūtra.) Jingde Record of the Transmission of the Lamp 景德傳燈錄, edited by Daoyuan 道原, T, vol. 51, no. 2076. Priest, Graham. 2002. Beyond the Limits of Thought. Oxford: Oxford University Press. Priest, Graham. 2014. “Speaking of the Ineffable …” In Liu JeeLoo and Douglas L. Berger, eds., Nothingness in Asian Philosophy. New York: Routledge. Priest, Graham, J. C. Beall, and Bradley Armour-Garb, eds., 2004. The Law of Non-Contradiction: New Philosophical Essays. Oxford: Oxford University Press. Recorded Sayings of Ancient Worthies 古尊宿語錄, edited by Yicangzhu 頤藏主, X, vol. 68, no. 1315. Recorded Sayings of Master Fenyang Wude 汾陽無德禪師語錄, edited by Chuyuan 楚圓, T, vol. 47, no. 1992. Shinsan Dainihon Zokuzōkyō 卍新纂大日本續藏經 (abbreviated as X). In CBETA Chinese Electronic Tripiṭaka Version April 2014. Taipei: Chinese Buddhist Electronic Text Association. Taishō Shinshū Daizōkyō 大正新脩大藏經 (abbreviated as T). In CBETA Chinese Electronic Tripiṭaka Version April 2014. Taipei: Chinese Buddhist Electronic Text Association. Treatise of Sengzhao肇論. By Sengzhao. T, vol. 45, no. 1858. (A collection of important essays written by the Chinese Mādhyamika thinker Sengzhao.) Treatise on the Essential Gate of Entering the Way through Sudden Awakening 頓悟入道要門論. Attributed to Dazhu Huihai. X, vol. 63, no. 1223. (This text probably records the discourses of Dazhu’s preceptor, Daozhi 道智, who belonged to Shenhui’s lineage.) Wang, Youru. 2003. Linguistic Strategies in Daoist Zhuangzi and Chan Buddhism: The Other Way of Speaking. London/New York: RoutledgeCurzon. (A rich and thoughtful work that deals with the linguistic strategies of Chan Buddhism as well as Daoist Zhuangzi.) Wanling Record of Chan Master Huangbo Duanji 黃檗斷際禪師宛陵錄. Compiled in the Recorded Sayings of Ancient Worthies, X 68: 16b20–23a6. (A record of the sayings of the Chan master Huangbo Duanji.) Wright, Dale S. 1992. “Rethinking Transcendence: The Role of Language in Zen Experience.” In Philosophy East and West 42.1: 113–138. Yang, Zengwen 楊曾文, ed., 1996. The Record of the Chan Discourses of the Monk Shenhui 神會 和尚禪話錄. Beijing: Zhonghua Shuju. Yang, Zengwen 楊曾文, ed., 2001. The Platform Sūtra of the Sixth Patriarch: A New Recension Based on the Manuscripts from Dunhuang 新版敦煌新本六祖壇經. Beijing: Zongjiao Wenhua Chubanshe. (A recommendable, well-collated edition of the Platform Sūtra of the Sixth Patriarch.) Ho, Chien-hsing  is Associate Research Fellow in the Institute of Chinese Literature and Philosophy at Academia Sinica, Taiwan. He received his doctorate in philosophy from the University of Delhi, India in 1999. Afterwards, he taught for many years in the Graduate Institute of Religious Studies at Nanhua University, Taiwan, before joining Academia Sinica in 2017. He specializes in Indian and Chinese Madhyamaka, Buddhist epistemology, and Buddhist philosophy of language, with additional research interests in Chan Buddhism and comparative philosophy. He is currently planning a book in English on Chinese Madhyamaka.

Part III

Logical Thought Transplanted from India and the West

Chapter 22

Yin Ming 因明 in Chinese Buddhism Mingjun Tang

What would happen when a zoologist steps into a pristine forest? She finds there a special kind of fresh excrement. By examining it scientifically, she knows with certainty that it belongs to a certain kind of raven. However, no raven was reported to exist in that area – this seems due to our poor knowledge of that forest. So, this zoologist picks up her telescope, trying to find that special kind of raven. At that moment, a thought flashes through her mind that she must at first take attention to any flying figure that is black. This thought is natural although it is perfectly possible that the raven that she wants to observe turns out to be not a black one. From the viewpoint of yinming logic, what backs the above natural thought is an argument running as follows. Conclusion: Premise 1: Premise 2: Premise 3:

The bird x to which the above kind of excrement belongs must be black; Because it is a raven. Whatever is a raven is observed to be black, like a raven a which is well known. Whatever is not black is observed to be not a raven, like a white horse.

According to yinming logic, the acceptability of the above argument is grounded in the fulfillment of three conditions. In connection with the present case, the first condition demands that being a raven should be a property of this bird x. Now it is fulfilled on account of the zoologist’s scientific examination of the excrement that she finds.

M. Tang (*) School of Philosophy, Fudan University, Shanghai, China e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_22

407

408

M. Tang

The second condition demands that there exists at least one individual which exhibits both the property of being a raven and the property of being black and that this individual should be not the bird x, which has not been observed yet. It is ­fulfilled because the raven a mentioned in premise 2 is well known to be a raven and at the same time black. The third condition demands that there exists no individual which is not the bird x and which does not exhibit the property of being black while exhibits the property of being a raven. It is fulfilled – suppose no raven was observed to be not black until that moment. The truth of premise 1 is ensured by the fulfillment of the first condition. The truth of premise 2 is ensured by the fulfillment of both the second and the third conditions. The truth of premise 3 is ensured by the fulfillment of the third condition. When all the above three conditions are fulfilled, premises 1–3 will be all true. In this situation, it is acceptable, according to yinming logic, to derive from premises 1–3 the conclusion, or rather an assumption, that the bird x must be black. As a matter of fact, premises 2 and 3 are true only in terms of previous observation. Namely, they are not universally true, since the bird x which is known with certainty to be a raven is not taken into account in either of the context of premises 2 and 3. The bird x has not been observed yet. The problem whether this bird, being a raven, is in fact black or not is undecidable at that moment. Hence, the bird x can be neither a counterexample nor an affirmative case to the proposition that whatever is a raven is black. Hence, a basic presupposition of the working of the above argument is that: when a principle that whatever is a raven is black holds in the realm of things other than the bird x, the same principle should be also applicable to this bird x. Namely, this bird x is presupposed to be not a single exception or “a singular abnormal case” to a principle which holds everywhere outside of the bird x. Now, let us turn to yinming logic. A brief history of yinming will be given at first so that an interested reader could know where to start if she would like to study yinming by herself. Then, we will explain step by step the structure of an argument as elaborated in yinming logic and the three conditions warranting the acceptability of this form of argument.

1  A Brief History Yin ming 因明 is Chinese translation of the Sanskrit word hetuvidyā, which means literally the science (vidyā, ming 明)1 of reasons (hetu, yin 因). This is an appellation of logic, or rather the theory of reasoning and argument, in the famous Buddhist

1  As a convention of this chapter, Sanskrit terms are given in brackets and followed by their usual Chinese translation in hetuvidyā literature. Sanskrit words hypothetically restored from Chinese begin with an asterisk (∗).

22  Yin Ming 因明 in Chinese Buddhism

409

encyclopedic work Yogācārabhūmi (Yu jia shi di lun 瑜伽師地論, “Stages of Yoga Practice,” YBh) compiled by Asaṅga (Wuzhuo 無著, ca. 315–390 CE).2 Five vidyās are enumerated there. They are five basic disciplines in medieval India, according to the following account of Xuanzang 玄奘 (602–664 CE). [W]hen the children reach the age of seven, the great treatises of the five sciences (ming 明) are gradually imparted to them. The first is the science of grammar (sheng ming 聲明), which explains the meanings of words and classifies them into different categories. The second is the science of manual craft (gong qiao ming 工巧明), which teaches arts and mechanics, the principles of yin and yang (negativity and positivity), and calendrical calculation. The third is the science of medicine (yi fang ming 醫方明), including the application of incantation, exorcism, drugs, stone needles, and moxibustion. The fourth is the science of reasons (yin ming), by which correctness and incorrectness are ascertained, and truth is differentiated from falsehood. The fifth is the inner science (nei ming 內明), which thoroughly investigates the teachings of the Five Vehicles and the subtle theory of cause and effect. 七歲之後, 漸授五明大論:一曰聲明, 釋詁訓字, 詮目疏別。二工巧明, 伎術機關, 陰陽曆數。三醫方明, 禁呪閑邪, 藥石針艾。四謂因明, 考定正邪, 研覈真偽。五曰內 明, 究暢五乘因果妙理。(Da tang xi yu ji 大唐西域記, T51, no. 2087, 876c17–21; Li 1996: 55–56, with my modification)

Just as yin ming is classified separately from nei ming in the above list, logic, being different from philosophical and religious dogmas, is a tool common to all the schools of Indian philosophy and available for any intellectual activity in classical India. It has a long history of more than one and half millennia dating from the beginning of the Common Era. Various schools of Indian philosophy, e.g. Nyāya, Vaiśeṣika, Sāṅkhya, Jainism, and Buddhism etc., contributed to its development and it was studied under different titles like ānvīkṣikī (“investigation”), vāda (“debate”), nyāya (“method”), tarka (“reasoning”), and later pramāṇa (“means of valid cognition”), etc. Hetuvidyā was only one among them. The first systematic presentation of Indian logic is the Nyāyasūtra (“Aphorisms on Method,” ca. second century CE), which is the classic of the Nyāya school. The first systematic presentation of hetuvidyā is found in one chapter of YBh. The chapter was only one of many competing manuals on logic in its times. No substantial contribution was made to the historical development of Indian logic by this chapter. The author of this chapter only reorganized materials handed down to him.

2  The definition of hetuvidyā in YBh is: hetuvidyā katamā. parīkṣārthena yad vidyamānaṃ vastu. (Yaita 2005: 98, HV 0) 云何因明處?謂於觀察義中諸所有事 (T30, no. 1579, 356a11–12) “What is hetuvidyā? [It treats] whatever existent factors related to the aim of [critical] examination.” The definition of hetuvidyā/yin ming in Chinese literature is more complicated, see Tang 2015: 260–273. 3  Chen na 陳那 is the shortened form of chen na jia 陳那迦 (ZYS 4.27a), the latter being a transcription of the Sanskrit diṅ-nā-ga (another pronunciation of dignāga according to Sanskrit internal Sandhi rules). The ideogram 陳, pronounced chén in modern Mandarin, was pronounced /ȡ’i̭ěn / around 600 CE in Chang’an 長安 (Karlgren 1957: 106–107, 陳 373a). The current Min-nan dialect 閩南語 in southern Fujian province of China and in Taiwan still preserves a similar pronunciation of 陳, which also begins with a voiced dental. I owe this footnote to Dr. Kejun Yu.

410

M. Tang

Logic earned the qualification as Buddhist about one century later through the intellectual efforts of Dignāga (Chen na 陳那,3 ca. 480–540 CE). It was Dignāga who wrote a lot of important works on logic. Of them there were at first the Hetucakraḍamaru (“The Drum Wheel of Reason”), then the Nyāyamukha (“Entrance to Method,” NMu), and at last the Pramāṇasamuccaya (“Compendium on Means of Valid Cognition,” PS).4 In them he advocated a theory of reasoning which signified a remarkable movement towards the idea of formal logic. By formal logic I mean a theory where arguments can be evaluated as reliable or not in a general way. Scholars after Dignāga, whatever scholastic affiliation they might bear, could make no progress in this field unless they took into consideration Dignāga’s theories, either following or criticizing them, but absolutely not neglecting them. Hence Dignāga was praised by modern authors as “the father of mediaeval logic [in India; M.T.]” (Vidyabhusana 1921: 270). The hetuvidyā tradition in China was mainly established by Xuanzang and his disciples. Although there were at least two translated texts on logic in Chinese Buddhist literature before Xuanzang, i.e., Fang bian xin lun 方便心論 (∗Prayogasāra) and Ru shi lun 如實論 (∗Tarkaśāstra), 5 they had little influence in China. When Xuanzang arrived at Chang’an 長安 in 645, he brought back 657 works from India. Among them there were 36 works on logic (Li 1996: 394), but only two were translated into Chinese by him. They were the Nyāyapraveśa (“Introduction to Method,” NP) of Śaṅkarasvāmin (Shang jie luo zhu 商羯羅主, ca. 500–560 CE) and the Nyāyamukha of Dignāga, being translated in 647 and 649 respectively. In the same period, YBh was also translated by him from 646 to 648 (Lü 1980: 80, 84). 6 It has been surmised that the translation of NP and NMu was aimed at providing supplementary materials for understanding the aforementioned hetuvidyā chapter of YBh (Takemura 2011: 26–31). In fact, the appellation hetuvidyā can be found only in YBh among Indian literature on logic we know today. The term is not attested to have taken place in any work of Dignāga and Śaṅkarasvāmin we have today. The two characters yin ming leading the Chinese title of both NMu and NP, i.e., Yin ming zheng li men lun 因明 正理門論 and Yin ming ru zheng li lun 因明入正理論, were added by Xuanzang himself perhaps with the aim of showing the doctrinal affinity of these two works to the hetuvidyā chapter of YBh, although the theory elaborated in NMu and NP and that in the hetuvidyā chapter of YBh were remarkably different. Nevertheless, YBh was one of the most authoritative pieces of the Faxiang School (Fa xiang zong 法相 宗), i.e., the East Asian tradition of Yogācāra Buddhism. Xuanzang was actually the initiator of this school. YBh abided a central place in Xuanzang’s career as a

 For Dignāga’s career and the sequence of his works, see Frauwallner 1959.  They were translated in 472 and 550 respectively, see Lü 1980: 84. 6  Luo 1981 argues that the date of Xuanzang’s translation of NMu should be 655 but not 649. We await further study on this issue. 4 5

22  Yin Ming 因明 in Chinese Buddhism

411

preacher of Yogācāra Buddhism. Therefore, it might be safe to say that the appellation hetuvidyā/yin ming was deliberately chosen by Xuanzang as a general title given to the Buddhist doctrines of reasoning and argument. In this sense, the Chinese tradition can be regarded as a hetuvidyā tradition in contrast with the pramāṇa tradition in India and Tibet. When he studied in India, Xuanzang learned NMu, PS and NP for many times under different teachers on different occasions (Yang 2011: 17–18; Zheng 2007: 94–97). Moreover, before his leaving for China, he successfully established the Yogācāra idealist tenet by means of an inference in a great debate held by King Śīlāditya (Jie ri wang 戒日王) in Kanyākubja (Qu nü cheng 曲女城). The famous inference, i.e., the inference of consciousness-only (wei shi bi liang 唯識比量), was later renowned as a truthful inference of consciousness-only (zhen wei shi liang 真 唯識量). Now, at the same time when Xuanzang translated NMu and NP, he had oral explanations of the texts to his translation team and disciples (see ZYS 1.2a2– b4; Zheng 2007: 86–90). The first generation of authors in China on hetuvidyā emerged out of the audience of Xuanzang’s oral explanations. They wrote commentaries on these two treatises on the basis of Xuanzang’s teachings. Taking into account Xuanzang’s learning experiences in India, we could imagine how substantial his oral explanations of Buddhist logic would be. There were about 20 works on logic out of the first generation. However, only four survived. They are Shentai 神 泰 (seventh century CE)’s commentary on NMu (YZMS, incomplete), Wengui 文 軌 (ca. 615–675 CE, Shen 2008: 3)’s commentary on NP (ZYS), and Jingyan 淨眼 (seventh century CE)’s two works which, combined together, form a complete commentary on NP (see Shen 2008). They are first hand materials for us to know Xuanzang’s teachings of Buddhist logic. Kuiji 窺基 (632–682 CE), the renowned son-in-Dharma of Xuanzang, belonged to the second generation. He based his Commentary on Nyāyapraveśa (Yin ming ru zheng li lun shu 因明入正理論疏) on not only works of the first generation, but also what Xuanzang later taught him separately.7 This made his work a grand synthesis of hetuvidyā doctrines handed down by Xuanzang. Hence, his Commentary was renowned as The Great Commentary on Hetuvidyā (Yin ming da shu 因明大疏, YMDS) in the hetuvidyā tradition.8 Most works on hetuvidyā after him were either in the form of commentary on it or as studies on certain topics in it. This is the case for not only his disciple and grand-disciple, Huizhao 慧沼 (650–714 CE) and Zhizhou 智周 (668–723 CE), but also the followers of the hetuvidyā tradition in Japan. In comparison with works before Kuiji, YMDS is far more comprehensive. The interpretations herein are more sophisticated. It bears much ambition of building a system of Buddhist logic. The works before him, sorted as “old commentar-

7  See Song gao seng zhuan 宋高僧傳, T50, no. 2061, 725c24–726a1; cf. Zheng 2007: 156–159, 2010: 4–7. 8  However, YMDS was not completed during Kuiji’s lifetime. The remaining part, about one sixth, was written by his disciple Huizhao 慧沼 (650–714 CE) after his death.

412

M. Tang

ies” (gu shu 古疏) in contrast with Kuiji’s “great commentary,” are less systematic and more like notes of Xuanzang’s oral explanations. However, we cannot safely attribute to Kuiji himself the elaborations which are not found in the “old commentaries” but only in the “great commentary,” because Kuiji was said to be taught by Xuanzang separately. As a matter of fact, Kuiji’s innovation alone cannot reasonably explain all the differences between the “old commentaries” and the “great commentary” if we read them more closely. It still remains a tough task to discriminate what is Xuanzang’s and what is Kuiji’s in YMDS. Nevertheless, to read YMDS together with the “old commentaries” will prove meaningful for us to appreciate different historical layers of the hetuvidyā tradition. Furthermore, the hetuvidyā tradition is not only a Chinese one, but also a tradition of Buddhist logic in the East Asian World. On one hand, when Xuanzang arrived in Chang’an, there had already been brilliant monk-scholars from Korea, e.g. the famous Wǒnch’ǔk 圓測 (613–696 CE). He became Xuanzang’s disciple and was among the first generation of authors in China on hetuvidyā. Tojǔng 道證 (seventh–eighth century CE) came from Korea, studied under Wǒnch’ǔk, and came back then. T’aehyǒn 太賢 (active during 735–744) was Tojǔng’s disciple in Korea. Wǒnhyo 元曉 (617–686 CE), having never been in China, was an insightful scholar on the subject. He was famous for his counterargument against Xuanzang’s inference of consciousness-only. All the four scholars had works on Buddhist logic. On the other hand, hetuvidyā was brought to Japan through Dōshō 道昭 (629–700 CE) and Genbō 玄昉 (?–746 CE). Both studied in China. The former was a disciple of Xuanzang, the latter of Zhizhou. They initiated respectively the South Temple Tradition (Nanji den 南寺伝) and the North Temple Tradition (Hokuji den 北寺伝) of hetuvidyā studies in Japan (Takemura 2011: 58–62).9 Of the hetuvidyā scholars in medieval Japan, the most famous two were Zenju 善珠 (723–797 CE) and Zōshun 藏俊 (1104–1180 CE). The former authored the Inmyō ronsho myōtō shō 因明論疏 明燈抄 (“Lamp of the [Great] Commentary on Hetuvidyā,” IRMS), the latter the Inmyō daisho shō 因明大疏抄 (“Notes on the Great Commentary on Hetuvidyā,” IDS). These two voluminous works are indispensable aids for us to understand not only Kuiji but also numerous hetuvidyā authors before and after Kuiji whose works have already lost and only fragments are preserved as citations here and there in these two works as well as in many other works composed by monk-scholars in medieval Japan. Tang dynasty (618–907 CE) witnessed the golden age of hetuvidyā in China. There were about 78 works on Buddhist logic composed in Tang dynasty; of them, a few were written by Korean scholars. Although most of these works do not survive today, we can see from the above number how Chinese were fascinated by the art of reasoning and argument. However, the introduction of Buddhist logic and its elaborations in China had neither relation to nor interaction with the indigenous ideas of logic as mainly proposed by philosophers in early China. The study of Buddhist

9  A summary of different views between these two traditions on Yogācāra philosophy and hetuvidyā is the Hossō tōmyō ki 法相燈明記, T71, no. 2310.

22  Yin Ming 因明 in Chinese Buddhism

413

logic declined together with the decline of the Faxiang School after the suppression of Buddhism by Emperor Wuzong 武宗 in 845–846 (Takemura 2011: 56). Texts written under the guidance of Xuanzang got lost in China gradually. Till the Ming dynasty (1368–1644 CE) when the interest in hetuvidyā revived among literati, there was no work composed by masters in Tang dynasty at their hands. This poor situation made their writings seriously imperfect. There were about 18 works composed in Song dynasty (960–1279 CE), and about 6 works in Ming dynasty. However, the hetuvidyā tradition in Japan has never been interrupted. Works by masters in Tang dynasty were preserved and studied there meticulously. There were about 289 works composed by Japanese monk-scholars from the beginning to the early twentieth century.10 Thanks for their efforts, hetuvidyā received its true revival in China when YMDS was recovered from Japan and reissued in 1896 by the Jinling Sūtra Publishing House (Jingling kejing chu 金陵刻經處).11 Xuanzang was not the only translator of Buddhist texts on logic in medieval China, but he was the only successful preacher. In medieval China, he was the main source of all the things about Indian Buddhist logic besides a limited number of translated texts, most of which also came from his translation team. All the writings on Buddhist logic in Tang dynasty were based on his teachings either directly or indirectly. Without making reference to writings in Tang dynasty, later scholars in China, Korea and Japan could not help to be blind before the subject. It still remains a principle today that the writings out of the first and second generations of hetuvidyā scholars in China are most valuable sources for a serious scholar to understand hetuvidyā, although discrepancies are found in them from time to time. As mentioned above, the two fundamental texts of this tradition are NMu and NP. Śaṅkarasvāmin, the author of NP, was said to be a disciple of Dignāga (YMDS 13; 91c26–92a1). His Nyāyapraveśa, following the framework of NMu, was a summary of Dignāga’s theory of logic. However, what was taught by Xuanzang was not limited to the theories elaborated in these two treatises. He reinterpreted the texts even according to later views expressed in Dignāga’s last magnum opus, the PS, and expounded the texts from the perspective of new developments even after Dignāga (see Tang, forthcoming). Meanwhile, around the seventh century in India, Buddhist logic was transformed by Dharmakīrti (Facheng 法稱, ca. 600–660 CE) in the form of reinterpreting Dignāga. Through his transformation various dialectical considerations were removed from Buddhist logic, while more ontological and epistemological bearings were given to it. It was Dharmakīrti’s logic that prevailed among later Indian Buddhists till the disappearance of Buddhism in India around the thirteenth century. His works were transmitted to Tibet and formed the basis of Tibetan Buddhist logic. He was the major figure in the Indian and Tibetan pramāṇa tradition. Scholars after him were used to understand Dignāga through his eyes. However,

 All the above calculations are based on the list in Takemura 2011: 356–346.  For the revival of Buddhist logic in modern China, see Zheng 2007: 292–358 and Aviv 2015; for the scholarship in contemporary China, see Zheng 2007: 359–506.

10 11

414

M. Tang

his name was unknown to the Chinese until Yijing 義淨 (635–713 CE). He had no influence on the hetuvidyā tradition. Therefore, as a working hypothesis, the hetuvidyā tradition can be regarded as mainly a tradition following the Indian interpretation of Dignāga before Dharmakīrti. As a matter of fact, most of the monographs on Buddhist logic after Dignāga and before Dharmakīrti do not survive today. They were shadowed by the success of Dharmakīrti. Our knowledge about what happened to Indian Buddhist logic during this period depends to a great extent on what we can learn from the hetuvidyā tradition. Nevertheless, as a precaution, we are not justified in attributing what was reported in the tradition to Indian thinkers without textual evidence of any kind from the Indian side, nor can we safely speak of a new development of Buddhist logic in China even when we fail to collect evidence from extant literature to prove its Indian authorship. We cannot know exactly what is Indian and what is Chinese in many cases. The boundary is unclear.

2  The Framework The basic framework of NMu and NP is different from that of PS and many other works in the pramāṇa tradition. The former two adopt a framework which is called “two goals with eight topics” (er yi ba men 二義八門) in the hetuvidyā tradition. The latter group adopts a framework of two pramāṇas (“means of valid cognition”). The former framework is sketched in the opening verse of NP as follows. Demonstration (sādhana), refutation (dūṣaṇa) and their false forms (ābhāsa) are for the understanding of others. Perception (pratyakṣa), inference (anumāna) and their false forms are for the understanding of oneself. sādhanaṃ dūṣaṇaṃ caiva sābhāsaṃ parasaṃvide | pratyakṣam anumānaṃ ca sābhāsaṃ tv ātmasaṃvide || 能立與能破 及似唯悟他, 現量與 比量 及似唯自悟。(NP 1, cf. Tachikawa 1971: 120)

The two goals are (i) the goal of bringing knowledge to others (parasaṃvid, wu ta 悟他) and (ii) the goal of bringing knowledge to oneself (ātmasaṃvid, zi wu 自悟). Under each title, there are four topics to be investigated. Under the former goal, there are (1) demonstration (sādhana, neng li 能立), (2) refutation (dūṣaṇa, neng po 能破), (3) false demonstration (sādhanābhāsa, si neng li 似能立) and (4) false refutation (dūṣaṇābhāsa, si neng po 似能破). Under the latter goal, there are (5) perception (pratyakṣa, xian liang 現量), (6) inference (anumāna, bi liang 比量), (7) false perception (pratyakṣābhāsa, si xian liang 似現量) and (8) false inference (anumānābhāsa, si bi liang 似比量). Hence there are eight topics in total. The framework of two pramāṇas consists of (i) perception and (ii) inference. The latter one, i.e., inference, is further divided into two types, i.e., inference for oneself (svārthānumāna) and inference for others (parārthānumāna). The former type signifies inference as a mental process of reasoning yet unverbalized, which is termed “inference” (anumāna) in the above framework. The latter type signifies the linguistic expression of an inference, which is termed “demonstration” (sādhana) in the

22  Yin Ming 因明 in Chinese Buddhism

415

above framework. Inference for others or “demonstration” comes to what we mean by argument today, i.e., “a structured group of propositions, reflecting an inference” (Copi and Cohen 2005: 7). Both the svārthānumāna/parārthānumāna distinction in the framework of two pramāṇas and the anumāna/sādhana distinction in the framework of two goals correspond approximately to the inference/argument distinction in today’s logic.12 However, the emphasis in each framework is different. In the framework of two pramāṇas, only inference for oneself is regarded as genuine inference. Inference for others is called “inference” (anumāna) only in a secondary sense, because the linguistic expression is not directly inference. It is the “cause” of an inference the speaker wants to invoke in the other’s mind through that expression. Only the “result,” i.e., an inference for oneself thus invoked in the other’s mind, can be called “inference” proper. Here, “the name of the result is metaphorically transferred to its cause” (kāraṇe kāryopacārāt, NB 3.2). The cause is named after its result.13 In the framework of two goals, “demonstration” (sādhana) or argument is the foremost concern, but not “inference” (anumāna). The section on “demonstration” and “false demonstration” is the most extensive part in both NMu and NP. Furthermore, as is explained by Kuiji as follows, only argument is regarded as genuine means of proof, while inference is one of many preparatory steps for constructing a genuine means of proof. The methods of proving [the ascription of] a [certain] attribute (artha, yi 義)14 [to a certain property-possessor (dharmin, you fa 有法) include] at first a genuine demonstration (zhen li 真立), for it proves [the ascription of] the attribute in a direct way; And second, instruments for demonstrating (li ju 立具), for they are the basis of a demonstration. A true reason (hetu, yin 因) and [a true] example (dṛṣṭānta, yu 喻) are named [collectively] genuine demonstration. Perception and inference, as two means of valid cognition (pramāṇa, liang 量), are named instruments for demonstrating. Therefore, various masters before [Dignāga also] called them [i.e., perception and inference] literally demonstration. [However,] they are no longer [regarded as] genuine demonstration after Dignāga, but only instruments for demonstrating, since they are [only] preparatory steps for [constructing] a demonstration. 立義之

 By inference we mean nowadays “a process of linking propositions by affirming one proposition on the basis of one or more other propositions.” By argument we mean “a structured group of propositions, reflecting an inference” (Copi and Cohen 2005: 7). The working of an inference does not necessitate the medium of certain linguistic expressions. An inference is only a process of pure thinking, regardless of whether or not it is expressed linguistically. However, the working of an argument necessitates the medium of certain linguistic expressions. An argument should spell out an inference in certain linguistic forms which are suitable to be understood by others. 13  The same idea has already been expressed by Dignāga in PS, see Potter 2003: 343. Thanks to Prof. Tom J. F. Tillemans for explaining this idea to me. 14  Yi 義 in this sense, translated as “attribute,” is in contrast with ti 體 “entity.” The ti 體/yi 義 distinction, as well as the distinction between property-possessor (dharmin, you fa 有法) and property (dharma, fa 法), corresponds to the subject/predicate distinction today. In addition, yi 義, as well as the corresponding Sanskrit artha, has many other meanings like “meaning,” “goal” and “object” etc. 12

416

M. Tang 法:一者真立, 正成義故。二者立具, 立所依故。真因喻等, 名為真立。現比二量, 名為 立具。故先諸師, 正稱能立。陳那以後, 非真能立, 但為立具, 能立所須。(YMDS 83; 96b11–15)

Here, Kuiji makes reference to the literal meaning of “demonstration” (sādhana), i.e., means (-ana) of proof (SĀDH). In logic texts before Dignāga, like the hetuvidyā chapter in YBh, there are various means of proof including perception and inference. But in the context of NMu and NP only an argument comprising three statements is regarded as sādhana. The other means of proof mentioned in earlier texts are now only regarded as preparatory steps for constructing a sādhana. This is not only the case for perception but also for inference. Both are now outside of “means of proof.” At the same time, refutation (dūṣaṇa) is only the application of the method of sādhana in refuting others in a debate. It is also a topic appended to the exploration of sādhana. Therefore, it is argument that constitutes the focus of the framework of two goals, but not other topics like inference or refutation. As this framework is universally accepted in the hetuvidyā tradition, argument also becomes the foremost concern of their logical investigation. On numerous occasions in hetuvidyā literature where “inference” (bi liang 比量) or simply “pramāṇa” (liang 量) is mentioned, what is actually meant is “demonstration” (neng li 能立), or rather argument.

3  A Three-Membered Argument According to the Indian custom, an argument or “demonstration” always takes place in the context of a debate between a proponent (vādin, li 立) and an opponent (prativādin, di 敵) with a witness (sākṣin, zheng 證) or a group of witnesses of some kind. A stock example of good argument can be fully expressed as follows (see NP 2.4, cf. Tachikawa 1971: 121–122). In the hetuvidyā tradition, an argument comprises three statements, i.e., “thesis” (pakṣa, zong 宗), “reason” (hetu, yin 因) and “example” (dṛṣṭānta, yu 喻). They are three members (trayāvayava, san zhi 三支) constituting a three-membered argument (san zhi zuo fa 三支作法) as we call it today. In addition, the “example” is divided into “positive example” (sādharmyadṛṣṭānta, tong [fa] yu 同[法]喻) and “negative example” (vaidharmyadṛṣṭānta, yi [fa] yu 異[法]喻). In practice, only positive example is indispensable while negative example is frequently omitted. So the positive example can also be called simply “example.” In contrast to Western logic, the Indian form of argument is always expressed in the reversed order. Namely, the conclusion, i.e., the “thesis,” is stated at first so as to indicate the subject matter of a debate, and then the premises, i.e., the “reason” and the “example,” are stated to support the proponent’s assertion of the truth of the “thesis.” Of the above argument, the thesis “sound is impermanent” can be considered as a subject-predicate statement where the predicate qualifies the subject. This kind of qualification should be upheld by the proponent and be negated by the opponent at

22  Yin Ming 因明 in Chinese Buddhism

417

the same time. That is why debate takes place. The subject “sound” can also be called pakṣa in a metaphorical sense. Another name of it is “property-possessor” (dharmin, you fa 有法) since it is upheld by the proponent to possess the property (dharma, fa 法) signified by the predicate “impermanence.” The predicate-property is called “the property to be proved” or “the inferable property” (sādhyadharma, suo li fa 所立法), or simply “the inferable” (sādhya, suo li 所立). If we have P standing for the subject (pakṣa), and S standing for the predicate (sādhya), we obtain the form (x)(Px → Sx) as a logical expression of the thesis. In order to persuade the opponent of the above qualification or possession relation, the proponent should cite another property which is accepted by both the proponent and the opponent to reside in the subject. The property “producedness” (kṛtakatva, suo zuo xing 所作性), cited for that aim, is called the “reason.” The second member “because of being produced,” or rather “because sound is produced,” using that property as predicate, is also called the “reason.” We call it a reason-statement (yin yan 因言) so as to differentiate it from the reason-property (yin fa 因法) “producedness.”15 We have H standing for the reason (hetu), and obtain the logical form (x)(Px  →  Hx) for the reason-statement. In addition, the reason-­ statement implies an existential import to the effect that there exists something which instantiates not only the subject (P) but also the predicate (H = reason) of that statement. Hence, we have the existential claim (∃x)(Px ∧ Hx).16 Putting it in conjunction with the above form, we obtain (x)(Px → Hx) ⋀ (∃x)(Px ∧ Hx) as a complete expression of the reason-statement. Then, in order to explain the relevance of the reason-property to the inferable property, the proponent should put forward the example. The term “example” (dṛṣṭānta, yu 喻) is inherited from early Indian logic. In early Indian logic, at the place of “example,” there is no proposition explicating the general relation between producedness and impermanence. Only the individual case, “a pot,” is mentioned. Such an individual case is also called “illustration” (udāharaṇa) in early Indian

 In the Sanskrit original and its Chinese counterpart, the reason-statement is expressed only by means of an ablative form of the reason-property. The subject of the reason-statement is always omitted. Therefore we have “because of being produced” instead of “because sound is produced” in Fig. 22.1. 16  See NP 3.2.1(4): dravyam ākāśaṃ guṇāśrayatvād ity ākāśāsattvavādinaṃ praty āśrayāsiddhaḥ || 虚空實有, 德所依故, 對無空論, 所依不成。“Ether is a substance because [ether] is a substratum of qualities. [When the argument is spoken] to an upholder of the non-existence of ether, the substratum (āśraya, i.e., subject) is not established.” And NP 3.2.1(1): śabdānityatve sādhye cākṣuṣatvād ity ubhayāsiddhaḥ || 如成立聲為無常等, 若言是眼所見性故, 兩俱不成。“When [one] is to prove that sound is impermanent, [the reason] ‘because [sound] is visible’ is not established for both (ubhayāsiddha).” Cf. Tachikawa 1971: 123–124. See also NMu 2.3; Tucci 1930: 14. NP 3.2.1(4) demands that (∃x)Px should be in place for the reason-statement to hold. Namely, there should be something which instantiates the subject (P) of the reason-statement so as to ensure the “substratum” to be “established.” NP 3.2.1(1) demands that (∃x)(Px ∧ Hx) should be in place for the reason-statement to hold. Namely, there should be something which instantiates not only the subject (P) but also the predicate (H) of the reason-statement. I owe this clarification to discussion with Prof. Yiu-ming Fung. 15

418

M. Tang

logic. Now in the yinming form of logic, both the positive example and the negative example are general statement followed by individual evidence. In Chinese literature, the statement is named yu ti 喻體 (“example-itself”), the individual evidence yu yi 喻依 (“example-basis”), while in Sanskrit both are called dṛṣṭānta (“example”). The general statement of either example is assigned to show the “invariable concomitance” (avinābhāva, bu xiang li xing 不相離性) of the reason (H) to the inferable (S). The avinābhāva relation means that the reason, e.g. producedness, is absent wherever the inferable, e.g. impermanence, is absent. Therefore the relation is in the form (x)(¬Sx →  ¬ Hx). The positive example-statement states this from the perspective of positive evidence, e.g. a pot etc., which affirm that wherever there is the reason there is the inferable. It is in the form  (x)(Hx  →  Sx). The negative example-­statement states this from the perspective of negative evidence, e.g. ether, which negate the possibility that the reason could take place where the inferable is absent. It is in the form (x)(¬Sx →  ¬ Hx). However, the above formulization needs slight but important revisions if we look into the matter more closely. Without the following revisions, one might easily confuse a three-membered argument in yinming logic with the Barbara form of syllogistic reasoning. At first, both of the above general statements are restricted to the realm of objects of which the proponent and the opponent have no different opinion with regard to their being impermanent or not. Therefore, both statements, although stating the affairs in a general way, are not truly universal statements. The positive example-statement is not to state the same thing as a major premise of a Barbara syllogism is normally expected to state that: Whatever produced is impermanent, but it is to state that: Whatever produced is observed to be impermanent, and the like (Tang 2015: 304–306). In the connection of the above argument, the last thing on which the opponent cannot agree with the proponent is that sound is impermanent. This is why debate begins. Hence, if an object fulfills the statement (x)(Hx → Sx), and both sides in debate can agree on that, it cannot be the subject (P) “sound” (Zheng 2015: 161–162). To express this logically, we have (x)(¬Px ⋀ Hx → Sx) for the positive example-statement, and (x)(¬Px  ⋀    ¬  Sx  →    ¬  Hx)  for its negative counterpart. Second, the positive example-statement, on one hand, involves an existential claim that there must be at least one object, e.g. a pot, which exhibits both the reason and the inferable, e.g. both producedness and impermanence. The claim is to confirm the possibility of the co-occurrence of these two properties in reality. In addition, that the object in question exhibits impermanence should also be a state of affairs both sides in debate agree (see NMu 5.3; NP 3.3.1(3)). Therefore it cannot be the subject (pakṣa) itself. So we have (∃x)(¬Px  ⋀  Hx  ⋀  Sx) for this existential claim. It should be put in conjunction with the logical form of positive example-­ statement as described above. On the other hand, the negative example-statement does not involve any existential claim (see NMu 5–5.1). Now, we arrive at a full expression of the positive example-statement as  (x)(¬Px  ⋀  Hx  →  Sx)  ⋀  (∃x) (¬Px  ⋀  Hx  ⋀  Sx), and that of the negative example-statement as (x) (¬Px ⋀  ¬ Sx →  ¬ Hx) (cf. Oetke 1994: 24, ES+eva4). Accordingly, the whole process of argument can be expressed as:

22  Yin Ming 因明 in Chinese Buddhism

419

Indeed, the whole process reflects no idea of deduction. For the entailment from the conjunction of the (positive) example with the reason to the thesis to be in place, what is presupposed is that the subject matter denoted by P should not be “a singular abnormal case” with regard to the truth of the avinābhāva relation, i.e., (x) (¬Sx →  ¬ Hx) (see Oetke 1996: 474). But this is only presupposed. It cannot be approved on account of either/both of the above example-statements. They can prove at most the relation to hold in the realm of things apart from what exhibits P. The presupposition could be invalidated when the subject matter proves to be “a singular abnormal case.” The idea behind this kind of argument is explained in Chinese literature as follows. There are two kinds of properties with regard to the property-possessor “sound.” The first is the property that is not established (asiddha, bu cheng 不成), i.e., “impermanence.” The second is the property that is well established (prasiddha, ji cheng 極成), i.e., “producedness.” On account of that the property that is well established (“producedness”) is present in the sound, [one] proves that the property that is not established, i.e., “impermanence,” [is also present] in the sound, so that [he] makes [the latter property] also well established. 有 法「聲」上有二種法:一不成法, 謂「無常」; 二極成法, 謂「所作」。以極成法在聲 上故, 證其聲上不成無常亦令極成。(ZYS 1.13b–14a, cf. Tang 2015: 292; see also YMDS 177; 102c3–11, and YZMS 1.15b)

That sound is produced is accepted by both sides in debate. Hence, producedness is a property that is well established (or simply established) with regard to sound. That sound is impermanent fails to be accepted by both, especially by the opponent. Hence, impermanence is a property that is not well established (or simply not established) with regard to sound. However, for an entity to be available for checking the avinābhāva relation between producedness and impermanence, a minimal requirement is that either impermanence or its negation, non-impermanence, should be well established with regard to it. For instance, a pot is an entity where not only producedness but also impermanence is well established.17 Ether (ākāśa), according to Indian metaphysics, is an entity where not only non-producedness but also non-­ impermanence is well established.18 Hence, both of them fall in the realm of ­discourse of both the positive and negative example-statements. Both statements are meant to reflect the avinābhāva relation. At any rate, the subject “sound” has no place in the realm in question, because neither impermanence nor its negation is well established with regard to it. Therefore, the realm of discourse of both the positive and negative example-statements is said to be a “domain of all the existent and

 ZYS 1.15b:「所作」兩處俱成就…「無常」瓶成、聲不成。“Producedness is established with regard to both loci [i.e., the sound and a pot]. … Impermanence is established with regard to a pot while not established with regard to sound.” YMDS 129; 100a6–7 ad NMu 2.2: 故知因、喻 必須極成。“Hence, we know the reason and the example should be well established.” 18  This is the case even for those who do not recognize the existence of ether, since in Buddhist logic, a non-existent entity is regarded as devoid of any property, let alone producedness or impermanence. See NMu 5.1. 17

420

M. Tang

non-existent entities other than the subject (pakṣa)” (一切除宗以外有、無法處, YMDS 253; 109b14–17).19 Now, the proponent is to persuade the opponent to accept that sound is impermanent, i.e., to make impermanence also well established with regard to sound. The train of thought runs as follows. The idea is that arguments should be based on what both sides in debate have already agreed. One arrives at a piece of new knowledge through inference actually by means of extending what has already been known to him. It is deemed that one cannot argue in a vacuum.

4  The Criteria for a Good Argument As is mentioned earlier, Dignāga develops a theory of reasoning which signifies a remarkable movement towards the idea of formal logic. He clarifies the structure of a three-membered argument on one hand – as I have shown above; and he determines on the other hand the exact meaning of the three conditions of a correct logical reason (trairūpya, [yin] san xiang [因]三相, “three characteristics [of reason]”). It is from Dignāga that the three conditions become the criteria by which Indian logicians can evaluate an argument as good or not in a general way. However, we will see after my explanation of the exact meaning of the three conditions that Buddhist logic, at least in both Dignāga's version and the yinming version, does not base the account of a cogent argument on the logical form of a three-membered argument. Dignāga’s own expression of the three conditions is as follows. Again, in an inference only the following rule is to be observed: (1) when the [inferential] sign (liṅga, xiang 相 = hetu reason) is ascertained in the subject of inference (anumeya, suo bi 所比), and in cases other than [the subject], we remember (2) that [the inferential sign] is certainly (ding 定) present in cases similar to that [subject in possessing the inferable property], and (3) that [it] is pervasively (bian 遍) absent where that [inferable property] is absent, then knowledge of this [subject] is generated. 又比量中唯見此理:若所比處此相 審定, 於餘同類念此定有, 於彼無處念此遍無, 是故由此生決定解。rjes su dpag pa la yaṅ tshul ’di yin par mthoṅ ste | gal te rtags ’di rjes su dpag par bya ba la ṅes par bzuṅ na | gźan du de daṅ rigs mthun pa la yod pa ñid daṅ | med pa la med pa ñid dran par byed pa de’i phyir ’di’i ṅes pa bskyed par yin no || (NMu 5.5 = PSV 4, K 150b5–7 [Tibetan translation, Kitagawa 1985: 521,8–13], cf. Tucci 1930: 44; Katsura [4]: 74)

One feature of the above trairūpya formulae is that each is stated with an epistemic requirement (cf. Oetke 1994: 77–78). Formula (1) prescribes what is to be ascertained. Formulae (2) and (3) prescribe what is to be remembered. If we disregard the different ways of apprehension between ascertaining and remembering, both of them can be understood in the context of debate as expressing the same thing as being accepted by both sides in debate. Hence, being ascertained or being remem-

 See also YMDS 269; 111a7–9, and Zheng 2015: 161–162. The domain is called “induction domain” in Hayes 1988: 113 and Katsura 2004: 125.

19

22  Yin Ming 因明 in Chinese Buddhism

421

bered amount to being well established. The trairūpya formulae in NP drop all these words indicating an epistemic requirement, but the epistemic requirement should be also presupposed in that context for a correct understanding of the text.20 The reason is of three characteristics. What are the three characteristics? [They are (1) the reason’s] being pervasively (bian 遍) a property of the subject, (2) being certainly (ding 定) present in similar instances, and (3) being pervasively (bian 遍) absent from dissimilar instances. hetus trirūpaḥ | kiṃ punas trairūpyam | pakṣadharmatvaṃ sapakṣe sattvaṃ vipakṣe cāsattvam iti || 因有三相。何等為三?謂遍是宗法性, 同品定有性, 異品遍無 性。(NP 2.2, cf. Tachikawa 1971: 121)

By comparing the Chinese translation of the above two passages with the corresponding Sanskrit and Tibetan texts, we know that the underlined adverbs “certainly” (ding 定) and “pervasively” (bian 遍) are added by the Chinese translator Xuanzang, and reflect his interpretation of the three conditions. The interpretation might have an Indian origin, about which we have no clear knowledge today (Tang 2015: 335–336). According to Dignāga, the three conditions deal with the extensional relations of a logical reason, e.g. “producedness,”21 to three sets of things involved in an argument. They are pakṣa (zong 宗, the subject), sapakṣa (tong pin 同品, similar instance) and vipakṣa (yi pin 異品, dissimilar instance). (1) Pakṣa are things with regard to which neither the inferable property nor its negation is well established, i.e., accepted by both sides in debate to reside in them; (2) Sapakṣa are things with regard to which the inferable property is well established;22 (3) Vipakṣa are things with regard to which the negation of the inferable property is well established.

20  This can be seen when we take into consideration the NP list of faults in violation of the trairūpya formulae, esp. NP 3.2.1 and 3.2.2(6). 21  In what follows we make reference to the argument that “sound is impermanent because of being produced” as a paradigm (see Fig. 22.1), when there is no special indication. 22  NMu 3.3: 若爾, 同品應亦名宗。不然, 別處說所成故。因必無異, 方成比量, 故不相 似。“[Objection:] If so, sapakṣa should be also called pakṣa. [Reply:] This is incorrect, because [a property] is said to be what is to be established/proved (sādhya, suo cheng 所成) with regard to cases other than [sapakṣa, while with regard to sapakṣa, the same property is said to be what has already been established]. [In comparison,] for an inference to be possible, there cannot be any difference of a [logical] reason [when it resides in pakṣa and when it resides in sapakṣa]. Therefore, there is no resemblance [between the property which is to be proved and the property which is used to be a logical reason].” Cf. Tucci 1930: 26; Katsura [2]: 126–128; and Katsura 2004: 122–123. YZMS 2.18b on this passage: 謂聲望瓶是瓶家別處, 於此別處成立無常。其聲上無常, 由敵論 人不許是無常, 今以因成立, 即說聲上無常為所成立。…其瓶上無常, 立、敵先成共許, 不須 成立。既不須成立, 何名所成?“That is to say, sound is a case other than a pot when it is compared with a pot. One is to establish impermanence in this other case. As to the impermanence of

422

M. Tang

In a debate where sound is claimed by the proponent to be impermanent and by the opponent to be not impermanent, sound is the pakṣa, i.e., the subject matter under debate. In this connection, a pot, lightning, and all the other things which have already been accepted by both to be impermanent are sapakṣa (lit. “pakṣa-like”). They are things similar to the pakṣa in that they possess impermanence, which the proponent wants to prove also in the case of pakṣa. Ether, atoms (paramāṇu, ji wei 極微) and all the other things which have already been accepted by both to be not impermanent are vipakṣa (lit. “pakṣa-unlike”). They are things dissimilar to the pakṣa in that they do not possess impermanence. Pakṣa, sapakṣa and vipakṣa, in an ideal manner where pakṣa is the only thing which has not yet been ascertained by both with regard to its possession of the inferable property or not, make up the complete universe of discourse. Hence, we have: (1) pakṣa ={x  |  Px}, (2) sapakṣa ={x | ¬ Px ∧ Sx}, and (3) vipakṣa ={x | ¬ Px ∧  ¬ Sx}.23 The first condition concerns the relation of a logical reason to the pakṣa. It prescribes that a correct logical reason should be a property of the pakṣa, i.e., the subject of an inference. The Chinese translation emphasizes that it should be pervasively a property of the pakṣa. That is to say, it should be a property pervading all the members of the pakṣa, but not only part of them. As is explained by Kuiji, the reason is meant to prove the pakṣa as a whole, and therefore it should be a property exhibited by all the pakṣa members. Objection: It has already been a reason [fulfilling the first condition] when being called pakṣadharma (zong fa 宗法, “a property of the subject”). Why the word pervasively (bian 遍) is necessary? … Reply: If the reason does not pervade [all] the pakṣa [members], then [the members] the reason does not pervade cannot be proved by the reason. There will be some [pakṣa member] which is unproved. In order to show that all [of them] are proved by the reason, the word pervasively should be mentioned. 問:稱為「宗法」, 即已是因, 何須 言「遍」?…答:若因不遍宗有法上, 此所不遍, 便非因成, 有所不立。顯皆因立, 是故 稱「遍」。(YMDS 186; 103a28–b3)

This amounts to the idea that: Since the subject is distributed in the thesis, as is the case with all three-membered arguments, it should also be distributed in the reason-­ statement. In this way, Buddhist logic is intended to avoid an “illicit minor” (Copi

sound, since the opponent does not accept that [sound] is impermanent, now [the proponent] is to establish it by means of a [logical] reason. In this sense we say that the impermanence of sound is what is to be established (sādhya). … As to the impermanence of a pot, since both the proponent and the opponent have already accepted [that a pot is impermanent], there is no need to establish [it again]. Since there is no need to establish [it again], how can it [i.e., the impermanence of a pot] be called what is to be established?” Cf. YMDS 200; 104b6–13. 23  For the “tripartitionism” of the universe of discourse in Indian and Tibetan Buddhist logic, see Hayes 1988: 113–118; Tillemans 1990, 2004; Oetke 1994: 33–51; Katsura 2004; and Hugon 2004. For the “tripartitionism” in Chinese Buddhist logic, see Chen 1997: 70–76; Zheng 1996: 48–50; Franco 2004: 207–211; and Zheng 2015: 155–163. In addition, although qualifications like “well established,” “accepted” and the like are of an epistemic nature, I avoid using epistemic logic throughout this chapter, for this kind of treatment will invite more problems. For pioneering studies in this direction, see Oetke 1994: 77–113.

22  Yin Ming 因明 in Chinese Buddhism

423

and Cohen 2005: 238–239). Kuiji continues to give two examples where the first condition is violated. (1) There is a property of the subject but it is not a pervading one (有宗法而非遍). … For example, a Vaiśeṣika, against a propounder of [the theory] that sound is produced (sheng sheng zhe 聲生者), sets forth [an argument] that: All kinds of sound are impermanent – thesis, because [they] result immediately from effort (qin yong wu jian suo fa xing 勤勇無 間所發性, prayatnānantarīyakatva) – reason. Both sides, i.e., the proponent and the opponent, accept only the internal sound (nei sheng 内聲, i.e., the sound of living creature) as resulting immediately from effort, but not the external sound (wai sheng 外聲, i.e., sound naturally caused). Both the proponent and the opponent claim that the reason is present in half of the pakṣa, while being absent from the remaining half. … (2) There is a property neither of the subject nor is it a pervading one (有非遍非宗法).24 … For example, a Sāṅkhya, against a Buddhist, sets forth [an argument] that: Sound is permanent – thesis, because [it] is visible – reason. Both [sides] claim that the reason is [completely] absent from sound. 「有宗法而非遍」者, …如勝論師對聲生者, 立「一切聲皆是無常宗, 勤 勇無間所發性因」。立、敵二宗, 唯許內聲有勤勇發, 外聲非有。立、敵俱說此因於 宗半有半無。…「非遍非宗法」者, …如聲論師對佛弟子, 立「聲為常, 眼所見故」, 俱說此因於聲無故。(YMDS 189–191; 103b19–c2)

In example (2), the reason “visibility” fails to be a property of the subject. The reality is that there is nothing that is sound (P) and is visible (H). We have ¬(∃x) (Px ∧ Hx) for this scenario. Being zong fa is a negation of it, namely (∃x)(Px ∧ Hx) – there exists some individual which exhibits both pakṣa and hetu (cf. footnote 16). In example (1), the reason “resulting immediately from effort,” though being a property of the subject, fails to be a pervading one. The reality is that there is something that is sound (P) but not resulting immediately from effort (¬H). We have (∃x) (Px ∧  ¬ Hx) for this scenario. Being bian is a negation of it, namely (x)(Px → Hx) – any individual, if exhibiting pakṣa, exhibits hetu. In Kuiji’s view, only a reason being zong fa and being bian at the same time (有是遍亦宗法) fulfills the first condition. Therefore, the first condition comes to be a conjunctive formula, i.e., (x) (Px → Hx) ⋀ (∃x)(Px ∧ Hx). This is exactly a definition of the truth of the reason-­ statement in a three-membered argument (see Fig. 22.2), in the sense that the reason-­ statement is true, if and only if the reason is pervasively a property of the subject. The thought that the reason should pervades all the pakṣa members can be regarded as a natural and implicit requirement of Dignāga’s logic.25 It was indeed spelt out by Dharmakīrti in a later period comparable to that of Xuanzang.26

 Here, being a pervading one means pervading the pakṣa. There are four possible combinations of zong fa and bian, i.e., 有宗法而非遍, 有是遍亦宗法, 有非遍非宗法, and 是遍非宗法 (being a pervading one but not a property of the subject). The last one does not exist according to Kuiji. For Kuiji’s complete discussion, see YMDS 186–193; 103a28–c10. For similar discussions, see ZYS 1.14a–15a, and IRMS 264a29–265a19. 25  Not surprisingly, about Dignāga’s pakṣadharma (zong fa) without qualifications like pervasively, a logician may arrive at the same conclusion as what is expounded in the Chinese tradition. See Hayes 1988: 122, “PH>0 & P~H=0.” 26  See NB 2.5: liṅgasyānumeye sattvam eva, “The [inferential] sign (liṅga = hetu) is universally (eva) present in the subject of inference.” 24

424

M. Tang

However, the addition of bian (pervasively) to the first condition as well as the emphasis thereupon was laid by Chinese scholars independently of Dharmakīrti. Perhaps Xuanzang and Dharmakīrti had a common source in this regard. The second and third conditions concern the relation of a logical reason to the sapakṣa and to the vipakṣa respectively. The second condition prescribes that a correct reason should be present in the sapakṣa, and the third condition that it should be absent from the vipakṣa. The Chinese translation emphasizes correspondingly that a correct reason should be certainly present in the sapakṣa, and pervasively absent from the vipakṣa. Kuiji’s explanation of the necessity of certainly (ding 定) and pervasively (bian 遍) in this connection is related to the structure of positive and negative example-­ statements on one hand and to Dignāga’s theory of hetucakra (“wheel of reason”) on the other hand. According to Dignāga, the structure of the two example-­statements is as follows. [The positive example states that] the reason is followed by the inferable (zong 宗 = sādhya); [the negative example states that] wherever the inferable is absent, the reason is [also] absent. … The former [statement] is negation with affirmative implication (paryudāsa, zhe quan 遮詮), the latter [statement] is only negation with no more implication (prasajyapratiṣedha, zhi lan 止濫), because one infers about an object from positive ­concomitance (anvaya, he 合) and negative concomitance (vyatireka, li 離). 說因宗所隨, 宗無因不有。(NMu k.11ab ≈ PS 4.2ab: sādhyenānugamo hetoḥ sādhyābhave ca nāstitā.) …前是遮詮, 後唯止濫, 由合及離比度義故。(NMu 5–5.1, cf. Tucci 1930: 36–37; Katsura [4]: 62–65)

Keeping the above passage in mind, Kuiji explains at first the necessity of pervasively in the third condition. The negative example is to negate with no more implication [the reason’s presence in the absence of the inferable]. If the negation is not a pervasive one, [the reason’s presence in the absence of the inferable] cannot be [sufficiently] excluded. 異喻止濫, 不遍止者非 遮。(YMDS 201; 104b15–16)

A negative example-statement is to negate the presence of the inferable in an entity at first, and then to negate the presence of the reason also in that entity, like: “whatever is permanent (or not impermanent) is observed to be not produced.” It claims that everywhere the inferable is negated, the reason is negated also (see Fig. 22.1). In addition, the negative example-statement involves no existential claim, since the negation in this statement is said to be “negation with no more implication.” Neither being permanent (i.e., not impermanent), nor being not produced, implies the existence of something not impermanent or something not produced. Now, the third condition is understood by Kuiji as a definition of the truth of the negative example-statement. The negation of the presence of the reason in the case of the third condition should be a pervasive one (bian zhi 遍止) to the effect that the reason should be not exhibited by any vipakṣa member, vipakṣa being the set of things where the inferable is negated. The adverb pervasively in the case of the third condition is to explicate this kind of pervasive negation. It excludes the possibility that the third condition could be fulfilled even when the reason is absent from only some of the vipakṣa members but not all of them. In the case of our paradigm argu-

425

22  Yin Ming 因明 in Chinese Buddhism

Thesis:

Sound is impermanent; anityaḥ śabdaḥ

Reason:

Because of being produced. kṛ takatvāt

Positive example:

Whatever is produced is observed (dṛṣṭa, jian be impermanent, like a pot, etc.;

) to

yat kṛtakaṃ tad anityaṃ dṛṣṭaṃ yathā ghaṭādiḥ Negative example:

Whatever is permanent is observed to be not produced, ). like ether (ākāśa, xu kong yan nityaṃ tad akṛtakaṃ dṛṣṭaṃ yathākāśam

Fig. 22.1  A three-membered argument

ment (Fig.  22.1), the third condition demands that producedness should be not exhibited by anything permanent. Hence, the third condition is of the form ¬(∃x) ((¬Px ∧  ¬ Sx) ∧ Hx) – there exists nothing which does not exhibit the pakṣa nor the sādhya (inferable), while exhibits the hetu (reason). This form is logically equivalent to the logical form of a negative example-statement: (x)(¬Px ⋀  ¬ Sx →  ¬ Hx) (see Fig. 22.2). After explaining the third condition, Kuiji then explains the necessity of certainly in the second condition. He begins with the following discussion of the structure of a positive example-statement. The positive example is to [show that] the inferable (zong 宗 = sādhya) will be established when [it] follows (shun 順) [the reason]. As long as the inferable is established [in this way], a positive example will be called [correct]. [Therefore,] how could a positive example prove an attribute of the pakṣa (zong yi 宗義 = the inferable) [only] when [the reason] pervades the inferable (suo li 所立, sādhya) [in this positive example]? [What is required is only that the inferable] follows [and pervades the reason, but not that the reason follows and pervades the inferable]. [We] only want to prove the inferable by means of the reason, [in the sense that] where the reason is present, the inferable will definitely follows (sui zhu 隨 逐). [We] do not want to prove the reason by means of the inferable, since where the inferable is present the reason is not to be certainly present. Therefore, the reason is not necessitated to be pervasively present in [all] the sapakṣa [members] (zong tong pin 宗同品) [i.e., in all the things exhibiting the inferable]. 同喻本順成宗, 宗成即名同喻, 豈由喻遍能 順所立方成宗義?但欲以因成宗, 因有宗必隨逐。不欲以宗成因, 有宗因不定有。故 雖宗同品, 不須因遍有。(YMDS 201; 104b18–21)

426

M. Tang

Thesis:

( )(

→

)

Reason:

( )(

→

) ⋀ (∃ )(

Positive example:

( )(¬ ⋀ → ) ⋀ (∃ )(¬

Negative example:

( )(¬

⋀ ¬

∧

⋀ →¬

⋀

)

)

)

Fig. 22.2  Logical form of a three-membered argument. (The symbolization, i.e., P for pakṣa, e.g. “sound,” S for sādhya, e.g. “impermanent,” and H for hetu, e.g. “produced,” is used throughout this chapter. In addition, (x)(¬Px ⋀ Hx → Sx) and (x)(¬Px ⋀  ¬ Sx →  ¬ Hx) are logically equivalent. Thanks for the reminder of Prof. Yiu-ming Fung. Hence, the positive example implies the negative one, but the negative example does not imply the positive one. With regard to a pot and ether etc. as mentioned in Fig. 22.1, they are individual cases instantiating either the truth of the positive example or that of the negative one. Therefore, for describing them, we do not need to invent new expressions other than what we have already obtained in Fig.  22.2. For example, in order to describe a pot as mentioned in the positive example, we only need to denote it by an individual constant a. Then, we obtain ¬Pa ⋀ Ha ⋀ Sa. It reads: a pot is not the subject, and it exhibits the reason and the inferable.)

According to Dignāga, a positive example is to affirm the presence of the reason (hetu, H) in an entity at first, and then to affirm the presence of the inferable (sādhya, S) also in that entity, like: “whatever is produced is observed to be impermanent.” We affirm at first the presence of producedness, and then the presence of impermanence (see Fig.  22.1). In contrast to the non-implicative negation in the negative example-statement, the negation as made in the positive example-statement is said to be a “negation with affirmative implication” (see above NMu 5–5.1). That is to say, being impermanent, or rather not permanent, implies that there exists something which is not permanent. Hence, the positive example-statement involves an existential claim. When one asserts that “whatever is produced is observed to be impermanent,” this person, in the same statement, has also asserted that there exists something which is both produced and impermanent. As we have shown above (see Fig.  22.2), a logical expression of the positive example-statement is: (x) (¬Px ⋀ Hx → Sx) ⋀ (∃x)(¬Px ⋀ Hx ⋀ Sx). The latter conjunct reflects this kind of existential claim. In the above passage, Kuiji is quite aware of the fact that the fulfillment of the second condition has some contribution to the truth of the positive e­ xample-­statement. The second condition, as presented in the above cited Sanskrit text of NP 2.2, is that: S0:

The reason is present in the sapakṣa (sapakṣe sattvam, tong pin you xing 同品有性).

According to Kuiji, there could be two possibilities for explicating the above S0 proposition. They are,

22  Yin Ming 因明 in Chinese Buddhism

S1: S2:

427

The reason is present in all the sapakṣa members. The reason is present in some sapakṣa members.

S1 means that the reason is “pervasively present” (bian you 遍有) in all the sapakṣa members, or simply that the reason “pervades” (bian 遍) the inferable. The fulfillment of S1 warrants the truth of the statement that wherever the inferable is present the reason is also present, namely that the reason (H) “follows” (shun 順 / sui zhu 隨逐) the inferable (S). S1 has the logical form (x)(Sx → Hx),27 and expresses the “invariable concomitance” (avinābhāva) of the inferable (S) to the reason (H). Kuiji’s above discussion contains three objections to this S1 version of the second condition. At first, the avinābhāva relation of the S1 kind is not the one that an argument needs for getting a starting point. As a matter of fact, an argument is to prove the pakṣa’s possession of the inferable on account of its possession of the reason. The right avinābhāva relation it needs is the “concomitance” of the reason (H) to the inferable (S), i.e., (x)(Hx → Sx), not that of the inferable (S) to the reason (H), i.e., not (x)(Sx → Hx). Therefore, if the second condition means S1, it will be irrelevant to the actual aim of an argument. Second, the “concomitance” an argument really needs, i.e., (x)(Hx → Sx), does not necessarily hold when the “concomitance” of S1 kind, i.e., (x)(Sx → Hx), holds. Therefore, the fulfillment of S1 has no contribution to the truth of a correctly structured positive example-statement, which is essentially to show a “concomitance” of the form (x)(Hx → Sx). Third, there is certain kind of argument which does not fulfill S1 but is obviously good argument. Hence, if the second condition means S1, it will be too much demanding as a necessary condition for an argument to be good. In the hetuvidyā tradition, there are two kinds of good argument. They are, A1: A2:

Sound is impermanent, because sound is produced. Whatever is produced is observed to be impermanent, like a pot. The internal sound is impermanent, because the internal sound results immediately from effort. Whatever results immediately from effort is observed to be impermanent, like a pot. 28

 This proposition should be also restricted to the realm of all the things outside of the pakṣa, and implies an existential claim. We skip all these restrictions here in order to make the point clear. 28  See YMDS 205; 104c19–21: 「同品有、異品非有」, 如勝論師立「聲無常, 所作性故, 喻如 瓶等」。無常之宗, 空為異品。所作性因, 於同品有, 於異品無。“With regard to [the reason’s being] present in [all the] similar instances and absent from [all the] dissimilar instances, [an argument, as an example for this scenario, is] like: A Vaiśeṣika (sheng lun shi 勝論師) sets forth that sound is impermanent, because of being produced, for example, like a pot, etc. [Here,] the subject (zong 宗 = pakṣa), [which is now claimed to be] impermanent, has ether as a dissimilar instance. The reason, i.e., ‘being produced,’ is present in [all the] similar instances and absent from [all the] dissimilar instances.” And YMDS 207; 105a10–14: 「同品有非有、異品非有」, 如勝論師立「 内聲無常, 勤勇無間所發性故, 喻若電、瓶」。此無常宗, 空為異喻。勤勇之因, 於同品瓶等 上有、電等上無, 異品空中一向非有。“With regard to [the reason’s being] present in [some of the similar instances] and absent from [the other] similar instances and [being] absent from [all 27

428

M. Tang

A1 has already been used as a paradigm of argument in Fig. 22.1. Both the reason of A1, “being produced,” and that of A2, “resulting immediately from effort,” are not exhibited by anything which does not exhibit impermanence and is outside of the pakṣa, i.e., “sound” or “the internal sound” in their respective argument. Both of them are absent pervasively from their respective vipakṣa, and hence fulfill the third condition. The difference relies in the fact that, although all the things impermanent are produced according to Indian metaphysics – impermanence and producedness are coextensive, only some of them result immediately from effort, i.e., produced through the effort of living creature, like a pot, which is produced through the effort of a potter. Things being produced through natural causes, like a flash of lightning and the blow of wind, although being produced and impermanent, do not result from the effort of living creature. Therefore, in the case of A1, the reason “being produced” is present in all the sapakṣa members, i.e., things being impermanent and not identical to sound. In the case of A2, the reason “resulting immediately from effort” is present in only some of the sapakṣa members, i.e., in only some of the things being impermanent and not identical to the internal sound. Although both A1 and A2 are obviously good arguments, A1 fulfills S1 and S2, but A2 fulfills only S2. Hence, Kuiji says that 雖宗同品不須因遍有 “the reason is not necessitated to be pervasively present in [all] the sapakṣa [members].” S1 is too much demanding. S2 seems to be a suitable choice for an interpretation of the second condition. However, there are still two possibilities concerning the exact meaning of S2. They are, S2(1): S2(2):

The reason is present in some sapakṣa members, and is absent from all the vipakṣa members. Namely, the reason is present only in the sapakṣa. The reason is present in some sapakṣa members, and it is not necessary that the reason is at the same time absent from all the vipakṣa members. Namely, the reason, when being present in the sapakṣa, could be present in the vipakṣa as well, with the second condition being still fulfilled.

S2(1) means: There is some individual which exhibits both the reason (H) and the inferable (S), and which is not the pakṣa (P), and there does not exist any individual which is not the pakṣa and does not exhibit the inferable, and which exhibits the reason. S2(1) is of the form (∃x)(¬Px ⋀ Hx ⋀ Sx) ⋀  ¬ (∃x)((¬Px ∧  ¬ Sx) ∧ Hx). It is equivalent to (x)(¬Px ⋀ Hx → Sx) ⋀ (∃x)(¬Px ⋀ Hx ⋀ Sx). However, S2(2) is only of the form (∃x)(¬Px ⋀ Hx ⋀ Sx) (cf. Oetke 1994: 24, ES+eva4 and ES-eva2). Therefore, S2(1) is obviously a definition of the truth of the positive example-­ statement, while S2(2) expresses only the existential claim implied in this statement (see Fig. 22.2). In order to make a decision upon whether S2(1) or S2(2) is the exact meaning of the second condition, let us look at Kuiji’s discussion which immediately follows the above cited passage in YMDS. He says that, In case when a reason is present in the vipakṣa, and [at the same time] is present in half of the sapakṣa and absent from the remaining half, the reason, although being inconclusive at any rate, is faulty on account of its presence in the vipakṣa, but not on account of [its] failure to pervade [all] the sapakṣa [members]. In the [schema of] nine reasons, the first and last

22  Yin Ming 因明 in Chinese Buddhism

429

forms [of reason] in the last three forms correspond to the [above] case. The middle form in the last three forms is included in the [category of] correct reason, since [the reason in this form] negates with no more implication [its presence in the vipakṣa] in an exhaustive way. Both the first and last forms [in the last three forms] are included in the [category of reasons which are] faulty [because] of being inconclusive, since both [of the reasons in these two forms] fail to negate with no more implication [their respective presence in the vipakṣa] in an exhaustive way. Therefore, with regard to the sapakṣa, [a correct reason] is said to be “certainly” present [in the sapakṣa]. With regard to the pakṣa and to the vipakṣa, [a correct reason] is said to be “pervasively” [a property of the pakṣa], and to be “pervasively” [absent from the vipakṣa] respectively. 若因於異品有, 同品半有半無, 雖並不定, 由因於異有故 成過, 非因於同不遍為失。即九句內, 後三句中初、後句是。後三中句正因所攝, 於 異品中止濫盡故。初、後二句不定過收, 皆止異品濫不盡故。由此同品說「定」有 性, 宗、異品中皆說「遍」也。(YMDS 201; 104b21–26)

Here, Kuiji is making reference to the famous theory of nine forms of reason (jiu ju yin 九句因). This theory was established for the first time in Dignāga’s Hetucakraḍamaru. It is known in the Indian context as the “wheel of reason” (hetucakra). This theory provides a list of all the possible relations of the reason to the sapakṣa and to the vipakṣa – on the presupposition that neither sapakṣa nor vipakṣa are empty set. The nine forms are: (1) The reason is present in all the sapakṣa members and is present in all the vipakṣa members, i.e., 同品有, 異品有. (2) The reason is present in all the sapakṣa members and is absent from all the vipakṣa members, i.e., 同品有, 異品非有. (3) The reason is present in all the sapakṣa members, and is present in some of the vipakṣa members and absent from the other vipakṣa members, i.e., 同品有, 異 品有非有. (4) The reason is absent from all the sapakṣa members and is present in all the vipakṣa members, i.e., 同品非有, 異品有. (5) The reason is absent from all the sapakṣa members and is absent from all the vipakṣa members, i.e., 同品非有, 異品非有. (6) The reason is absent from all the sapakṣa members, and is present in some of the vipakṣa members and absent from the other vipakṣa members, i.e., 同品非 有, 異品有非有. (7) The reason is present in some of the sapakṣa members and absent from the other sapakṣa members, and is present in all the vipakṣa members, i.e., 同品有 非有, 異品有. (8) The reason is present in some of the sapakṣa members and absent from the other sapakṣa members, and is absent from all the vipakṣa members, i.e., 同品 有非有, 異品非有. (9) The reason is present in some of the sapakṣa members and absent from the other sapakṣa members, and is present in some of the vipakṣa members and absent from the other vipakṣa members, i.e., 同品有非有, 異品有非有. (Cf. YMDS 203–209; 104b26–105b5; Ganeri 2004: 344–350; Tang 2015: 285–287 and 311–323)

430

M. Tang

Forms (1)–(3) are called the first three forms (chu san ju 初三句), (4)–(6) the middle three forms (zhong san ju 中三句), and (7)–(9) the last three forms (hou san ju 後三句). Hence, we know that in the above passage Kuiji is talking about forms (7), (8) and (9). Among all the nine forms, only forms (2) and (8) reflect the relation a correct reason should have with regard to sapakṣa and vipakṣa. A1 corresponds to form (2), and A2 to form (8). Forms (7) and (9), besides forms (1), (3) and (5), are forms in possession of which a reason will be accused to be inconclusive (anaikāntika, bu ding 不定). Examples of argument for forms (7) and (9) are as follows. A3: A4:

Sound does not result immediately from effort, because sound is impermanent. Whatever is impermanent is observed to not result immediately from effort, like lightning. Sound is permanent, because sound is incorporeal. Whatever is incorporeal is observed to be permanent, like ether.29

A3 corresponds to form (7), and A4 to form (9). The reason of A3 is “being impermanent,” and that of A4 is “being incorporeal.” Their relations to the pakṣa (“subthe] dissimilar instances, [an argument, as an example for this scenario, is] like: A Vaiśeṣika sets forth that the internal sound (nei sheng 内聲, i.e., the sound of living creature) is impermanent, because of resulting immediately from effort, for example, like lightning and a pot. [Here,] the subject, [which is now claimed to be] impermanent, has ether as a negative example [i.e., dissimilar instance]. The reason, i.e., ‘[resulting immediately from] effort,’ is present in [some of the] similar instances [like] a pot, etc., and absent from [the other similar instances like] lightning, etc., and is absolutely absent from [all the] dissimilar instances [like] ether.” Cf. NP 2.2 and Tang 2015: 311–319. 29  See YMDS 207; 105a7–10: 「同品有非有、異品有」, 如聲論師立「聲非勤勇無間所發, 無 常性故, 喻若電、空」。此非勤宗, 瓶為異喻。無常性因, 於同品電上有, 空上非有, 異品瓶 中一向是有。“With regard to [the reason’s being] present in [some of the similar instances] and absent from [the other] similar instances and [being] present in [all the] dissimilar instances, [an argument, as an example for this scenario, is] like: A Śābdika (sheng lun shi 聲論師, i.e., a propounder of the permanence of sound) sets forth that sound does not result immediately from effort, because of being impermanent, for example, like lightning and ether. [Here,] the subject, [which is now claimed to] not [result immediately from] effort, has a pot as a negative example [i.e., dissimilar instance]. The reason, i.e., “being impermanent,” is present in [some of the] similar instances [like] lightning, and absent from [the other similar instances like] ether, and is absolutely present in [all the] dissimilar instances [like] a pot.” And YMDS 207; 105a14–18: 「同品有非有、異品 有非有」, 如聲論師對勝論師立「聲為常, 無質礙故, 喻若極微及大虚空」。此中常宗, 以 瓶、樂等而為異喻。無質礙因, 於其同品虚空上有、極微上無, 亦於異品瓶等上無、樂受等 有。“With regard to [the reason’s being] present in [some of the similar instances] and absent from [the other] similar instances, and [being] present in [some of the dissimilar instances] and absent from [the other] dissimilar instances, [an argument, as an example for this scenario, is] like: A Śābdika, against a Vaiśeṣika, sets forth that sound is permanent, because of being incorporeal (amūrta, wu zhi ai 無質礙), for example, like atom and ether. Here, the subject, [which is now claimed to be] permanent, has a pot, [the feeling of] happiness, etc. as negative examples [i.e., dissimilar instances]. The reason, “being incorporeal,” is present in [some of] its similar instances [like] ether, and absent from [the other similar instances like] atom, and is absent from [some of its] dissimilar instances [like] a pot, etc., and present in [the other dissimilar instances like] the feeling of happiness, etc.” Cf. NP 3.2.2(3), NP 3.2.2(5) and Tang 2015: 311–319.

22  Yin Ming 因明 in Chinese Buddhism

431

ject”), sapakṣa (“similar instances) and vipakṣa (“dissimilar instance”) of their respective argument are as follows. A3 pakṣa: sapakṣa: vipakṣa: A4 pakṣa: sapakṣa: vipakṣa:

Sound (impermanent). Things, apart from sound, which do not result immediately from effort, e.g. lightning (impermanent) and ether (permanent). Things, apart from sound, which do result immediately from effort, e.g. a pot (impermanent). Sound (incorporeal). Things, apart from sound, which are permanent, e.g. ether (incorporeal) and atom (corporeal). Things, apart from sound, which are not permanent, e.g. a pot (corporeal) and the feeling of happiness (incorporeal).

According to Kuiji’s above discussion, A3 and A4 are regarded as unreliable because the truth of their respective positive example-statement meets a counterexample. In the case of A3, a pot, which results immediately from effort and is impermanent, constitutes a counterexample to the statement that: Whatever is impermanent is observed to not result immediately from effort. In the case of A4, the feeling of happiness, which is not permanent and is incorporeal, constitutes a counterexample to the statement that: Whatever is incorporeal is observed to be permanent (Ganeri 2004: 346, Fig. 22.3). In the language of Buddhist logic, A3 and A4 are unreliable because their respective reason fails to be absent from all the vipakṣa members, and fails to fulfill the third condition. Nevertheless, the second condition is said by Kuiji to have been fulfilled in the case of A3 and A4, as is said by Kuiji that: “In case when a reason is present in the vipakṣa, and [at the same time] is present in half of the sapakṣa and absent from the remaining half, the reason, although being inconclusive at any rate, is faulty on Positive example:

For any entity other than sound, if producedness is well established with regard to it, impermanence is also well established with regard to it, like a pot etc.

Negative example:

For any entity other than sound, if the negation of impermanence is well established with regard to it, the negation of producedness is also well established with regard to it, like ether.

Reason:

Now, producedness is well established with regard to sound.

Thesis:

Impermanence must be also well established with regard to it.

Fig. 22.3  Epistemic paraphrase of a three-membered argument

432

M. Tang

account of its presence in the vipakṣa. … Therefore, with regard to the sapakṣa, [a correct reason] is said to be ‘certainly’ present [in the sapakṣa]” (若因於異品有, 同品半有半無, 雖並不定, 由因於異有故成過…由此同品說「定」有性). Therefore, we know that the second condition is considered to hold when the reason is present not only in the sapakṣa but also in the vipakṣa. The above S2(1) version of the second condition is not intended in the hetuvidyā tradition.30 What is intended is only the S2(2) version. Hence, the second condition is of the form (∃x)(¬Px ⋀ Hx ⋀ Sx) – there exists something which does not exhibit the pakṣa and does exhibit the sādhya (inferable) and the hetu (reason). The above two cited passages from Kuiji can be regarded as two steps towards an accurate explication of the second condition. The first passage proclaims that the reason does not need to be present in all the sapakṣa members, rejecting the S1 reading and adopting the S2 reading. The second passage proclaims that the reason does not need to be present only in the sapakṣa realm, rejecting the S2(1) reading and adopting the S2(2) reading. Through these two steps it is made clear that the reason's being certainly present in the sapakṣa realm means exactly its being exhibited by at least one existent sapakṣa member. To sum up, we now arrive at the following list of the three conditions of a correct logical reason in the hetuvidyā tradition. By comparing Fig. 22.4 with Fig. 22.2, we know that: (1) The reason statement is true if and only if the first condition is fulfilled; (2) The negative example-­ statement is true if and only if the third condition is fulfilled; (3) The positive example-­statement is true if and only if both the second and third conditions are fulfilled. The three conditions, being put in conjunction, are conditions for the truth of all the premises of a three-membered argument, although the truth of the conclusion, i.e., the “thesis,” cannot be warranted even all the three conditions are fulfilled and all the corresponding premises are true (see above Fig. 22.2 and my explanation thereof). There are many other stipulations of a good argument in Buddhist logic, but they are of minor importance (cf. NP 3.1 and 3.3). The three conditions are the major criteria for Buddhist logicians in the medieval East Asian World  – and ­probably their predecessors in the sixth and the early seventh centuries’ India – to 30  For similar discussion, see ZYS 1.16a: 初相為主, 正為能立。藉伴助成, 故須第二。雖有主 伴, 其濫未除, 故須第三異品無相。“The first characteristic is the chief one [among the three characteristics, because it corresponds to the expression of the reason. The reason] is the major means of proof. [However, the reason cannot complete the argument only by means of itself. Some evidence other than the subject, showing the co-occurrence of the reason and the inferable, should be] used to accompany [with the reason]. Therefore, the second [characteristic] is necessary so far as helping [the reason] complete the argument. [Furthermore,] although the chief and the accompanying [characteristics] are [fulfilled], the overflow [of the reason from the sapakṣa realm] has not been eliminated. Therefore, the third characteristic, i.e., [the reason’s] absence from vipakṣa, is necessary.” Cf. Tang 2015: 323–336, where the same conclusion is obtained from a different angle. However, what is exactly meant by Dignāga himself with regard to the second condition, esp. whether the S2(1) version or the S2(2) version was actually meant by him, is still a difficult problem. See Oetke 1994: 55–73, Katsura 2000 and Lasic 2009. My conclusion about the second condition, therefore, concerns only what was thought by hetuvidyā scholars, and is valid only to this extent.

433

22  Yin Ming 因明 in Chinese Buddhism

1.

“[The reason] is ( )( pervasively a property of the subject.”

2.

“[The reason] is certainly (∃ )(¬ present in similar instances.”

3.

“[The reason] is ¬(∃ )((¬ pervasively absent from dissimilar instances.”

) ⋀ (∃ )( ∧ )

→

⋀

⋀

∧¬

)∧

)

)

Fig. 22.4  The three conditions and their logical form

evaluate an argument in a general way. Being the major criteria for a good argument, the three conditions are actually aimed at excluding various arguments based on one or more false premises, whether a false reason-statement or a false examplestatement or both. At the same time, both the kind of arguments rejected from the viewpoint of the three conditions and the kind of arguments fulfilling all the three conditions are of the same logical form. This could be easily seen if we compare the above good arguments A1 and A2 with the bad ones A3 and A4. Therefore, when proposing a theory of argument like the theory of three conditions, the Buddhist logicians do not base their account of a cogent argument on its logical form. They do not intend to make differences between various logical forms, distinguishing valid ones from invalid ones. They aim to theorize our basic intuition that a good argument should begin with true premises (see Tang, forthcoming).

5  Conclusion The hetuvidyā tradition, which flourished from the seventh century onwards in the East Asian World and which was called yin ming in China, can be said to be mainly a tradition of Buddhist logic following the sixth and the early seventh centuries’ interpretation of Dignāga’s logic in India. As an outcome of cultural interflows between India and China, the yin ming literature witnessed the efforts of classical monk-­scholars, at first Chinese, then Korean and Japanese, in representing Indian logic by means of classical Chinese, a language deemed to have significant differences from the Indo-European languages. They invented a new language of logic in classical Chinese. According to our knowledge today, there is no evidence for their debt to indigenous thoughts in early China, nor evidence for their impacts on the Chinese philosophy from Tang dynasty onwards. In the above pages, I have made a sketch of the history and literature of yin ming, of the basic way of argument and of the basic criteria for a good argument according to this tradition of logic. I have tried to base my judgments on the texts. The decipherment of yin ming texts demands not only knowledge about classical China but

434

M. Tang

also knowledge about classical India. Although the international scholarship on the Indo-Tibetan pramāṇa tradition of Buddhist logic is rich and fruitful, the studies in the East Asian heuvidyā tradition are, properly speaking, a new area.

References Aviv, Eyal. 2015. “A Well-Reasoned Dharma: Buddhist Logic in Republican China.” Journal of Chinese Buddhist Studies 中華佛學學報 28: 189–234. Chen, Daqi 陳大齊. 1997. Observations on the Great Commentary on Hetuvidyā 因明大疏蠡測. Reprint, Tainan: Zhizhe chubanshe. Copi, Irving M. & Cohen, Carl. 2005. Introduction to Logic (Twelfth Edition). New Jersey: Pearson Education, Inc. Franco, Eli. 2004. “Xuanzang’s Proof of Idealism (Vijñaptimātratā).” Hōrin 11: 199–212. Frauwallner, Erich. 1959. “Dignāga, sein Werk und seine Entwicklung.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 3: 83–164. Ganeri, Jonardon. 2004. “Indian Logic.” In D. M. Gabbay and J. Woods, eds., Handbook of the History of Logic, Vol. 1. North Holland: Elsevier BV, 309–395. Hayes, Richard P. 1988. Dignāga on the Interpretation of Signs. Dordrecht: Kluwer Academic Publishers. Hugon, Pascale. 2004. “Interpretations of the trairūpya in Tibet.” Hōrin 11: 95–117. IRMS = Lamp of the Commentary on Hetuvidyā 因明論疏明燈抄 of Zenju 善珠. T68, no. 2270. (A running commentary on YMDS.) Karlgren, Bernhard. 1957. Grammata Serica Recensa. Stockholm: Museum of Far Eastern Antiquities, Bulletin 29. Katsura, Shōryū 桂紹隆. 2000. “Dignāga on trairūpya Reconsidered: A Reply to Prof. Oetke.” In Culture and Logic in India. A Felicitation Volume for Prof. H.  Tosaki on His Seventieth Birthday インドの文化と論理:戸崎宏正博士古稀記念論文集 (241–266). Fukuoka: Kyūshū University Press. Katsura, Shoryu. 2004. “Pakṣa, Sapakṣa and Asapakṣa in Dignāga’s Logic.” Hōrin 11: 119–128. Kitagawa, Hidenori 北川秀則. 1985. A Study of Indian Classical Logic. Dignāga’s System インド 古典論理学の研究—陳那(Dignāga)の体系. Reprint, Kyoto: Rinsen Book. Lasic, Horst. 2009. “Dignāga’s Description of a Logical Mark in Pramāṇasamuccaya 2.5cd.” International Journal of Buddhist Thought & Culture 13: 7–22. Li, Rongxi (trans.). 1996. The Great Tang Dynasty Record of the Western Regions. Berkeley: Numata Center for Buddhist Translation and Research. Lü, Cheng 呂澂 (ed.). 1980. A New Catalogue of Chinese Tripiṭaka 新編漢文大藏經目錄. Jinan: Qilu shushe. Luo, Zhao 羅炤. 1981. “A Study of the Date of Xuanzang’s Translation of Nyāyamukha玄奘譯《 因明正理門論本》年代考”. Studies in World Religions 世界宗教研究 1981.2: 29–36. NB = Nyāyabindu of Dharmakīrti. In Paṇḍita Dalsukhbhai Malvania, ed., Paṇḍita Durveka Miśra’s Dharmottarapradīpa. Patna: Kashiprasad Jayaswal Research Institute, 1971. NMu = Nyāyamukha of Dignāga. In Katsura, Shōryū 桂紹隆. 1977–1987. “A Study of the Nyāyamukha 因明正理門論研究.” The Hiroshima University Studies, Faculty of Letters 広島 大学文学部紀要, [1] vol. 37, 1977, 106–126; [2] vol. 38, 1978, 110–130; [3] vol. 39, 1979, 63–82; [4] vol. 41, 1981, 62–82; [5] vol. 42, 1982, 82–99; [6] vol. 44, 1984, 43–74; [7] vol. 46, 1987, 46–85. NP = Nyāyapraveśa of Śaṅkarasvāmin. In Tachikawa 1971: 140–144; Chinese translation, T32, no. 1630.

22  Yin Ming 因明 in Chinese Buddhism

435

Oetke, Claus. 1994. Studies on the Doctrine of Trairūpya. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien. Oetke, Claus. 1996. “Ancient Indian Logic as a Theory of Non-Monotonic Reasoning.” Journal of Indian Philosophy 24: 447–539. Potter, Karl H. (ed.). 2003. Encyclopedia of Indian Philosophies. Volume IX: Buddhist Philosophy from 350 to 600 A.D. Delhi: Motilal Banarsidass Publishers Pvt. Ltd. PS(V) = Pramāṇasamuccaya(vṛtti) of Dignāga. Tibetan texts in Kitagawa 1985: 441–579. Shen, Jianying 沈劍英. 2008. A Study on Hetuvidyā Manuscripts in Dunhuang 敦煌因明文獻研 究. Shanghai: Shanghai guji chubanshe. T / Taishō = Taishō Shinshū Daizōkyō 大正新脩大藏經, ed. J.  Takakusu 高楠順次郎 & K. Watanabe 渡邊海旭. Tokyo: Taisho shinshu daizokyo kankokai, 1924–1935. Tachikawa, Musashi. 1971. “A Sixth-Century Manual of Indian Logic: A Translation of the Nyāyapraveśa.” Journal of Indian Philosophy 1: 111–145. Takemura, Shōhō 武邑尚邦. 2011. Hetuvidyā: Origin and Transition 因明学—起源と変遷. New edition, Kyoto: Hōzōkan. Tang, Mingjun. 2015. “A Study of Gomyō’s ‘Exposition of Hetuvidyā,’ Text, Translation and Comments (1).” In G. Paul, ed., Logic in Buddhist Scholasticism (255–350). Lumbini: Lumbini International Research Institute. Tang, Mingjun. Forthcoming. “The Concept of sādhana in Chinese Buddhist Logic.” In Proceedings of the Fifth International Dharmakīrti Conference, Heidelberg, August 26-30, 2014. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Tillemans, Tom J. F. 1990. “On Sapakṣa.” Journal of Indian Philosophy 18: 53–79. Tillemans, Tom J. F. 2004. “The Slow Death of the trairūpya in Buddhist Logic: A Propos of Sa skya Paṇḍita.” Hōrin 11: 83–93. Tucci, Giuseppe (trans.). 1930. The Nyāyamukha of Dignāga: The Oldest Buddhist Text on Logic, after Chinese and Tibetan Materials. Heidelberg: Materialien zur Kunde des Buddhismus, 15. Heft. Vidyabhusana, Satis Chandra. 1921. A History of Indian Logic: Ancient, Mediæval and Modern Schools. Calcutta: Calcutta University. Yaita, Hideomi 矢板秀臣. 2005. Three Sanskrit Texts from the Buddhist Pramāṇa-Tradition 仏教 知識論の原典研究. Narita: Naritasan Shinshoji. Yang, Tingfu 楊廷福. 2011. A Chronicle of Xuanzang 玄奘年譜. Shanghai: Shanghai guji chubanshe. YMDS = Great Commentary on Hetuvidyā 因明大疏 of Kuiji 窺基. In Zheng 2010; also T44, no. 1840, Yin ming ru zheng li lun shu 因明入正理論疏. (Reference made to both editions, separated by a semicolon [“;”]). YZMS = Commentary on Nyāyamukha 因明正理門論述記 of Shentai 神泰. Nanjing: Zhina neixue yuan, 1923. Zheng, Weihong 鄭偉宏. 1996. A General Introduction to Buddhist Logic 佛家邏輯通論. Shanghai: Fudan daxue chubanshe. Zheng, Weihong 鄭偉宏. 2007. Studies on Chinese Buddhist Hetuvidyā 漢傳佛教因明研究. Beijing: Zhonghua shuju. Zheng, Weihong 鄭偉宏. 2010. The Great Commentary on Hetuvidyā. Critical Text with Notes, Modern Translation and Investigation 因明大疏校釋、今譯、研究. Shanghai: Fudan daxue chubanshe. Zheng, Weihong. 2015. “Dignāga and Dharmakīrti: Two Summits of Indian Buddhist Logic.” In G.  Paul, ed., Logic in Buddhist Scholasticism. Lumbini: Lumbini International Research Institute, 135–167. ZYS = Commentary on Nyāyapraveśa 因明入正理論疏 (abbr. Zhuang yan shu 莊嚴疏) of Wengui 文軌. Nanjing: Zhina neixue yuan, 1934.

436

M. Tang

Mingjun Tang  is Junior Associate Researcher in the Department of Religious Studies, School of Philosophy, Fudan University, China. He studied philosophy and logic in the Department of Philosophy, East China Normal University, and Chinese Hetuvidyā (Yin Ming) under the supervision of Prof. Weihong Zheng at Fudan University, where he received his PhD in 2011. He was assistant researcher at the Institute of Philosophy, Shanghai Academy of Social Sciences from 2011 to 2015. His research interests include East Asian Buddhist logic, Indian Buddhist epistemology and logic, and classical Indian philosophy.

Chapter 23

Proper Relations of Association (Zheng 正) vs. Logical Validity of Syllogism: A Case Study of Shared Practices of Matteo Ricci, S. J and Chinese Mathematicians in Seventeenth Century Jinmei Yuan

1  Introduction Thinking through the discourse between Jesuits and Chinese scholars in seventeenth century China, one can discover some significant progress in building a mutual understanding between the West and the East at a level of logical practices. One of successful examples is the translation of Euclid’s Elements Books 1-VI, Jihe yuanben 幾何原本, done by Matteo Ricci, S. J. (1552–1610) and Chinese Confucian scholar, mathematician Xu Guangqi 徐光啟 (1562–1633) in 1607. The remaining nine books [Elements Books VII-XV] of Elements were translated by the Protestant missionary Alexander Wylie and Chinese mathematician Li Shannan 李善蘭 (1811–1882), another Confucian scholar, in 1857. Their efforts and contributions provide a good case for us to study how a mutual understanding between two very different language games1 can arise. The contemporary philosopher Richard Rorty suggests that it would be extremely hard to introduce a notion, even one as simple as the concept of “pain,” to another group of people who lack it in their language game. The concepts in a language are related, and one cannot grasp one concept without understanding a whole set of other concepts in the language game. Although it is a fact that the discourse between Chinese people and Westerners often encounters problems and misunderstandings, Matteo Ricci’s and Xu Guangqi’s effort in bridging the East and the West serves to clarify how accurately one can understand the rules in another language game, and further to understand  Language game, which I am using here, is a Wittgenstein’s term, which holds that language is like a co-ordinate system, with concepts functioning in linguistic, historical, and cultural context. 1

J. Yuan (*) Department of Philosophy, Creighton University, Omaha, NE, USA e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_23

437

438

J. Yuan

another unfamiliar way of thinking. My modest ambition in this paper is to clarify what difficulties that Fr. Ricci and XU might face when translating Aristotelian logic into Chinese concepts, which gaps between Aristotelian and Chinese logics that they need to bridge, what logical principles or rules they might have to negotiate on, and which common logical rule that they finally used for carrying out their translation project. First of all, I argue that logical validity is a concept that is lacking in Chinese logic. I shall demonstrate the gaps between the effort of finding proper relations of association (Zheng 正)in Chinese ways of thinking and the requirement of logical validity in Aristotelian syllogistic reasoning. Zheng 正 is a term regularly used by ancient Chinese scholars in the Pre-Qin Dynasty, such as Confucius and Moists, in doing logical reasoning and geometrical problems solving. The term, like some other Classic Chinese characters, functions as both verb and noun. As a noun, it refers to pointing out a right direction or an implicit standard (e.g. Mohist Canons A36 zheng nan 正南 “due South” and B 31, 70 zhengti 正體 “hold the limbs right”). (Graham 1978: 169–170) As a verb, it refers to a practice of clarification and regulation, which confirms if the standard and the conducts/objects are properly associated (e.g. Mohist Canons B 68 zhengming 正名 “get the names right”; Confucius’s Analects: 13.3. Bi ye zhengming hu 必也正名乎 “to insure that names are used properly”).2 In logical contexts, zheng examines whether proper relations can be found in different associations. But what can be defined as “proper” often time varies from case to case. A good reasoning reveals the proper relation of a particular kind of association. The goal of proving logical certainty by using valid forms in syllogism plays no role in Chinese reasoning. These gaps cause major difficulties in introducing Aristotelian logic to Chinese readers. And this kind of difficulties could also be encountered by contemporary comparative philosophers when they deal with the text translations. Thus, to explore how the co-work of Matteo Ricci and Chinese mathematicians was done in Jihe yuanben will not only show a successful case in bridging Aristotelian and Chinese logics, but also provide some interpretive approaches for contemporary comparative philosophers to consider. Secondly, I shall review selected literature from the work done by modern comparative philosophers, and introduce their discoveries on the gaps between Chinese and western thinkers who think with different sets of logic rules. A common logical pattern of Chinese people doing geometry and mathematics, which was recorded in one of the oldest Chinese mathematics texts, Jiuzhang suanshu 《九章算術》, will serve as an example to show that how ancient Chinese mathematicians solve geometrical and mathematical problems without any steps of proofs. Comparing the logic pattern in Jiuzhang suanshu to the theory of Syllogism, I argue that they are two kinds of different logical practices. This literature review provides background knowledge for me to further clarify the gaps that Fr. Ricci and Xu must bridge in their translation project. 2  Translation is from Roger Ames and Henry Rosemont (Ames and Rosemont 1999: 162). They also translate the zheng in Analects as “proper,” “proper relations,” or “proper in personal conducts.”

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

439

Thirdly, I shall do four investigations for revealing a list of serious mismatches between Chinese and Aristotelian logics. The first investigation is on the gap between the concept of proper relations of association (zheng 正) in Chinese logic and the concept of logical validity in Aristotelian deductive logic. The former deals with the everlasting changes in an empirical world, which is taken as the real one, while the latter is based on the First Principle and can only be discussed in a deductive system or a purely logical world. My second and third investigations focus on the lack of standard-forms of syllogism3 and the First Principle in Chinese logic, which could be the major cause of other mismatches between Chinese and Aristotelian logics. Ancient Chinse logicians and mathematicians had no interest in starting an argument from the First Principles, which are self-evident propositions or assumptions according to an Aristotelian tradition. The logical relations presented by Confucius in his famous zhengming argument in Analects will be analyzed and compared to the rules of Syllogism. The fourth investigation studies why there is no distribution requirement in Chinese logic. This gap is caused by the lack of universal affirmative and negative propositions in ancient Chinese logic. Classical Chinese has neither the subject-predicate structure (being structure) nor clear distinction between singular and plural nouns and verbs. Copula “being” is not required grammatically. Categorical syllogisms are arguments based on the relations of classes or the membership in categories, with the subject-predicate structure, distributions are required between terms in any type of propositions. This fourth gap explains why logical validity fails to work in Chinese logic. These are the major gaps which Matteo Ricci and XU Guangxi faced when introducing Aristotelian logic to Chinese people. Fourthly, along with my four investigations on the mismatches between Chinese and Aristotelian logics, I have two major discoveries. The first discovery is how Matteo Ricci and XU Guangqi solved the problem of lacking the First Principle in Chinese reasoning. Matteo Ricci, as a traditional Aristotelian thinker, tried hard to adopt Confucian ways of thinking for introducing Aristotelian deduction to Chinese people. Having mastered Chinese language and with a clear understanding of Chinese ways of thinking, Fr. Ricci gives up the requirement of the First Principle in doing syllogistic reasoning. He allows the roles of Aristotelian Frist Principle to be replaced with the wisdom of ancient Chinese Sage-Kings and ancestors. In doing so, Fr. Ricci becomes one of the earliest Western scholars to try to interpret Western logic and mathematics in Chinese concepts. At the same time, Confucian scholars, Xu Guangqi and other Chinese mathematicians also tried hard to demonstrate how deductive rules work, for example, in geometry. They confirmed that Aristotle’s First Principle was hold by Chinese Sage-Kings and ancestors, but Chinese people forgot them. With this replacement, Chinese learners’ 3  A standard-form syllogism must have three terms, major term, minor term and middle term. Major premise contains major term and must be the first premise; minor premise contains minor term and is the second premise. The conclusion follows. The predicate of the conclusion is the major term, the subject of the conclusion is the minor term. Middle term must be repeated in both premises and must be distributed at least once.

440

J. Yuan

worry of “when names were not used properly, language will not be used effectively” (Ames and Rosemont 1999: 162) was removed. My second discovery is that, with their compromise on the universal principles or axioms in Elements, Fr. Ricci and XU found at least one common logic rule, which they both clearly understood and practiced in translating Elements to Jihe yuanben. For Fr. Ricci, it is one of inductive rules when giving definitions, ostensive definition. For Chinese mathematicians, it is a popular rule in doing reasoning since ancient time, Pointing Out, zhi 指. One points out an exemplary case to illuminate the concept that one wants to clarify. As long as proper relations between the exemplary case and concept is found, the argument is a good one. This common practice finally helps Fr. Ricci and Chinese mathematicians bridge the gaps and open a discourse at a logical level. Practicing the rule of ostensive definition or Pointing Out, Fr. Ricci and Chinese mathematicians learned from one another and made it possible for Chinese learners to understand syllogistic patterns through particular cases. As a result, valid syllogistic forms were introduced to Chinese learners. The conclusion of this paper is that Chinese logic and Aristotelian logic are very different. The effort of seeking for proper relations of association (zheng 正) is not comparable to seeking logical validity of syllogism. However, attempts to understand the unfamiliar and the novel are common to all human beings. Whether a mutual understanding is possible, the effort must be made from both sides. Xu Guangqi’s strong curiosity in accurate proofs echoes Fr. Ricci’s effort. With some compromising, Aristotelian deductive logic is finally introduced to Chinese people, not via a text of logic but via Elements, a geometrical text. The shared practices of Matteo Ricci and Chinese mathematicians in the seventeenth century made it possible to introduce Aristotelian logic to a Chinese audience through geometry. Their work could be a revelation to today’s researchers. The discourse between Jesuits and Chinese mathematicians in the seventeenth century offers an excellent example of having an open mind and a sense of wonder. This is the hope for human knowledge and mutual understanding.

2  A  Few Major Difficulties When Introducing Aristotelian Logic to Chinese Language Speakers To understand how deduction works, one must understand the concept of “logical validity.” Traditional Aristotelian deduction employs categorical propositions, which must be either true or false, to construct arguments. A valid argument is a deductive argument whose premises, if they were all true, would provide conclusive grounds for the truth of its conclusion. Logical validity guarantees that truth value can be passed from premises to the conclusion. A categorical proposition is either universal or particular, and either affirmative or negative. The distribution of terms in a proposition must be very clear. One of the major tasks in deductive logic is to separate valid arguments from invalid ones. For example, to prove a geometry

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

441

p­ roblem, the deductive steps must be valid. How to develop valid steps is guided by a set of syllogistic forms. Syllogistic rules can work in the deduction because they are universal forms, which start with the primary First Principle. The Greek term logos associates with a principle of order and knowledge. The order in the world is presupposed. It is necessary to have a pre-existing order, such as genus and species, for building up steps to reach a valid argument in Aristotelian deductive logic. When Aristotle introduces the primary First Principles, he confirms his acknowledgement of the presumption of the pre-existing order and claims that the First Principles are “the first basis from which a thing is known” (Met. 1013a14–15).” He says, “In every systematic inquiry (methodos) where there are first principles…for the things known to us are not the same as the things known unconditionally (haplôs)” (Phys. 184a10–21). Aristotle’s separation of deduction from induction shows that the logical validity of syllogism can only happen in the deductive logical world. It has nothing to do with induction. There are fifteen valid forms of categorical syllogism, which guarantee that reasoning from universal premises to a particular or univeral conclusion is certainly valid. Syllogistic forms are tools of thinking that represent the order in a logical world, where the truth values of every premises will not change in the subjective time and during any empirical process. Euclidean geometry holds that the truth values of a few axioms are self-evident, and theorems follow logically from its axioms are as truth as the axioms. Therefore, the proofs of geometrical problems, from giving definitions to developing deductive steps and then reaching a conclusion, could be the best examples which demonstrate the logical validity in a deductive system. In contrast, ancient Chinese thinkers are interested in discovering proper relations among names, objects, and/or events in an empirical world which is changing. Finding out the proper relations of association in one’s changing life and world turns to be a way of clarification in doing reasoning. The proper relations of association are neither pre-determined by any unchanged Platonic Form, nor presupposed by Aristotelian First Principle or Euclidean axioms. Thus, ancient Chinese thinkers’ discoveries can hardly be done in a pure abstractive or deductive way. Instead, their logical reasoning often times could happen in a creative process. Their investigation of logic and mathematics tends to be grounded in their empirical world and guided by their practical concerns, rather than to be based axioms and guided by the valid forms in deduction.4 Evidence can be found in Moists’ work, Xiao Qu 小取, which is considered as one of the ancient Chinese logic texts. The pattern of some arguments in Xiao Qu might look like Aristotelian syllogisms, but they are not. For example, Mozi 墨 子 says, 5 4  I argue that Chinese logic is not a deductive logic. However, I also hesitate to lab Chinese Logic as inductive logic. It is because that Aristotelian deductive reasoning aims at proving the certainty of thoughts, while Aristotelian inductive reasoning aims at collecting evidences to offer the high probability of a pre-existing order. Chinese logic does not associate with the order but changes. 5  The following translations from Xiao Qu are based A. C Graham 1978.

442

J. Yuan

獲, 人也; 愛獲, 愛人也。(小取)

Huo is a person. Loving Hou is loving a person (Xiao Qu). 臧, 人也; 愛臧, 愛人也。(小取)

Zang is a person. Loving Zang is loving a person. (Xiao Qu) But, right after making these statements, Mozi says again: 盜人, 人也, 殺盜非殺人也。(小取)

The robber is a person. Killing the robber is not killing a person. (Xiao Qu) From an Aristotelian perspective, the argument, “The robber is a person. Killing the robber is not killing a person,” is obviously invalid. It is a self-contradiction and no deductive rule supports the steps to the conclusion. But the above statements all are acceptable to Moists, according to a logic which allows for the involvement of subjective time, as well as permits changing perspectives when judging situations. If this is the case, then one has to say that syllogistic forms of validity have no authority over Chinese logician’s thinking. In a changing process, self-contradiction can disappear in time. The proper relation of association between “a robber” and “a person” cannot be confirmed. A robber is associated with a crime, which a person should never do, so that “[K]illing the robber is not killing a person.” Similarly, Moist logic teaches Chinese people, “If this horse’s eyes are blind we say that this horse is blind; when this horse’s eyes are big we do not say that this horse is big. 之馬這目眇, 則謂之馬眇。之馬之目大, 而不謂之馬大。” (Graham 1978:142) In the first case about “being blind,” the proper relation between “being blind” and “horse” is found because “blindness” can associate with both “eyes” and “horse.” In the second case about “being big,” the proper relation between “big” and “horse” is not found, because “bigness” here associates with “eyes” in a different scale of manner when it associates with “horse.” As long as proper relations of association can be found, different kinds 類 lei can be determined case by case. Moist logic emphasizes “accept according to the kind, present according to the kind. 以類 取, 以類予。” (Graham 1978:151)6 Thus, there are no concepts and logical relations to match Aristotelian “logical validity” to a traditional Chinese way of thinking. Chinese logic is based on a presumption that everything is changing. Seeking for proper relations of association amongst many relations for the time being is what a good reasoning aims for. A proper relation could be examined from different perspectives for practical purposes and involves subjective judgements. While Aristotelian thinkers clearly separate the deductive logical world from the empirical one, Chinese thinkers show no such attempt. Logical validity was a completely alien concept to Chinese thinkers before Matteo Ricci, S. J and Xu Guangqi published Jihe yuanben. As pioneering thinkers who put much effort in finding a right perspective to bridge Aristotelian logic and Chinese logic, Fr. Ricci and Chinese mathematicians in the seventeenth century had to find a way to reach a mutual understanding on  Further discussions on kinds, lie 類, see Jinmei Yuan 2005: 181–199.

6

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

443

both logical validity and proper relations of association. Without understanding logical validity, there is no way to understand Aristotelian deductive logic. Similarly, without understanding the tricks of looking for proper relations in a changing world, there is no way to understand Chinese logic. Chinese logic and Aristotelian logic are two different ways to represent the structure of logical relations.

3  L  iterature Review on Two Different Ways to Structure Logical Relations The goal of opening a discourse between Chinese and western philosophies and religions in order to build a mutual understanding between Chinese and westerners has been pursued by many philosophers and theologians. However, misunderstandings constantly happen not only at the natural language level, but also in the ways of thinking. The frontline research in Comparative philosophy nowadays pertains to how to choose an effective methodology and inquire into the roots of both Chinese and western ways of thinking from the correct perspective. Looking back at some shared practices of Matteo Ricci, S. J and Chinese mathematicians in the seventeenth century when translating Euclid’s Elements into Chinese, philosophers today might gain some important messages on how two very different language games communicate. The aim of a translation is to pass on as much original information as possible, but no translation can be exact in conveying the meaning of the original. The boundaries between concepts are shaped by underlying logical relations. Different languages might structure logical relations in different ways. These are gaps that I want to look at. Some gaps are minor, some are crucial. The gap between proper relations of association and logical validity is a crucial one that one cannot overlook. When discussing the gaps between different language games, American contemporary philosopher Rorty suggests that what we are able to say is “a topic that provides the greatest temptation to think that explanations of how language works will also help us see how ‘language hooks onto the world’ and thus how truth and knowledge are possible” (Rorty 1980:265). Reviewing the research on the gaps between Aristotelian and Chinese logics done by previous comparative philosophers, one will learn that their reports, to some degree, sketch out what the logical relations of the two cultures might look like. This review provides background knowledge for me to further clarify the gap between the concepts of proper relations of association and logical validity. Jacques Gernet also points out that the difficulties in reaching a common understanding between Chinese and Western people arise not only because of “different intellectual traditions but also of different mental categories and modes of thought” (Gernet 1985:3). He says: All that we, from our Western point of view, regard as anomalous stems, on the contrary, form a faultless logic. In Chinese concepts, everything depends upon everything else, all

444

J. Yuan

things are connected, and this coherence makes the historical character of mental schemata that have prevailed in the West stand out more clearly in contrast. (Gernet 1985:4)

That “everything depends upon everything else” and that “all things are connected” might be evidence of complicated overlapping relations, but this would not be sufficient to constitute a logical system from an Aristotelian point of view. However, from a point of view of Chinese logic, this picture reflects the way things are. In his paper “Kosmos in the Acosmotic Han Dynasty,” Ames makes strong points regarding the deep structural differences that can hinder accurate interpretation of logical relations The most influential strains of classical Chinese thought illustrate the primary characteristics of first problematic thinking—namely, the absence of belief in a single-ordered world and the employment of aesthetic over logical senses of order. Correlative thinking is an implication of these assumptions. (Ames and Hall 1998:12)

“[T]he absence of belief in a single-ordered world and the employment of aesthetic over logical senses of order” is a presupposition for Chinese thinkers to explain the world. Chinese ways of thinking represent how everything connects. Logical relations, for them, are neither objective nor pre-existing a priori, but changing in the process and can be found from different perspectives subjectively. The following comparison of a Chinese classical mathematic text to Aristotelian syllogism can demonstrate what the different logical structures might look like: The most logical thinking is mathematical thinking. Chinese classical mathematical texts, such as Jiuzhang suanshu 《九章算术》, provide good mathematical and geometrical problems and the resolutions to the problems. But the steps of proofs were skipped. The pattern in Jiuzhang suanshu often is: 1 ) Giving a problem to solve (qiu 求), 2) Providing an answer (da 答). Sometimes, a discussion of tricks (shu 术) follows, the other times has no discussion. For example: 1) A giving problem is that there is a piece of land. Its width is one and a half bu. Asking how large one mu will be. Jiyou tian guang yibu ban, qiu tian yimu, wen cong jihe? 今有田广一步半, 求田一亩, 问从几何? 2) The answer is: one hundred sixty bu. Da yue: yibai liushi bu. 答曰:一百六十步。 Neither deductive proofs are given, nor the processes of thinking are demonstrated in solving this geometric problem. It seemed that when a proper relation between the question and the answer was found, Chinese mathematicians’ job was done. Here, the mathematical problems are solved without the First Principle as base. Ancient Chinese mathematicians do not need the concepts such as “universal forms” and “general principles” to discuss mathematics. They are good at solving

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

445

practical problems one after the other, but had no interest in generalizing principles, formulas and theorems. Turning to discuss the structure of deductive relations, it is crucial to meet the requirements of valid forms. Aristotelian logic presupposes the existence of an order in the universe, which can be demonstrated through reason. If one allows the truth value of propositions change in the subjective time while doing a deductive proof, the certainty that deductive logic guarantees will be gone. In his paper “The Language of Being: Between Aristotle and Chinese Philosophy,” Comparative philosopher Jiyuan Yu indicates that “[I]n Topics 1.9, Aristotle establishes his doctrine of categories (ketegoria) directly from the predication form ‘S is P’” (Yu 1999:441). According to Aristotelian logic, every syllogism has its mood and figurer. Mood is a characterization of categorical syllogisms, which is determined by the standard-­ form categorical propositions it contains. A, E, I, O are only four forms of propositions. A is universal affirmative, “all S is P.” E is universal negative, “no S is P.” I is particular affirmative, “some S is P.” O is particular negative, “some S is not P.” Figure is the position of the middle term. The Mood and Figure determine the validity of a syllogism. If one allows the mood (A, E, I, O)7 and figure of a set of propositions change during a deductive proof, then the logical validity disappears immediately. The nature of Aristotelian logic is a logic of terms with strict classification and rules. The key logical relations with which it is concerned are the opposition between (contrary) terms and the hierarchy of classification—from genus down to species, sub-species and so on. As contemporary philosopher Mary Tiles describes: The kind of rational structure which is given prominence in Aristotle’s works and which came to dominate Scholastic thought is the structure of a classificatory system—a hierarchical structure of names for kinds of things organized into genera and species. To identify an object—give it its correct name—is to locate it in a classificatory system. This location is determined by the definition of the term (a definition which gives an account of what it is to be a thing of that kind, an account of its essence). (Deutsch 1997:212)

The function of logic is to get rid of ambiguity of thoughts/language, so that the system of knowledge can be validly demonstrated. In a logical world, deductive logic as a neat and beautiful tool of thinking, deals with the validity of pure logical forms. Aristotelian logicians believed that these forms are hidden behind different native languages. They can be rules. Thinking according to deductive rules guarantees the conclusion is a logical consequence of its premises, no matter what native language is used. In the logical world, these forms are assumed to be objective. If thoughts move according to these forms, the argument is valid, otherwise, it is invalid. Validity proves that no mistake happens in the process of referring. While Jiuzhang suanshu can skip proofs, Aristotelian syllogism requires them. In the Aristotelian tradition, sciences are divided into different branches or fields. Each of those fields, such as physics, mathematics, chemistry, biology, etc. has its

 There are only four forms of proposition, so that there are exactly 64 moods in syllogism.

7

446

J. Yuan

own primary first principle, which is a self-evident proposition or assumption and is beyond doubt. In Euclid’s Elements, hundreds of propositions can be deduced from a few axioms. A set of given propositions are building blocks of an argument. They must be either true or false. During a deductive process, as long as logical forms are obeyed, one can construct an argument without ambiguity. Logical certainty is what Aristotelian deduction aims at. Although there are changes, it is not the job of deductive logic to test whether a proposition is true or false in the empirical world. Aristotelian logic leaves it with induction. Induction allows logical truth to be examined by data or experiences. But, inductive logic still tries to get rid of ambiguity. Inductive rules help people to reach high probabilities, which I will not be able to fully discuss in this paper. Both certainty and probability aim at stability or order.

4  T  he First Investigation: The Gaps between Proper Relations of Association and Logical Validity This section is my first investigation into the gaps between proper relations of association (Zheng 正) and logical validity The Chinese ways of thinking were considered as very alien to the Jesuits during Matteo Ricci’s time. Du Halde (1674–1743) summarized the shock of Jesuit missionaries when they saw the work of Chinese mathematicians in 1735: “While they do manage to resolve certain problems, this is by induction rather than by any guiding principle” (Martzloff 1997:3). The concept of logical validity does not exist in the Chinese language game due to the absence of a presupposition of a universal order. No hierarchical system of classes exists to structure memberships into genus and species, Chinese logicians and mathematicians show no interest in seeking for the First Principle that should be beyond doubt. To differentiate a good argument from a bad one, Chinese scholars normally employ the concept of proper relations of association. Proper relations of association (zheng 正) cannot be taken as a pure logical term, although it is used to justify arguments. The practice of finding proper relations of association can be also found in ethics, politics, aesthetics, mathematics, etc. Zheng 正, according to Xu Shen’s Shuo Wen Jie Zi 《說文解字》, an early Chinese etymology text, was written as

. Xu Shen says:

Zheng is shi. It associates with stopping arguments. When the unity is reached, we stop (Zheng, shiye. Chong zhi, yi yi zhi. 正, 是也。从止, 一以止。). Shi means “straighten.” The Sun is right at the top, which is the position of zheng (Shi zhi ye. Chong ri, zheng. 是直也。从日、正。). (Xu Shen 121: Book II B, re-print 1996)

Shuo Wen also records another form of zheng,

. It explains:

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

447

Zheng comes from “one” and “feet.” The feet also associate with stopping (Zheng cong yi, zhu. Zhu zhe yi zhiye. 正从一、足。足者亦止也。). (Xu Shen 121: Book II B, re-print 1996)

Chinese characters, according to philologists, are divided into six categories (liushu 六書). The first category is pictographs or imitative drafts. According to L. Wieger, they are “rough sketches representing the object; 畫成其物, 隨體詰詘.” (Wieger 1965:10). The ancient Shuo Wen dictionary contains 364 of these. These basic pictographic characters are building blocks or indicators among the 5000–7000 most useful Chinese characters. The second category is indicative symbols. They are generally called “ideographs.” According to Henry Rosemont, they supposedly represent “more abstract ideas incapable of being pictured directly but they do have much illustrative content” (Rosemont 2015).8 Thinking with a pictographic language, Chinese reasoning does not require a pure logical world, in which right forms guarantee the validity, so that the hidden order can be represented. When moving, stopping, unifying, man’s feet are involved, in other words, subjective time and perspectives are involved. From the connotation of zheng 正, one can interpret: a) the concept relates to subjective experiences; b) as a settlement, when the proper relations are found, no further reasoning is needed; c) the proper relations of association distinguish good thinking processes from the bad ones; d) pointing out the proper relations of association is the best hope for Chinese thinkers when dealing with a changing world. To illuminate how zheng, as an approach, works, I will employ a famous discussion on zheng in Confucius’ Analects. Instead of giving a definition, Confucius gives examples of the practice of zheng for strengthening relationships to the names. When Zilu, a disciple of Confucius, asked, “What would be your first priority?” Confucius replied, “Without question it would be to insure that names are used properly (Bi ye zhengming hu 必也正名乎).” When Zilu asked further, “What is it for names to be used properly anyway (xi qi zheng, 奚其正)?” (Confucius: 13. 3; Ames and Rosemont, Jr. 1999:162), Confucius gave his famous teaching on zheng ming as follows: When name is not used properly, language will not be used effectively; language is not used effectively, matters will not be taken care of; when matters are not to be taken care of, the observance of ritual propriety (li 礼) and playing of music will not flourish; when the observance of ritual propriety (li 礼) and playing of music do not flourish, the application of laws and punishment will not be on mark; when the application of laws and punishment is not on mark, the people will not know what to do with themselves. Thus, when the exemplary person puts a name to something, it can certainly be spoken, and when spoken it can certainly be acted upon. There is nothing careless in the attitude of an exemplary person toward what is said.” (Confucius: 13.3; Ames and Rosemont, Jr. 1999: 162)

The above dialogue is considered as good reasoning. Confucius employs the concept of zheng to demonstrate its function in the case of clarifying ming 名, names, and the necessity of using it properly. Ames and Rosemont translate zheng 正 in Analects as “proper.” It is a translation that allows subjective judgment to determine  See also Yuan 2019.

8

448

J. Yuan

proper relations. When the things/events and names associate properly, a convictable logical relation is found in a changing world. If logical relations can be properly discovered through a subjective judgment, then they are associative. Confucius emphasizes neither deductive rules, nor assumes that objective forms9 (i.e. syllogism forms) exist. What are the criteria for a relationship to be identified as a “proper” one? It seems that for Confucius and other Chinese thinkers, criteria are found in the wisdom and good deeds of their ancestors. Asking why it is the case that if “language is not used effectively, matters will not be taken care of” is out of the background context. In his book, Aristotle in China, Robert Wardy comments on Chinese empiricist criteria as follows: What is true and important is that the exclusively normative trust of 正名 [zhengming] is not meditated by semantic theory…but rather because the original meaning of the true word derives from the wisdom of the ancient Sage-Kings, a wisdom best characterized in sociopolitical rather than linguistic terms.” (Wardy 2000:58)

Trusting the ancient Sage-Kings’ wisdom and deed is not the undoubtable First Principle to develop an argument, but it is a passionate choice based on the best resources of human experiences and past data in the system of knowledge. Trust does not require re-examination. If wisdom from the ancient Sage-Kings reveals that if “language is not used effectively, matters will not be taken care of,” then one should use language effectively, which depends on how to use names properly. Therefore, a legitimate logical relation among “taking care of matter,” “using language effectively,” and “using names properly” is found. In a changing world, one can only do this much. Confucius and Chinese scholars do not expect a perfect sounding process. If following the Sage-Kings’ wisdom does not work in one case at this time, one attributes it to bad luck. However, he/she still must trust the Sage-­ Kings’ wisdom. I want to bring in an Aristotelian term, logical truth, as a comparison. According to Aristotelian categorical logic, truth values must be assigned to each proposition when structuring a deductive argument. People can surely ask why a proposition is true or false. But testing whether the proposition is true or not is not a job that belongs to deductive logic theoretically. Truth and/or falsity is merely a logical value for Aristotelian logicians. Others who work in the real empirical world are in charge of testing the truth value of these propositions. One can doubt these test results but not syllogistic forms. If one hypothesis does not work, the proposition is false. But the valid syllogistic forms remain unchanging. Within a logical world, logical patterns guarantee that the truth value of every proposition can be validly passed on to the conclusion, so that a set of propositions can construct an argument without leaving ambiguities. An argument can be valid if all the premises and the conclusion are glued in syllogistic forms; an argument can also be sound, which requires that all the premises and conclusion are true proposi-

 In the Platonic tradition, Forms are objective and a priori.

9

449

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

tions, and that the argument is valid as well. The following table shows the unmatched places that I discussed above: Relations of association (proper or improper) Yes. The proper relations between two statements are important Logical validity Not related

Logical relations in an argument Proper relations of association (zheng 正)

Logical soundness

Not related

Proposition (true or false) Not important.

Criteria for a good argument Wisdom and deeds of ancient Sage-Kings and ancestors

Syllogistic Validity forms Not None related

Either True or The first principle Yes False and syllogistic forms True The first principle Yes and syllogistic forms

Must follow Must follow

The effort of removing ambiguity from thoughts is not the major concern for Chinese thinkers. The major concern is to find proper relations case by case. Therefore, the building blocks for an argument in a Chinese language game can hardly be defined as propositions without empirical context to back it up. One of the most significant gaps shown in the table is the lack of syllogistic forms in Chinese logic, while these forms must be followed for validity in Aristotelian deduction. I believe that it is one of the major gaps that Fr. Ricci and Chinese mathematicians faced in translating Elements into the Chinese Jihe yuanben, because the logical proofs in the cases of geometry in Elements must follow the valid syllogistic forms precisely.

5  T  he Second Investigation: The Gaps Caused by the Lack of Syllogistic Forms in Chinese Logic My second investigation reveals that no syllogistic form required in Chinese thinker’s reasoning. Taking Confucius’s quote (Analects 13.3) from section four as an example again, I shall further investigate the gaps between proper relations of ­association and logical validity due to the lack of syllogistic forms in Chinese logic. Although the thought flows smoothly in Confucius’ quote from section four, and proper relations of association are found in the conclusion; although the conditional statements are followed one after another, like falling in a hypocritical syllogism, Confucius’s argument has nothing to do with syllogism. A further study on Syllogism will be helpful to clarify the gaps caused by different rules in Chinese and Aristotelian logics.

450

J. Yuan

Let me give symbolic definitions to Confucius statements in section four, and then reexamine them. If the following meanings are assigned to N, L, M, R, A and P: N: name is used properly, L: language will be used effectively; M: matters will be taken care of; R: the observance of ritual propriety (li 礼) and playing of music will not flourish; A: the application of laws and punishment will be on mark; P: the people will not know what to do with themselves. If the statement above could be written in Aristotelian propositional form, then it would be: ~N ⊃ ~L ~L ⊃ ~M ~M ⊃ ~R ~R ⊃ ~A ~A ⊃ ~P _____________ N

But, unfortunately, such an argument in the Chinese way of thinking is regarded as invalid and contains syllogistic fallacies from the perspective of Aristotelian logic. A short truth table, an effective tool to test validity, could prove its invalidity: A T

L T

M T

N F

R T

P T

~N ⊃ ~L T

~L ⊃ ~M T

~M ⊃ ~R T

~R ⊃ ~A T

~A ⊃ ~P T

N F

There is at least one possibility to make the conclusion of the argument false but all the premises True, when the above truth values are assigned to A, L, M, N, R, and P

a

Being a syllogistic form, there are six rules which I will list out soon. Disobeying any one of them, an argument will be invalid. Taking the first three conditional statements above as three categorical propositions and examining in a syllogistic pattern, one could have the following argument: ~N ⊃ ~L ~L ⊃ ~M ____________ N

(Read as: If naming is not used properly, then language will not be used effectively. If language is not used effectively, then matters will not be taken care of. Therefore, naming should be used properly).

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

451

The thinking pattern found in Confucius’s argument from section four is obviously allowed for when doing reasoning in the Chinese way of thinking. Confucius used it and so did other ancient Chinese scholars. ~N and N can co-exist or can be both right if a proper connection is found in time. But when applying syllogistic rules, more gaps show immediately: Hypothetical syllogism as one of the rules of reference in propositional logic, which allows either a single or a compound statement to be a proposition, is not governed by the rules of categorical syllogism in term of logic, which deals with distributive relationship among terms involved with standard syllogistic forms. Distribution is “a characterization of whether a categorical proposition refers to all members of the class designated by the term” (Copi and Coben 2005:184), which I shall examine in the section seven. The propositional logic allows the statements above to work as a particular proposition, when doing deduction. Modern logicians add Universal or Existential Quantifiers to deal with membership between universal and particular propositions. “N,” “L” and “M” are propositions, not terms. So, there is no problem of “distribution.”

Six rules of syllogism Rule 1. Avoid fourth term Rule 2. Distribute middle term in at least of one premise

Rule 3. Any term distribute in the conclusion must be distributed in the premises

Satisfy the rules √ X

X

X Rule 4. Avoiding two negative premises X Rule 5. If either premise is negative, conclusion must be negative.

The gaps between Zheng and valid syllogistic forms shown in Confucius’ argument Propositions, N, L and M are three particular propositions, not terms. No fourth term involved. a) Due to no interest in distinguishing “universal” and “particular” terms in Chinese logic, the middle terms are not distributed. b) Viewing it from an Aristotelian perspective, it asserts a connection between “When name is not used properly” and “matters will not be taken care of,” by relying upon the links between each of these “classes” (in Aristotelian sense) and the class of “language will not be used effectively”—but “not using language effectively” is an undistributed term in both of the premises. a) Distributions in both conclusion and premises are unclearly stated. Confucius does not study the distributive relations between “universal” and “particular” but “proper relations of association.” b) Membership in a class is signed to the hierarchical system of genus and species. There is no such system to classify name, language and matter, although they can be associatively connected In Confucius’ reasoing, no confirmative premise. Both premises are negative. Conclusion is affirmative.

452

Six rules of syllogism Rule 6. From two universal premises no particular conclusion may be drawn

J. Yuan Satisfy the rules ?

The gaps between Zheng and valid syllogistic forms shown in Confucius’ argument Confucius does not have universal affirmative and universal negative propositions in his argument. It is unclear how the conclusion is drawn, unless there is clear understanding of seeking proper relations of association in his minds.

The above table shows the gaps between the concepts of proper relations of association and logical validity not only at the conceptual level but also at the level of forms. Logical validity can only be achieved through syllogistic forms, while proper relations of association can be achieved by finding at least one effective connection between conducts/events and names. I sum up the gaps between proper relations of association (zheng 正) and logical validity as the following: A) Proper relations of association does not require a truth value to be signed to every proposition, while Aristotelian deduction employs propositions with truth values as building blocks to structure arguments. Propositions must be either true of false, which leaves no ambiguity while thoughts flow. B) Proper relations of association lead to neither a valid argument, nor a sound argument. Seeking proper relations of association in a changing world, subjective experiences are allowed. To translate Elements, both Fr. Ricci and Chinese mathematicians had to find ways to seal these gaps. Otherwise, this translation work would not be possible. And the effort must be made from both sides.

6  T  he Third Investigation and My First Discovery: Lacking the First Principle and Fr. Ricci’s Efforts and His Compromises to Proper Relations of Association My third investigation studies the lack of the First Principle in Chinese ways of thinking and Fr. Ricci’s compromises on the requirement of the Frist Principle. Turning to a discussion on the translation of Elements, Peter Engelfriet has carefully done a significant study and research on Fr. Ricci’s and Xu’s Jihe yuanben. In his book, Euclid in China: The Genesis of the First Chinese Translation of Euclid’s Elements Books I-VI (Jihe yuanben; Beijing, 1607) and its Reception up to 1723, Peter Engelfriet says: Aristotle set out the criteria a science must meet in order to count as a demonstrative science yielding absolutely certain knowledge. Geometry was offered as a paradigm, but the theory was certainly not restricted to mathematics. These criteria were partly of a logical nature: they required that demonstrations be cast in syllogistic form, partly of a metaphysical nature, pertaining to ontological status of the objects considered. (Engelfriet 1998:38)

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

453

Geometry requires that demonstrations be cast in syllogistic form. The steps of each proof in Geometry demonstrate the forms in Aristotelian logic. Logical validity indicated by syllogistic forms represents the beauty of thoughts. Fr. Ricci and Xu Guanqi worked on Clavius’s edition of the Elements. Clavius believed that mathematical proofs in essence are syllogistic (Engelfriet 1998:44). Introducing Aristotelian logic via translating a geometry text is legitimate in demonstrating the syllogistic forms and validity. But the gaps that I briefly listed in the previous sections are barriers for the Chinese to understand Aristotelian logic. One fact that deserves mentioning is that Aristotle’s Categories, a western logic classic text, was translated by Jesuit Francisco Furtado 傅汛济 (1587–1653) and Li Zhizhao 李之藻(1565–1563) and published in 1631. The title of the translated book is Ming Li Tan 《名理探》. Chinese readers struggled to understand Aristotelian logic in this text, especially deduction. Unfortunately, due to lacking understanding of categorical presuppositions and no experience of separating the logical world from the empirical world in Chinese culture, Ming Li Tan was not well received in China. No historical data shows that any traditional Chinese logicians pay attention to Ming Li Tan at the time. The second fact that deserves a mentioning is that XU Guanqi and LI Shanlan are famous Confucian scholars in Ming Dynasty, mingru, 明儒. They respect ancient SageKings and ancestors, and sincerely follow Confucian tradition of zhengming. When Fr. Ricci decided to wear a Confucian scholar’s gown in 1593, as suggested by Confucian scholar Qu Taisu 瞿太素10, he committed to understanding Confucius’s way of thinking. Working with Confucian mathematician Xu Guangqi to translate Elements I-IV, according to Fr. Ricci, is the first test to see whether their project can be acceptable to Chinese people.11 There is no doubt that Fr. Ricci puts a lot of effort into introducing the Aristotelian way of thinking, which tightly relates to his Jesuit training and mission. As Wardy points out: The Ratio Studiorum of the Jesuit order (1585) obliges them to promulgate Aristotelian philosophy in general and his logic in particular, and the obligation is complete in their Constitution: “the teaching of Aristotle is to be followed in logic, natural and moral philosophy and metaphysics.” Thus if the Jesuits were to teach the Chinese anything, at least in these areas, they were from the outset officially committed to starting with Aristotle (Wardy 2000:77).

Translating Elements to a language game which has neither concept of Form nor understanding of syllogism is not an easy task. Fr. Ricci must know well the gaps between demonstrating logical validity by practicing syllogism and looking for proper relations of association in Chinese ways of thinking. He says: Knowledge can only be called thus if there is no more room for doubt, while, by their very nature, subjective guesses go together with doubt. Even when ideas which are not or only partially based on reason contain elements of truth, doubts remain, and it is always possible to refute them with other principles. They are able to lead people into believing they are right, but they are not capable of bringing people to the point where they let go of all doubts. 10 11

 Qu Taisu actually translated book I of Elements with Fr. Ricci and stopped.  Fr. Ricci indicated this testing project in the preface by Ricci in Jihe Yuanben.

454

J. Yuan

Only fully rational and crystal-clear principles relieve the heart from doubt and convince people that it is impossible to come to false conclusions, that that there are no other undermining principles. (Engelfriet 1998:77)12

What Fr. Ricci states, to “lead people into believing they are right” and also to “let go of all doubts,” is his description of the nature of a valid and/or a sound argument. The arguments, which “are able to lead people into believing they are right, but not that they are not capable of bringing people to the point where they let go of all doubts,” have true premises but incorrect deductive forms, so these arguments are invalid. The problem of an invalid argument is that it is unable to relieve all doubts in the thinking process, even assuming that one does not have to doubt the truth or falsity of each premise (the above Confucian argument in the section four falls in this category).13 It could be the case that every statement on name, language, matter, ritual propriety…are true in one way or the other, but lacking logical validity could still make an argument invalid or an argument that cannot “let go of all doubts.” Fr. Ricci said that from the very first moment he set his feet on Chinese soil, he noticed that many Chinese people, “even the most eminent scholars were not capable of explaining how they had reached their conclusion” (Engelfriet 1998:459). For bridging the gaps, Fr. Ricci and his co-translator Xu Guangqi (for Book I– VI) took the following constructive steps, which set up a positive example of how to open a discourse between two different language games. First of all, both Fr. Ricci and Xu Guangqi took a step to compromise two different presuppositions about the world. A pre-existing order is presupposed in Aristotelian tradition, while no pre-existing order in a changing world is presupposed by ancient Chinese thinkers. The former requires the First Principle, which is taken as an undoubtable universal law, to start deductive thinking, while the latter seeks for proper relations in associations. The compromise between the two is to re-interpret the undoubtable First Principle needed in geometry, as if it is from the wisdom of the ancient Chinese Sage-Kings and ancestors. There is a reasoning still commonly shared by both Chinese logic and Aristotelian inductive logic. It is analogical reasoning. For the Chinese, finding proper relations itself is a practice of doing analogical reasoning. Aristotelian induction includes analogical reasoning as well. To claim that the wisdom of ancient Chinese Sage-­ Kings and ancestors can be analogical to the undoubtable First Principle could make good sense to Chinese people. Fr. Ricci walked across the boundary between induction and deduction. He practiced finding proper relations of association with Chinese Confucians and claimed  Fr. Ricci wrote the preface of Jihe yuanben in Chinese. Engelfiet translates it into English and includes it in his book. 13  Even all the premises can be true, and conclusion is also true, it can still be an invalid argument if failing to follow the correct syllogistic forms. For example: 12

It is raining (T) King likes the sun (T) Therefore, Honolulu is in the US (T) (No middle term)

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

455

that the Aristotelian First Principle is analogical to the wisdom of ancient Chinese Sage-Kings and ancestors. Starting from there, he introduced the Aristotelian understanding of truth value and validity to Chinese mathematicians. In the preface of Jihe yuanben, Fr. Ricci did not try to define or explain syllogistic forms one after another. He talked about the usefulness of solid and undoubtable knowledge. For example, he says that it is good if a large group of soldiers could bravely defeat a small group of enemies. But it is even greater if a small group of soldiers could bravely defeat a large group of enemies. In his homeland, there were brave warriors who were able to, With a small and weaken group of soldiers to withhold against an army that was ten times as numerous, and to defend isolated cities under siege, and to fend off attacks over sea and over land; in a manner similar to that, as it is told, in China Gong Shu and Mozi repulsed each other’s attacks nine times…it was only because they brought mathematical methods into practice and in such a way became familiar with mathematical theory. (Engelfriet 1998:459)14

With a respect to Chinese ancient sages and logicians, Fr. Ricci confirms that a relation between smart western soldiers and mathematical methods is discovered, and these western soldiers in his hometown did things in the same manner that Chinese sages and ancestors, such as Gong Shu Ban公輸般and Mozi 墨子, did, and, of course, shared the sages’ wisdom. That is why one should learn geometry. Fr. Ricci’s argument for learning geometry is neither according to deductive reason, nor for demonstrating validity, but according to proper relations of association. He impresses the Chinese emperor and Chinese learners with this kind of proper logical relations. No syllogism or logical validity was applied in his own argument above. This is his compromise. Chinese mathematicians support Fr. Ricci immediately, by indicating that the knowledge of forms and numbers could have existed in the ancient time but lost practice after the Han Dynasty. In the preface of Jihe yuanben, a quote from a Confucian sage, Mencius, is given as the justification: All things have come into being with pattern, form and number…Mencius says: [When the sages had used the vigor of their eyes], they call into their aid the compass, the square, the level and the line, to make things square, round, level, and straight. The use of the ­instruments is inexhaustible.” I think that in antiquity, people like Gongshu [Ben] and Mo Di, always have concentrated their minds on these matters... (Engelfriet 1998:454)15

According to the above statement, with the sages’ wisdom and deeds as the source of knowledge, the Chinese ancestors used to practice with forms and patterns as well. Sage scholar Mencius provides evidence, which should not be doubted. Chinese people could go back and pick up that wisdom and learn the knowledge provided in Jihe yuanben. A common ground for them to understand the forms in syllogism is found because it existed in the ancient time.

 Fr. Ricci’s preface of Jihe yuanben. Tran. Engelfiet.  Chinese scholar Wu Xuehao 吴学灏, also wrote a Preface of Jihe yuanben. Engelfiet translates it into English and includes it in his book.

14 15

456

J. Yuan

A bridge between the wisdom of the ancient Chinese Sage-Kings and ancestors and the Aristotelian primary First Principle is possibly built by associating them properly in a Chinese style.

7  T  he Fourth Investigation and My Second Discovery: The Lack of Distribution and Common Practice of the Rule of “Pointing Out” My fourth investigation examines the lake of distribution in Chinese ways of thinking, as well as demonstrate how Fr. Ricci and Chinese mathematicians employ the rule of “Pointing Out” to overcome the difficulties in their translation project. As I pointed out in the section five, categorical syllogism are arguments based on the relations of classes or the membership in categories. The validity of a syllogism requires that the logician “distribute the middle term in at least one of premises” (Rule 2) and that “any term distributed in the conclusion must be distributed in the premises” (Rule 3). According to the definition of distribution given in the section five, if all members in a class are involved, it is a universal proposition. There are four standard forms of categorical propositions: two universal and two particulars. However, Classical Chinese has neither the subject-predicate structure (being structure) nor clear differentiation of singular and plural nouns.16 For example, in Classical Chinese, an affirmative statement, “Huo is a person.” (Xiao Qu), is written as “Hou ren ye” 獲, 人也。(小取). “S P ye” is a typical affirmative statement in Moist logic, but the sentence pattern does not fit in a subject-­ predicate structure, “S is P.” The last word, ye, 也 has no function of “being,” but an associative marker, which confirms relations. According to the Shuo Wen Jie Zi 17 is a pictograph representing “intercourse.”18 Whether or not this again, ye etymology is accurate, it is clear that to make an affirmative statement or to locate an object in a relationship in Classical Chinese (which is done with the ye construction) is to confirm an association between Huo and a person, or two particulars. If an “intercourse” is found, the association between the two parties is confirmed, and the statement stands as affirmative. A negative statement, “Killing the robber is not killing a person” (Xiao Qu) is written as “Shadao fei sharen ye” 殺盜非殺人也(小

 Without distinguishing singular and pleural terms in Classical Chinese language, the traditional Square of Oppositions (A, E, I, O) does not work. Without a “being” structure in Classical Chinese language, the Law of Identity is lacking. Without the Law of Identity, the Law of excluded middle and the Law of non-contradiction do not exist. For these reason, Chinese logic can hardly be a logic of membership. Validity is a product of categorical syllogism. It is a membership related concept. 17  Ye 也 is written in an early Chinese script, zhu shu 篆書 18  Xu Shen says, “‘也’ is the female sexual organ. It is pictographic. (Ye, nuyin ye. Xiang xing. 也, 女阴也。象形。)” Shuo Wen Jie Zi, Xun Zhen. 12B. 16

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

457

取). Fei 非 is also demonstrative character.19 The connection between two individual events, “Killing the robber” (S) and “Killing a person” (P) is blocked by 非, which means separating or no connection according to Shuo Wen. The sentence structure “S fei P” does not fit in a subject-predicate form neither. Using the patterns of “S P ye,” and “S fei P,” to think, Aristotelian Universal Affirmative and Negative propositions (A and E) are lacking. Distribution relations among the terms cannot be found in patterns of “S P ye” and “S fei P,” As a fact, classical Chinese has no clear distinction between singular and plural nouns. The differences between universal and particular statements are not strictly separated from the perspective of membership. The following table shows the gaps between syllogism and the Chinese ways of thinking on distribution, which Ricci and Xu had to bridge: Proposition form All S is P All S is not P Some S is P

Some S is not P

Name and type A-Universal affirmative E-Universal negative I-Particular affirmative

Distribution Subjective term distributes Both subject and predicate distribute Neither subject and predicate distributes

O-Particular negative

Predicate term distributes

Finding proper relation between terms in Chinese logic No Distribution involved No strict “Universal Affirmative” No Distribution involved No strict “Universal Negative” No Distribution involved, but relation between subjective and predicate terms exists between particulars. No Distribution involved, but relation between subjective and predicate terms exists between particulars.

Distribution clarifies the membership in a hierarchical system of classes. It represents the relation between two classes involved in a categorical proposition. Using the above A, E, I, O four types of categorical propositions, one can construct a categorical syllogism. Whether an argument is valid can be proved in Aristotelian logic. The steps of a formal proof demonstrate why the argument is valid. Without an understanding of distribution, one can hardly understand universal affirmative (A) and negative (E). Without universal form, syllogism does not function. Logical validity loses meaning in translation. With the nature of distribution between genus and species, Elements demonstrates a deductive chain, which Fr. Ricci meant to teach the Chinese. He says: Preceding the propositions and proofs the definition has been laid down. After that the general principles have been formulated on which the propositions and proofs rest. Next, the propositions follow. They give an explanation of the problem, and a manner of construction or proof. What comes later, is founded upon results that have been obtained before …

 Xu Shen says, “‘非, 违也’ Fei means separation, cut off a connection.” (Shuo Wen Jie Zi, Book 11).

19

458

J. Yuan

Nowhere can the order be reversed; it is one unbroken chain. The undoubtable principles at the beginning are extremely simple and clear. (Engelfriet 1998:458)20

This chain must be valid and can be proved by syllogistic rules. But, looking for proper relations of association, one’s focus is on discovering proper relations among names, objects and/or events, instead of serious studies on distributions between universal and particular terms. As long as a proper relation is found, which means that the problem is solved, then it is done. No further steps are required. Bridging these gaps is challenging. If the distribution issue can be solved, syllogistic forms should be able to work accordingly, and then logical validity can be discussed. Fr. Ricci, of course, wanted to teach Chinese people the valid forms. But, some common efforts had to be made if there are no concepts, such as, form and distribution in Chinese logic. In Aristotelian logic, if a singular proposition, such as “Socrates is a philosopher,” has only one member in the class of “Socrates,” it is customary to assert that “All S is a philosopher” (S is the class of “Socrates”). In other words, for the singular proposition, the only member in the unit class (one membered-class) can be understood as a universal proposition (A or E). “Socrates,” although a single member, is distributed. If an individual “Socrates” is used to present all the members in the unit class of “Socrates,” then it is a universal proposition. To understand the universal concept “Socrates,” one can Point Out this particular philosopher to demonstrate “all the member(s)” in this unit class. Namely, demonstrating a universal proposition by pointing out the only member in a single member set is acceptable in Aristotelian logic. It seems that this might open a possibility to translate universal propositions (A and E) to a language which has no understanding of universal form and distribution. Some meaning will be lost in the translation, but it might be practical to try. My second discovery is that in translating Elements into Jihe yuanben, a demonstrative rule, ostensive definition, or “Pointing Out,” was used in the process to prove almost every geometry problem. Ostensive Definition is “a kind of denotative definition in which the objects denoted by the term being defined are referred to by means of pointing, or with some other gesture” (Copi and Coben 2005:673). According to Aristotelian inductive logic, an ostensive definition is a demonstrative definition. However, for Chinese mathematicians, using Pointing Out to demonstrate associations is an important practice of logical rule, zhi,指, Pointing Out, which has been practiced be Chinese ancient philosophers and logicians, such as Confucius, Zhuang Zi, ang Gongsun Long Zi. Pointing Out examples, conducts, or cases, Chinese thinkers can reach proper relations of association. Thus, Pointing Out becomes a rule, which opens possible paths for Chinese thinkers to understanding logical validity and for Fr. Ricci to introduce Aristotelian logic vie Elements. Assuming in the set of “point,” there is only one “point.” Then, using a single-­ proposition, this only point, p would be “all p.” If one wants to define what is a point, or dian 點 in Chinese, one can make a dot on the paper and then point to the  Fr. Ricci’s preface of Jihe yuanben. Tran. Engelfiet.

20

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

459

dot and say, “This is a dian.” That would be an Ostensive Definition or “Pointing Out” of “dian.” The connotation of Chinese character dian 點 in Chinese is a black dot. The left radical of dian is black, hei 黑; the right radical is a dot. Dian is a particular object, like a particular apple for eating. The following is the first definition among 36 definitions in Fr. Ricci and Xu’s first book of Jihe yuanben Book I-VI21 (my comments are in parentheses): 1. 點者無分 Definition 1:

A point is that which has no part. (Heath’s translation) A point: it has no part. (Engelfriet’s translation) (Engelfriet 1998:155)

[Commentary:] It has no length and breath, nor thickness. As the following diagram shows: (This is the alternative definition that Ricci and Xu give to describe “a point”). 甲 (A) (This is the diagram in Chinese Jihe yuanben). If definition 1 reads: “a point is that which has no part,” it is a definition from the form of a point. But dian 點 is a property of an object, dot. Engelfriet comments on this very first translated definition in Jihe yuanben: The character used for ‘point,’ dian 點, actually means ‘dot.’ It is a conspicuous physical mark, not really suggestive of a dimensionless object… The Chinese says literally, “A point: it has no part,” or “A point: There are no parts.” The expression does not really seem to render the idea of a definition. It merely states a property of the object “point”. It lacks the notion that only such an object—with the exclusion of all others—is to be considered a point.” (Engelfriet 1998:155–6)

I think that the gap that Engelfriet points out exists from the first definition to the last one in Jihe yuanben, if there is a lack of understanding of universal Form in Chinese language game. If the “dot” as the diagram shows is a point, then how could other smaller or bigger dots not be a point? A dot 點 has parts. If there is no Form of a point, how can this very first definition be taken as a universal proposition to start the learning of geometry? The above diagram is actually a denaturation of a point by using Ostensive Definition to point to a “dot.” But, on the other hand, assuming that this definition is a singular proposition, which has one member only, then the “dot” could be all d in the collective kind/set. By pointing out the dot, one shows an example of a point. Turning to discuss a particular object/case, Chinese ways of determining proper  I italicized the commentaries, which are clearly not from the original text but Fr. Ricci and Xu add-ins or quotes from other sources, and include the diagrams that they used to illuminate the definitions.

21

460

J. Yuan

relations in logical associations work just fine. In Jihe yuanben, Fr. Ricci and XU use a Chinese character “fan, 凡” to indicate “in any case of a kind,” such as “in any triangle” (Engelfriet 1998:149). Although “in any case of a kind” is not about the Form of a kind, but it is a collective term. Therefore, the Form and distribution issues turn to be an issue of Pointing Out which individual case/object to be an example to discuss. In this way, Aristotelian hierarchical relation of membership between terms in universal or particular proposition is replaced by the study of the proper relation of particular individuals in one kind or different particulars in different kinds. The diagram of a dot, dian is demonstrative, and a black dot associates with a name of point or dian 點 in a proper relationship. If a proper name, point, or dian 點 matches the above black dot, then a black dot properly associates with its name dian. So, does all other dots collectively. Allowing Chinese mathematicians to use diagrams to interpret definitions of Form and axioms creates a Chinese style of Elements. Jihe yuanben includes many diagrams which original Elements does not have. The follows are more examples: Euclid’s Definition 2 reads: “A line is breadless length” (Engelfriet 1998:156). It is about the form of line. Following the same pattern to definition 1, point, in Jihe yuanben, a solid line is drawn under the definition of a line in Chinese: 線有長無廣。

It does not matter how thin a solid line is. It has breadth in a drawing. But, the above diagram of line associates with a name of line or xian 線 in a proper relationship. Euclid’s Definition 5 reads: “A surface is that which has length and breadth only”. (Engelfriet 1998:159). Some demonstrative explanations were added in Jihe yuanben. One of them is: “想一線橫行所留下之跡即成面 If you imagine a line that moves transversely the trace that remains form a surface” (Engelfriet 1998:159). And then, a rectangle shape drawing, with the trace left by a moving line, follows:

Chinese leaners might still have no concept of universal forms of point, line and surface, but deduction can go on and rules for validity can be applied if one can avoid distribution issues and simply use a singular proposition as universal affirmative or universal negative. Deductive rules can be introduced. “Discussing by pointing out” was, in fact, what Ricci and Xu did in translating Jihe yuanben. Starting with the first definition of a “dian 點 (point),” all the way until the end of the book (both six- and fifteen-books versions), diagrams went

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

461

along with almost every proposition and problem. When LI Shannan did the fifteen-­ book version, he even added more diagrams in Jihe yuanben Book I-XV. With the help of diagrams, Jihe yuanben gives definitions to geometrical terms, introduces axioms, proves propositions, adds explanations and discusses problems, without worrying much about the lack of universal Forms. With the assistance of demonstrative images, although universal forms were dropped, the deductive rules could be taught and understood with the practice of particular affirmative (I) and negative (O) propositions. These two types of propositions are about the relationships of particulars. When solving geometry problems in particular cases, proper relations of association (zheng 正) can also be reached and agreed upon.

8  Conclusion Chinese and Aristotelian logics hold different presuppositions when demonstrating the world. Chinese ways of thinking never involve a clear separation between a logical world and an empirical one, but Aristotelian deductive logic does. Exploring the logical structure of the Chinese ways of thinking, it is a fact that the lack of Western propositional logic in Chinese thought is one of the major difficulties in seeking for a mutual understanding. There is no attempt to seek the First Principle and universal Forms in dealing with logical relations. Chinese philosophers are interested in finding proper relations amount associations, while Aristotelian syllogism requires valid logical steps for thoughts to move on. They are two very different ways of thinking. For translating Euclid’s Elements into the Chinese Jihe yuanben, Fr. Ricci and Chinese mathematicians in the seventeenth century creatively dealt with these challenges at not only the level of natural language but also the level of logic. Matteo Ricci, S. J tried very hard to make Chinese people understand Western logic. He spoke Chinese and even claimed that he was a “Chinese” among Chinese. Chinese mathematicians also tried hard to understand a whole set of new concepts and rules. Not only did the major translator, Xu Guangqi, contribute to this project of making the translation acceptable by Chinese learners, but also other Chinese mathematicians. One of the most famous Chinese mathematicians in the seventeenth century, Mei Wending 梅文鼎(1633年–1721), first tried to put Euclid’s Elements in the traditional Chinese framework of mathematics. It did not work well. Then, the later Chinese mathematician, Li Sanlan, clearly distinguished two different kinds of methodologies, “Zhong Fa (Chinese methodology) and “Xi Fa (western methodology),” and claimed that both were important. Similarly, with respect, Fr. Ricci had to compromise. He might not be able to teach the Chinese the First Principle and universal form, but the approaches of doing deduction were introduced to Chinese people via learning Geometry case by case. As a result, the translated Elements, Jihe Yuanben, became an understandable Chinese text which introduces both geometry and deductive proofs in particular cases. Looking back at the work done by Ricci

462

J. Yuan

and Chinese mathematicians, one finds that when western logic meets Chinese logic, the conflicts are negotiable. And the process of negotiation turns tout to be a process of creativity. This successful case suggests that having a discourse open between two very different language games will be more fruitful than either shutting off the conversation or forcing others to give up their own rules. Matteo Ricci’s and the Chinese mathematicians’ work proves a significant example of trying hard to understand the unfamiliar.

References Ames, Roger T. and Hall, David L. 1998. Thinking from the Han: Self, Truth. and Transcendence in Chinese and Western Culture. Albany: State University of New York Press. Ames, Roger T. and Rosemont, Jr, Henry. trans. 1999. The Analects of Confucius: A Philosophical Translation. New York: Ballantine Books. Copi, I. M. and Coben, C. 2005. Introduction to Logic. New Jersey: Prentice Hall. Engelfriet, Peter. 1998. Euclid in China: The Genesis of the First Chinese Translation of Euclid’s Elements Books I-VI (Jihe yuanben; Beijing, 1607) and its Reception up to 1723. Leiden: Brill. Gernet, Jacques. 1985. China and the Christian Impact. Cambridge: Cambridge University Press. Graham, A. C. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: The Chinese University Press Martzloff, Jean-Claude. 1997. A History of Chinese Mathematics (trans. Stephen S.  Wilson). New York, Berlin: Springer Verglag. Originally published as Histoire des mathématiques chinoises. Paris: Masson. Rosemont, Henry Jr. 2015. “Translating and Interpreting Chinese Philosophy” (Stanford Encyclopedia of Philosophy, Oct. 27, 2015). https://plato.stanford.edu/entries/ chinese-translate-interpret/. Rorty, Richard. 1980. Philosophy and the Mirror of Nature. New Jersey: Princeton University Press. Tiles, Mary. 1997. “Images of Reason in Western Culture.” In Eliot Deutsch. ed., Introduction to World Philosophies. Upper Saddle River: Prentice Hall. Xu, Shen (121). (re-print) 1996. Shuo Wen Jie Zi Book II B (說文解字卷二下). Beijing: Chinese Book Store Press. Yuan, Jinmei. 2005. “‘Kinds, Lei 类’ in Chinese Logic—A Comparison to ‘Categories’ in Aristotelian Logic” History of Philosophy Quarterly 22.3:181–199. Yuan, Jinmei. 2019. “Exploring an Alternative Pre-Qin Logic.” In Jim Behuniak and Ian M. Sullivan eds., One Corner of the Square: Forward-Looking Reflections from the Students of Roger T. Ames. Honolulu: University of Hawaii Press. Yu, Jiyuan. 1999. “The Language of Being: Between Aristotle and Chinese Philosophy.” International Philosophy Quarterly, 39.4: 439-454. Wardy, Robert. 2000. Aristotle in China: Language, Categories and Translation. Cambridge: Cambridge University Press. Wieger, L, S.J. 1965. Chinese Characters: Their Origin, Etymology, History, Classification and Signification. New York: Paragon Book Reprint Corp and Dover Publication, Inc. Jinmei Yuan  graduated from the University of Hawaii at Manoa with PhD in Comparative Philosophy (1999). She is currently Professor in the Department of Philosophy, Creighton University, USA. Her areas of specialty are Comparative Logic, Chinese Philosophy and Chinese Literature. Publications include: “Recognizing the Existence of Chinese Logic” (co-author with

23  Proper Relations of Association (Zheng 正) vs. Logical Validity …

463

Mary Tiles), Journal of International Communication of Chinese Culture, Vol 3, No. 2 (2016):305–323; “The Role of Time in the Structure of Chinese Logic,” Philosophy East and West (January 2006):136–152; “ ‘Kinds, Lei 类’ in Chinese Logic—A Comparison to ‘Categories’ in Aristotelian Logic,” History of Philosophy Quarterly (July 2005): 181–199.

Chapter 24

Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic in China Rafael Suter

1  Introduction The present chapter traces the adoption of logic in Chinese scholarly discourse during the final years of the empire and its ensuing establishment as an academic discipline in the Republican period. Intended as an overview on various aspects of the complex and multifaceted developments that accompanied  this process, it disregards detailed analyses of more technical questions. Starting with the struggle of early Chinese pioneers both to achieve an appropriate understanding of logic and to establish an accurate logical terminology in Chinese, it ends with the first contributions of Chinese logicians to the field of mathematical logic. This was a watershed which indicates that by the early 1950s – only 50 years after logic had begun to take roots in China – a noticeable number of young Chinese logicians were working at the cutting edge of logical research. The survey starts with a sketch of the increasing interest in logic after the First Sino-Japanese war (1894–95) and the subsequent Hundred Day Reforms (June– September 1898) which for the first time led to a broad attention for the discipline in Chinese elite discourses. Heavily drawing on Kurtz’s (2011) seminal book-length study on the topic, it focuses on early translations and studies in logic and presents an account on the integration of logic into the educational system of the Qing and the accompanying development of a market for textbooks. It also shows how the early establishment of academic logic was accelerated by the availability of textbooks, studies and teachers from Japan. This not only ensured a fast transfer of I would like to thank Fung Yiu-ming, Wolfgang Behr, and an anonymous reviewer for their valuable comments on a previous version of this text. R. Suter (*) Chinese Studies, Institute of Asian and Oriental Studies, University of Zurich, Zürich, Switzerland e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_24

465

466

R. Suter

l­ogical knowledge, but also offered an almost ready-made terminology. With a few exceptions – most notably the term for ‘logic’ itself – this terminology has survived to the present day. The second section deals with some milestones in the development of logic during the Republican period. While the discourse on logic in the final years of the imperial period was often marked by a limited understanding of the technicalities of the discipline, this began to change in the 1920s. On the one hand, an increasing number of Chinese studied in the United States or in Europe and were familiar with the most recent developments. On the other, there was a specialization in China itself, where logic was now widely taught at high schools and universities. Moreover, a modest but growing number of indigenous scientific journals offered an outlet for specialized publications. The academic institutions which took shape in this period provided the context for the first Chinese specialists in logic like Jin Yuelin 金岳霖 (1895–1984), Wang Dianji 汪奠基 (1900–1979) or Shen Youding 沈有鼎 (1909–1989). Chinese conceptions of logic in the 1920s were strongly influenced by the pragmatism of John Dewey (1859–1952) who had personally visited China in 1919/20. The seemingly more practical concerns of pragmatist logic, but also the seminal works of Dewey’s student Hu Shi 胡適 (1891–1962), secured it a dominant position in academic discourse, at least with a majority of intellectuals more concerned with the practical utility of logic than its technical intricacies. In the late 1920s, yet another type of logic arrived in China. An increasing number of intellectuals, dissatisfied with the political and cultural development of the young Republic, were looking for a philosophy more intimately related to the pressing concerns of the time. They turned to Marxist theories reaching China from Japan and later from the Soviet Union. The ‘materialist dialectics’ propagated by these leftist intellectuals promised an entirely new kind of logic fundamentally different from the “socially irrelevant” formal logic they claimed to have overcome. The 1920s and 1930s also mark the time when a fourth kind of logic was more widely adopted in China. In 1922/23, Zhang Junmai 張君勱 (a.k.a. Carsun Chang, 1887–1969), later-on one of the founding figures of twentieth century New Confucianism, had invited the German philosopher Hans Driesch (1867–1941) to visit China. Driesch’s lecture at Qinghua University on the epistemology of Kant and his predecessors marked the beginning of an ever-growing awareness of transcendental logic in Republican China. It was popularized by the writings of Zhang Dongsun 張東蓀 (Chang Tung-Sun, 1886–1973) (original name: Wantian 萬田). The 1940s witnessed a step towards a new quality of Zhang’s reflections: Turning his interest to culture rather than purely philosophical epistemology, he conceived the idea of a linguistic relativism that allegedly also affected the core of reasoning, namely logic. Zhang Dongsun championed the idea of multiple logics – thus completing a development that had begun somewhat earlier with thinkers who emphasized the particular nature of Chinese logic without necessarily sharing Zhang’s particularistic perspective. Another intellectual strand leading back to the philologist Zhang Binglin 章炳 麟 (a.k.a. Taiyan 太炎, 1868–1936) took up the renaissance of Buddhist scholarship

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

467

that had begun in the late years of the nineteenth century and tried to revive the traditional Buddhist ‘art of reasoning’ (yinming 因明, Sanskrit hetuvidyā) which is sometimes labeled “Buddhist logic”1 and once flourished  – for a short time  – in Tang China and subsequently spread to Korea and Japan. From the very beginning of the adoption of Western logic in the early twentieth century, this ‘science of reasons’ was compared to logic and regarded as its more or less powerful counterpart (for an overview on Zhang’s studies on “Buddhist logic” see Aviv 2015: 214–219.) Philologists soon noticed striking parallels between a small corpus of Chinese texts from the pre-imperial period, Buddhist yinming studies and newly introduced publications on “Western” logic. These observations, prepared by Japanese scholars like Kuwaki Gen’yoku 桑木嚴翼 (1874–1946), marked the beginning of a development that led to the discovery of what was considered a “genuinely Chinese” logic. However, both the terminology and the understanding of this term changed over time. While Chinese logic was  first identified in a rather limited corpus of ancient Chinese sources  that seemed to contain theoretical reflections on logical problems, there was an increasing tendency to accept ever more texts as witnesses of its existence. Importantly, this process implied a tacit shift from explicit reasoning about logic to implicit applications of logic. The final section of the chapter resumes an anomaly in the development of Chinese logic. Less than two decades after the first modern translations of logical works and the recognition of logic as an important scientific discipline, when the basics of traditional syllogistics had just gained ground in China, the lectures of Bertrand Russell (1872–1970) in 1920/21 marked the beginning of the adoption of an entirely new kind of logic in China. Although Russell and his teacher Alfred North Whitehead (1861–1947) had started their project of a logical foundation of mathematics with the aim to reconcile the logic of philosophers with that of mathematicians, their axiomatization of logic and mathematics marked the definitive end of the dominance of traditional syllogistics. It amplified a tendency towards the mathematization of logic which had started in the late nineteenth century and was eventually accomplished within a few years after the publication of the Principia Mathematica. Some of the pioneers of this discipline in China and their students soon belonged to the international vanguard of mathematical logic which has changed the face of logic forever.

2  The Institutionalization of Logic In the early times of its appropriation, logic  – the “science of sciences”  – was regarded as a means to reform society and culture (Wang 2000: 1; Zeng 2000: 38; Kurtz 2011: 119, 149), but also as a key to renew Chinese scholarship (Kurtz 2011:

1  For arguments in support of the common characterization of hetuvidyā as “Buddhist logic” see Aviv 2015: 192.

468

R. Suter

150). One of the most influential adherents of logic in late nineteenth century China, Yan Fu 嚴復 (1853–1921), advanced what Joachim Kurtz has called an “inflated notion” of logic – a new guiding ideology that seemed to allow for integrating political, social, moral, and spiritual dimensions; and it was precisely this view on logic that facilitated the subsequent popularization of the interest in this foreign discipline (Kurtz 2011: 165). While this conceptualization had accelerated its general acceptance, some think that it eventually blocked a more thoroughgoing occupation with logic in the long run (see, e.g., Zeng 2000: 38). Transmission of logic to China was incidental and unsystematic, more often than not suffering both from a considerable degree of arbitrariness in the choice of works to be translated and a rather poor quality of the resulting translations (Zeng 2000: 39). Many of the Chinese pioneers who in the early twentieth century contributed to the establishment of logic had studied abroad, but only a minority were trained specialists in logic (Wang 2000: 2). The first society for the study of logic in China, the Logical Society (Mingxuehui 名學會), was founded by Yan Fu in 1900 (Wang 2000: 4; Zeng 2002: 38; Kurtz 2011: 167). There were also local societies for the study of logic, e.g. the Logical Society of Guizhou Province (Guizhou Lunli Xuehui 貴州論理學會) which was among the first in the country (Kurtz 2011: 201). Private colleges and academies were the first Chinese schools that included logic into their curriculum (Kurtz 2011: 194).2 Through his ties to the imperial administration, the afore-mentioned Yan Fu contributed to the integration of logic into the curriculum of normal schools during the educational reforms of the first years of the twentieth century (Kurtz 2011: 168). The second draft of the regulations of the Imperial University (Jingshi Daxuetang 京師大學堂), explicitly intended as precedents for the reform of schools nationwide, introduced logic in the school’s ‘government subdivision’ (zhengzhike 政治 科) (Kurtz 2011: 196–197). Since 1904 logic (then translated bianxue 辯學)3 figured as an elective in the ‘department of (Chinese) classics’ (jingxueke 經學科). Schools had difficulties to recruit teachers for logic, and many hired foreign instructors, mostly from Japan (Kurtz 2011: 200). Almost all textbooks in logic published during the last decade of the Qing dynasty were based on Japanese models (Kurtz 2011: 204).4 The sheer number of translations from Japanese combined with the fact that some 37,000 Chinese had spent their studies in Japan between 1901 and 1909 led to an unprecedented adoption of Japanese neologisms (Wang 2000: 2; Lippert 1979; cf. Masini 1993). In the early years of the twentieth century, when Chinese 2  In 1897, the Yucai Shushu 育才書墅 (School for nourishing talent), the precursor of today’s Jiaotong 交通 University in Shanghai, included courses in logic (bianxue 辯學, lit. ‘science of disputation’) in its English curriculum. In 1898, courses in logic (bianxue 辯學; mingxue 名學, lit. ‘science of names’) were planned at Beijing’s Tongyi Xuetang 通藝學堂 (College of comprehensive arts) (Kurtz 2011: 194). After 1903, the Jesuit-supported Université de l’Aurore (Zhendan Xueyuan 震旦學院) in Shanghai offered logic courses (Kurtz 2011: 202). 3  That bianxue included lessons in phonology and rhetoric and that it was not accessible for students of law and the natural sciences shows the narrow limits of the understanding of logic (cf. Kurtz 2011: 198–199). 4  For details see Kurtz 2011: 207–224, esp. his Appendix A. “Textbooks on Logic Adapted from Japanese, 1902–1911”, in Kurtz 2011: 367–371.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

469

i­nterest in logic arose, there existed already a well-established logical terminology in Japanese which was borrowed into Chinese almost in its entirety. In the Republican period, societies for the study of logic were established at various universities like Qinghua (Tsing Hua) 清華 or Yanjing (Yenching) 燕京 in Beijing (Zeng 2000: 38).5 The Chinese Society for Philosophy (Zhongguo Zhexuehui 中國哲學會) founded by the editors of the journal Philosophical Review (Zhexue Pinglun 哲學評論) in 1935, was instrumental in the dissemination and establishment of logic in China (Zeng 2000: 38, 2002: 38). A Committee for the translation of important philosophical and logical works (Zhongguo Zhexuehui Xiyang Zhexue Mingzhu Fanyi Weiyuanhui 中國哲學會西洋哲學名著 翻譯委員會,  Committee of the Chinese Society of Philosophy for the Translation of Famous Works of Western Philosophy) followed only in 1941 (Zeng 2000: 38). According to He Lin  賀麟 (1902–1992) it ensured, for the first time, a “strict and reliable, systematic and well-­ planned” process of adoption of Western philosophy in editions published by experts who ensured “translation and presentation on an adequate scholarly level” (He 1959: 26).6 Since the very beginning of the thriving interest in logic in the late nineteenth century, newly established magazines and journals played a crucial role in propagating knowledge about the discipline. Liang Qichao’s  梁啟超 (1873–1929) Yokohama-based Xinmin congbao 新民叢報 (The New People’s Gazette, 1902–1907) was among the first to write about Aristotelian syllogistics and Baconian inductionism (Wang 2000: 3). In contrast, the first Chinese publications on mathematical logic were only to appear around 1920 when Qingnian Zazhi 青年雜誌 (La jeunesse) (1915–1922) printed Zhang Shenfu’s 張申府 (1893–1986) portray of Russell along with selected translations of the British philosopher’s work (Shi/Zeng 1998: 49; Wang 2000: 4). Later on, academic journals like The Philosophical Review (Zhexue Pinglun 哲學評論, established in 1927) offered an ever more professional platform for logical discussion and research (Wang 2000: 4). Between 1915 and 1949, some 130 articles on logic were published in Chinese journals, more than 30 of them in The Philosophical Review (Zeng 2002: 37).

3  Traditional Formal Logic The first work on Western logic available in Chinese, the Minglitan 名理探 (Exploration of names and principles), a partial translation of Jesuit commentaries on the Aristotelian organon accomplished in the seventeenth century, remained largely unnoticed by Chinese scholars7 (Standaert 1994: 417; Wardy 2000; Kurtz

5  A National Society for Logic (Zhongguo Luoji Xuehui 中國邏輯學會) was only founded in 1979 (Zeng 2000: 38, 2002: 38). 6  Quoted in Zeng 2000: 38. 7  On Jesuit adaptations of logical writings and notions to Chinese see, e.g., Kurtz 2011, ch.1.

470

R. Suter

2011: 78–79). Unlike their Jesuit predecessors, protestant missionaries in the second half of the nineteenth century, who played an important role for the spread of natural and mathematical science to China, were largely uninterested in questions of logic. They produced only a few textbooks, the ones by William Stanley Jevons (Yefangsi 耶芳斯, 1835–1882) probably being the most influential (Wang 1979: 406; cf. Kurtz 2011: 107–111). Kurtz notes that St. John’s College in Shanghai was the only school in nineteenth century China to teach logic (Chin. bianshixue 辯實 學) in Chinese (Kurtz 2011: 120). Before the twentieth century, virtually no Chinese author ever mentioned European logic, even in the debates about the Chinese origins of Western knowledge that were crucial for popularizing Western sciences (Kurtz 2011: 138). One exception is Tan Sitong 譚嗣同 (1865–1898) who traced the origin of Western logic back to the pre-imperial debaters Hui Shi 惠施 (tr. ca. 370–310 BC) and Gongsun Long 公孫龍 (tr. ca. 325–250 BC) (Kurtz 2011: 140). Modern Chinese logic begins with Yan Fu.8 After his return from Britain, where he had studied seamanship from 1877 to 1879, he won fame as a translator. Influenced by the social Darwinism of Thomas Henry Huxley (1825–1895) and Herbert Spencer (1820–1903) who he had translated at the turn to the twentieth century, he became convinced of the importance of logic for China’s struggle for survival. To him, logic, the alleged core of the “new” scientific knowledge imported from abroad, was the key to China’s ability to cope with the imperial powers (Kurtz 2011: 151; Wang 2000: 2). It was mainly for this extrinsic motive that he translated John Stuart Mill’s (1806–1873) A System of Logic (Mile mingxue 彌勒名學 [Lit. “Mill’s science of names”], 1903–1905)9 into Chinese. In 1909 he published Jevons’s Primer of logic (Mingxue qianshuo 名學淺說 [Lit. “Cursory explanations on the science of names”])10 (Wang 1979: 406; Wang 2000: 3; Sun 2001: 32).11 For his translations Yan Fu mainly resorted to existing Chinese vocabulary: in his Chinese rendering of Jevons’s work, he used the term ci 詞/辭 for ‘proposition’ and ming 名 for ‘term’  – a choice in fact suggested by Mill’s use of ‘name’ in this sense.12 Coining the Chinese term mingxue for ‘logic’ allowed him to appeal to the age-old Confucian topic of the ‘correction of names’ (zhengming 正名) (for an overview see Makeham 1992: 35–95) in order to convince his Chinese public of the importance of logic (Sun 2001: 33; Kurtz 2011: 156–157). Adopting Mill’s emphasis on the value of inductive reasoning, Yan stressed the role of experiments and tests in modern natural sciences. Although Yan Fu showed no interest in the technicalities of the discipline, his active propagation of logic – not least due to his connections to the educational authorities – were crucial for its spread in China (Sun 2001: 33).  For an English overview on Yan Fu see Kurtz 2011, ch.3: 147–192.  Yan published the first part of his translation consisting of Book I of Mill’s Logic in 1903. The full text of his translation, comprising more than half of the original, came out in 1905 at the Commercial Press in Shanghai (Kurtz 2011: 171–172). 10  This text had already been translated into Chinese by the protestant missionary and philologist Joseph Edkins (1823–1905) in the 1880s (Kurtz 2011: 172). 11  Li 2004: 19 gives 1908 for the Mingxue qianshuo. 12  Cf. Kurtz 2011: 156. 8 9

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

471

Independently from Yan Fu, but roughly at the same time, other scholars started to engage in the study of Western logic. During his studies in Japan, Wang Guowei 王國維 (1877–1927) in 1902 had started to read Jevons (Sun 2001: 33). He eventually published a translation of Jevons’s Elementary Lessons in Logic under the title Bianxue 辯學 (lit. “Science of disputation”) in 1908 (Sun 2001: 34; Li 2004: 19), one year before Yan Fu’s translation of the same text. Already in 1906, Hu Maoru 胡茂如 (1883–1907)13 during his studies in Japan had translated Ōnishi Hajime’s ( 大西祝, 1864–1900) Logic (Ronrigaku 論理學) as Lunlixue 論理學 (Wang 2000: 3; Sun 2001: 34; Li 2004: 19). Classicist scholars like Zhang Binglin, Liu Shipei 劉師培 (style Guanghan 光 漢, 1884–1919), Liao Ping 廖平 (style Xugai 旭陔, 1852–1932), and others were considerably influenced by the empiricist logic of these works. Both Liu Shipei’s Xunzi mingxue fawei 荀子名學發微 (lit. “Detailed explanations of Xunzi’s science of names”) and Liao Ping’s Shi ming 釋名 [lit. “On the analysis of names”] resorted to the methods presented in Yan Fu’s mingxue (Wang 1979: 406). In his Yuanming pian 原名篇 [lit. “Investigating the origin of names”], published in 1909, Zhang Binglin even went a step further and constructed his own method conjoining the newly introduced logic with traditional “Buddhist logic” (yinming) and the indigenous thought of ‘names and principles’ (ming li 名理) to shape his own method of “analyzing reasons” jie gu 解故 (Wang 1979: 406; Xiang 1992: 60–62; Li 2004: 20). Wang Dianji 汪奠基 lists a number of other works on logic published between 1903 and 1911, among them Lunlixue gangyao 論理學綱要 (Essentials of logic, 1903) translated from Japanese by Tian Wuzhao 田吳炤 (1870–1926) and Lunlixue jiepou tushu 論理解剖圖書 (Logical analysis illustrated and explained, 1906) translated by Tang Zuwu 湯祖武 (no dates available). These publications were widely sold and commonly used in normal schools of the time, while Wang Guowei’s and Yan Fu’s above-mentioned translations of William Jevons’s Lessons were rather adopted in higher education (Wang 1979: 406; Wang 2000: 3). The first logic textbook written by a Chinese appears to have been Lin Kepei’s 林可培 (no dates available) Lunlixue tongyi 論理學通義 (Comprehensive introduction to logic). Published in 1909, it relied heavily on Japanese precedents (Wang 2000: 3). By the 1910s, more than ten monographs on traditional logic had been published (Zeng 2002: 37), the earliest among them being Wang Yanzhi’s 王延直 (1872–1947) Putong yingyong lunlixue 普通應用論理學 (General applied logic, 1912) (Xiang 1992: 60; cf. Su 2008), Jiang Weiqiao’s 蔣維喬 (1873–1958) Lunlixue jiaokeshu 論理學教科書 (Textbook in logic, 1912), Tang Zuwu’s Lunlixue jiepou tuli 論理 學解剖圖例 (Logical analysis in tables and examples, 1912), and Zhang Zihe’s 張 子和 (a.k.a. Chuntang 春堂, 1885–1974 [?]) Xin lunlixue 新論理學 (A new logic, 1914).14 13  According to Hu’s obituary by Yang Du 楊度 (1875–1931) in the 7th issue of Zhongguo Xinbao 中國新報 in 1907, Hu Maoru (style Cipu 次樸) had died in the same year with only 25 years of age. 14  Xiang Rongxian (1992) further mentions Xing Bonan 邢柏南 (no dates available), Lunlixue (Logic, 1914); Zhang Yucong 張毓聰 (no dates available), Lunlixue (1914); Fan Bingqing 樊炳

472

R. Suter

Zhou Yunzhi 周雲之 characterizes the time after the events of May 4th 1919 as the “period of popularization of traditional logic” in China (Zhou 1987: 41). Between 1920 and 1940 roughly 30 books on traditional Western logic were published in China, many of them textbooks translated from foreign languages. In 1926, the Chinese logician Liu Qi 劉奇 (no dates available) translated James Edwin Creighton’s (Chin. Kuleidun 枯雷頓) (1861–1924) An Introductory Logic, a textbook common at universities in the United States at the time. The revised version of this book, co-authored by Harold Robert Smart (Chin. Sima Te 司馬特, 1892–1979), was translated by Qiu Jinzhang 丘瑾璋 (style Gaoshan 高山, no dates available) in 1934 (Sun 2001: 34). Also Tu Xiaoshi’s 屠孝實 (1898–1932) Mingxue gangyao 名 學綱要 (Essential logic, 1925) was used as a university textbook and exerted a certain influence. Wang Zhangyang’s 王章炴 (no dates available) Lunlixue daquan 論 理學大全 (Great compendium of logic, 1930) did not only cover traditional syllogistics, but also included traditional “Chinese logic” (ming bian 名辯) and “Buddhist logic” (yinming) (Sun 2001: 34). By the late 1940s, around 80 monographs on traditional logic by Chinese authors had been published (Wang 2000: 3; Sun 2001: 34). A key figure for the popularization of logic in Republican China was Zhang Shizhao 章士釗 (style Xingyan 行嚴, 1881–1973). He studied in Japan and Britain, where he took classes in logic with the Scottish philosopher William Leslie Davidson (1848–1929). After his return he started to teach courses in logic at Beijing University in 1918. His Luoji zhiyao 邏輯指要 (Essentials of logic), published in 1939, was based on these lectures which tapped the rich and varied resources of traditional texts to make available China’s logical heritage (Wang 2000: 3). He Lin, who in the 1940s wrote one of the first overviews on the history of philosophy in late nineteenth and early twentieth century China, equalled the importance of this path-breaking work to Ma Jianzhong’s 馬建忠 (1845–1900) systematic account on the grammar of classical Chinese (Shi/Zeng 1998: 50). As early as 1909, Zhang had published a text titled “Lun fanyi mingyi” 論翻譯名義 (On the meaning of translated terms) where he famously and influentially argued against the current Chinese designations for logic, both those invoking traditional Chinese notions like mingxue 名學 (lit. “science of names”) or bianxue 辯學 (lit. “science of disputation”) and those imported from Japan like lunlixue 論理學. Instead he favored the use of the bare transliteration luoji 邏輯, as it appeared to be free of suggestive but eventually misleading connotations (Xiang 1992: 62). However, although certainly aware of the merely transliterational origin of luoji, Zhang Shenfu, one of the Chinese pioneers in mathematical logic, had suggested a semantic motivation of the neologism luoji early on, when he related it to the meaning of the Greek original, a proposal arguably motivated by didactic considerations.15 Nonetheless it is due to Zhang 清 (1876/7–1929/31), Lunlixue yaoling 論理學要領 (Essential outline of logic, 1914), Han Shuzu 韓述組 (no dates available), Lunlixue jiangyi (Lectures in logic, 1918) (Xiang 1992: 60; cf. also Zuo 2016: 162). 15  Zhang traced Greek logikón back to the verb legeĩn which in turn he related to Latin legere, equivalent, in its meaning of ‘to gather’, with the second component word of the Chinese compound luoji ‘logic’, i.e. ji 輯 ‘to gather/to collect’, also used as a variant of 緝 ‘to connect’. In

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

473

Shizhao’s intervention and his highly influential lectures that – in spite of the fact that modern Chinese logical terminology is almost entirely adopted from Japanese – the very expression for ‘logic’ itself differs from Japanese. Although Zhang’s lectures exemplified the concepts and rules of logic by a variety of traditional Chinese texts, he emphasized their universal character. This notwithstanding, his work was also a crucial motivation for the search for a genuinely “Chinese” logic that flourished during the period when he held his lectures (Xiang 1992: 62). Larger debates about logic in the 1920s and 1930s concerned the relationship between form and content – a discussion informed by the then popular pragmatist understanding of logic –, and the relation of logic to dialectics and epistemology raised by Marxist scholars (Zhou 1987: 41; cf. below). Both pragmatist and Marxist positions were attacked by logicians insisting on the essentially formal character of their discipline (Zhou 1987: 42). Beyond these strongly ideological debates, the Republican period also saw some genuine work on traditional logic (Zeng 2000: 39), although most authors concentrated on introducing the fundamentals of traditional logic, often closely following Western models which they simply paraphrased in Chinese (Zeng 2000: 39). This alleged lack of innovation on the part of their coevals was harshly criticized by people like Mou Zongsan 牟宗三 (1909–1995) (Mou 1934: 93; Zeng 2000: 38) and Xie Youwei 謝幼偉 (1905–1976) (Zeng 2000: 39).16 Zeng Zhaoshi 曾昭式 lists a number of misunderstandings common in publications from the 1930s about logic, especially with regard to the relation of logical intension and extension (e.g. in Wang Tefu’s 王特夫 (no dates available) Lunlixue tixi 論理學體系 [A system of logic]) and the laws of thought (e.g. in Yang Peisun’s 楊芾蓀 (1914–1990) Zhongguo luoji sixiangshi jiaocheng 中國邏輯思想史教程 [A textbook on the history of Chinese logical thinking]). He also mentions Lin Zhongda’s 林仲達 (no dates available) introduction of an “unified” or “comprehensive” logic (lit. “synthetical logic”, zonghe luoji 綜合邏輯) which tried to integrate the whole of logic, pragmatism and dialectics, but was marred with misunderstandings and simplifications (Zeng 2000: 39; Zhang 2001: 380, see below). The following lines by Wang Tefu not only document the widespread dissatisfaction with what was felt to be a lamentable state of the field in the early 1930s, they also illustrate the exaggerated expectations from the popularization of logic: “Although our intellectual circles have recognized the importance of this science [of logic] in recent years, and although the ministry of education has also established logic as an educational standard and as a mandatory course from high school to university level, the stagnation of the societal production and of the thinking in material sciences has not yet been surmounted, and the discipline of logic is not yet recognized in its importance by most common people.”17

combination with luo 邏, literally ‘to inspect’, this yields something like the science of “verifying”, lit. “checking [the correctness of] connections” [between propositions] (Yan 2007: 51). 16  Xie Youwei 1997, qtd. in Zeng 2000: 39. 17  Wang Tefu 1933: 2; qtd. in Zeng 2000: 38.

474

R. Suter

One of the less ideological controversies about traditional logic pertained to the necessity of the copula in judgments or propositions. It involved people like Pan Zinian 潘梓年 (pseudonym: Zai Mu 宰木, 1893–1972), Song Zijun 宋子俊 (no dates available), Zhu Zhaocui 朱兆萃 (no dates available), the Steyler missionary Fr. Albert Czech (Chai Xi 柴熙, 1902–1993) and others (Zhou 1987: 43–44) and played a central role for Zhang Dongsun’s logical relativism which claimed that Chinese – marked, among other things, by the lack of a proper copula – was inapt for expressing “formal logic” (see below). Another debate, situated in the field of traditional syllogistics, concerned the controversial relevance of the existence of the referent of the subject term for the validity of direct inferences between universal and particular propositions of different quality. This discussion was initiated by Jin Yuelin who, in an article published in 192718 and somewhat later in his Luoji of 1935, had proposed a complex solution considering different existential characterizations of the subject term (categorical, hypothetical, non-hypothetical) (Xiang 1992: 64). It has been suggested that Jin’s contribution anticipated, by some 15 years, Peter Strawson’s (1919–2006) seminal discussion of the relevance of presupposition for questions of reference (Le 1988: 50; see Strawson 1950). While Jin’s papers about the relation of existence and direct inference are explicitly resuming a problem going back to Russell, others of Jin’s publications more immediately reflect contemporary debates inside China. A short paper on “self-contradiction” (1927)19 in a way forestalled the heated debate on dialectics and its relationship to formal logic starting in the late 1920s; “Incompatible logical systems” (1934)20 was an intervention in the larger context of a discussion about the compatibility of competing conceptions of logic, while “On different logics” (1941)21 was a reply to Zhang Dongsun’s concept of multiple logics (Le 1988: 48). In a sense, Jin’s work represents the transition from traditional formal logic to mathematical logic: At home in both quarters, his publications often were of an intermediary character.

4  Experimental Logic22 During his visit to China in May 1919, John Dewey (1859–1952) gave three lectures, one dedicated to his theory of logic (Zhao 2000: 21; Dong 2003: 100). Only 3  years earlier Dewey had published a collection of essays about what he called  Jin Yuelin, “A, E, I, O de zhijie tuilun” A、E、 I、 O的直接推論 (Direct inferences between A, E, I, and O), Zhexue Pinglun 3.3, 1930, repr. in Zhou, s.a., vol. 1: 473–501. 19  Jin Yuelin, “Lun zixiang maodun” 論自相矛盾 (On self-contradiction), Zhexue Pinglun 1.3, 1927, repr. in Zhou, s.a., vol. 1: 283–304. 20  Jin Yuelin, “Buxiangrong de luoji xitong” 不相容的邏輯系統 (Incompatible logical systems), Qinghai Xuebao 9.2, 1934, repr. in Zhou, s.a., vol. 1: 607–624. 21  Jin Yuelin, “Lun butong de luoji” 論不同的邏輯 (On different logics), Qinghua Xuebao 13.1, 1941, repr. in Zhou, s.a., vol. 2: 378–402. 22  Cf. Dong 2003: 100–102. 18

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

475

“experimental logic” which was translated into Chinese by Liu Boming 劉伯明 (1887–1923) in 1920 (Dong 2003: 100). Criticizing the formalist bent of traditional syllogistics, Dewey underlined the importance of the subject-matter of thought. He defined logic as the way humans think and distinguished it from psychology only by a difference in focus: Whereas psychology is said to investigate the natural course of thought, logic has to deal with the problem of effective investigation (Zhao 2000: 21). In How We Think Dewey offered a five-step analysis of inquiry which for him begins with (1) a felt difficulty, which is then (2) located and defined. (3) This is supposed to suggest a set of possible solutions (4) whose bearing is subsequently developed by reasoning until, in a final step, (5) further observation and experiment lead to the acceptance or rejection of the suggestion (Dewey 1910: 72–77; see also Dong 2003: 101). In his misgivings about the lop-sidedness of formal logic and his criticism of its focus on deduction, Dewey echoed Mill’s Baconian disapproval of traditional syllogistics that had been imported to China by the first advocators of logic at the beginning of the twentieth century. Characterizing formal logic as subjective, Dewey compared it to a spider’s web incapable of withstanding the high pressure required for the solution of concrete problems. Dewey’s disregard for logic is pointedly captured by a remark of his contemporary Arnold Reymond (1874–1958, Chin. Lei Meng 雷孟), whose survey on the state of the art in logical studies was translated into Chinese in 1936 by He Zhaoqing 何兆清 (1907–1969),23 and who laconically states that “scholars propagating pragmatism always neglect logic, especially formal logic” (Zhao 2000: 22). Many commentators agree that Dewey’s “experimental logic” is no logic in the genuine sense of that word (Zhao 2000: 22; Dong 2003: 101). Even if we leave open the question whether Dewey’s analysis of thought rightfully claims the name of logic, it is evident that the methodology of inquiry with which he identifies the term is much broader in scope than both traditional syllogistics and modern mathematical logic. The most famous propagator of pragmatist ideas in Republican China, Dewey’s student Hu Shi, both popularized experimental logic and attacked traditional formal logic for its ineffectiveness in teaching people correct and efficient reasoning (Dong 2003: 101; Zhao 2000: 22). His pragmatist understanding of logic as an efficient method of inquiry was also pivotal for his work on the “logical method” in pre-modern China, as it allowed him to include texts hardly resembling traditional works on logic (Kurtz 2011: 348–349; see also Xiang 1992: 62). In the 1920s, many Chinese scholars regarded Dewey’s experimental logic as the most advanced variant of logic: Chen Xianwen 陳顯文 (no dates available) remarks that “from a contemporary perspective, experimental logic is the most perfected logic. Extremely helpful both for science and philosophy, it is full of merit for people’s everyday lives” (Zhao 2000: 23); Wu Junsheng 吳俊 昇 (1901–2000) notes that “logic textbooks in bookstores are all biased towards its formal aspect, they are not used anymore” (Zhao 2000: 23; Dong 2003: 101). And

 His translation of Reymond’s Les Principes de la logique et la critique contemporaine (Paris: Boivin, 1932, Chin. 邏輯原理及現代各派之評述 “The principles of logic and its critique by contemporary schools of thought”) was republished in Taiwan in 1975.

23

476

R. Suter

in spite of its title, Jiang Hengyuan’s 江恒源 (1885–1961) Lunlixue dayi 論理學大 意 (The essential meaning of logic) exclusively deals with Dewey’s experimental logic (Zhao 2000: 23). Scholars like Zhu Zhaocui, Yi Rongxu 伊榮緒 (no dates available), and Chen Gaoyong 陳高佣 (1902–1976) agreed in their preference for pragmatist logic. This sometimes went hand in hand with more or less sharp refutations of formal logic, e.g. when Wu Junsheng criticized formal logic for its neglect of the actual inquiry of the content of propositions and for its lofty abstractness (Dong 2003: 101; Zhao 2000: 23); Zhu Zhaocui in his Lunlixue ABC (The ABC of logic) sees a need to “resort to comparisons with psychology” in order to “complement the shortcomings of formal logic and to elucidate the essence of logic” (Zhao 2000: 23). Similarly the psychologist and philosopher Chen Daqi 陳大齊 (1886–1983), who later won fame for his work on the Analects and other Confucian classics (e.g. in Chen 1958, 1970), in his Eight lectures on applied logic (Shiyong lizexue ba jiang 實用理則學八講, 1945) characterizes “traditional logic” as a misguided discipline that leads thought into wrong directions. Song Zijun, in a somewhat more reconciliatory tone, propagated a “dialectical unification of form and content” (Zhao 2000: 23) and in a sense anticipated later developments of the late 1920s and 1930s. In opposition to these pragmatist pleas for content, Xie Youwei emphatically remarked that “logic needs to be formal and whatever is not formal is not logic” (Zhao 2000: 24). In the same vein, Wang Zhen 汪震 (style Bolie 伯烈 1901–?) claimed that experimental logic in the sense of a methodology guiding human actions – just like psychology – is not logic (Zhao 2000: 24). Also for Shen Youding experimental logic belongs to psychology rather than logic (Zhao 2000: 24). The American authors Frank Miller Chapman and Paul Henle, whose work The Fundamentals of Logic was translated into Chinese in 1937 by Yin Haiguang 殷海 光 (Fusheng 福生, 1919–1969), only 4 years after its first American edition, likewise noted that experiment adds nothing to the validity of logic and that therefore no logic can be a matter of experimental inquiry (Zhao 2000: 24).

5  Dialectical Logic i. Materialist Dialectics The modern Chinese term for ‘dialectics’, bianzhengfa 辯證法, imported from Japanese, was not uncontroversial in the formative period of Western-style philosophy in China. In the Republican period, competing Chinese translations were proposed. He Lin favored maodunfa 矛盾法, i.e. “method of contradiction”, Zhang Dongsun used bianyanfa 辯演法 “method of discursive evolution” in an early phase, later-on turning to duiyanfa 對演法 “method of oppositional evolution” (Ma 2002: 75). In the 1930s, when Marxist theories gained a footing in China, the expression ‘dialectical logic’ (bianzhengfa luoji 辯證法邏輯) was used as a synonym of dialectics (Zeng 2009: 13).

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

477

The violent conflict between right- and left-wing nationalists on the position towards the Communist Party culminating, in spring 1927, in the so-called “Shanghai massacre” represents the concrete historical background which is mirrored in the fierce intellectual debate about materialist dialectics. It was in 1927 that the Marxist theoretician Ai Siqi 艾思奇 (1910–1966) noted that “the storm of materialist dialectics sweeps through the whole country and the immensity of its power is unprecedented by any philosophical trend of the past twenty two years” (Ai 1991: 66). The ensuing years saw a fierce polemic of Marxist intellectuals against formal logic which they blamed for its alleged metaphysical underpinnings. Regarding it as a relic of the bourgeois epoch, formal logic in their view should entirely cede to materialist dialectics. This storm of critique, sometimes called the “materialist dialectics fever” (weiwu bianzhengfa re 唯物辯證法熱) or, in reference to the place where it arose, the “Shanghai debate on logic” (Shanghai luoji lunzhan 上海邏輯論戰) (Zhang 2001: 379), reached its peak in the mid-1930s (Lu 2007: 45; Sun 2001: 34; Wu 1999: 21). In contrast, Hegel, the idealist pioneer of the critique of formal logic (Yan/Yan 2003: 27; Zhang 2001: 377), never attracted an attention comparable to his materialist heirs. The introduction of Hegelian logic to China thus was partly promoted by adherents of Marxist dialectics, partly advanced by their adversaries intending to correct what they considered a Marxist distortion of the concept. In his Shorter Logic, Hegel characterized logic as the “thinking of thinking” (§19) (Hegel 1991: 46). He proposed a tripartite process of cognition including the three subsequent stages of sensation, understanding and reason. He identified formal logic as the cognitive logic of the understanding, thus distinguishing it from reason which he considered to be manifest in the dialectical movement of thinking. While formal logic is supposed to maintain the distinction of opposites, the dialectics of reason is regarded as a movement of thought between opposites which it is allegedly able to subsume (Zhang 2001: 277). “The mere logic of the understanding [i.e. formal logic, RS] is contained in the speculative Logic and can easily be made out of the latter; nothing more is needed for this than the omission of the dialectical and the rational.” (Hegel, Shorter logic, §82.3 [Hegel 1991: 131–132]; Zhang 2001: 277). Engels lauded the merits of Hegel’s dialectical logic, claiming that, unlike formal logic, it does not content itself with simply “enumerating the forms of motion of thought”, but rather “develops the higher forms out of the lower” (Engels, Dialectics of Nature, Dialectics of Nature – Notes, Engels 1949: 237). In his “Notes to Anti-Dühring”, he eventually conceives “dialectics as the science of the most general laws of all motion”, whose “laws must be equally valid for motion in nature and human history and for the motion of thought. […] without the metaphysical philistine being clearly aware that it is one and the same law that he has come to know.” (Engels 1940: 314): What Hegel had still identified as the laws of the movement of thought thus for Engels had become the laws of movement in general. Eventually, Lenin famously noted that although Marx had not left a “Logic” behind, he took “everything valuable in Hegel and developed it further” to apply, in his analysis of the Capital, “to a single science logic, dialectics and the theory of knowledge of materialism”, adding in brackets that these “three words are not needed”, as

478

R. Suter

“they are one and the same thing” (eto odno i to zhe). (Lenin 1958: 301, 1976: 318; Zhang 2001: 377). This concise sketch makes it clear that formal logic in the Hegelian tradition is related to the abstract forms of thought, while dialectics deals with the concrete operation of the laws of logic as manifest in thinking. Logic is furthermore regarded in terms of its role in cognition, a lop-sidedness inherited from Kant and his focus on transcendental logic. But in spite of their esteem for dialectics, neither Hegel nor Lenin simply discarded formal logic. This only changed when Soviet philosophers from the 1920s to the 1940s interpreted Lenin’s short note as a refusal of formal logic in favor of dialectics (Zeng 2009: 14). Thus, the Soviet Short Philosophical Dictionary (Kratkiy filosofskiy slovar’) which came out in 1930 and was translated into Chinese immediately after its publication states that “the rules of formal logic and the rules of dialectical logic are opposed to each other […] formal logic has no content, is impoverished, and abstract, because the rules and categories which it determines do not correspond to objective reality” (Zeng 2009: 14). More specifically, materialist dialecticians referred back to the Stalinist critique of Aleksey Moseevich Deborin’s (1881–1963) understanding of formal logic. Deborin’s Introduction into the Philosophy of Dialectic Materialism (Vvedenie v filosofiyu dialekticheskogo materializma), published in 1915, soon became part of the philosophical canon of Soviet Russia (Dong 2003: 102), but his adversaries like Mark Borisowich Mitin (1901–1987) and Pavel Fyodorovich Yudin (1899–1968) criticized Deborin and his followers for the allegedly formalistic and metaphysical bent of their formal logic which they considered as opposed to dialectics. Many Chinese Marxists swiftly adopted this critique, emphasizing the allegedly reactionary and conservative implications of formal logic which they condemned as a variant of the flawed metaphysics and subjective idealism of an outdated bourgeois ideology (Zhao 2000: 25; Dong 2003: 102). Like pragmatist criticism of formal logic before it, dialectic materialism insisted on the importance of the content in logic: Whereas logic was considered subjective in form, it was said to be objective in content. A concept, conceived as both abstract in form and concrete in content, on this view admitted for differences of content in what is formally identical. For materialist dialecticians, the mediation of form and content is realized in a dialectic process, and logic therefore becomes historical in character (Zhao 1999b: 33). A first anthology in Chinese of writings on the dialectical method by Marx, Engels, and Lenin, compiled by the Japanese Marxist Kawakami Hajime 河上肇 (1879–1946), was made available in 1930 in a translation by Cheng Shiren 程始仁 (sobriquet of Gao Yuhan 高語罕, 1888–1948) (Lu 2007: 47). In 1933, the Soviet-­ trained Marxist Shen Zhiyuan 沈志遠 (1902–1965) published his Dictionary of the new philosophy (Xin zhexue cidian 新哲學辭典), the first dictionary of Marxist philosophy to appear in Chinese (Hu 2014: 31). Shen’s dictionary heavily relied upon the Short Philosophical Dictionary (1930) by Timofey Semenovich Ishchenko (1903–1938), the first philosophical dictionary to appear in the Soviet Union (Hu 2014: 31). In 1936, Shen Zhiyuan issued two other classics of Soviet philosophy, Mark B. Mitin’s Historical materialism (Istoricheskiy materializm, 1932) and his Dialectical materialism (Dialekticheskiy materializm, 1934) (Shi 1997: 6; Zhao

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

479

1999b: 34). Lu Yi 盧毅 lists no less than 22 monographs on materialist dialectics that were translated into Chinese and published between 1929 and 1931 alone (Lu 2007: 49). By 1937, authors like Zhu Ming 朱明 (sobriquet of Zhang Xiaoxian 張 小嫻, no dates available), Zhang Shu’an 張恕安 (sobriquet Zhang Ruxin 張如心, 1908–1976), Guo Zhanbo 郭湛波 (1905–?), Chen Weishi 陳唯實 (1912–1974), Ai Siqi, Li Da 李達 (1890–1966) and others24 had eventually published no less than 18 works on materialist dialectics (Lu 2007: 50). In a critique of formal logic entitled “‘Deduction’, ‘induction’ and ‘dialectical materialism’”, published in the 1929 volume of the journal Minduo 民鐸 (The People’s Bell), Xu Xingkai 許興凱 (1900–1952) blamed both induction and deduction – two opposites that had marked the main frontline in logical controversies so far – as static and inflexible methods unable to capture the dynamic character of natural and social phenomena (Wu 1999: 21; Sun 2001: 34). In a monograph on dialectic logic that appeared only 1 year later, Guo Zhanbo attacked traditional logic along the same lines.25 In 1932, the journal Dongfang Zazhi 東方雜誌 (The Eastern Miscellany)26 eventually published a text in which Yi Ying 亦英 (a.k.a. Zhu Baiying 祝百英, 1902–?) proclaimed the end of formal logic (Zhou 1988: 21; Sun 2001: 34). In the same year, Li Da, one of the central figures for the translation and dissemination of Marxism in China, and Lei Zhongjian 雷仲堅 (no dates available) translated Ivan Mikhaylovich Shirokov (1899–1984) and Arnol’d Samuilovich Ayzenberg’s (1896–1936) A Course in Historical Materialism (Materialisticheskaya dialektika, 1932)27 into Chinese, most probably from Japanese (Knight 1996: 133; Zhao 1999b: 34). This text was read and annotated by Mao Zedong 毛澤東 (1893–1976) and is taken to be instrumental for the composition of his own “Maodunlun” 矛盾論 (On contradiction)28 (Knight 1996: 114; Shi 1997: 7). In his above-mentioned Lunlixue tixi 論理學體系 (A system of logic), Wang Tefu in 1933 used the Soviet Marxist terminology of Shirokov and Ayzenberg to denigrate formal logic as the thinking method of subjective idealism (Sun 2001: 34), and Qiu Ruiwu 邱瑞五 (no dates available)29 condemned it as “metaphysical” in 1934 (Sun 2001: 34). Marxist-­ inspired critics of formal logic also include Ren Zhuoxuan 任卓宣 alias Ye Qing 葉

24  Lu Shun’ang盧舜昂 (no dates available), Liu Jianheng 劉劍橫 (no dates available), Zhao Yiping 趙一萍 (1905–1986), Wen Jiangong 溫健公 (1908–1938), Li Hengzhi 李衡之 (no dates available) (Lu 2007: 50). 25  Bianzhengfa yanjiu 辯證法研究 (Investigations in dialectics) (Sun 2001: 34). 26  Dongfang Zazhi 29.6. 27  Bianzhengfa weiwulun jiaocheng 辯證法唯物論教程. This textbook had been prepared for use in the Soviet Communist party school. For a detailed discussion see Knight 1996: 133–139; cf. also Fogel 1987: 68. 28  Shi Mingde (Shi 1997: 8) notes that the original version of the “Maodun lun” contained a paragraph about the connection between the law of identity in formal logic and the dialectical law of contradiction. In the 1950s, this passage, reflecting the debate about dialectical logic of the 1930s, was removed from the official version of the text, allegedly by Mao himself. 29  “Xingshi luoji yu bianzhengfa de luoji” 形式邏輯與辯證的邏輯 appeared in the journal Wenli 文理 (vol. 3, 1933).

480

R. Suter

青 (1896–1990) who popularized materialist dialectics in his article “Xin zhexue de liang tiao zhanxian” 新哲學的兩條戰線 (Two frontlines of new philosophy, 1934) and in his monograph Lunlixue wenti 論理學問題 (Problems of logic, 1937) (Sun 2001: 34). In the 1936 volume of Shidai luntan 時代論壇 (Christian Times), Lin Meinan 林美南 alias Miao Nan 繆南 (1909–1955) demanded nothing less than the “destruction of formal logic” (Sun 2001: 34), and in his 1936 work Dazhong zhexue 大眾哲學 (Philosophy for the masses), Ai Siqi heralded that formal logic had long been sentenced to death (Wu 1999: 22; Sun 2001: 34; Dong 2003: 102). Allegedly reaching a higher level of development, dialectic logic, in his view, had overcome formal logic and rendered it dispensable (Yan/Yan 2003: 39). In 1937, Pan Zinian remarked in his Luoji yu luojixue 邏輯與邏輯學 (Logic and the discipline of logic) that formal logic is to be distinguished from a genuine “discipline of logic”, which he conceived in Marxist terms as a ‘method’ (fangfa 方法) of the dialectical development of knowledge (Sun 2001: 34). Considering traditional logic as a mere technical formalism, Pan judged it unsuitable as a genuine method of thought and denied its scientific character. There is some tension in Pan’s approach, though: While he refutes the system of formal logic, he still intends to keep some of its technical apparatus in service of dialectics (Wu 1999: 22). In his Shehuixue da gang 社會學大綱 (Outline of sociology) the above-mentioned Li Da takes a similar stance towards formal logic. Eventually, a Chinese translation of M. B. Mitin’s and T. S. Ishchenko’s Dictionary of materialist dialectics (Weiwulun bianzhengfa cidian 唯物論辯證法辭典) appeared in 1939. It was accomplished by (Zhao) Pingsheng (趙) 平生 (1903–1990), (Liu) Zhizhi (劉) 執之 (no dates available), Nai Gang 乃剛 (no dates available), and Maiyuan 麥園 (a.k.a. Song Jiaxiu 宋家修 1907–1993) on the basis of Jōkichi Hiroshima’s 定吉宏島 (1896–1964) Japanese version (Hu 2014: 32). One year later, the Soviet Short Philosophical Dictionary just 1 year after its publication by Mark Moiseevich Rozental’ (1906–1975) and Pavel Federovich Yudin (1899–1968) came out in two separate translations by Sun Yefang 孫冶方30 (1908–1983) and Hu Ming 胡明31 (1907–1966) respectively (Hu 2014: 33). As materialist critics of formal logic like Xu Xingkai, Ye Qing, Wang Tefu, Fan Shoukang 范壽康 (1896–1983)32 or Ai Siqi first and foremost blamed its purportedly static nature and propagated their dynamic dialectics as an antidote (Zeng 2009: 15), Zhang Dongsun, one of their most prominent opponents, published an article titled “Is a dynamic logic possible?” (“Dong de luoji shi keneng de ma?” 動 的邏輯是可能的嗎?) in 1933. Relegating dynamicity to the realm of the factual, Zhang raised doubts about the need of anything like a dynamic logic. He held that even if facts continuously change, the rules determining our thought about them 30  Luojingtai’er 落靜泰爾 (= Rozental’) and Youqin 猶琴 (= Yudin), Jianming zhexue cidian 簡明 哲學辭典, Beiping, Xinzhi Shushe 新知書社. 31  Luozengta’er 羅曾塔爾 (= Rozental’), M. and Youjin尤金 (= Yudin), H., Zuixin zhexue cidian 最 新哲學辭典, Guangming Shuju 光明書局. 32  Fan Shoukang, “Xingshi lunli yu bianzhengfa” 形式論理與辯證法, in: Ye Qing, ed., Zhexue Lunzhan 哲學論戰, Shanghai, Xinken Shudian辛垦書店, 1935.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

481

nonetheless remain constant (Wu 1999: 22; Zhao 2000a: 26–27; Yan/Yan 2003: 28). This intervention in turn provoked fervent protests by Ai Siqi and Ye Qing and urged the latter to publish a book-length summa against Zhang33 in which he reproached him to misconceive “things as something unchangeable and static” (Wu 1999: 22; Yan/Yan 2003: 28). Directed at such dialectical critics of formal logic, Zhang Shenfu, an early Chinese proponent of mathematical logic, who had adopted Russell’s identification of the “formal” with “logic”, declared the expression “formal logic” to be an outright oxymoron (Yan 2007: 52), a view shared by other influential philosophers like Jin Yuelin or Feng Youlan 馮友蘭 (1895–1990). The materialist critique of the purported metaphysical implications of formal logic is arguably based on fundamental misunderstandings and obviously driven by ideological motives, a situation which readily explains its often aggressive and sometimes rather superficial character (Wu 1999: 22). In this regard, Feng Qi’s 馮 契 (1915–1995) assessment of the debate about logic in the 1930s is sobering: “Generally speaking, the participants of the debate did not reach a high level in logic. Even the basic principles of logic and methodology […] lacked genuine and in-depth investigation.”34 Like in the early days of its adoption, when logic was primarily regarded as a powerful weapon for China’s self-strengthening, the Marxist propagators of dialectical logic once again regarded logic as a tool to overcome the notorious weaknesses of China. Formal logic was criticized not for genuinely logical reasons, but because it was considered an obstacle for to the country’s development (Sun 2001: 38; Zeng 2002: 39; Li 2004: 22–23). ii. Hegelian Logic and Dialectics The dialectical materialists were not the only ones in Republican China to call formal logic into question: He Lin, a specialist of Hegel who had sketched his own brand of “neo-idealism”(xin xinxue 新心學), remarked that “if in playing with symbols one abstracts from their content and ignores the entire system of thought, if one shares the excessive preoccupation with the trivia of medieval scholastics, getting lost in quibbling about this term or that proposition, this inevitably serves the ostentation of exceptional intellectual skills, but it ignores the genuine task of logic” (Li 2004: 22; cf. He 1990: 211). In spite of repudiating Marx’s “putting dialectics back on its feet”, Hegelians like He still adopted their master’s critical account of formal logic. Characteristically, the first Chinese monograph on Hegel’s dialectical logic, appearing in 1930, Heige’er yu bianzhengfa 黑格爾與辯證法 (Hegel and the dialectical method), was published by a Marxist, namely Shen Zhiyuan (Hu 2014: 31). In the same year, Zhou Gucheng 周谷城 (1898–1996), then president of the department of sociology at Zhongshan University (Zhongshan Daxue 中山大學), began 33  Ren Zhaoxuan (= Ye Qing) published a book-length summa against Zhang: Zhang Dongsun zhexue pipan: dui guannianlun, eryuanlun, zhezhonglun de jiantao 張東蓀哲學批判:對觀念論二 元論折衷論的檢討 (A critique of Zhang Dongsun’s philosophy: An examination of idealism, dualism, and ecclecticism), Shanghai, Xinken Shudian 辛墾書店, 1934. 34  “Feng Qi xu” 馮契序 in Peng Yilian 1991; qtd. in Zeng 2000: 39. See also Sun 2001: 34.

482

R. Suter

his studies on Hegel’s philosophy. Among the earliest Chinese intellectuals who seriously engaged with the German idealist’s logic, he issued translations of William Torrey Harris’s (1835–1909) Outlines of Hegel’s Logic and William Wallace’s (1844–1897) The Logic of Hegel in 1930, before completing, 3 years later, his own Chinese translation of Hegel’s Shorter Logic. His plan to publish it in several issues of Siwei Yuekan 思維月刊 (lit. “Thinking Monthly”) failed because the journal soon ceased publication (He 2005: 448). A Chinese version of the Shorter Logic was only published in 1950, almost 20 years later (He 2005: 460). In the same year 1933, Zhou published an introductory article on Hegel’s logic,35 adding, 1 year later, another paper on the role of quality in Hegel’s logic36 (He 2005: 448). Probably the first public debate about Hegelian logic in China took place in 1931, when Zhang Yi 張頤 (1887–1970) who had studied in the United States, Britain and Germany, defended his peculiar understanding of the role of categories in Hegel’s Logic against the criticisms of Zhang Junmai, equally trained in Germany (He 2005: 444–445). He Lin calls Guo Bendao’s 郭本道 (1901–1948) 1934 monograph Heige’er 黑格爾 (Hegel) the first systematic introduction of Hegel’s philosophy in China. The second part of this study was dedicated to Hegel’s logic and presented a detailed introduction into the Shorter Logic (He 2005: 446). In 1934, Guo additionally published an article on Hegel’s dialectic (He 2005: 447),37 while Zhou Gucheng had completed his above-mentioned Chinese version of the Encyclopaedia Logic. According to He Lin, this was among the earliest translations of Hegel’s logic into Chinese (He 2005: 448).38 During the debate about materialist dialectics, traditionalist scholars like Zhang Junmai, Zhang Dongsun, but also the leading liberal thinker Zhang Foquan 張佛泉 (1907–1993) criticized what they considered Marxists’s materialist usurpation of Hegelian dialectics (He 2005: 451–452). iii. “Unified Logic” There were also some more reconciliatory voices in the 1930s controversy about dialectical materialism and formal logic. They argued for the need to adjust the various conflicting conceptions of logic current at the day and to establish a “unified logic”, e.g. Lin Zhongda in a monograph by this title in 1936. Somewhat earlier, Guo Zhanbo in his Lunlixue shiliu jiang 論理學十六講 (Sixteen lectures on logic) published in 1933 had already propagated a unified concept of logic covering not only traditional formal logic and mathematical logic, but also dialectical and experimental logic (Zhang 2001: 380–381). Altogether some twenty books dedicated to this new trend in logic were published, about two thirds of them designed as textbooks (Zhou 1988: 18). Generally speaking, proponents of a “unified logic” tended  Zhou Gucheng, “Heige’er luoji yinlun” (黑格爾邏輯引論), Xin Zhonghua 1.18, 1933.  Zhou Gucheng, “Heige’er luoji zhong de ‘zhiliang’ lun” (黑格爾邏輯中的‘質量’論) , Shidai Gonglun 時代公論 3.8, 1934. 37  Guo Bendao, “Duiyu Heige’er bianzhengfa de jidian yijian” (對於黑格爾辯證法的幾點意見), Xingjian Yuekan 行健月刊 4.5, 1934. 38  It was published as a special issue of the Shanghai-based Zhengli Bao 正理報 and reprinted in 1952 by Commercial Press. 35 36

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

483

to acknowledge the value of more practically oriented variants of logic like dialectics or experimental logic, but they were reluctant simply to dismiss the entirety of classical formal logic. Not least, their reconciliation of these hardly compatible notions of logic also aimed at maintaining the unity of the discipline. Zhou Yunzhi has criticized these attempts as rather coarse and widely inconsistent (Zhou 1988: 19). He distinguishes three periods in the development of “unified” logic, the first arising in the mid-1920s, when scholars tried to reconcile pragmatist experimental logic with a more traditional understanding of logic. Zhou mentions Wu Junsheng and Zhu Zhaocui as representatives of this early period (Zhou 1988: 19–20). The second period starts in the early 1930s and is marked by the intention to adjust traditional formal logic to the dialectical logic of the materialists. Unlike their successors, authors of this period like Shi Ming 石明 (sobriquet of Hu Qiuyuan 胡秋原, 1910–1989),39 Zhu Zhangbao 朱章寶 (1888–1962), Feng Pinlan  馮品蘭 (also Zelan 澤蘭, style Huitian 蕙田, 1894–1984),40 or Guo Zhanbo41 treat dialectical logic, sometimes dubbed “philosophical logic”, as just one variant of logic among others. The following statement by Zhang Dainian 張岱年 (1909–2004) nicely illustrates this position: “In recent years, the dialectical method has matured and in fact become an important type of logic so that together with the deductive and the inductive logic it can be called one of the three important types of logic” (Zhang 1993: 21). With Lin Zhongda’s publications from the mid-1930s, the development of a “unified logic” reached its third phase. Authors now acknowledged the determining role of dialectical logic for the systematic unity of the entirety of logic. Zhou mentions Ye Qing as another representative of this third period. Attempts to construe a unified logic came to an end in the late 1930s (Zhou 1988: 20; cf. also Zeng 2002: 16). In spite of its theoretical limitations, unified logic, which accepted the period’s predominant discourse about dialectical materialism without sharing the materialists’s depreciation of formal logic, is considered to have contributed to the dissemination of traditional logic in the 1930s (Zhou 1988: 22).

6  Transcendental Logic in China In spite of early references to Kant by Zhang Binglin, Liang Qichao, Wang Guowei or Cai Yuanpei 蔡元培 (1868–1940),42 it was probably a lecture series by the German biologist and philosopher Hans Driesch (1867–1941) in 1923 that was dedicated to Kant’s epistemology (“Epistemology before Kant and Kant’s 39  Shi Ming, Lunlixue changshi 論理學常識 (Common knowledge about logic), 1931 (Zhou 1988: 20). 40  Zhu Zhangbao and Feng Pinlan, Lunlixue gangyao 論理學綱要 (Essentials of logic), 1932 (Zhou 1988: 20). 41  Guo Zhanbo, Lunlixue shiliu jiang 論理學十六講 (Sixteen lectures on logic), 1933 (Zhou 1988: 20). 42  For a detailed account cf. He 2005: 425–438.

484

R. Suter

Scholarship”, He 2005: 438) which marked the beginning of a more systematic introduction of the work by the German idealist in China. Driesch was invited by Zhang Dongsun and Zhang Junmai, who also served as his interpreter. In the same year, Zhang Yi started to teach Kant and Hegel at the philosophy department of Beijing University. According to He Lin, the most virulent phase of reception of Kant’s work in Republican China was between 1924 and 1925 (He 2005: 442). In 1925, the journal Minduo published a special issue on Kant, 3 years later an entire issue of the renowned Philosopical Review was dedicated to Kant. One of its contributions was an introduction to Kant’s Critique of Pure Reason by Qu Junong 瞿菊 農 (style Shiying 士英) (1901–1976), who had studied at Harvard with the New Hegelian William Ernest Hocking (1873–1966) (He 2005: 439). In 1933, the first Chinese translation of the Critique of Pure Reason by Hu Renyuan 胡仁源 (1883–1942) was published (He 2005: 442). Although Zhang Dongsun, a very active publisher, journalist and politician, can hardly be regarded as a genuine specialist for Kant,43 the propagation of his own “epistemological pluralism” (renshi duoyuanlun 認識多元論) was essential for the popularization in China of Kant’s epistemology in general and his transcendental logic in particular (He 2005: 440). This is especially true for the work of Mou Zongsan, an early protegé of Zhang who later broke ties with his former mentor. Mou tried to save a peculiar variant of Kant-­ inspired apriorism, although he concentrated on more recent developments in mathematical logic in his logical studies, especially on the Principia Mathematica by Russell and Whitehead. Although there was a considerable number of people among the intellectuals dominating the debates about logic in the Republican era who had studied in Germany, no one seems to have worked on the transcendental logic of Edmund Husserl (1859–1938), whose Logical investigations after its publication in 1900 had become widely read in German speaking academia. Only Shen Youding, who had personally met with the father of phenomenology during his studies in Freiburg from 1931 to 1934 (Jin 2004: 13), referred to Husserl’s analysis of logical syntax and semantics in a number of articles published in the early 1940s (Jin 2004: 18).44 Unlike his disciple Mou Zongsan, whose identification of universal logic with reason as the transcendental legislator of thinking remained in a sense true to Kant’s idea that human reason determines the possible structure of thought and cognition, Zhang Dongsun, in spite of his role as the most prominent popularizer of Kant’s philosophy, doubted that there is any foundation of logic outside language: Logic reflects neither human reason nor the structure of reality. At the same time, Zhang still preserves another aspect of Kantianism, namely the idea that there are certain  Zhao Baoguo (2000a: 25) identifies Zhang as a proponent of a modified Neo-Kantianism.  “Zhenli di fenye” 真理底分野 (Domains of truth, 1940; see Shen 1992: 111–126), “Yuyan, sixiang yu yiyi – yizhi fenxi de di yi zhang” 語言、思想與意義 — — 意志分析的第一章 (Language, thought and meaning – an analysis of intentionality, first part, 1943, see Shen 1992: 127–149), “Yiyi de fenlei  – yishi fenxi de di er zhang” 意義的分類  —  — 意志分析的第二章 (A Categorization of Meaning  – an analysis of intentionality, second part, 1944, see Shen 1992: 150–175).

43 44

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

485

structural restrictions that predetermine the way we think and recognize. But for Zhang these constraints rest on the merely contingent regularity of language. Like the late Wittgenstein, Zhang compares the rules of language to the rules of a game (Ma 2007: 51), and it is the entirety of these basic rules of language which he identifies with logic (Zhao 2000a: 27; Cheng 2006: 119–120; Ma 2007: 50). This lays the ground for some of Zhang’s more specific theses, e.g. that the Chinese language, claimed to lack an exact equivalent of the general verb ‘to be’, was unsuitable as a means for expressing formal logic (Cheng 2006: 120). While Zhang had fervently defended formal logic in the 1930s, he slightly revised his position when, in the 1940s, he shifted his focus towards culture: Regarding language and its structure as deeply entwined with logic, Zhang assumed that different linguistic contexts translate into different kinds of logic: “Logic is brought about by a cultural need and it advances in lockstep with philosophical thought. This means that logic is neither universal nor fundamental. Moreover, there is no ‘single logic as such’, but there are only various different kinds of logic” (Zhang 1997: 20; cf. Zeng 2002: 14). Zhang himself distinguished four types of logic, namely traditional formal logic, mathematical logic, metaphysical logic, and a logic of political thought (Zhao 2000a: 29–30; Zeng 2002: 14; Cheng 2006: 118–119; Ma 2007: 51). In spite of Zhang’s innovative focus on culture, this quartering of the field clearly reflected current intellectual debates about logic that had also fueled the attempts to establish a “unified logic” (Cf. Zeng 2002: 15, 16) mentioned above. Zhang Dongsun was criticized by Jin Yuelin for his failure to acknowledge a universal logic underlying the different explicit logical systems like, e.g., syllogistics and its rules of inference (Jin 1998: 379). This is not surprising if one takes into account that Jin himself had identified necessity as the central issue in logic and that this precludes comparing the implicit laws of logic with the explicit rules of a game. Unlike the latter, the former do not rest on convention and do not allow conscious violation. From a similar universalist perspective, Mou Zongsan tried to rescue the idea of a universal logic as the a priori structure of thought and refused contemporary attempts to reduce logic to the rules of linguistic syntax (Mou 2003: 120–126; Suter 2017).

7  The Revival of Buddhist Logic45 Yang Wenhui 楊文會 (style Renshan 仁山) (1837–1911), sometimes called the father of the modern Buddhist renaissance in China (cf. Makeham 2014), also is a key figure for the beginning of modern research on Chinese “Buddhist logic”, called “science of reasons” (Chinese yinming 因明, Sanskrit hetuvidyā; on Yang see Aviv 2015: 208–211). During his journey to England, the lay Buddhist had met with Nanjō Bunyū 南條文雄 (1849–1927), who later was one of the editors of the Taishō Tripiṭaka, a comprehensive edition of the Chinese Buddhist canon.

45

 For a more comprehensive discussion in English see Aviv 2015.

486

R. Suter

Nanjō provided Yang with some 300 Buddhist texts that had allegedly been lost in China. Among them was the Yinming Dashu 因明大疏 [“Large subcommentary on the science of reasons”] (Yinming ru zhengli lun shu 因明入正理論疏 [“Subcommentary about the śāstra on the introduction into the methods of the science of reasons”], T. 1840) by the Tang dynasty yogācāra scholar monk Kuiji 窺基 (632–682). This copious commentary is a compilation of other early Tang commentaries to the Yinming zhengli men lun ben 因明正理門論本 (“Commentary on the introduction into the methods of the science of reasons”, ∗Nyāyamukha/ ∗Nyāyadvāra, T. 1628 [1629])46 by Dignāga (ca. 480–540, Chin. Chenna 陳那) and to the Nyāyadvārata rakaśāstra (Yinming ru zhengli lun 因明入正理論, “Śāstra on the introduction into the methods of the science of reasons”) by Dignāga’s disciple Śaṅkarasvāmin (sixth century. AD) (Chin. Shangjieluozhu 商羯羅主, Tianzhu 天 主) (Yao 1994: 142; Yao 2008: 85, 87, 106; for a concise English overview on “Buddhist logic” and its reception in China see Aviv 2015: 203–204). Early in 1908, when Yang established a Buddhist academy he accorded Buddhist logic a prominent position in the curriculum, recommending third year students to study Kuiji’s comprehensive commentary on Dignāga’s ∗Nyāyamukha/-dvāra mentioned above (Aviv 2015: 209). It was also under Yang’s  influence that Wang Xiaoxu 王小徐 (1875–1948), a leading scientist and lay Buddhist formulated a systematic critique of Western science from the perspective of Buddhist hetuvidyā (Aviv 2015: 211–212, referring to Hammerstrom 2010). Another early advocate of “Buddhist logic” was Song Shu 宋恕 (1862–1910), an acquaintance both of Yang Wenhui and Zhang Binglin. Eyal Aviv credits him for pointing out for the first time the similarities of Western logic with the Buddhist “science of reasons”, although he recognizes Song’s merits in his role as a “spokesman for the cause of making Buddhist logic part of the Buddhist curriculum in China” rather than in his studies on the subject (Aviv 2015: 213–214): His “most crucial contribution to the promotion of Buddhist studies was the introduction of Buddhism to […] Zhang Taiyan [a.k.a. Binglin]” (Aviv 2015: 214). Other pioneers of the Buddhist renaissance at the turn from the nineteenth to the twentieth centuries to show a pointed awareness of the importance of “Buddhist logic” were people like the yogācāra revivalist Ouyang Jingwu  歐陽竟無 (1871–1943) as well as  the charismatic teacher and prolific writer Taixu 太虛 (1890–1947) (Aviv 2015: 219–221). In 1928, Lü Cheng 呂澂 (1896–1989), Ouyang Jingwu’s “major disciple” (Aviv 2015: 221), and the Buddhist monk Shi Yincang 釋印滄 (1906–1943) edited a critical edition of Dignāga’s ∗Nyāyamukha/-dvāra, on the basis of which Qiu Bo 丘檗 (a.k.a Qiu Ximing 邱晞明, no dates available) wrote the first modern commentary on the text (Yinming zhengli men lun jiaoshu, 因明正理門論斠疏 [Notes on the Commentary on the introduction into the methods of the science of reasons], 1930) (Yao 1994: 144). Additionally, the 1930s and 1940s saw the composition of no less than eight commentaries on Śaṅkarasvāmin’s Nyāyadvārata rakaśāstra, 46  The text was presumably translated by both Xuanzang (title: Yinming zhengli men lun ben 因明 正理門論本) and Yijing 義淨 (635–713) (title: Yinming zhengli men lun 因明正理門論) whose translation only survives in fragments (Yao 2008: 85).

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

487

among them the lecture notes by Zhou Shujia 周叔迦 (1899–1970) (Yao 1994: 144).47 Eventually, three comprehensive commentaries on Kuiji’s Yinming dashu were accomplished during the Republican period: In 1925, Mei Guangxi 梅光羲 (1880–1947) published his notes and explanations on the text (Yinming ru zhengli lun shu jie lu ji zhu 因明入正理論疏節錄集疏 [Collected notes on extracts from the Subcommentary on the introduction into the methods of the science of reasons]), and only 1 year later Xiong Shili’s 熊十力 (1885–1968) commentary came out. In 1938, Chen Daqi eventually published his magisterial Yinming dashu lice 因明大 疏蠡測 (Desultory fathomings on the Large subcommentary on the science of reasons) (Yao 1994: 145). The first book-length study on “Buddhist logic” after Zhang Binglin’s pioneering but rather idiosyncratic work48 was Xie Meng’s 謝蒙 (Wuliang 無量) (1884–1964) Fojiao lunlixue 佛教論理學 (Buddhist logic, 1916) which in its very title adopted the Chinese neologism for logic to refer to Buddhist “science of reasons” (hetuvidyā). Ten years later, Lü Cheng issued his Yinming gangyao 因明綱要 (Essentials of hetuvidyā; for a more detailed account on Lü Cheng’s work on Buddhist logic see Aviv 2015: 222–227). Chen Wangdao’s 陳望道 (1891–1977) widely read Yinming xue gailüe 因明學概略 (Comprehensive introduction to the study of hetuvidyā), which consisted of the notes of a lecture held at Fudan University in 1928, followed in 1931 (Yao 1994: 145). Another book-length study titled Luoji yu yinming 邏輯與因明 (Logic and hetuvidyā) by Gong Jiahua 龔家驊 (Yunbo 雲伯, no dates available) appeared in 1935 (Yao 1994: 146). One year later, Taixu published his lecture notes of a course taught at Wuchang University in 1922 as Yinming gailun 因明概論 (Introduction to hetuvidyā), and Yu Yu 虞愚 (1909–1989) gave out his Yinmingxue 因明學 (Studies on hetuvidyā) – one of the

47  In addition to Zhou Shujia’s Yinming ru zhengli lun shi 因明入正理論釋 (published in 1989), Yao (1994: 144) mentions an undated Yinming ru zhengli lun yuezhi 因明入正理論約旨 (Summary of the principal ideas of the Introduction into the methods of the science of reasons) by a certain master of Beihua 悲華, Wang Jitong’s 王季同 (1875–1948) Yinming ru zhengli lun moxiang 因明 入正理論摸象 (Tentative interpretations of the Introduction into the methods of the science of reasons, 1930), a master of Beiyuan’s 悲園 Yinming ru zhengli lun jiangyi 因明入正理論講義 (Lectures on the Introduction into the methods of the science of reasons, 1932), the Yinming ru zhengli lun 因明入正理論 (On the Introduction into the methods of the science of reasons, undated) by the Japanese Jōdo monk Hayashi Genmyō (Lin Yanming) 林彥明 (1868–1945), Mi Lin’s 密林 (dharma master Chisong 持松, 1894–1972) Yinming ru zhengli yijie 因明入正理論易 解 (Easy explanations of the Introduction into the methods of the science of reasons, 1940), the Yinming ru zhengli lun shi 因明入正理論釋 (An analysis of the Introduction into the methods of the science of reasons, 1943) by the yogācāra specialist Han Kezhong 韓克忠 (1884–1949, a.k.a layman Qingjing 清淨居士), and the Buddhist scholar monk Changxing’s 常惺 (1896–1939) Yinming ru zhengli lun yaojie 因明入正理論要解 (Essential explanations on the Introduction into the methods of the science of reasons, 1948). 48  Zhang had shortly mentioned the yinming rules of inference in the preface of his comprehensive study on the so-called masters-literature (Zhuzi lunxue 諸子論學) published in 1906 and further developed and systematized his reflection about yinming and its relations to classical Chinese thought and modern logic in later publications, famously his 1909 study Yuanming 原名 (Investigating the origin of names).

488

R. Suter

first studies systematically using logical terminology for elucidating the Buddhist “science of reasons” (Yao 1994: 145). Zhou Shujia’s Yinming xin li 因明新例 (New case studies in hetuvidyā) appeared in the same year 1936. Chen Daqi’s Yindu lizexue 印度理則學 (Indian logic), which came out in 1939, is another seminal study on hetuvidyā composed in the Republican period (Yao 1994: 146). When early discussions attempted to capture the relationship of Buddhist logic to Aristotelian syllogistics, authors resumed an opposition that had become current since the very beginning of the adoption of modern logic in early twentieth century China: In contrast to the deductive nature of Western formal logic, Buddhist hetuvidyā was often characterized as “inductionist” – a fact which approximated it to the scientific paradigm broadly identified, in the first decades of the twentieth century, with Baconian inductionism. Generally, scholars who tried to present the “science of reasons” with its highly specialized vocabulary in recently established logical terminology also had to specify the exact relation between Buddhist and Aristotelian inferences. Given the fact that Buddhist logic seemed to emphasize the role of analogy and exemplification, one major problem was the clarification of the precise relationship between the three branches of ‘thesis’ (zong 宗, Sanskrit pratijñā), ‘reason’ (yin 因, hetu), and ‘analogy’ (yu 喻, udāhārana) in Dignaga’s hetuvidyā and their logical equivalents in the syllogism, namely conclusion, minor premise and major premise. Besides their genuine preoccupation with Buddhist scholastics and the logical reconstruction of the “science of reasons”, many twentieth century scholars of Chinese Buddhist logic also recognized in hetuvidyā a traditional tool of systematic thought and argument which else seemed to be missing from the historical record of Chinese philosophy. Thus, Xiong Shili, who after intensive studies in Buddhism eventually turned into one of the most influential Confucian revivalists of his generation, at one point bemoaned that “the works of Chinese philosophy lacked systematicity because they emphasized personal experience without dealing with logic” (Li 2004: 21). Elsewhere, Xiong underscored the indispensability for philosophy of logic which – arguably informed by his deep engagement with the Buddhist “science of reasons” – he identified as the art of “careful thinking and clear distinction” (shen si ming bian 慎思明辨).

8  Research in the History of Logic in China Inspired by the adoption of Western style logic, Chinese scholars in the twentieth century set out to uncover traditional Chinese “logic”, which is customarily referred to by the meanwhile established term ‘studies of names and disputation’ (mingbianxue 名辯學) since the 1980s. According to a well-established narrative, this logic had been lost after the Chinese medieval period (third–sixth century AD) (Sun 2001: 29) so that, for want of a continuous indigenous logical tradition, research on ­traditional Chinese logic had to rely on comparative approaches since its beginnings (Sun 2001: 30). Discussion about the possible existence of logic in traditional China

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

489

emerged in Meiji Japan where “Chinese philosophy” had become fashionable among Japanese scholars thinking about the “Eastern” contribution to universal philosophical problems (Kurtz 2011: 283–289). In 1895, Matsumoto Bunzaburō 松本分三郎 (1869–1944) related the skillful argumentation of Gongsun Long to the Greek sophists. Only 2 years later, Kanie Yoshimaru 蟹江義丸 (1872–1904) identified a “logical spirit” in the ‘Zhengming’-chapter (Rectification of names) of the pre-imperial Xunzi. Kanie’s insufficient analysis was complemented by Kuwaki Gen’yoku in his 1898 essay “Junshi no ronrisetsu” (Xunzi’s logical theories) (Kurtz 2011: 284–285). Another Japanese pioneer in the study of Chinese logic was Murakami Senjō 村上專精 (1851–1929) (Makeham 2012: 103). These Japanese origins of modern research on logic in traditional Chinese texts have provoked Joachim Kurtz’s pointed remark that “if it were necessary to identify a discoverer of Chinese logic, Kuwaki Gen’yoku would have to be named as the most serious contender” (Kurtz 2011: 289). In China, the search for texts about logic in the Chinese tradition started in the first years of the twentieth century, roughly at the same time as the very adaptation of the discipline of logic itself (Wang 2000: 4; Zeng 2002: 36). Early attempts to look for parallels of the newly imported science of logic in China’s own intellectual heritage by people like Liu Shipei,49 Zhang Binglin,50 or Liang Qichao51 in the first decade of the twentieth century were often marred by more or less obvious misunderstandings. While Liu is said to have poured “old philological wine into new logical bottles” (Kurtz 2011: 299), Zhang, heavily influenced by yogācāra Buddhism, regarded logic as an “art of reasoning” (Kurtz 2011: 304). Another pioneering study, Liang Qichao’s first investigation of the dialectical chapters of the Mozi 墨子, published in 1904, presented an “exuberant account of scores of logical ideas” purportedly anticipated by the ancient contender of the Confucius (Kurtz 2011: 314). Unlike Liu Shipei and Zhang Binglin before him, Liang Qichao eschewed redefining the understanding of logical concepts before applying them to Mohist sources. It has been speculated that Liang’s studies may have been directly inspired by the work of Kuwaki Gen’yoku (Kurtz 2011: 319–320). In 1904, Wang Guowei, another founding figure of the research in traditional Chinese logic, translated Kuwaki Gen’yoku’s article on the logic of Xunzi (Kurtz 2011: 327). Wang considered Mozi (Mo Di 墨翟, trad. ca. 490–381 BC) as the originator of Chinese logic (Kurtz 2011: 329; 333), but he argued that the tradition of Chinese logic had been interrupted since the Western Han dynasty. As a consequence, Chinese logic for him was only of archival interest (Kurtz 2011: 329). Sometimes, Sun Yirang 孫 詒讓 (1848–1908), who spent almost 30 years of his life editing and commenting the Mozi, is also counted among the forerunners of the research in premodern Chinese logic. In the wake of Yan Fu’s popularization of logic in China, he had noticed that the ‘Mohist Canons’ (Mojing 墨經) somewhat resembled Aristotle’s

 On Liu Shipei’s adoption of the term mingxue 名學 see Kurtz 2011: 295–301.  On Zhang Binglin (Taiyan) cf. Kurtz 2011: 302–312. Cf. Makeham 2012: 103. 51  Cf. Kurtz 2011: 313–327. 49 50

490

R. Suter

deductionist logic, Bacon’s inductionism and Indian hetuvidyā. Although Sun never specified these purported resemblances, his observations still stand at the very beginning of a growing interest for China’s indigenous logical heritage (Yang 2005: 19). Only a decade after more or less specialized studies on logic had reached China, the works of Liu Shipei, Zhang Binglin, Liang Qichao and Wang Wei thus had already established a distinct discourse of “Chinese logic” (Kurtz 2011: 335). According to Sun Zhongyuan 孫中原, however, serious academic research on ancient Chinese logic only began in the 1920s, when the more mature logical works of Liang Qichao and Hu Shi provided the first detailed studies on the topic (Sun 2001: 30). Liang Qichao, who published a number of works on the ‘Mohist Canons’,52 was the first to identify particular Mohist expressions with logical terms of the Western tradition. Most remarkably, he equated Mohist bian 辯 (‘distinction; dialectic’) to then current Chinese equivalents of ‘logic’ (lunlixue 論理學, luoji 邏輯) (Sun 2001: 30; Zeng 2002: 37; Yang 2005: 19; cf. also Zhang 2004: 23). In his view, the ‘Mohist Canons’ dealt with the same principles and rules like Western and Buddhist logic, though on a less elaborate level (Yang 2005: 19). Consequently, he considered them as “one of the world’s oldest books on logic (mingxue 名學)” (Sun 2001: 30). With his remark that, “with the fall of the Qin and Han dynasties, this invaluable treasure was hidden away in a basement for two millennia” (Sun 2001: 30), Liang Qichao established the standard narrative about the demise of Chinese logic during imperial times and its rediscovery in the twentieth century. In his studies, Liang moreover noticed a remarkably close relationship between Buddhist hetuvidyā and Mohist logic, the main purpose in both cases consisting in “making reasons explicit”, only that the Buddhists used the term yin 因 while the Mohists spoke of gu 故 (Sun 2001: 30; Yang 2005: 19).53 In 1909, Zhang Binglin published his Guogu Lunheng 國故 論衡 (Evaluation of our national heritage). It contained a chapter titled Yuanming pian 原名篇 (“Investigating the origin of [the philosophy of] names”) which compared old Chinese theories of names and argument to Buddhist yinming and Western logic. Focusing on their establishment of proofs, he noticed that all three are similar in advancing in three steps, while they vary both in terms of the order of arguments and of the range of elements affected. Zhang considered Buddhist logic as the most advanced among the three (Yang 2005: 19–20). Tan Jiefu  譚戒甫 (1887–1974) likewise interpreted the principles of reasoning in Mohist dialectic in terms of “Buddhist logic”, identifying the concepts of gu 故 ‘reason’, li 理 ‘pattern’, and 類 lei ‘kind’54 in the Mohist ‘Great Pick’ (Daqu 大取) with the yinming notions of zong 宗 ‘thesis’, yin 因 ‘reason’, and yu 喻 ‘analogy’. In contrast to these assimilations of Mohist logic to the Buddhist art of reasoning, Zhang Shizhao put forward a syllogistic interpretation of the Mohists’s thinking in three steps (Zhang 2004: 23).

 Mozi zhi lunlixue 墨子之論理學 (The logic of the Mozi); Mozi xue’an 墨子學案 (Case studies on Mohism); Mozi jiaoshi 墨子校釋 (Collation and interpretation of the Mozi). See Yang 2005: 19. 53  Cf. also Zhou Yunzhi 1991: 9–10, cit. in Yang 2005: 19. 54  Cf. For the translations of li and lei cf. Graham 1978: 480. 52

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

491

Gao Yuan’s 高元 (style Chengyuan 承元) (1892–1980) long essay “Bianxue gu yi” 辯學古遺 (Our ancient logical heritage, 1916) is the first of a wealth of publications about the nature and essential elements of Chinese logical thought. An outstanding contribution among these was Liang Qichao’s Mojing jiaoshi 墨經校釋, published in 1920, which suggested a reconstruction of the relationship between the ‘Mohist Canons’ (short and concise guidelines) and Explanations (corresponding explanations or illustrations). Liang’s reestablishment of the text since has become widely acknowledged as the most adequate reading (Kurtz 2011: 345). Other early contributions to the understanding of Mohist dialectic include seminal works like Wu Feibai’s 伍非百 (1890–1965) Mojing jiegu 墨經解故 (Explanations on the ‘Mohist Canons’, 1922) and Luan Tiaofu’s 孿調甫 (1889–1972) Mobian taolun 墨辯討論 (Discussion of the dialectical chapters of the Mozi, 1926) (Kurtz 2011: 345–346). Inspired by Liang Qichao’s studies, Hu Shi, who had studied at Columbia University from 1915 to 1917, wrote his doctoral dissertation about “The Development of the Logical Method in Ancient China”, according to Kurtz the first monograph on Chinese logic written in any language (Kurtz 2011: 345). Informed by a rather inflated pragmatist concept of logic, Hu presented an analysis of the logical methods found in various pre-imperial master texts. Although he laid special emphasis on the Mohist notion of ‘explaining knowledge’ (shuo zhi 說知) and the methods of argumentation found in the ‘Mohist Canons’ (Yang 2005: 19), his broad understanding of logic allowed him to include the thought of Confucius, Laozi, Zhuangzi and others into his study (Kurtz 2011: 349). Famously and controversially he ranked the “sophists” Hui Shi and Gongsun Long among the “Later Mohists”, rejecting the category ‘school of names’ (mingjia 名家) which he regarded as a retrospective and misguided invention (Kurtz 2011: 349). Hu’s doctoral dissertation, the English version of which was only published in 1922, was the basis of his comprehensive history of Chinese philosophy (Zhongguo zhexueshi dagang 中國哲 學史大綱, 1919) (Kurtz 2011: 346). Underscoring the value of the comparative method for his studies, his interpretation of the allegedly technical vocabulary of the ‘Mohist Canons’ is obviously influenced by the then virulent debate about the relation of inductive and deductive logic (Sun 2001: 30–31). Characterizing the later Mohists as excellent “logicians” (luojixuejia 邏輯學家), he emphasized their exceptional status in Chinese intellectual history. For him, Mohism represented the only autochthonous intellectual movement which developed a scientific logic, an achievement which in his view earned them an important position in the development of the discipline on a global scale (Sun 2001, 31; on Hu see also Xiang 1992: 62–63). Among other things, Hu Shi argued that the ‘Mohist Canons’ stated the laws of non-contradiction and of the tertium non datur, a view that was accepted by Zhang Shizhao later on, in spite of his corrections of many of Hu Shi’s individual observations (Zhang 2004: 23–24). Zhang Shizhao’s above-mentioned Luoji zhiyao 邏輯指要 (Essentials of logic) was the second monograph on Chinese logic to be composed. Unlike Hu Shi’s history of Chinese logic, Zhang’s book was conceived as a general course in logic (Kurtz 2011: 350). Zhang argued from an explicitly universalist stand, emphasizing

492

R. Suter

that in spite of the fact that the word logic had first made its appearance in Europe, the rules which this science investigates are universal. Unwittingly or not, Zhang, who drew on a broad array of examples from Chinese texts to illustrate the logical rules which he discussed, blurred the distinction between explicit theoretical statements in ancient Chinese texts and arguments with implicit logical structures lending themselves to analyses in logical terms (Kurtz 2011: 352). His mention of a distinctly Chinese ‘science of names’ (mingxue 名學) was mistaken by several authors as a claim of an independent identity of Chinese logic as a whole (Kurtz 2011: 354). In his Xian-Qin bianxue shi 先秦辯學史 (A history of the science of disputation in the pre-Qin era), published in 1932, Guo Zhanbo thus had argued that the emphasis of ancient Chinese logic lay on “disputation” rather than “naming”, a perspective which secured the “morally dubious” dialecticians Gongsun Long and Hui Shi a central position in its history. This view was basically shared by Wang Zhanghuan, although in his Lunlixue daquan 論理學大全 (Great compendium of logic) published in 1930 he maintained that the traditional term mingxue designated a subfield of the art of disputation specifically devoted to the logic of names. In his renowned study Zhongguo mingxue 中國名學 (The Chinese science of names) published in 1937, Yu Yu eventually suggested that the term lunlixue – science of reasoning – should be adopted as the general name of a field further subdivided into luoji  – Western logic –, yinming  – Indian logic –, and mingxue  – Chinese logic (Kurtz 2011: 354). Sun Zhongyuan regards Shen Youding, who had started his career in the Republican era as a specialist in modern logic and semantics, as the most eminent specialist on ancient Chinese logic. His research on the Mohist logic – conducted mostly after 1949 – has left its imprint on the academic discourse about the topic in China. In her overview on the development of research on pre-Buddhist logic in China, Yang Wen 楊文 traces the now-current notion mingbian 名辯 for Chinese logic back to the Mohist concept of bian and the notion of ming taken from the Xunzi, claiming that all discussions on logic in ancient China can be identified with either of the two (Yang 2005: 19). Kurtz points out that this composite term appeared no earlier than the 1940s and that it was only advocated as a positive designation for Chinese logic since the 1980s by Zhou Yunzhi and Liu Peiyu 劉培育 who based themselves on Zhang Dainian (Kurtz 2011: 355). Zhang Qing 張晴 observes that research on logic in pre-modern China in the early twentieth century focused on the quest for similarities between Chinese sources on the one hand and Western or Buddhist Indian sources on the other. This position is prominently expressed in Zhang Shizhao’s statement that he “considers European logic as the warp and China’s investigations on concepts and principles as the weft” of his comparative endeavor of reconstructing the texture of Chinese logic (Zhang 2004: 25). In Zhang Qing’s opinion, early research about Chinese logic was not only restricted to textual material directly connected to questions of logic and largely neglecting these texts’s historical and cultural background, but it was often additionally hampered by an insufficient or even inadequate understanding of Western logic (Zhang 2004: 25).

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

493

9  Mathematical Logic Mathematical logic reached China in the early 1920s, when Bertrand Russell upon his visit introduced its basics in various lectures throughout the country. His lectures at Beijing University, which started in October 1920, were dedicated to introducing his own propositional and class calculus (Li 1989: 2; Shi/Zeng 1998: 49; Song 2000: 12; Sun 2001: 36).55 Russell’s translator was the America-trained physicist and mathematician Zhao Yuanren 趙元任 (Chao Yuen Ren, 1892–1982), later to become one of the leading Chinese linguists of the twentieth century. The general prominence of Russell earned the discipline an exceptionally broad attention among Chinese intellectuals. Shortly after Russell’s arrival, a Russell Society was founded, and in January 1920 a Journal of Bertrand Russell Studies (Luosu yuekan 羅素月 刊) was launched (Shi/Zeng 1998: 50). Wu Fanhuan’s 吳范寰 (1897–1978) notes of Russell’s Beijing lectures on mathematical logic appeared as a monograph jointly published by Beijing University and Commercial Press in 1921 (Xu 2010: 185). The lecture notes by Wang Shiyi 王世毅 (no dates available) appeared in volume 3 of the short-lived The Mathematico-Physical Journal of Beijing University (Beijing Daxue Shuli Zazhi 北京大學數理雜誌, 1919–1921) under the title “An introductory guide to Russell’s Mathematical Philosophy” (Luosu de shuli zhexue daolun 羅 素的數理哲學導論). In 1922, Fu Zhongsun 傅種孫 (1898–1962) and Zhang Bangming 張邦銘 (no dates available) translated Russell’s Introduction to Mathematical Philosophy (Luosu suanli zhexue 羅素算理哲學).56 In 1930, when the work had won world-wide fame, it was republished, now translated as Suanli zhexue 算理哲學 (Song 2000: 12). The establishment of mathematical logic in China was promoted by people like Zhang Shenfu, Jin Yuelin, Wang Dianji, and others (Song 2000: 12; Sun 2001: 36).57 Zhang Shenfu had published a number of Russell’s texts on logic and mathematics before and after the latter’s visit to China. For many, Zhang’s translations and introductory articles on mathematical logic marked the first encounter with the new face of the discipline. In 192758, Wang Dianji published the first Chinese monograph touching upon mathematical logic, Luoji yu shuxue luoji lun 邏輯與數 學邏輯 論 (On logic and mathematical logic). It was the first introduction to logic written in Chinese to include more recent advances in mathematical logic (Song 2000: 12; Wang 2000: 3). Delineating the development of the discipline from Leibniz over de Morgan and Boole, Wang staged Russell as the one who had brought logic to perfection. In 1935 Jin Yuelin published his Luoji, a comprehensive intro-

 For an English overview on Russell’s role in the introduction of mathematical logic in China see Xu 2010. 56  According to Sun Zhongyuan (Song 2000: 36), the lecture notes of Russell were previously published by Beijing University in 1921. 57  Sun Zhongyuan (Sun 2001: 36) calls Jin the most representative propagator of “modern logic” in Republican China. 58  Sometimes 1926 is given as the year of publication. The book has no impressum, but the preface is dated to 1926 (Minguo 15). 55

494

R. Suter

duction to logic which included hundreds of axioms and proofs from Whitehead’s and Russell’s Principia Mathematica (Shi/Zeng 1998: 50; Song 2000: 12; Wang 2000: 3). Presenting students with the complete logical derivations, this book proved pivotal for the spread of mathematical logic in China. In the course of the 1930s, a series of other publications on mathematical logic in Chinese similarly presented basic knowledge about Russell’s propositional and predicate calculus as well as his calculus of classes (Shi/Zeng 1998: 50; Zeng 2000: 39–40). Zeng mentions Shen Youqian’s (Eugene Shen) 沈有乾 Xiandai luoji 現代 邏輯 (Modern logic, 1933), Wang Dianji’s Xiandai luoji (1934), Mou Zongsan’s Luoji dianfan 邏輯典範 (The norm of logic, 1941), and two articles by Zhu Yanjun 朱言均 (Gongjin 公謹, no dates available) (“Shuli luoji gangyao” 數理邏輯綱要 [Essentials of mathematical logic], 1934; “Shuli luoji daoyin” 數理邏輯導引 [Introduction to mathematical logic], 1936).59 Other early publications on mathematical logic in China were written by Zhang Shenfu and Xiao Wencan 肖文燦 (1917–1963)60 (Shi/Zeng 1998: 50). These latter works focused on the introduction of set theory, predicate logic, and the philosophy of logic (Song 2000: 12). However, most recent developments in the studies of mathematical logic abroad were not presented in any Chinese publication, and innovative research on mathematical logic was as rare as people understanding it (Zeng 2000: 39). Yet, as Eyal Aviv observes, Russellian logistics in spite of these obvious shortcomings still was “part of the zeitgeist” of the Republican period (Aviv 2015: 195). Only a small number of scholars engaged in proper research on mathematical logic. He Lin lists Jin Yuelin, Hu Shihua 胡世華, Shen Youqian, Shen Youding, Wan Zhuoheng 萬卓恒 (1902–1947),61 Wang Dianji, Wang Xianjun王憲鈞 (1910–1993), Yu Dawei 俞大維 (1897–1993)62, and Zhang Yinlin 張蔭麟 (1905–1942)63 (Shi/Zeng 1998: 50; Zeng 2000: 40, 2002: 37). Very few Chinese publications of the Republican period presented independent or even outstanding

59  These papers appeared in the philosophical quarterly of Wuhan University Wuhan Daxue Like Jikan 武漢大學理科季刊 5.2 and in Shuxue Zazhi 數學雜誌 (“Mathematical Revue”) 1.1 respectively (Shi/Zeng 1998: 50). 60  In 1932, Xiao Wencan had published a paper “The theory of irrational numbers” (Wulishu zhi lilun 無理數之理論, Wuhan Daxue Like Jikan 武漢大學理科季刊 2.4) (Shi/Zeng 1998: 50). 61  After graduating from Harvard, Wan Zhuoheng from 1936 to 1947 was professor for educational philosophy at Wuhan University where, among other things, he taught classes in “mathematical logic” on the basis of Whitehead’s and Russell’s Principia Mathematica. 62  Yu Dawei (Yu Ta-wei, David Yule) after his graduation from Shanghai St. John’s college in 1918 studied mathematical logic and Indic philology at Harvard. Among his teachers were Henry Maurice Sheffer and Clarence Irving Lewis (1883–1964). In 1921, he graduated with a PhD thesis entitled “Theories of abstract implication”. In the same year, he received a Sheldon Traveling Fellowship and went to Berlin. In 1926, he published an article “On the foundation of the calculus of classes” (Yule David, “Zur Grundlegung des Klassenkalküls”, Mathematische Annalen 95, 446–452) in German. He later served as a lieutenant general in the Republican Army. 63  Zhang Yinlin had graduated from Qinghua University in 1929. During his ensuing stay at Stanford, he took classes in mathematical logic. His only published text about logic appears to be a short article about ‘inductive logic’.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

495

research. Exceptions include the mathematician Tang Zaozhen 湯藻真 (Tang Tsao-­ Chen, 1893–1951)64 who was able to show that every provable theorem in Lewis’s calculus of strict implication corresponds to a topological identity,65 and Hu Shihua who designed an infinite-valued propositional calculus (Zeng 2000: 40; on the beginnings of mathematical logic in China see also Xiang 1992: 63–64). Zhang Shenfu (style Songnian 崧年) was one of the pioneers of mathematical logic in China and probably the first Chinese specialist on Russell’s philosophy. Although he taught mathematical logic at Beijing and Qinghua universities (Yan 2007: 51), his published translations of philosophical works by Russell included only a few texts about logic (Shi/Zeng 1998: 52). Nonetheless, his publications have doubtless accelerated the spread of mathematical logic in China. Zhang moreover has the credit to have been the first translator of Ludwig Wittgenstein’s (1889–1951) Tractatus Logico-Philosophicus from Ogden’s authorized English version in 1922. It is thus his merit that Chinese was the third language in which the Tractatus was available. Another pivotal figure for the propagation of mathematical logic in Republican China was Jin Yuelin. Having graduated from Qinghua University in 1914 (Sun 2001: 36), he received a PhD in political science from Columbia University in 1920 (Song 2000: 12; Sun 2001: 36). During his studies at London University in 1921,66 he became familiar with the work of Russell (Song 2000: 12; Sun 2001: 36).67 In 1926, he was nominated professor at Qinghua University,68 offering courses in logic and philosophy. Teaching Russell’s mathematical logic since 1927, he had published a number of introductory articles in journals like the Qinghua Xuebao 清華 學報 (Tsing Hua Journal) or Zhexue pinglun 哲學評論 (Philosophical Review) after 1933 (Zeng 2002: 37). In 1935, he eventually gave out his above-mentioned monograph on logic laconically titled Luoji (Logic). Designed as a textbook in logic, this work proved extremely influential in the following years (Le 1988: 48; Song 2000: 12; Sun 2001: 37). He Lin attributes it a qualitative standard previously

 Tang had published, in the 1930s, a number of English papers, among them one on the “The Nine Circle Theorem and the Enlarged Geometry” in The American Mathematical Monthly 45.7, 1938, 430–433. 65  The Chinese sources refer to an article entitled “代數公設和路易士嚴格蘊含演算的一個幾何 解釋”, literally “An Algebraic Postulate and a Geometrical Explanation of Lewis’s Calculus of Strict Implication”, but they fail to provide complete bibliographical references. However, Tang published a paper with the title “Algebraic Postulates and a Geometrical Interpretation of the Lewis Calculus of Strict Implication” in the Bulletin of the American Mathematical Society in 1938, where he had already published a paper on a paradox in Lewis’s calculus of strict implication two years before (Tang 1936). It is well possible that these English publications were elaborating on the above-mentioned Chinese article. 66  Sun Zhongyuan (Sun 2001: 36) gives 1922. 67  Sun Zhongyuan (Sun 2001: 36) quotes the introduction to Jin Yuelin’s Lun dao (“On Dao”, 1940), published in 1940, where Jin mentions the crucial role which the philosophy of Russell’s and Whitehead’s Principia Mathematica played for his understanding of philosophy – identifying himself with an analytical style of philosophy. 68  Sun Zhongyuan (Sun 2001: 37) gives 1927. 64

496

R. Suter

unknown in Chinese introductions to the topic (Le 1988: 48; Sun 2001: 37). The logician Mo Shaokui 莫紹揆 (1917–2011) and the philosopher and historian of philosophy Feng Qi share this assessment (Le 1988: 50). In Song Wenjian’s 宋文 堅 view, Jin’s Luoji was at the edge of the discipline’s international development and thus contributed to the comparatively high level of the instruction of logic at Chinese universities in the Republican period (Song 2000: 12; cf. also Shi/Zeng 1998: 50). In characteristic modesty, Jin called himself a layman without a firm standing in the discipline in the book’s preface (Shi/Zeng 1998: 52). Emphasizing that traditional logic and mathematical logic belong to the same field, Jin presented the latter as an organic and necessary development out of the former (Song 2000: 12; Sun 2001: 37). Later logicians like Shen Youding, Wang Xianjun, Wang Hao 王浩 (1921–1995) and Zhou Liquan 周禮全 (1921–2008) were among Jin’s students at Qinghua University.69 He also discovered the extraordinary talent of Yin Haiguang whom he invited to Beijing and who later in 1938 joined the National Southwestern Associated University (Xinan Lianhe Daxue 西南聯合大學) in Kunming to become one of his students.70 According to Song Wenjian, it was under Jin’s influence that a number of these young scholars graduating from Qinghua University, but also Hu Shihua and Mo Shaokui from Beijing University, studied in the United States, Germany, Austria, and Switzerland under the guidance of internationally renowned specialists in mathematical logic like Henry Maurice Sheffer (1882–1964), Alfred North Whitehead, Kurt Gödel (1906–1978) and Paul Bernays (1888–1977) (Song 2000: 12–13). Upon their return to China, representatives of this second generation of Chinese mathematical logicians contributed to the further development and consolidation of the field (Song 2000: 13). Jin’s student Wang Hao who wrote his dissertation under the supervision of Willard Van Orman Quine (1908–2000) stayed abroad and became a leading figure in mathematical logic and computer science (among other things working for IBM), publishing cutting edge research in many areas of the field (see Grossi et al. 2012). Sun Zhongyuan credits Jin for his innovative treatment of the law of identity. In contrast to Russell who had conceived it as pertaining to individuals, Jin distinguished identity with respect to different levels such as predicates, concepts, relations, and universals, defining identity as either self-implication or self-equivalence (Sun 2001: 37). As he identified logical systems as representing what is necessary rather than simply possible, Jin regarded modality as the central problem addressed in logical research. Holding a monistic view, he interpreted different logical systems as different tools for explicating the same normative foundation. He refuted the idea of essentially distinct logics in the plural: logical systems can be said to be particular and manifold, logic however is in itself both essentially general and unitary (Sun 2001, 37). Jin also worked on the problem of axiomatization: He established three criteria for axioms of logical systems: (1) sufficiency (gouyong 夠用); (2)

69 70

 He Lin, Dangdai Zhongguo zhexue; cit. in Sun 2001: 37.  He 2013: 13.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

497

­independence (duli 獨立); (3) consistency (yizhi 一致). (1) Sufficiency requires that the set of axioms allows for deriving the required theorems. The sufficiency of axioms thus corresponds to the completeness of a logical system. (2) Axioms have to be independent of each other. They must not imply each other. Independence thus means non-derivability. It guarantees that no axiom can be inferred from another one by means of the defined rules of derivation of the system. (3) Consistency requires that the axioms must not be contradictory to each other. According to Jin Yuelin, completeness, independence and consistency of a logical system point toward the meta-­level of a fundamental logic mentioned above which lies beyond the domain of any particular logical system and hence cannot be proved within any of these (Sun 2001, 37). In 1937, still as a high school student, Yin Haiguang, the later liberal critic of both Maoist totalitarianism and Guomindang authoritarianism, published a translation of Champan and Henle’s The Fundamentals of Logic.71 His introduction, entitled “What, after all, are logic and the science of logic?” (“Luoji yu luojixue daodi shi shenme? 邏輯與邏輯學到底是甚麼”), was published as a separate paper on the recommendation of Jin Yuelin in the same year. In 1938 he studied at the department of philosophy of the National Southwestern Associated University in Kunming where he also took courses in logic with Jin. Five years later, he published a monograph entitled Lectures on Logic (Luoji jianghua 邏輯講話) – a collection of articles part of which he had previously published independently.72 In 1956, a revised version of this book appeared under the title A New Introduction into Logic (Luoji xin yin 邏輯新引) in Hong Kong.73 It was to become one of the standard textbooks on contemporary logic on Taiwan. After the end of World War II, Yin became professor for logic and philosophy at the Jinling University in Nanjing, before he moved to Taiwan in 1949. Still on the mainland, he published a translation of Rudolf Carnap’s Philosophy and Logical Syntax in 1946.74 In the department of philosophy at the National Taiwan University he taught classes on logic, logical empiricism, logical semantics, philosophy of science, but also on the philosophy of Bertrand Russell.75 In the course of the 1950s he translated half a dozen of publications related to problems of logic in the broad sense.76 In his last years, the focus of his studies changed toward intellectual history until his radical criticism of alleged auhoritarian tendencies in traditional Chinese thinking eventually cost him his post at the university and exacerbated the health problems that led to his early death in 1969.77

 The book was published in Nanjing by Zhengshu Shuju (He 2013: 265).  He 2013: 13. 73  Republished as part of his collected works in 2011 (Yin 2011a). 74  The book came out at Commercial Press in Chongqing (Yin 2011b: 171). 75  He 2013: 265. 76  Cf. Yin 2011b. 77  He 2013: 267. 71 72

498

R. Suter

Hu Shihua 胡世華 (Hu, Shih-hua, also Hu Zihua 胡子華, Hoo Tsu-Hua) (1912–2011)78 studied philosophy at Beijing University, where he graduated and began to teach mathematics in 1935. From 1936 until 1940 he went to Europe where he enrolled at Vienna University, the Westfälische Wilhelms-Universität in Münster, as well as in Switzerland and France where he studied and researched on mathematical logic and the foundation of mathematics. He received a PhD from the University of Münster in 1939 with a dissertation (1939) about Topological Foundations of Pseudo-Boolean Algebra (supervisors Heinrich Scholz [1884–1956] and Gottfried Köthe [1905–1989]) (Lu et al. 2006). After his return to China in 1941, he became Associate Professor at the Institute of Mathematical Astronomy of Zhongshan University in Guangzhou until 1943. From 1943 till 1946 he was professor for mathematical logic and the foundations of mathematics at the Department of Philosophy at Zhongyang University (Zhongyang Daxue 中央大學) in Chongqing. From 1946 onwards, he was professor for these disciplines at Beijing University. Hu later played a crucial role in establishing computer science in China (Lu et al. 2006; Sun 2001: 38). His topological dissertation remained unpublished, but a summary appeared in Xueshu jikan 學術季刊 (“The Academic Quarterly”). In 1943, Hu published a first paper on artificial language (“Lun renzao de yuyan” 論人造的語 言, Xueshu jikan 學術季刊 1.3), 2 years later an article on the logical calculus (“The new system and logical constants of yet another calculus [“Zai xian suanshu xin xitong ji qi luoji changci” 再現算數新系統及其邏輯常詞], Xueyuan 15) (Lu et al. 2006). At the end of the 1940s, his research on multivalued logic and its applications in mathematics found broad international attention. In 1949, he published an English article on an “m-valued subsystem of (m+n)-valued propositional calculus” (The Journal of Symbolic Logic 14.3), in 1950 another paper on “The construction of an Aleph-zero-valued propositional calculus”, and in 1951 a joint paper with Chen Qiangye 陳強業 (no dates available) on “A four-valued propositional calculus and the four color problem” (Lu et al. 2006). Shen Youqian 沈有乾 (Eugene Shen) (1900–1996) was the older brother of the much more prominent Shen Youding (Gao 2008: 94). In 1922, he had received a Boxer Rebellion Indemnity scholarship and studied experimental psychology, statistics and mathematical logic at Stanford, Harvard and Columbia universities. At Harvard, he was a student of Henry Maurice Sheffer (1883–1964) (Gao 2008: 97). In 1926, he received a PhD degree in philosophy from Stanford. He published in English and American journals already as a student (Gao 2008: 94). After his return to China, he became a professor of logic and psychology at Shanghai Guanghua University (光華大學 Guanghua Daxue). He then changed to the National Zhejiang University, where he taught pedagogy. After accepting a post as a secretary at the United Nations in 1948 he moved to New York, quitting all teaching duties in China. After his retirement in 1975, he wrote a primer in logic which was later-on published in Taiwan (Gao 2008: 95). Shen Youqian’s publications on logic in Chinese include his monograph Xiandai luoji (Modern logic, 1933), as well as three

78

 For a short overview see, e.g. Sun 2001: 38–39.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

499

t­extbooks on logic, among them Lunlixue (Logic, 1936,21958; English review by Wang Hao in 1948; cf. Grossi et  al. 2012: 2), and the above-mentioned primer Chuji lizexue gangyao (An introduction to basic logic, 1975) (Gao 2008: 96). During his studies in the United States he had published an article on the antilogism of his teacher Christine Ladd-Franklin (1847–1930) under his English name Eugene Shen (Shen 1927). Shen Youding (Shen Yuting, style Gongwu 公武) (1909–1989) was Shen Youqian’s younger brother. After developing an interest for logic already in high school, he studied at Qinghua University in Beijing as a student of Jin Yuelin. Upon graduation from the department of philosophy at Qinghua University in 1929 (Sun 2001: 37; Zhang 2008: 108), he was granted a scholarship for the United States in the same year. In 1931, he received an M.A. degree from Harvard (Zhang 2008: 108). From 1931 through 1934 he studied in Germany, from 1945 until 1948 he was at Oxford (Sun 2001: 37). After his return to China, he was first professor at Qinghua University, then at Beijing University. His English publications on mathematical logic won him international fame. In the early 1950s, he published on the problem of paradoxes (“Paradox of the class of all grounded classes”, 1953; “Two semantical paradoxes”, 1955) (Sun 2001: 37; Zhang 2008: 108). Shen’s “elementary calculus”, presented in 1957, is said to have been the first calculus developed by a Chinese scholar (Sun 2001: 38). The “paradox of the class of all grounded classes” is known in English publications as the “Shen Yuting paradox” (Chin. Shen shi beilun 沈氏悖 論) (Zhang 2008: 108). In China, he is moreover renowned for his later research on Mohist logic (Zhang 2008: 108). Wang Xianjun 王憲鈞 (1910–1993) graduated in philosophy at Qinghua University in 1933 and continued his graduate studies there until, in 1936, he moved to Europe for research in Austria (Vienna ) and Germany (Münster). In 1938, he returned to China and became a lecturer at the National Southwestern Associated University, professor of philosophy at Qinghua University and Beijing University, acting dean of the science department, and head of the Research Institute for Logic (Kang 2013). Among his early works is a paper on implication (“Lun hanyun” 論含蘊) which appeared in the Tsing Hua Journal in 1940 and an article on semantic necessity which came out in 1944 in Zhexue Pinglun (Philosophical Review). In the 1950s he published on logical semantics and its relation to propositional truth (Wang 1956, 1959). After graduating in mathematics from the Southwestern Associated University in 1943, Wang Hao entered Qinghua University where he studied at the philosophy department that had been established by JIN Yuelin. Two years later he earned a Master’s degree with a study “On the basis of empirical knowledge” (Lun jingyan zhishi de jichu 論經驗知識的基礎). During his undergraduate studies he heard lectures in formal logic by Wang Xianjun and took classes on Wittgenstein’s Tractatus taught by Shen Youding. As a student in Kunming he autodidactically studied the proofs from the first volume of Principia Mathematica contained in Jin Yuelin’s Logic, as well as David Hilbert (1862–1943) and Wilhelm Ackermann’s (1892–1962) Grundzüge der theoretischen Logik (Principles of mathematical logic), David Hilbert and Paul Bernays’s (1888–1977) Grundlagen der Mathematik (Foundations

500

R. Suter

of mathematics), and Rudolf Carnap’s (1891–1970) The logical syntax of language. In 1946 he went to Harvard where he wrote his doctoral dissertation under the supervision of Quine. Wang’s first publications dealing with the relationship of metaphysics and language were in Chinese and appeared in Zhexue Pinglun in 1944 and 1945. Soon after his arrival in the United States he started to publish in internationally renowned journals like the Journal of Symbolic Logic or Mind. Between 1947 and 1950, he issued around a dozen articles about a wide range of topics like induction, quantification, axiomatization of set theory, etc. (see Grossi et al. 2012: 1–2). Wang, who was to stay in the United States, became one of the leading scholars in formal logics of his generation, publishing on topics as diverse as set theory, number theory, the theory of induction, Turing machines, the mechanization of mathematical proofs, and, in his later years, the philosophy of Kurt Gödel (1906–1978) and the late Ludwig Wittgenstein (1889–1951). Informed by his reading of Wittgenstein and Gödel, he remained critical of what he considered as reductionist tendencies in contemporary analytic philosophy (Floyd 2012). The present chapter is but a short survey on the establishment and development of logic as an academic discipline in China during the first half of  the twentieth century. Although space only allows for a very limited and highly selective account, it is hopefully still enough to illustrate some of the dynamics that pushed the “naturalization” of logic ahead and which also soon led to the general acceptance of the idea of an indigenous Chinese logical tradition. Well into the Republican period, intellectual debates about logic were more often than not led by extrinsic ideological motives rather than genuinely logical concerns. However, in sharp contrast with this rather ambivalent picture, the present chapter likewise documents the remarkable fact that it took less than two generations from the first serious but still cautious engagement of Chinese scholars with the exotic subject of logic to its establishment in an institutional academic setting which trained people, some of whom soon were among the world’s leading experts in the field.

References Ai, Siqi 艾思奇. 1991. Collected Works of A i Siqi 艾思奇文集, vol. 1. Beijing: Renmin Chubanshe. Aviv, Eyal. 2015. “A Well-Reasoned Dharma: Buddhist Logic in Republican China.” Journal of Chinese Buddhist Studies 28.2015: 189–234. Ayzenberg, A., I. Shirokov, and R. Yankovskiy. 1932. Materialist Dialectics (Materialesticheskaya dialektika: posobie dlya komvuzov i sotsial.-ekon. vuzov). Moskva, Leningrad: Partizdat ZK VKP (b). Chen, Daode 陳道德. 2007. The Development of Semantics in the 20th century and the Rise of Language Logic 二十世紀意義理論的發展與語言邏輯的興起. Beijing: Zhongguo Shehui Kexue Chubanshe. Chen, Daqi 陳大齊. 1958. Collected Essays on Confucius Studies 孔子學說論集. Taibei: Zhengzhong Shuju 正中書局. Chen, Daqi 陳大齊. 1970. Collected Commentaries on the Analects 論語集釋. Taibei: Shangwu Yinshuguan.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

501

Cheng, Zhongtang 程仲棠. 2006. “A Critique of Zhang Dongsun’s Culturalist View on Logic 評 張東蓀的文化主義邏輯觀.” History of Chinese Philosophy 中國哲學史 2006.3: 117–124. Deng, Wenchu 鄧文初. 2007. “Shen Youding and Shen Youqian. Logic, a Chinese Story 沈有鼎與 沈有乾——邏輯, 中國的故事.” Authors and Books 書人書事 2007.9: 94–99. Dewey, John. 1910. How We Think. Lexington, MA.: D.C. Heath. Ding, Dongwen 丁東文. 2009. “One Hundred Years of Kantian Studies in China 百年康德哲學研 究在中國.” World Philosophy 世界哲學 2009.4: 32–42. Dong, Zhitie 董志鐵. 1995. “Twentieth Century Research on [the] Chinese [Theory of] Names and Disputation (Logic) 二十世紀中國名辯(邏輯)研究.” History of Chinese Philosophy 中 國哲學史 1.1995: 111–117. Dong, Zhitie 董志鐵. 2003. “A Bitter Fate – Traditional Logic in China 命運多舛——傳統邏輯 在中國.” Journal of Beijing Normal University 北京師範大學學報 (Social Sciences Edition 社會科學版) 177.3: 99–105. Dumitriu, Anton. 1977. History of Logic, 4 vols. Tunbridge, Wells, Kent: Abacus Press. Engels, Frederick [Friedrich]. 1940. Dialectics of Nature (trans. and annot. Dutt, C.P., and J.B.S. Halsdane). New York: International Publishers. Floyd, Juliet. 2012. “Wang and Wittgenstein.” In Parsons, C. and M.  Link, eds., Hao Wang, Logician and Philosopher (143–190). London: College Publications, . Fogel, Joshua A. 1987. Ai Ssu-ch’i’s Contribution to the Development of Chinese Marxism. Cambridge, MA.: Harvard Contemporary China Series, no. 4. Frankenhauser, Uwe. 1996. “Logik und nationales Selbstverständnis in China zu Beginn des 20. Jahrhunderts.” In Hammer, Christiane and Bernhard Führer, eds., Chinesisches Selbstverständnis und kulturelle Identität – “Wenhua Zhongguo” (69–80). Dortmund: Projekt-Verlag. Gao, Shanshan 高山杉. 2006. “The Intellectual Background of Yu Dawei’s Studies in Mathematical Logic and Sanskrit – an Academic Fate in the “Golden Age” of the Philosophy Department at Harvard University 俞大維學習數理邏輯和梵文的學問背景——哈佛大學哲學系“黃金時 代”一段學術因緣.” World Philosophy 世界哲學 2006.5: 61–67. Gao, Shanshan 高山杉. 2008. “Sheffer and Shen Youqian 謝佛和沈有乾.” Eastern Philosophy 東 方哲學 2008.2: 94–101. Graham, Angus Charles. 1978. Later Mohist Logic, Ethics and Science. Hong Kong: Chinese University Press. Grossi, Marie, Montgomery Link, Katalin Makkai, and Charles Parsons. 2012. “A Bibliography of Hao Wang.” Philosophia Mathematica (III) 2012: 1–17. Guo, Qiao 郭橋. 2006. Logic and Culture: Studies on the Spread of Western Logic in China in the Modern Era 邏輯與文化: 中國近代時期西方邏輯傳播研究. Beijing: Renmin Chubanshe. Guo, Zhanbo 郭湛波. 1932. History of Chinese Logic (Studies on Disputation) 中國辯學史. Shanghai: Zhonghua Shuju. Guo, Xiuyun 郭秀云. 2013. “An Overview on Zhang Shenfu’s Studies 張申甫研究綜述.” Sociological Dynamics in Higher Education 高教社科動態 2013.5: 34–38. He, Lin 賀麟. 1959. “Philosophy in Contemporary China 當代中國哲學.” Materials for the Critical Investigation of the Academic Thought of the Capitalist Class 資產階級學術思想批 判參考資料, vol. 4. Beijing: Shangwu Yinshuguan. He, Lin 賀麟. 1990. Collected Essays on Philosophy and the History of Philosophy 哲學與哲學 史論文集. Beijing: Shangwu Yinshuguan. He, Lin 賀麟. 2005. “The Propagation of the Philosophies of Kant and Hegel in China – With a Personal Review on the Introduction of Kant and Hegel 康德、黑格爾哲學在中國的傳 播――兼論我對介紹康德黑格爾哲學的回顧.” In Zhang, Xuezhi 張學智, ed., Selected Works by He Lin 賀麟選集 (424–461). Changchun: Jilin Renmin Chubanshe.” He, Zhuo’en 何卓恩. 2013. The Heritage of Liberalism: Political, Economical and Cultural Studies of Yin Haiguang, Fu Daoping, Xu Fuguan 自由主義的新遺產:殷海光、復道平、徐 傅觀政治經濟文化論說. Bejing: Jiuzhou Press 九州. Heige’er 黑格爾 (= Hegel, Georg Wilhelm Friedrich). 1980. Shorter Logic 小邏輯 (trans. He Lin 賀麟). Beijing: Renmin Chubanshe.

502

R. Suter

Hegel, Georg Wilhelm Friedrich. 1991. The Encyclopaedia Logic (with the Zusätze), Part I of the Encyclopaedia of the Philosophical Sciences with the Zusätze (trans. Geraets, T. F.; Suchting, W. A.; Harris, H. S.). Indianapolis, Cambridge: Hackett Publishing Company. Hu, Shi 胡適. 1919. An Outline of the History of Chinese Philosophy 中國哲學史大綱. Shanghai: Shanghai Yinshuguan. Hu, Shih. 1922. The Development of the Logical Method in Ancient China. Shanghai: Yadong Tushuguan (The Oriental Book Co.). Hu, Shih-hua. 1949. “m-valued Subsystem of (m+n)-valued Propositional Calculus.” The Journal of Symbolic Logic 14.3: 177–181. Hu, Weixiong 胡為雄. 2014. “A Cursory Examination on Shen Zhiyuan and the Edition of China’s First Marxist Philosophical Dictionary 沈志遠與中國首部馬克思主義哲學辭典之編著考 略.” Philosophical Trends 哲學動態 2014.1: 31–36. Ishchenko, Timofey Semenovich. 1930. Short Philosophical Dictionary (Kratkiy filosofskiy slovar’). Moskva: Izdatel’stvo kommunisticheskogo universiteta im. Ya. M. Sverdlova. Jin, Xiping 靳希平. 2004. “Shen Youding and Husserl’s Phenomenology 沈有鼎與胡塞爾的現象 學.” Philosophy and Culture 哲學與文化 2004: 12–23. Jin, Yuelin 金岳霖. 1936. Logic 邏輯. Beiping: Shangwu Yinshuguan 商務印書館. Second edition. Jin, Yuelin 金岳霖. 1998. “On Different Logics 論不同的邏輯.” In Jin Yuelin Xueshu Jijinhui Xueshu Weiyuanhui, ed., Jin Yuelin wenji 金岳霖文集, vol. 2 (378–402). Lanzhou: Gansu Renmin Chubanshe (First publication: Tsing Hua Journal 清華學報 13.1, 1941). Kang, Hongkui 康宏逵. 2013. “My Teacher Xianjun 吾師憲鈞.” Scientific and Cultural Critique 科學文化評論 10.5: 113–126. Knight, Nick, ed. and trans. 1990. Mao Zedong on Dialectical Materialism. Armonk, NY: M.E. Sharpe. Knight, Nick. 1996. Li Da and Marxist Philosophy in China. Boulder, CO: Westview Press. Kurtz, Joachim. 2011. The Discovery of Chinese Logic (Modern Chinese Philosophy, vol. 1, series ed. John Makeham). Leiden/Boston: Brill. Le, Yiou 樂逸鷗. 1988. “Fifty Years of Research on Jin Yuelin’s Logic 金岳霖邏輯思想研究50 年.” Philosophical Trends 哲學動態 6.7 (Research Materials 研究資料): 48–51. Lenin, Vladimir Ilyich. 1976. “Plan of Hegel’s Dialectics (Logic) (1915).” In: Lenin, Collected Works, 4th ed., vol. 38. (trans. Dutt, C.P. 315–318). Lenin, Vladimir Il’ich. 1958. “Plan dialektiki (logiki) Gegelya.” In: Polnoe sobranie sochineniy, 5th ed., vol. 29 (297–302). Moskva: Izdatel’stvo Politicheskoy Literatury. Li, Chunyong 李春勇. 2004. “Achievements and Shortcomings of Logical Studies in Twentieth Century China 二十世紀中國邏輯學的成就與不足.” Journal of Shanghai Administration Institute 上海行政學院學報 5.6: 19–25. Li, Chunyong 李春勇. 2004a. “The Evolution of Twentieth Century Chinese Scholars’s View on Logic” 二十世紀中國學者邏輯觀的演進. Journal of East China Normal University 華東師 範大學學報 (Philosophy and Social Sciences 哲學社會科學版) 36.4: 109–114. Li, Kuangwu 李匡武. 1989. History of Chinese Logic 中國邏輯史, 5 vols. Lanzhou: Gansu Renmin Chubanshe. Lin, Xiashui 林夏水, and Zhang, Shangshui 張尚水. 1983. “Mathematical Logic in China 數理邏 輯在中國.” Studies in the History of Natural Sciences 自然科學史研究 2.2: 175–182. Lippert, Wolfgang. 1979. Entstehung und Funktion einiger chinesischer marxistischer Termini: Der lexikalisch-begriffliche Aspekt der Rezeption des Marxismus in Japan und China (Münchener Ostasiatische Studien, vol. 19). Wiesbaden: F. Steiner. Liu, Bangfan 劉邦凡. 2007. “The Renaissance of Chinese Logic in the Modern Period and Its Future Development 中國邏輯的近代復興與未來發展.” Journal of Yanshan University 燕山 大學學報 (Philosophy and Social Sciences 哲學社會科學版) 8.1: 1–7. Liu, Peiyu 劉培育. 2001. “Comparative Research on the Chinese Theory of Naming and Disputation (= Chinese Logic), (Western) Logic and Buddhist hetuvidyā in the Twentieth Century 20世紀名辯與邏輯、因明的比較研究.” Journal of Social Sciences and Philosophy 人文及社會科學集刊 (Philosophical Research 哲學研究) 2001.3: 14–18.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

503

Liu, Qi 劉奇. 1941. Ancient Examples of Logic 論理古例. Chongqing: Longmen Shudian 龍門 書店. Liu, Zongtang 劉宗棠. 1986. “Wang Yanzhi and the Study of Logic 王延直先生與邏輯學.” Research Materials on the Local Gazeteer of Guiyang 貴陽志資料研究 10. Lu, Yi 盧毅. 2007. “The ‘materialist dialectics fever’ of the Thirites of the Twentieth Century 20世 紀30年代的“唯物辯證法熱””. CPC History Research and Teaching 黨史研究與教學 196.3: 45–53. Lu, Zhongwan 陸鍾萬, Wang, Shiqiang 王世強, Xu, Kongshi 許孔時, and Yang, Dongping 楊東 屏. 2006. “Hu Shihua: The Founder of Mathematical Logic in China 胡世華: 中國數學邏輯 奠基人.” Guangming wang 光明網: http://news.ustc.edu.cn/kdr/200611/t20061127_109941. html (04-08-2015) Ma, Qiuli 馬秋麗. 2007. “A Critique of Zhang Dongsun’s View on Logic 張東蓀邏輯觀述評.” Theoretical Exploration 哲學問題——理論探索 168.6: 50–53. Ma, Yanbo 馬彥波. 2009. “On Jin Yuelin’s Contributions to Chinese Traditional Logic from the 1930s 論20世紀30年代金岳霖對我國傳統邏輯學貢獻.” Journal of Liaoning Education Administration Institute 遼寧教育行政學院學報 26.1: 17–19. Ma, Yongkang 馬永康. 2002. “Discussion of the term ‘dialectic’: focussing on He Lin and Zhang Dongsun “dialectic” 譯名討論——以賀麟、張東蓀為中心.” World Philosophy 世界哲學 2002.1: 75–79. Makeham, John. 1992. Name and Actuality in Early Chinese Thought. Albany, NY: State University of New York Press. Makeham, John. 2012. “The Role of Masters Studies in the Early Formation of Chinese Philosophy.” In Makeham, John, ed., Learning to Emulate the Wise – the Genesis of Chinese Philosophy as an Academic Discipline in Twentieth-Century China (73–101). Hong Kong: The Chinese University Press. Makeham, John, ed. 2014. Transforming Consciousness: Yogācāra Thought in Modern China. Oxford: Oxford University Press. Masini, Federico. 1993. The Formation of Modern Chinese Lexicon and its Evolution toward a National Language: The Period from 1840 to 1898 (Journal of Chinese Linguistics. Monograph series, vol. 6). Berkeley, CA: University of California. Miller Chapman, Frank, and Paul Henle. 1933. The Fundamentals of Logic. New York/Chicago.: Charles Scriber’s Son. Mou, Zongsan 牟宗三. 1934. “Logic and Dialectical Logic 邏輯與辯證邏輯.” In Zhang, Dongsun 張東蓀, ed. Debates on Materialist Dialectics 唯物辯證法論戰. Beijing: Minquan Shuju 民 權書局. Mou, Zongsan 牟宗三. 2003. The Norm of Logic 邏輯典範 (Complete Works of Mou Zongsan 牟宗三先生全集, vol. 11). Taibei: Lianjing. (First publication: Hong Kong: Shangwu Yinshuguan, 1941). Moh, Shaw-Kwei (Mo Shaokui 莫紹揆). 1954. “Logical Paradoxes for Many-Valued Systems.” Journal of Symbolic Logic 19: 37–39. Peng, Yilian 彭漪漣. 1991. On the Intellectual History of Logic in China 中國邏輯思想史論. Shanghai: Renmin Chubanshe. Peng, Yilian 彭漪漣. 2006. On the Intellectual History of Logic in China in the Modern Period 中 國近代邏輯思想史論. Shanghai: Shanghai Renmin Chubanshe. Prior, Arthur Norman. 1955. Formal Logic. Oxford: Oxford University Press. Reymond, Arnold. 1932. Les Principes de la logique et la critique contemporaine. Paris: Boivin. Reymond, Arnold (= Lei Meng 雷孟). 1975. The principles of logic and its critique by contemporary schools of thought 邏輯原理及現代各派之評述. He Zhaoqing, tr. Taibei: Shangwu Yinshuguan. First edition 1936. Robinson, Abraham. 1961. “Non-standard analysis.” Indagationes Mathematicae 23: 432–440. Rozental’, M., and P.F.  Yudin. 1939. Short Philosophical Dictionary [Kratkiy filosofskiy slovar’]. Moskva: Ob”edinenie gosudarstvennykh knizhno-zhurnal’nykh izdatel’stv (OGIZ), Gosudarstvennoe izdatel’stvo politicheskoy literatury.

504

R. Suter

Russell, Bertrand. 1921 Mathematical Logic 數理邏輯 (trans. and ed. Wu Fanhuan 吳範寰). Beijing: Beijing Daxue Xinzhi Shushe 北京大學新知書社. Russinoff, I. Susan. 1999. “The Syllogism’s Final Solution.” The Bulletin of Symbolic Logic 5: 451–469. Schulz Zinda, Yvonne. 2001. “Li” and “Shi” as an Ontological Solution of Hume’s Problem of Causality within the Problem of Induction in Jin Yuelin’s Work Lun Dao [„Li“ und „Shi“ als ontologische Antwort auf das Humesche Kausalitätsproblem innerhalb des Induktionsproblems in Jin Yuelins Werk Lun Dao]. Doctoral dissertation, University of Hamburg. Schulz Zinda, Yvonne. 2012. Jin Yuelin’s Ontology: Perspectives on the Problem of Induction. Boston/Leiden: Brill. Sha, Qing 沙青, Xiaoyan Zhang 張小燕, and Yanjing Zhang 張燕京. 2002. The Disruption of Analytical and Dialectical Reason: A Historical Reflection on the Development of Twentieth Century Debates on Logical Thinking in China 分析性理性與辯證理性的裂變——二十世 紀中國邏輯思想論爭的歷史反思. Baoding 保定: Hebei Daxue Chubanshe. Shen, Eugene. 1927. “The Ladd-Franklin formula in logic: The antilogism.” Mind 37: 54–60. Shen, Jiru 沈驥如. 1985. “Shen Zhiyuan’s Studies on Hegel in the 1930s – A Complement of “A Critique of Chinese Studies on Hegel” 沈志遠三十年代對黑格爾的研究——對《我國的黑 格爾研究評述》一文的補充.” Fudan Journal 復旦學報 (Social Sciences Edition 社會科學 版) 1985.4: 81–82. Shen, Youding 沈有鼎. 1957. “A Primitive Calculus 初基演算.” Acta Mathematica Sinica 數學 學報 7.1: 132–143. Shen, Youding 沈有鼎. 1992. Collected Works of Shen Youding 沈有鼎文集. Beijing: Renmin Chubanshe. Shen, Youqian 沈有乾. 1933. Modern Logic 現代邏輯. Shanghai: Xinwen Shuju 新聞書局 (or: Xinyue Shudian 新月書店). (Gao 2008: 96). Shen, Youqian 沈有乾. 1936. Logic 論理學 (Logic for Normal and Local Normal Schools 師範學 校及鄉村師範學校論理學). Nanjing: Zhengzhong Shuju 正中書局. (Gao 2008: 96). Shen, Youqian 沈有乾. 1975. An Introduction to Basic Logic 初級理則學綱要. Taibei: Shangwu Yinshuguan. Shen, Youqian 沈有乾. 1986. “Remembering Six American Teachers 懷念六位美國業師.” Biography Literature 傳記文學 49.1: 109–116. Shi, Mingde 時明德. 1997. “The Logical View in Mao Zedong’s Philosophical Notes 毛澤東哲學 批註中的邏輯觀.” Mao Zedong Thought Study 毛澤東思想研究 1997.4: 6–8, 17. Shi, Mingde 時明德, and Zhaoshi Zeng 曾昭式. 1998. “An Enquiry into the Reasons of the Stagnation of the Development of Mathematical Logic in China 數理邏輯在中國發展滯緩 的原因探析.” Journal of Xinyang Normal University 信陽師範學院學報 (Philosophy and Social Sciences Edition 哲學社會科學版) 18.2: 49–55. Song, Chunwei 宋春偉. 2014. “One Century of the Mathematics Department – One Generation of Members of the Democratic League 百年數學院——世代民盟人.” Popular Tribune 群言 2014.2. (http://www.dem-league.org.cn/qunyan/shouye/whmb/40390.aspx) Song, Wenjian 宋文堅. 2000. “Eighty Years of Chinese Mathematical Logic 中國數理邏輯八十 年.” Journal of Beijing University of Aeronautics and Astronautics 北京航空航天大學學報 13.1: 12–17. Standaert, Nicolas. 1994. “The Investigation of Things and the Fathoming of Principles (Gewu Qiongli) in the Seventeenth-Century Contact between Jesuits and Chinese Scholars.” In Witek, John W., SJ, ed. Ferdinand Verbiest, S.J. (1623–1688) (395–420). Nettetal: Steyler. Strawson, Peter Frederick. 1950, “On referring.” Mind 59: 320–344. Su, Yue 蘇岳. 2008. “Wang Yanzhi’s ‘General Applied Logic’ – the Earliest Logical Work in the Yunnan and Guizhou Region 王延直的《普通應用論理學》—— 雲貴地區發現的最早邏 輯專著”. Law and Society 法制與社会 2008.18: 270–271. Sun, Zhongyuan 孫中原. 2000. “An Outline of One Hundred Years of Research on the History of Chinese Logic 中國邏輯史研究百年玄覽.” Guang Ming Daily 光明日報 7, 06/06/2000. Sun, Zhongyuan 孫中原. 2001. “A Sketch of One Hundred Years of Research on Chinese Logic 中 國邏輯研究百年論要.” Southeast Academic Research 東南學術 2001.1: 29–39.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

505

Sun, Zhongyuan 孫中原. 2006. Chinese Logic 中國邏輯研究. Beijing: Shangwu Yinshuguan. Suter, Rafael. 2017. Logic and a priori between Perception and Cognition – an Investigation into the Early Works of Mou Zongsan [Logik und Apriori zwischen Wahrnehmung und Erkenntnis – eine Studie zum Frühwerk Mou Zongsans (1909–1995)]. Berlin: De Gruyter. Tang, Tsao-chen. 1936. “A Paradox of Lewis’s Strict Implication.” Bulletin of the American Mathematical Society 42: 707–709. Tang, Tsao-chen. 1938. “Algebraic Postulates and a Geometric Interpretation of the Lewis Calculus of Strict Implication.” Bulletin of the American Mathematical Society 1938.44: 737–744. Wang, Dahai 王大海. 2004. “Breakthroughs and Developments in the Research of Jin Yuelin’s Philosophy of Logic – a Critique of Dr. Zhang Xueli’s Research on the Logical Philosophical Thought of Jin Yuelin 金岳霖邏輯哲學思想研究的突破與進展 —— 評張學立博士的《金 岳霖邏輯哲學思想研究》一書.” Qiannan Normal College for Nationalities 黔南民族示範 學院學報 2004.5: 45–48. Wang, Dianji 汪奠基. 1937. Modern Logic 現代邏輯. Beiping: Shangwu Yinshuguan. Wang, Dianji 汪奠基. 1927. On Logic and Mathematical Logic 邏輯與數學邏輯論. Shanghai: Shangwu Yinshuguan. Wang, Dianji 汪奠基. 1961. An Analysis of Chinese Logical Thinking 中國邏輯思想分析. Beijing: Zhonghua Shuju. Wang, Dianji 汪奠基. 1979. An Intellectual History of Chinese Logic 中國邏輯思想史. Shanghai: Renmin Chubanshe. Wang, Kexi 王克喜. 2000. “The Role of Exchange Students in the Spread of Logic to the East 留學 生與邏輯學的東漸.” Journal of Xuzhou Normal University 徐州師範大學學報 (Philosophy and Social Sciences Edition 哲學社會科學版) 9.2000: 1–5. Wang, Tefu 王特夫. 1933. A System of Logic 論理學體系. Shanghai: Xinken Shudian 辛恳書店. Wang, Xianjun 王憲鈞. 1956. “On Truth Functions in Mathematical Logic as Logical Abstractions from Compound Propositions 數理邏輯的真質函項是複合命題的邏輯抽象. Journal of Peking University 北京大學學報 1956.2: 50–60. Wang, Xianjun 王憲鈞. 1959. “A Critique of the Semantics of Logical Positivism 批判邏輯實證 主義的意義理論.” Journal of Peking University 北京大學學報 1959.2: 97–104. Wang, Xiangqing 王向清, and Liping Liu 劉麗萍. 2007. “Zhang Dongsun on Language and Logic 張東蓀論語言與邏輯.” Journal of Shaoyang University 邵陽學院學報 6.3: 5–7. Wardy, Robert. 2000. Aristotle in China: Language, Categories and Translation (Needham Research Institute studies, vol. 2). Cambridge: Cambridge University Press. Wen, Gongyi 溫公頤, and Qingtian Cui 崔清田. 2001. A Course on the History of Chinese Logic 中國邏輯史教程. Tianjin: Nankai Daxue Chubanshe 南開大學. Witek, John W. JS, ed. 1994. Ferdinand Verbiest, S.J. (1623–1688): Jesuit Missionary, Scientist, Engineer and Diplomat. Nettetal: Steyler Verlag. Wu, Jiaguo 吳家國. 1999. “The Historical Fate and Modernization of Traditional Logic in China 傳統邏輯在中國的歷史命運合現代化.” Journal of Beijing University of Aeronautics and Astronautics 北京航空航天大學學報 12.4: 21–27. Xiang, Rongxian 向容憲. 1992. “An Overview on the Development of the Science of Logic in Modern and Contemporary China 略論近現代中國邏輯科學的發展.” Journal of Guiyang Teachers’ College 貴陽師專學報 (Social Sciences Edition 社會科學版) 27.1: 58–65. Xie, Youwei 謝幼偉. 1997. A Critique of Famous Works of Modern Philosophy 現代哲學名著述 評. Jinan: Shandong Renmin Chubanshe. Xu, Yibao. 2010. “Bertrand Russell and Mathematical Logic in China.” History and Philosophy of Logic 24.3: 181–196. Yan, Shaoyong 閆曉勇, and Huadong Yan 顏華東. 2003. “A Critique of Ai Siqi’s Logical Thinking 艾思奇邏輯思想述評.” Journal of Lanzhou Railway Institute 蘭州鐵道學院學報 (Social Sciences Edition 社會科學版) 22.2: 27–29. Yan, Zhongjun 顏中軍. 2007. “On the Logic of Zhang Shenfu 論張申府的邏輯觀.” Journal of Bijie University 畢節學院學報 25.5: 51–54. Yang, Lei 楊蕾. 2005. “Language, Thought and Logic 言語、思想與邏輯.” Academic Journal of Zhongzhou 中州學刊 146: 131–134.

506

R. Suter

Yang, Peisun 楊芾蓀. 1988. A Course in the Intellectual History of Chinese Logic 中國邏輯思想 史教程. Lanzhou: Gansu Renmin Chubanshe. Yang, Wen 楊文. 2005. “Contemporary Research on [the] Chinese [theory of] Naming and Disputation 中國名辯的現代研究.” Journal of Xinyang Normal University 信陽示範學院學 報 25.6: 19–22. Yao, Nanqiang 姚南強. 1994. “One Hundred Years of Research on Chinese Hetuvidyā-Studies 百年來的中國因明學研究.” Chinese Academy of Social Sciences 中國社會科學 1994.5: 142–151. Yao, Nanqiang 姚南強, ed. 2008. A Dictionary of Hetuvidyā 因明辭典. Shanghai: Shanghai Cidian Chubanshe. Yin, Haiguang 殷海光. 2011a. A New Introduction into Logic 邏輯新引. Taipei: Taida Chuban Zhongxin 臺大出版中心. (First publication: Hong Kong: Xianggang Yazhou Chubanshe 香港 亞洲出版社出版, 1956). Yin, Haiguang 殷海光. 2011b. English Writings and Translations 英文著述與譯作. Taipei: Taida Chuban Zhongxin 臺大出版中心. Yu, Shuo. 1993. “L’introduction de la philosophie de la logique en Chine”. Archives européennes de sociologie 34.1: 139–151. Yu, Yu 虞愚. 1937. The Chinese Theory of Naming 中國名學. Nanjing 南京: Zhengzhong Shuju 正中書局. Yuan, Caiyun 袁彩雲. 2002. “Research on Jin Yuelin’s Logical Thought 金岳霖邏輯思想研究. ” Jianghan Tribune 江漢論壇 2002.3: 37–40. Zhang, Dainian 張岱年. 1989. Collected Works of Zhang Dainian 張岱年文集. vol. 3. Beijing: Qinghua Daxue Chubanshe. Zeng, Zhaoshi 曾昭式. 2000. “Investigating the Fate of the Spread of Logic to the East 邏輯學東 漸的命運探析.” Chinese Academy of Social Sciences 社會科學: 38–42. Zeng, Zhaoshi 曾昭式. 2002. “A Preliminary Analysis of Zhang Dongsun’s Theory of Multiple Logics 張東蓀多元邏輯觀試析.” Journal of Shangqiu Normal University 商丘示範學院學 報 18.1: 14–16. Zeng, Zhaoshi 曾昭式. 2002. “The Spread of Logic to the East and the History of Modern Logic in China 邏輯學東漸與中國現代邏輯史.” Logical Thinking and the Spirit of the Age 哲學思 索與時代精神 (Social Sciences Edition 社會科學) 2002.8: 36–40. Zeng, Zhaoshi 曾昭式. 2009. “The Introduction of Materialist Dialectics in the 1920s and 1930s and the Dialectical Logics of Ma Pei 19世紀 (sic !) 20–30年代唯物辯證法的傳入與馬佩先 生的辯證邏輯觀.” Journal of Henan University 河南大學學報 (Social Sciences Edition 社會 科學版 49.4: 13–17. Zhang, Jialong 張家龍. 1997. “Important Advances of Logical Research in China 我國邏輯學研 究取得重要進展.” People’s Daily 人民日報, 15-11-1997. Zhang, Jialong 張家龍. 2008. “On the Status of Shen Youding’s Paradox in the History of Mathematical Logic 論沈有鼎悖論在數理邏輯史上的地位.” Philosophical Investigations 哲學研究 2008.9: 108–112, 119. Zhang, Qing 張晴. 2004. “Research on the History of Logic in 20th century China 20 世紀的中 國邏輯史研究.” Journal of Xinyang Normal University 信陽師範學院學報 (Philosophy and Social Sciences Edition 哲學社會科學版) 24.4: 23–26. Zhang, Qing 張晴. 2007. Research on the History of Logic in 20th century China 20 世紀的中國 邏輯史研究. Beijing: Zhongguo Shehui Kexue Chubanshe. Zhang, Qingyu 張清宇. 1993. “The Paradox of the Class of all Non-Infinite Classes 所有非ℵ類的 類的悖論.” Philosophical Investigations 哲學研究 1993.10: 43–44. Zhang, Shengbin 張盛彬. 2001. “A Critique of the Leftist Interferecne with Logical Research 評邏 輯研究中的極左干擾.” Journal of Anhui Normal University 安徽師範大學學報 (Humanities and Social Sciences Edition 人文社會科學版) 29.3: 376–381. Zhang, Shizhao 章士釗. 1943. Essentials of Logic 邏輯指要. Chongqing 重慶: Shidai Jingshen She 時代精神社. First edition 1939.

24  Logic in China and Chinese Logic: The Arrival and (Re-)Discovery of Logic…

507

Zhang, Xiaoyan 張小燕, and Zhao Geng 耿昭. 2004. “Zhang Dongsun on Linguistic Structure and Logic 張東蓀論言語結構與名學方式.” Journal of Hebei University 河北大學學報 (Philosophy and Social Sciences Edition 哲學社會科學版) 29.2: 26–28. Zhao, Baoguo 趙保國. 2000. “Experimental Logic and Its Critique and Refusal of Traditional Logic 試驗邏輯及其對傳統邏輯的批評和否定.” Journal of Xinyang Normal University 信陽師範學院學報 (Philosophy and Social Sciences Edition 哲學社會科學版) 20.4, 2000, 21–25. Zhao, Baoguo 趙保國. 2000a. “Perspectives on Zhang Dongsun’s Logical Thinking 張東蓀邏輯 思想透視.” Philosophical Investigations 哲學研究 130.5: 25–30. Zhao, Zongkuan 趙總寬. 1999. One Hundred Years of Logic 邏輯學百年. Beijing: Beijing Chubanshe. Zhao, Zongkuan 趙總寬. 1999a. “The Development of Dialectical Logic in 20th century China”. Journal of Beijing University of Aeronautics and Astronautics 北京航空航天大學學報 (Social Sciences Edition 社會科學版) 1999.4: 33–42. Zhou, Liquan 周禮全 et al., eds. s.a. Collected Works of Jin Yuelin 金岳霖文集. Jin Yuelin Xueshu Jijinhui Xueshu Weiyuanhui 金岳霖學術基金會學術委員會. s.l.: Gansu Renmin Chubanshe. Zhou, Wenying 周文英. 1979. A Sketch of the Intellectual History of Logic in China 中國邏輯思 想史搞. Beijing: Renmin Chubanshe. Zhou, Wenying 周文英. 1998. “The Rise and Development of Chinese Traditional Logic in the Early Modern, Modern and Contemporary Periods 中國傳統邏輯在今、現、當代的升華與 發展.” Journal of Jiangxi Institute of Education 江西教育學院學報 19.1: 1–6/19: 1–8. Zhou, Xuan 周璇. 2014. “A Comparative Analysis of the Logical Methods of Hu Shi, Feng Youlan and Jin Yuelin 對胡適、馮友蘭、岳霖邏輯方法的比較分析.” Logic and Scientific Methodology 28.1 邏輯與科學方法論 28.1: 36–42. Zhou, Yunzhi 周雲之. 1987. “Some Problems in Debates on Traditional Logic in China before Liberation 解放前我國傳統邏輯中爭論的若干問題.” Philosophical Trends 哲學動態 5.2 (Research Materials 研究資料): 41–44. Zhou, Yunzhi 周雲之. 1988. “A Preliminary Discussion on the Specificities of Comprehensive Logic in China’s 1930s, the Process and Reason of Its Rise 試論我國三十年代綜合邏輯的 特點及其興起的過程與原因.” Journal of Tianjin Normal University 天津師大學報 1988.1: 18–23. Zhou, Yunzhi 周雲之. 1991. Selected Materials on the History of Chinese Logic 中國邏輯史資料 選 ― The Modern Era 現代卷, vol. 2. Lanzhou 蘭州: Gansu Renmin Chubanshe. Zhou, Yunzhi 周雲之. 2004. History of Chinese Logic 中國邏輯史. Taiyuan 太原: Shanxi Jiaoyu Chubanshe 山西教育. Zuo, Yuhe 左玉河. 2016. “The Theories of Naming, Disputation and Logic: Translations of Logic in Late Qing and the Formation of the Discipline of Modern Logic in China 名學、辯學與論 理學: 清末邏輯學譯本與中國現代邏輯學科之形成.” Social Science Research 社會科學研 究 2016.6: 153–164. Rafael Suter.  Studies in Sinology, General Linguistics, Slavistics, and Political Philosophy at the University of Zurich (1996–2004). His PhD dissertation is on Mou Zongsan’s notion of logic (2011). He is Teaching and Research Fellow in Chinese Studies, Institute of Asian and Oriental Studies, University of Zurich, Switzerland. His research focus is on various aspects of Chinese philosophy and linguistics of Old Chinese. His publications are mainly in linguistics, Chinese intellectual history, and philosophy. His recent book, titled Logik und Apriori zwischen Wahrnehmung und Erkenntnis [Logic and a priori between perception and cognition] (Berlin: De Gryuter, 2017), is on the concepts of logic and apriori in the early work of Mou Zongsan.

Part IV

Logic Studies in Chinese Communities

Chapter 25

Logic Studies in Mainland China Guoping Du and Hongguang Wang

1  Logic Studies Before the Founding of New China Western logic was introduced into China for the second time by Chinese intellectuals such as Yan Fu 严复, Wang Guowei王国维, Hu Maoru胡茂如 and Liang Qichao梁启超 during the late Qing Dynasty. They made this contribution after fully recognizing the importance of logic to China, especially to Chinese academia. For example, in Mill’s Logic (《穆勒名学》), the Chinese translation of J. S. Mill’s A System of Logic, the translator YAN Fu rephrased Francis Bacon to state that “logic is the law of all laws and the knowledge of all kinds of knowledge”, (Yan 1981:2) In recognition of the contribution of these far-sighted scholars, western logic was ultimately included in university courses and even high school courses in Chinese modern education, from where it was diffused widely. In 1920, Bertrand Russell was invited to give lectures in China. Wu Fanhuan 吴范寰 organized and compiled Russell’s lectures on mathematical logic before the book Mathematical Logic was published by Peking University Xinzhi Press in 1921. Mathematical logic has since been introduced into China. Jin Yuelin金岳霖 began to teach logic in Tsinghua University in 1926 and Wang Dianji汪奠基’s Logic and Mathematical Logic was published by The Commercial Press in 1927 before Modern Logic, another book, was subsequently published in 1937 by the same publisher. Jin’s Logic, in which propositional and predicate calculus were introduced, and had also been published by The Commercial Press in 1936. These publications gradually dispersed modern logic in China. The return of scholars who had studying mathematical logic abroad, such as Shen Yuting沈有鼎, G. Du (*) Institute of Philosophy, Chinese Academy of Social Sciences, Beijing, China H. Wang Department of Philosophy, Peking University, Beijing, China e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_25

511

512

G. Du and H. Wang

Wang Xianjun王宪钧, Hu Shihua胡世华, Mo Shaokui莫绍揆, also made an essential contribution by helping to launch research on mathematical logic in China.

2  T  he Development of Logic in Contemporary Mainland China: Three Stages The development of studies in logic in contemporary mainland China can be roughly divided into three stages, which are as follows:

2.1  1949–1966 During this period, logic studies in mainland China failed to progress in a linear manner. In the early 1950s, academic circles in mainland China were influenced by the Soviet Union and were therefore predisposed to regard logic as a form of metaphysics and idealism. Furthermore, formal logic was regarded as elementary logic, while dialectic logic was regarded as advanced logic. But when Mao Zedong 毛泽 东entertained the intellectuals during the Spring Festival (the lunar new year) of 1956, he spoke to JIN Yuelin and pointed out that mathematical logic required no justification other than its own importance. One year later, Mao spoke about logic and emphasized two points: firstly, that there is no elementary/advanced distinction between formal and dialectic logic; secondly, that formal logic should be universally applicable and therefore falls outside of the orbit of class. In 1958, Mao called on cadres to “learn some grammar and logic” and this, in addition to Mao’s talks about studying logic, helped to dispel the misgivings of intellectuals and establish the basis for logic studies in mainland China: (1) Jin Yuelin, Wang Xianjun and Hu Shihua were responsible for the perspective plan on mathematical logic studies. (2) Many logic scholars wrote articles to propagate and popularize logic. (3) A large number of classic logic textbooks were translated and compiled. For instance, Mo Shaokui translated David Hilbert’s and Wilhelm Ackermann’s Grundzüge der theoretischen Logik (Principles of Mathematical Logic), while Zhou Liquan周礼全, Wu Yunzeng吴允曾 and Yan Chengshu晏成书 translated Alfred Tarski’s Introduction to Logic and the Methodology of Deductive Sciences. (4) Many universities and research institutions established specialized divisions that would teach and research logic, and undergraduate and graduate students majoring in logic were enroled. In 1960, Renmin University’s philosophy department set up a five-year bachelor’s degree program that majored in logic. (5) The research team employed almost one hundred people who were engaged in teaching and researching logic.

25  Logic Studies in Mainland China

513

(6) The Chinese logicians achieved substantial results. In 1949, Hu Shihua published his work on m-valued logic in The Journal of Symbolic Logic (Hu 1949). In 1960, the recursive function theory on natural number was generalized to character set. (Hu and Liu 1960). In 1950, Mo Shaokui constructed two logical systems that could practically avoid “implication paradoxes”. (Mo 1950) Shen Yuting carried out research into paradoxes before publishing his work in The Journal of Symbolic Logic. (Shen 1953, 1955)

2.2  1977–1999 During this period, logic studies in mainland China began to recover and develop. During the Great Proletarian Cultural Revolution (1966–1976), almost all of the academic activities in mainland China began to stagnate, and logic studies was no exception in this respect. In 1978, the Institute of Philosophy and the editorial department of Philosophical Researches, which operated within the Chinese Academy of Social Sciences, organized a National Symposium on Logic, which was held in Beijing. In 1979, a second event, entitled “the Second National Symposium on Logic” was held in Beijing, and the Chinese Association of Logic was founded in the same year. This reiterated that the teaching and researching of logic in mainland China had grown at an unprecedented rate. (1) The education and popularization of logic proceeded apace, with courses on logic being taught in colleges and universities. Many logic scholars discussed educational initiatives. For example, Ouyang Zhongshi 欧阳中石tried to implement logical education in Chinese middle school teaching, and obtained good results (Ouyang 1983), although here it should be recognized that the national self-taught examination has been suspended since 1989. Through the efforts of logic scholars such as Wu Jiaguo 吴家国, majors in a range of subjects instituted logic as a compulsory subject, including administration, economics, ideological/political education and law. Wu was the editor-in-­ chief of Outline of Common Logic for National Self-taught Examination and Principles of Common Logic, which were respectively published in 1985 and 1989. The latter (which, after 2010, was renamed Common Logic and operated under Du Guoping’s 杜国平 general editorship) was a textbook for national self-taught examination that made logical education available to millions of people. After the ‘Reform’ and ‘Opening-up’, the increased role of non-­ governmental actors in the running of schools created more educational opportunities, and many schools also established logic courses. The China Correspondence University of Logic and Language, which has enrolled more than 400,000 students since April 1982, is the largest educational actor that has strongly affected the popularization of logic in China. (2) Universities and research institutes resumed the recruitment of graduate students. Research teams engaged with logic in mainland China have therefore been enriched by the contribution/s of trained PhDs.

514

G. Du and H. Wang

(3) International exchanges and cooperation in logic were constantly expanding. Hong Jiawei 洪加威made an academic visit to Canada after being invited by Prof. S.  Cook in 1979. Various other Chinese academics (including Yang Dongping杨东屏, Huang Wenqi黄文奇, Zhu Hong朱洪, Shen Fuxing沈复 兴, Liu Xuhua 刘叙华 and Zhang Jinwen张锦文) were also invited to visit renowned foreign logic-focused research institutions. A number of Chinese scholars also returned to mainland China after obtaining their doctoral degrees in logic-­related majors, including Li Wei李未, Lo Libo罗里波, Ding Decheng 丁德成, Feng Qi冯琦, Wang Ju王驹, Fu Yuxi傅育熙, Zhou Qing周青 and Jiang Ying 蒋颖. Many worldwide renowned logicians were also invited to conduct academic exchanges in mainland China, including Ken Kunen, Thomas Jech, Anil Nerode, Gerald E. Sacks, G. Mueller, Hao Wang and Chong Chi Tat etc. The 3rd and 6th “Asian Logic Conference” were respectively held in Beijing in 1987 and 1996. (4) The Chinese logicians obtained a series of important working results. Examples include Yang Dongping’s work on α recursion theory (Yang 1984), Lo Libo’s findings on model theory and the decidability of free group (Lo 1983a, b), Zhou Haoxuan’s research on topology in terms of set-theoretic axioms (Zhou 1982, 1983, 1984), Zhang Jinwen’s enquiries into fuzzy set theory over fuzzy relations (Zhang 1982), Shen Yuting’s results relating to pure logic deductions (Shen 1984) and Hong Jiawei’s findings that relate to computational complexity theory. The impact of each of these contributions was substantially enhanced and reiterated when they were published in important academic journals (Hong 1979, 1982, 1984). Further contributions were forthcoming when Zhou Chaochen周巢尘 provided first-order temporal logic (Zhou 1979), Mo Shaokui simplified the axiomatic system of set theory (Mo 1987) and Ding Decheng managed to resolve C.  G. Jockusch and M.  A. Ingrassia Conjecture (Ding 1992). Meanwhile, Feng Qi offered a hierarchy of Ramsey cardinals (Feng 1990), Tang Zhisong唐稚松 designed the programming language XYZ system and provided the executable temporal logic language XYZ/E (Tang 1990) and Xu Jiafu and Yi Bo proposed analogy calculus (Yi and Xu 1992, 1993). Finally, Ying Mingsheng proposed a logic for approximate reasoning (Ying 1994), Li Wei gave the limit theory of formal theory sequence and open logic (Li 1993b) and Zhang Mingyi etc. offered the self-compatible default logic (Gottlob and Zhang 1994; Zhang 1996).

2.3  2000-Present Day Logic studies in mainland China have persisted into a new century that coincides with the expansion of China’s national power, providing a clear incentive for the government to engage basic science research. Accordingly, logic studies in mainland China has steadily developed and taken on a healthier appearance.

25  Logic Studies in Mainland China

515

(1) To take one example, logic studies are strongly supported at the national level. Logic-related research projects are annually supported by the National Nature Science Foundation and the National Philosophy and Social Science Foundation. Since 2010, there have been 11 major competitive projects that engaged various aspects of logic studies and that were set up by the National Philosophy and Social Science Foundation. These included: 2010.12 Zou Chongli 邹崇理 (Institute of Philosophy, Chinese Academy of Social Science): A study of logical semantics for natural language information processing. 2011.10 Huang Huaxin 黄华新 (School of Humanities, Zhejiang University): A study on cognition based on logic. 2012.10 Zhou Beihai 周北海 (Department of Philosophy, Peking University): A study on cognition based on multiple disciplines. 2012.10 Liu Peiyu刘培育 (Institute of Philosophy, Chinese Academy of Social Science): A Study on The Studies in Hetuvidya in China from 1896 to Today. 2013.11 Zhao Xishun赵希顺 (Institute of Logic and Cognition, Sun Yat-sen University): A study on the logic and computational simulation of social game. 2014.11 Zhai Jincheng翟锦程 (Faculty of Philosophy, Nankai University): History of logic in China (8 volumes). 2014.11 Du Guoping杜国平 (Institute of Philosophy, Chinese Academy of Social Science): A study on applied logic and the applications of logic. 2014.11 Liu Hu刘虎 (Institute of Logic and Cognition, Sun Yat-sen University): A study on the logic of information interaction and its application. 2014.11 He Xiangdong何向东 (School of Political Science and Public Administration, Southwest University): A study on logic, cognition and computation of information interaction. 2015.11 Cai Shushan蔡曙山 (Center for Psychology and Cognitive Science, Tsinghua University): A study on higher order cognition of language, thought and culture. 2015.11 Ren Xiaoming任晓明 (Faculty of Philosophy, Nankai University): A study on modern inductive logic: recent advance, frontiers and applications. (2) Influential research teams have emerged in various universities and institutes. These include Sun Yat-sen University, Peking University Zhejiang University, Nankai University, Southwest University, Nanjing University, Chinese Academy of Social Science, East China Normal University, Wuhan University, ­Tsinghua University, Renmin University of China, South China Normal University, Beijing Normal University and Fudan University etc. Sun Yat-sen University is particularly important as it organizes a logic research base within the Institute of Logic and Cognition that brings together more than 20 logic scholars. (3) International academic exchanges have become more frequent. The satellite meeting of Asian Logic Conference was held in Chongqing in 2002; The 13th International Congress of Logic, Methodology and Philosophy of Science was

516

G. Du and H. Wang

held in Beijing in 2007; CCR08 and SAT08 were respectively held in Nanjing and Guangzhou in 2008. (4) A fairly large number of Chinese logicians actively work in international academic circles. Zhang Yi 张弈 is the editor-in-chief of Logic and Algebra and Ying Mingsheng 应明生, Zhang Mingyi 张明义 and Zhao Xishun respectively serve on the editorial boards Fuzzy Sets and Systems and Journal of Computer Science and Technology, AI Communication and the Journal of Satisfiability and Modeling and Computation (Xishun is also the chair of SAT08). (5) Chinese logicians can point to clear research achievements. Feng Qi has produced important results in relation to axiomatic set theory (Feng 2000, 2009; Feng and Jensen 2004). Ding Yunlong and Su Gao have confirmed a conjecture put forward by G. Hjorth (Ding and Gao 2007) and Yu Liang喻良 and Ding Decheng have solved an important problem in complexity theory. In doing so, they have proven that there are many H-degrees in the random reals, and there are no SW-complete C.E. reals (Yu and Ding 2003, 2004a, 2004b; Yu et  al. 2004). Ying Mingsheng has constructed a computation theory based on quantum logic (Ying 2000, 2005), Wang Guojun has developed quantitative logic (Wang and Zhou 2009) and He Huacan has proposed a theory of universal logics (He et al. 2001).

3  R  esearch Findings in Logic in Contemporary Mainland China 3.1  Mathematical Logic (1) Model theory: Lo Libo has conducted research into Vaught’s conjecture, and obtained a series of results (Lo 1983b, 1998). Shen Yunfu defined the space of strong types, thereby proving that space is finite otherwise has a continuum (Shen 1997). Shen Enshao 沈恩绍et al have conducted research into various modal-theoretic properties of partition logics (Shen 1998; Shen and Tian 1996; Shen and Chen 2000) and Wang Shiqiang王世强, Wu Wangming, Shen Fuxing and Song Jie have researched the modal theory of lattice-valued logic (Wang and Wu 1981; Shen 1987; Song 1998). (2) Recursion theory: Hu Shihua and Mo Shaokui have pioneered the study of recursion theory in mainland China. Sun Zhiwei, Zhu Hong and Lu Xianjie, Li Lian, Li Huiling and Liu Yixun conducted research on the undecidability of some mathematical problems (Li et al. 1993; Sun 1992; Zhu et al. 1997), Li Angsheng, Wu Guohua and Zhang Zaiyue evidenced research achievements in relation to recursive enumerable degrees (Li et al. 2000; Zhang 2001) and Yang Dongping obtained important results after researching α recursion theory (Yang 1984).

25  Logic Studies in Mainland China

517

(3) Set theory: Zhang Jinwen conducted research on axiomatic set theory system based on weak predicate logic (Zhang 1983), constructed the axiomatic system ACG upon the basis of system GB (Zhang 1986) and created a hierarchy of axiomatic system ACG (Zhang 1990). In addition, he also constructed a transfinite sequence of axiomatic set theory system (Zα: α∈On) (Zhang 1988). Zhao Xishun demonstrated that in the case of an arbitrary natural number n, the consistency of ZFn is provable in ZF (Zhao 1990). Li Xuhua also showed that the consistency of ZFn is also provable in ZFn + 3 (Li 1993a). Feng Qi conducted research on Rado’s conjecture and, in working with R.Jensen, developed a new method that constructed a KC type model that uses supercomplete type 1 extenders. (Feng and Jensen 2004). Zhang Shuguo, in working with J. Brendle, resolved the problem of cardinal invariance of star relation structure amongst differentiated natural numbers (Brendle and Zhang 2006). Zhang Yi investigated the cardinal number of eventually different function sets, the maximum eventually disjoint function sets, before as well as the cardinal invariance involved cofinitary groups (Zhang 1999a, b; Hrušák et al. 2001). Zhao Xishun and Wang Ju conducted research into the existence of large cardinal in Chang’s model C (Zhao and Wang 1998) and Li Na built the Boolean-valued model of the axiom system of GB (Li 1991). (4) Formalized methods and automated reasoning: Hu Shihua, Mo Shaokui, Wu Yunzeng, Tang Zhisong and Zhou Xiaochen conducted early research into the applications of mathematical logic in computer science. Lu Ruqian generalized algebraic semantics by describing how static control structure became dynamic (Lu 1992a, b). Lin Huimin conducted long-term research into formal semantics and formalized methods (Lin 1993, 1995) and Fu Yuxi developed a higher order calculus (Fu 1999). The use of computers to conduct automated reasoning is also a key research preoccupation for Chinese researchers. Wang Xianghao et al. developed the generalized resolution method (Wang and Liu 1982) and Liu Xuhua et al. extrapolated the generalized-resolution to logic with identity connective (Liu and An 1986). Meanwhile, Xu Yang and Wang Wei et al. conducted research into α-automated reasoning method of lattice-valued logic (Wang et al. 2002) and Wang Guojun et al. provided a new algorithm for solving FMP (Wang et al. 2008).

3.2  Philosophical Logic (1) Modal logic: Studies of modal logic have a fairly long history in mainland China. In the 1950s, Mo Shaokui analyzed the basic properties of modal sequence, and constructed modal systems with finite modal sequences (Mo 1957, 1958). Gao Hengshan and Lv Jianan conducted research on modal logic during the 1960s and published a series of papers that systematically built algebraic semantics of quantified modal logic (Gao 1987, 1995, 2000; Lv and Gao

518

G. Du and H. Wang

1988; Gao and Lv 1998, 1999). Zhou Beihai provided a new semantic frame of modal logic, known as Grafted Frames, and demonstrated the completeness of modal logic system S1 (Zhou 1999). Liu Hu and Ju Shier offered a new model for non-logical omniscient belief logic (Liu and Ju 2007). (2) Multiple-valued logic: Studies of multiple-valued logic in mainland China have produced clear research contributions. Luo Zhukai and Liu Renren etc. conducted long-term research on the Sheffer function that emphasized the partial K-valued logical function, producing a clear set of research achievements (Liu et al. 2004; Liu 2008b; Luo 1963, 1980, 1984a, b, 2003; Luo et al. 1992). Zhu Wujia and Xiao Xi’an, who had been long-term research partners since the 1980s, systematically developed the theory of Medium Logic Theory (Zhu et al. 1996; Zhu 2012; Zhu and Xiao 1988). Meanwhile, Wang Guojun and his research team obtained many results related to fuzzy logic (Wang 1983, 2000, 2003, 2004). During the early twentieth century, Bertrand Russell revealed the paradox that could be derived from the principle of comprehension basing on classical logic, which became known as the widely acknowledged Set Theory Paradox. In 1956, Mo Shaokui drew on Łukasiewicz’s finite-valued logic Łn(3∙n∙ℵ0) to demonstrate that the paradox could also be derived from the principle of comprehension (Mo 1954). In 1985, Zheng Yuxin, Xiao Xi’an and Zhu Wujia offered evidence that the paradox could still be derived from the principle of comprehension by drawing on countable infinite-valued logic (Zheng et  al. 1985; Zhu 2008). In 2007, Du Guoping proposed a set {x: x∈x↔p} that, unlike Russell’s set {x: x∉x}, is only involved in identity connective “↔” as opposed to “negation”. In addition, Du demonstrated that the paradox could be derived from the principle of comprehension upon the basis of finite-valued, countably infinite-valued and uncountably infinite-valued logics (Du et  al. 2009; Du 2012). (3) Lattice-valued logic based on lattice implication algebras: In 1993 Xu Yang developed the concept of “lattice implication algebra”, which is the non-­ classical logic algebra that consists of lattice and implication algebra. Lattice-­ valued logic can be built upon the basis of lattice implication algebra. Since 1993, Xu has conducted a substantial amount of systematic and profound research on lattice implication algebra, lattice-valued logical system (which is based on lattice implication algebra) and theories and methods of uncertain reasoning and automatic reasoning (that are based on the lattice-valued logical system that has developed since 1993) (Xu and Qin 1993; Xu et al. 1997, 2003). (4) Paraconsistent logic: During the first half of the twentieth Century, Shen Yuting developed a paraconsistent logic system that could “localize contradictions” (Wang 1981: 158). Since 1980s, Zhang Qingyu has conducted research on paraconsistent logic, obtaining extensive results in the process (Zhang 1991, 1992). Gui Qiquan, Chen Zili and Zhu Fuxi also conducted research on paraconsistent logic and its applications (Gui et al. 2002). Du Guoping proposed that intuition could provide a new strategy to solve the problem of reasoning in inconsistent theories (Du et al. 2008; Du and Wang 2011; Wang et al. 2009).

25  Logic Studies in Mainland China

519

(5) Other philosophical logics: Liu Fenrong obtained important results from her research on preference logic (Liu 2008a, 2009, 2010, 2011). Li Xiaowu constructed variations of dynamic systems to characterize the concept of dynamic negation (Li 2008) and his introduction of corresponding semantics enabled him to demonstrate the soundness and completeness of the systems. Liao Beishui, Huang Huaxin and Gao Ji developed a flexible agent for non-monotonic reasoning (Liao et al. 2008), Zou Chongli conducted research on the logic of language (Zou 2015) and Gao Hengshan demonstrated the strong completeness theorem and soundness theorem of intuitionist quantified modal logic MIPC∗ (Gao 2000).

3.3  The Philosophy and History of Logic Mainland scholars (such as Shen Yuting, Wen Gongyi温公颐, Liu Peiyu, Sun Zhongyuan孙中原, Cui Qingtian崔清田, Zhou Yunzhi周云之, Dong Zhitie董志 铁, Zhou Shan周山, Zhang Zhongyi张忠义 et al) conducted comprehensive and far-reaching studies that engaged with the history of logic in China. Zhang Jialong张家龙 et al. extensively researched the history of western logic; Wang Lu 王路, Chen Bo陈波 and Zhang Jianjun张建军 et al. focused on the philosophy of logic, and paradox studies in particular. Due to limitations of space, their work cannot be described in extensive detail.

References Jörg Brendle and Zhang, Shuguo. 2006. “Converse Dual Cardinals.” In The Journal of Symbolic Logic 71.1: 22–34. Ding, Decheng. 1992. “The Distribution of The Generic Recursively Enumerable Degrees.” In Archive for Mathematical Logic 32.2: 113–135. Ding, Longyun and Gao, Su. 2007. “Graev Metric Groups and Polishable Subgroups.” In Advances in Mathematics. 123.2: 887–901. Du, Guoping 杜国平. 2012. “Progress in Research of Russell’s Paradox 罗素悖论研究进展.” In Journal of Hubei University (Philosophy and Social Science) 湖北大学学报(哲学社会科学 版). 39.5: 1–6. Du, Guoping and Wang, Hongguang. “Knowledge Implication Intuitionism Logic System.” In Eastern Social Science 2: 10–19, 2011. Du, Guoping, Wang, Hongguang and Li, Na. 2008. “The Completeness and Decidability of Intuitive Implication Logic System.” In Proceedings of the Fourth International Conference on Natural Computation 4: 573–577. Du, Guoping杜国平, Wang, Hongguan王洪光, Li, Na李娜, and Zhu, Wujia朱梧槚. 2009. “Paradox of Set Theory-Universal Logic集合论-泛逻辑悖论.” In Journal of Beijing University of Aeronautics and Astronautics北京航空航天大学学报. No.3: 376–379. Feng, Qi. 1990. “A Hierarchy of Ramsey Cardinals.” In Annals of Pure and Applied Logic 49.3: 257–277.

520

G. Du and H. Wang

Feng, Qi. 2000. “Strongly Baire Trees and a Cofinal Branch Principle.” In Israel Journal of Mathematics 117.1: 71–92. Feng, Qi. 2009. “A New Characterization of Supercompactness and Applications.” In Annals of Pure and Applied Logic 160.2: 192–213. Feng, Qi and Jensen, Ronald B. 2004. “Supercomplete Extenders and Type 1 Mice: Part I.” In Annals of Pure and Applied Logic 128.1: 1–73. Fu, Yuxi. 1999. “Variations on Mobile Processes.” In Theoretical Computer Science. 221:327–368. Doi: https://doi.org/10.1016/S0304-3975(99)00037-7. Gao, Hengshan 高恒珊. 1987. “Algebraic Semantics of Quantified Modal Logic (I) ── about S5 Condition量词模态逻辑的代数语义学(I)——关于S5情形.” In Acta Mathematica Sinica, Chinese Series 数学学报(中文版)30.6:838–846. Gao, Hengshan 高恒珊. 1995. “Algebraic Semantics of Quantified Modal Logic (III) ──about Regular Modal System Condition without Barcan Formula量词模态逻辑的代数语义学 (III)──关于不含Barcan公式的正规模态系统的情形.” In Acta Mathematica Sinica 38.4. Gao, Hengshan. 2000. “A Strong Completeness Theorem in Intuitionistic Quantified Modal Logic.” In Science China Technological Sciences 43.1: 60–70. Gao, Hengshan 高恒珊 and Lv Jianan 吕健安. 1998. “Algebraic Semantics of Quantified Modal Logic (IV, part one)-about Irregular Modal System Condition Contrain Barcan Formula 量 词模态逻辑的代数语义学(IV,上)──关于含Barcan公式的非正规模态系统情形.” In Philosophical Researches哲学研究12. Gao, Hengshan 高恒珊 and Lv Jianan 吕健安. 1999. “Algebraic Semantics of Quantified Modal Logic (IV, part two)-about Irregular Modal System Condition Contrain Barcan Formula” 量词模态逻辑的代数语义学(IV,下)──关于含Barcan公式的非正规模态系统情形. In Philosophical Researches 哲学研究 3. Gottlob, Georg and Zhang, Mingyi. 1994. “Cumulative Default Logic: Finite Characterization, Algorithms and Complexity.” In Artificial Intelligence. 69.1-2: 329–345. Gui, Qiquan 桂起权, Chen, Zil 陈自立, and Zhu, Fuxi 朱福喜. 2002. Paraconsistent Logic and Artificial Intelligence次协调逻辑与人工智能. Wuhan: Wuhan University Press. He, Huacan 何华灿 Wang, Hua 王华, and Liu, Yonghuai 刘永怀. 2001. Principles of Universal Logic泛逻辑学原理. Beijing: Science Press. Hong, Jia-wei. 1979. “On Lower Bounds of Time Complexity of Some Algorithms.” In Scientia Sinica 22.8: 890–900. Hong, Jia-wei. 1982. “On Some Deterministic Space Complexity Problems.” In SIAM Journal on Computing 11.3: 591–601. Hong, Jia-wei. 1984. “A Trade off Theorem for Space and Reversal.” In Theoretical Computer Scienc 32.1–2: 221–224. Hrušák, Michael, Steprans, Juris and Zhang, Yi. 2001. “Cofinitary Groups, Almost Disjoint and Dominating Families.” In The Journal of Symbolic Logic. 66.3: 1259–1276. Hu, Shihua. 1949. “M-valued Subsystem of (m+n)-valued Propositional Calculus.” In The Journal of Symbolic Logic 14: 177–181. Hu, Shi-hua and Liu, Zhong-Mo. 1960. “Kernal Functions, The Theory of Recursive Algorithms-II.” In Scientia Sinica. 9.7: 876–888. Li, Na. 1991. “The Boolean-valued Model of The Axiom System of GB.” In Chinese Science Bulletin 36.2: 99–102. Li, Xuhua. 1993a. “On the Consistency of ZFn in ZFn+3.” In Chinese Quarterly of Mathematics 8.3: 7–12. Li, Wei. 1993b. “An Open Logic System.” In Science in China Series A 36.3: 362–375. Li Xiaowu 李小五. 2008. “Logic of Described Dynamic Negation刻画动态否定的逻辑.” In Journal of Sun Yatsen University (Social Science Edition) 中山大学学报(社会科学版)48.2: 106–111. Li, Lian, Li, Huilin and Liu, Yixun. 1993. “A Decision Algorithm for Linear Sentences on A PFM∗.” In Annals of Pure & Applied Logic 59.3: 273–286.

25  Logic Studies in Mainland China

521

Li, Ansheng, Wu, Guohua, and Zhang, Zaiyue. 2000. “A Hierarchy for Cuppable Degrees.” In Illinois Journal of Mathematics 44.3: 619–632. Liao, Beishui, Huang, Huaxin and Gao, Ji. 2008. “A Flexible Agent Support for Autonomic Computing and Based on Defeasible Logic.” In Journal of Software 19.3: 605–620,. Lin, Hui-Min. 1993. “A Verification Tool for Value-Passing Process Algebras.” In IFIP Transactions C-16: Protocol Specification, Testing and Verification. North-Holland: 79–92. Lin, Huimin. 1995. “PAM: A Process Algebra Manipulator.” In Formal Methods in System Design. 7.3: 243–259. Liu, Fenrong. 2008a. Changing for the Better, Preference Dynamics and Agent Diversity. PhD Thesis. ILLC, University of Amsterdam. Liu, Renren. 2008b. “Decision and Construction of the Sheffer Functions in Partial Four-Valued Logic.” In Computer Engineering & Science 30.11: 75–76. Liu, Fenrong. 2009. “Diversity of Agents and Their Interaction.” In Journal of Logic, Language and Information 18.1: 23–53. Liu, Fenrong. 2010. “von Wright’s ‘The Logic of Preference’ Revisited.” In Synthese 175. 1: 69–88. Liu, Fenrong. 2011. “A Two Level Perspective on Preference.” In Journal of Philosophical Logic 40.3:421–439. Liu, Xuhua and An, Zhi. 1986. “Generalized Resolution Using Paramodulation.” In Chinese Science Bulletin 31.21: 1441–1444. Liu, Hu 刘虎 and JU, Shier 鞠实儿. 2007. “Logical Omniscience Problem of Belief Logic 信念逻 辑的逻辑全知问题.” In Seeking Truth 求是学刊34.6: 31–34. Liu, Renren, Chen, Jianer, and Chen, Songjiao. 2004. “The Decision on Sheffer Function in Partial Two-valued Logic.” In Computer Engineering 30.24: 19–21. Lo, Libo. 1983a. “The T-theory for Free Groups is Undecidable.” In The Journal of Symbolic Logic. 48.3: 700–703. Lo, Libo. 1983b. “On the Number of Countable Homogeneous Models.” In The Journal of Symbolic Logic. 48.3: 539–541. Lo, Libo. 1998. “Vaught’s Conjecture on Unitary Relations.” In Acta Mathematica Sinica 14.2: 277–284. Lu, Ruqian. 1992a. “P/R Nets and Process Concepts. I.” In Science in China Series A 35.1: 21–31. Lu, Ruqian. 1992b. “P/R Nets and Process Concepts. II.” In Science in China Series A 35.2: 148–157. Luo Zhukai. 1963. “All Maximal Closed Sets of Four-valued Logic.” In National Conference on Mathematical Logic. Luo, Zhukai. 1980. “Problem Identification of Maximal Closed Sets in Multiple-valued Logic.” In Acta Mathematica Sinica 1: 152–156. Luo, Zhukai. 1984a. “Complete Theory of Some Multiple-valued Logic Function.” In Mathematica Sinica 5: 676–683. Luo, Zhukai. 1984b. “Maximal Closed Sets of Some Multiple-valued Logic Fuction.” In Acta Mathematica Sinica 27.6: 795–800. Luo, Zhukai 罗铸楷. 2003. Research of Multiple-valued Logic and Its Application: Collected Works of Prof. Luo Zhukai多值逻辑理论及应用研究: 罗铸楷教授文集. Changsha: National University of Defence Technology Press. Luo, Zhukai 罗铸楷, Hu, Mou 胡谋 and Chen, Yanhuai 陈廷槐. 1992. Multiple-valued Logic and Its Application多值逻辑的理论及应用. Beijing: Science Press. Lv, Jianan 吕建安, and Gao, Hengshan 高恒珊. 1988. “Algebraic Semantics of Quantified Modal Logic (II) ── about Regular Modal System Condition Contrain Barcan Formula量词模态逻 辑的代数语义学(II)——关于含Barcan公式的正规模态系统的情形.” In Acta Mathematica Sinica, Chinese Series 数学学报(中文版)31 (5): 692–698. Mo, Shaokui. 1950. “The Deduction Theorems and Two New Logical Systems.” In Methodos 2: 56–75.

522

G. Du and H. Wang

Mo, Shaokui. 1954. “Logic Paradox for Many-valued Systems.” In The Journal of Symbolic Logic: 37–40. Mo, Shaokui. 1957. “Modal System of Finite Modal Word.” In Acta Mathematica Sinica 1. Mo, Shaokui. 1958. “Basic System Finite Modal System.” In Acta Mathematica Sinica 2: 153–180. Mo, Shaokui. 1987. “The Simple Reduction of Set-theoretic Axioms and the Power of Cardinal Number.” In Chinese Annals of Mathematics 2. Ouyang, Zhongshi 欧阳中石. 1983. “A Project of Reform in Chinese Education.” 一个语文教改 方案的设想In the Journal of Beijing Normal College北京师院学报 (Philosophy and Social Science) 2. Shen, Yuting. 1953. “Paradox of the Class of All Grounded Classes.” In The Journal of Symbolic Logic. 18.2: 114. Shen, Yuting. 1955. “Two Semantical Paradoxes.” In The Journal of Symbolic Logic. 20.2: 119–120. Shen, Yuting, 1984. “The Part of Independent Quantifier in Pure Logical Deduction.” In Acta Mathematica Sinica 5. Shen, Fuxing. 1987. “The Consistency Property In Lattice-valued Model Theory for Infinitary Logic.” In Acta Mathematica Sinica 2: 103–108. Shen, Yunfu. 1997. “The Space of Strong Types and An Open Map Theorem.” In Acta Mathematica Sinica 13.1: 59–63. Shen, Enshao. 1998. “Some Notes on Graph Automata, Tiling Systems and Partition Logic.” In Journal of Computer Science & Technology 13.6: 483–489. Shen, Enshao, and Chen Yijia. 2000. “The Downward Transfer of Elementary Satisfiability of Partition Logics.” In Mathematical Logic Quarterly 46.3: 477–487. Shen, Enshao and Tian, Qijia. 1996. “Monadic Partition Logics and Finite Automata.” In Theoretical Computer Science 166.1–2: 63–81. Song, Jie. 1998. “The Fundamental Theorem of Ultraproducts in Lattice-valued Model for Logic.” In Chinese Quarterly Mathematics 13.2: 5–9. Sun, Zhiwei. 1992. “Reduction of Unknowns in Diophantine Representations.” In Science in China, Ser. A 35: 257–269. Tang, Zhisong 唐稚松. 1990. “Design Thinking about the System of XYZ XYZ系统的设计思 想.” In Journal Of Software 软件学报1.1: 47–55. Wang, Hao 王浩.1981. Popular Lectures on Mathematical Logic 数理逻辑通俗讲话. Beijing: Science Press. Wang, Guojun. 1983. “A New Fuzzy Compactness Defined by Fuzzy Nets.” In Journal of Mathematical Analysis and Applications 94.1: 1–23. Wang, Guojun 王国俊. 2000. Non-classical Mathematical Logic and Approximate Reasoning非经 典逻辑与近似推理. Beijing: Science Press. Wang, Guojun 王国俊. 2003. Introduction to Mathematical Logic and Resolution Principle数理 逻辑引论与归结原理. Beijing: Science Press. Wang, Guojun. 2004. “Formalized Theory of General Fuzzy Reasoning.” In Information Sciences 160.1–4: 251–266. Wang, Xianghao 王湘浩 and Xuhua Liu 刘叙华. 1982. “General Resolution Principle广义归结.” In Chinese Journal of Computers 计算机学报2: 81–92. Wang, Shiqiang and Wu, Wangming. 1981. “On a Classification Problem for Complemented Lattices.” In Chinese Science Bulletin 26.4: 289–292. Wang, Goujun and Hongjun Zhou. 2009. “Quantitative Logic.” In Information Sciences 179.3: 226–247. Wang, Wei, Xu, Yang and Wang, Xuefang. 2002. “α-Automated Reasoning Method Based on Lattice-Valued Propositional Logic LP (X).” In Journal of Southwest Jiaotong University 10.1: 98–111. Wang, Guojun, Hui, Xiao-Jing and Song, Jian-She. 2008. “The R0-type Fuzzy Logic Metric Space and An Algorithm for Solving Fuzzy Modus Ponens.” In Computers and Mathematics with Applications 55.9: 1974–1987.

25  Logic Studies in Mainland China

523

Wang, Hongguang, Li, Na and Du, Guoping. 2009. “Intuitive Implication Logic System.” In Proceedings - The 6th International Conference on Fuzzy Systems and Knowledge Discovery 2: 252–256. Xu, Yang and Keyun Qin. 1993. “Lattice-valued Propositional Logic (I).” In Journal of Southwest Jiaotong University 2: 123–128. Xu, Yang, Qin Keyun and Song, Zhenming. 1997. “Syntax of Lattice-valued First-order Logic FM.”In Chinese Science Bulletin 42.16: 1337–1340. Xu, Yang, Ruan, Da, Qin, Keyun and Liu, Jun. 2003. Lattice-Valued Logic: An Alternative Approach to Treat Fuzziness and Incomparability. Springer. Yan, Fu 严复. 1981. Mill’s Logic 穆勒名学. Beijing: The Commercial Press. Yang, Dongping. 1984. “On the Embedding of-recursive Presentable Lattices into the -recursive Degrees below 0.” In The Journal of Symbolic Logic. 49.2: 488–502. Yi, Bo and Xu, Jiafu. 1992. “Analog Model and Analog Correspondence.” In Science in China Series A 35.3: 374–384. Yi, Bo and Xu, Jiafu. 1993. “Analogy Calculus.” In Theoretical Computer Science 113.2: 211–230. Ying, Mingsheng. 1994. “A Logic for Approximate Reasoning.” In The Journal of Symbolic Logic 59.3: 830–837. Ying, Mingsheng. 2000. “Automata Theory Based on Quantum Logic. II.” In International Journal of Theoretical Physics 39.11: 2545–2557. Ying, Mingsheng. 2005. “A Theory of Computation Based on Quantum Logic. I.” In Theoretical Computer Science 344.2-3: 134–207. Yu, Liang and Ding, Decheng. 2003. “An Extension of Harrington’s Noncupping Theorem.” In Science in China Series F 46.3: 199–209. Yu, Liang and Ding, Decheng. 2004a. “There are 2ℵ0 Many H-degrees in the Random Reals.” In Proceedings of the American Mathematical Society 132: 2461–2464. Yu, Liang and Ding, Decheng. 2004b. “There is No SW-Complete C. E. Real.” In The Journal of Symbolic Logic 69: 1163–1170. Yu, Liang, Ding, Decheng and Downey, Rodney. 2004. “The Kolmogorov Complexity of Random Reals.” In Annals of Pure and Applied Logic 129.1-3: 163–180. Zhang, Jinwen. 1982. “Between Fuzzy Set Theory and Boolean Valued Set Theory.” In Fuzzy Information and Decision Processes. North-Holland: Amsterdam-New York. 143–147. Zhang, Jinwen. 1983. “Fuzzy Set Structure with Strong Implication.” In Advances in Fuzzy Sets, Possibility Theory, and Applications. New York: Plenum. 107–136. Zhang, Jinwen. 1986. “Aggregation, Phase Sequence and Base Tensors.” In Acta Mathematica Sinica 29: 217–223. Zhang, Jinwen. 1988. “A Hierarchy of Axiom System ZF.” In Acta Mathematica Sinica 4.1: 72–75. Zhang, Jinwen. 1990. “A Hierarchy of Axiom System ACG.” In Science in China Series A 33.2: 129–134. Zhang, Qingyu, 1991. “A Week Paraconsistent Conditional Logic.” In The Journal of Non-classical Logic, 8.1: 45–57. Zhang, Qingyu. 1992. “Paraconsistent Modal Logics CnG′, n∈ω.” In Boletim da Sociedade Paranaense de de Matemática, 13.1–2: 33–41. Zhang, Mingyi. 1996. “A New Research into Default Logic.” In Information and Computation, 129.2: 73–85. Zhang, Yi. 1999a. “Towards a Problem of E. van Douwen and A. Miler.” In Mathematical Logic Quarterly 45.2: 183–188. Zhang, Yi. 1999b. “On a Class of M.A.D.  Families.” In The Journal of Symbolic Logic 64.2: 737–746. Zhang, Zaiyue. 2001. “Extending the Cooper Minimal Pair Theorem.” In Journal of Computer Science and Technology 16.1: 77–85. Zhao, Xishun. 1990. “Coordination Results about Replaceable of Axiom Schema for System of Z+Σn-formula.” In Chinese Science Bulletin 35.12:979–979

524

G. Du and H. Wang

Zhao, Xishun and Wang, Ju. 1998. “ω1-ordinal Definable Sets and Chang’s Model C.” In Chinese Quarterly Journal of Mathematics 13.1: 98–106. Zheng, Yuxin, Xiao, Xian and Zhu, Wujia. 1985. “Finite-Valued or Infinite-Valued Logical Paradoxes.” In Journal of Mathematical Research with Applications 5.3:16. Zhou, Chaochen 周巢尘. 1979. “Program Module and Predicate Calculus 程序模式和谓词演算.” In Chinese Journal of Computers 计算机学报3: 14–29. Zhou, Haoxuan. 1982. “On the Small Diagonals.” In Topology & Its Applications 13.3: 283–293. Zhou, Haoxuan. 1983. “A Conjecture on Compact Frechet Spaces.” In Proceedings of the American Mathematical Society 89.2: 326–328. Zhou, Haoxuan. 1984. “On Countable Compactness and Sequential Compactness.” In Proceedings of the American Mathematical Societ 90.1: 121–127. Zhou, Beihai. 1999. “Grafted Frames and S1-Completeness.” In The Journal of Symbolic Logic 64.3: 1324–1338. Zhu, Wujia 朱梧槚. 2008. Logical Foundation of Mathematics and Infinity数学与无穷观的逻辑 基础. Dalian: Dalian University of Technology Press. 107–123. Zhu, Wujia 朱梧槚. Logic Basis of Infinity and Medium in Mathematics数学无穷与中介的逻辑 基础. Beijing: Science Press, 2012. Zhu, Wujia and Xiao, Xian. 1988. “The System of Medium Axiomatic Set Theory MS.” In Science in China Series A 31.11:1320–1335. Zhu, Zhaohui, Shi, Qingsheng and Zhu, Wujia. 1996. “Program Compatible Set: Program Semantic of Medium Logic, Described by Force Approximation.” In Science in China Series E 6: 567–573. Zhu, Hong, Lun, Xianjie and Denis, Richard. 1997. “The Undecidability of Arithmetic Structures.” In Chinese Annals of Mathematics, Series A 18: 1–6. Zou, Chongli 邹崇理. 2015. “An Approach to Logical Grammar of Non-continuous Compound Quantitative Sentence非连续性的复合量化句逻辑语法处理方式.” In Journal of Yunan Normal University (Humanities and Social Sciences)云南师范大学学报(人文社会科学 版)47.1: 69–74. Guoping Du,  is a professor of philosophy at Institute of Philosophy, Chinese Academy of Social Sciences (CASS). He is also the deputy President and the Secretary-General of the Chinese Association of Logic (CAL), and the chairman of Applied Logic Studies Committee, CAL. He received his first Ph.D. in logic from CASS in 2000 and his second Ph.D. in computer science from Nanjing University of Aeronautics and Astronautics in 2009. He was on the faculty of the Department of Philosophy at Nanjing University from 2000 to 2010. His research interests include history of Chinese logic, applied logic and the application of logic. Hongguang Wang,  is currently a Postdoctoral Research Fellow at Peking University. She received her B.A., M.A., and Ph.D. in philosophy from Nanjing University in 2006, 2009 and 2015. She also taught philosophy and logic at Huaqiao University from 2016 to 2019. Her research interests include philosophical logics and philosophy of logic and language.

Chapter 26

Logic Studies in Taiwan Zeqiang Wu and Wen-fang Wang

1  S  tages and Characteristics of the Development of Logic in Taiwan The development of logic research in Taiwan is intertwined with Taiwan’s social and historical development. ZHENG Wenhui said: “The emergence and development of any logic theory can find its historical root in society.” (Zheng 1994: 4) We think that the development of logic research in Taiwan is a concrete example of this general rule. The present research shows that logic studies in Taiwan have been closely linked to changes in political and economical situations. Therefore, we take radical changes or major events that happened in political and economical situations to be the main cut-off points of the development of logic studies in Taiwan. Accordingly, we divide the development of logic research in Taiwan into four different historical stages and characterize them stage by stage. The infant stage is the period before 1945 during the Japanese occupation. The next stage of development starts in 1945 with the retrocession of Taiwan to the R.O.C. and ends in 1964, the year before the “economic take-off” of Taiwan. In the third stage, running from 1964 to 1987, Taiwan underwent rapid economic growth and the enforcement of martial law, allowing logic research to have a very good opportunity to develop. Finally, with the democratization of Taiwan from 1987 until now, logic in Taiwan has prospered fruitfully. Each stage has its own characteristics, as stated below. The First Stage  The infant stage starts with the Japanese occupation and ends in 1945. FAN Shoukang, who studied the educational environment in Taiwan during Z. Wu Zhengzhou Chenggong University of Finance & Economics, Henan, China W.-f. Wang (*) Institute of Philosophy of Mind & Cognition, National Yang Ming University, Taipei, Taiwan e-mail: [email protected] © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7_26

525

526

Z. Wu and W.-f. Wang

Japanese occupation, revealed that there was only one university in Taiwan at that time  – “Taiwan Imperial University” (TIU; now called “National Taiwan University”). There were 394 students at TIU back then, including 111 Taiwanese students and 283 Japanese students. Besides TIU, there were also three “higher normal schools” with 2888 students in total, including 523 Taiwanese students and more than 2000 Japanese students. (Fan 1931: 114) By analyzing existing literature kept at the library of National Taiwan University, we also found some early (maybe the earliest) logic works in Taiwan. One of them is particularly important; its Japanese edition was published by a Japanese publisher (松邑三松堂) at the beginning of 1926 (the 15th year of Taisho, January 9th). This book was written by four Japanese authors, Tanaka, Kitazawa, Hihama, and Tomin (田中寬一、北澤種一、 日填權一、島田民治) and was written as a textbook for the “new education” at higher normal schools and TIU. The book has three main parts: introduction, main body, and an appendix. The introductory part includes two chapters. The first chapter describes the main concern of logic, while the second chapter emphasizes the importance of logic. The main body of the book is divided into two parts. The first part contains four chapters and is about the essential elements of logic. The first chapter is on principles of thinking, the second chapter on concepts, the third chapter on assertion, and the last chapter is on reasoning. The second part is about methodology and includes three chapters. The first chapter is on the meaning of methodology and on classification of methods, the second chapter is on methods of inquiry, and the third chapter is about categories of integration methods. The appendix also has two parts; one of them is on the method of association while the other contains a few exercises. Overall, this is a book about basic concepts and methods of Aristotelian logic and Mill’s inductive methods, such as direct inference and syllogisms. It does not contain much material relevant to modern logic. As can be clearly seen from these data, even though Taiwanese people of this period rarely had a chance to be educated, it doesn’t mean that logic studies did not exist at all. Rather, logic studies in Taiwan of this period were influenced by the logic studies developed in Japan. This infant stage was deeply affected by the special history of Japanese occupation, and has the following characteristics. First, logic works in this period were mainly written in Japanese. During the Japanese occupation, especially in 1937 after Japan launched an all-round aggressive war against China, Japan began to implement Japanization campaigns, one of which was to “ban the use of Taiwanese and Mandarin and stop the publication of Chinese journals”. (Qi 2004: 55) This was true of books in logic as well as all other textbooks; publishing houses no longer received Chinese versions of manuscripts, and published books in Japanese only. Second, logic works were written or translated by Japanese authors and translators only, while Taiwanese authors or translators were nowhere to be found. For instance, Iwanami Shorten (岩波書店; a Japanese publisher) published more than 80 logic works related to Japanese scholars (including translated works) in the period, none of which were written or translated by Taiwanese. As a result, most nowadays scholars are under the wrong impression that there was no logic study in Taiwan before 1945.

26  Logic Studies in Taiwan

527

Third, this stage focused on introducing western logic to Taiwanese students and exploring basic logic concepts and methods of logic, with contents mainly involving Aristotelian logic and philosophical logic. Ancient Western logic scholars such as Aristotle and historically influential thinkers such as Kant, Hegel, and so on were introduced to Taiwanese students through the translation of their works. Logic concepts discussed in this stage include the subject matter of logic, categorical judgments, subjects and predicates, direct inference, syllogism, fallacies and subsumed concepts. Basic logic methods introduced, on the other hand, include scientific methods vs. popular methods, systematical methods vs. partial methods, analytic methods vs. synthetic methods, reasoning methods vs. dialectical methods, inference methods and diagram methods, etc. The Second Stage  The second period of logic development in Taiwan starts with the retrocession of Taiwan to the R.O.C. in 1945 and ends in 1964, the year before the “economic take-off” of Taiwan. During this period, Taiwan was in a state of reconstruction, and the declining economy hindered the development of logic studies. It was not until 1949 that genuine logic research began to appear in Taiwan, when Mainland China logic scholars CHEN Daqi (陳大齊), MOU Zongsan (牟宗 三), and YIN Haiguagn (殷海光) came to Taiwan, restructuring the environment of logic research and bringing it to a new stage. This stage has the following characteristics. First, logic works were written in Chinese. After Japan surrendered in 1945, the R.O.C. government took over 86 public libraries, 1 educational museum, 51 tutorial schools for industry, and 38 women’s vocational schools. (Fan 1948: 29) Taiwanese and Chinese language publications have been restored and books were all published in Chinese. Second, scholars who immigrated from Mainland China after WWII dominated logic research in Taiwan, becoming a guiding force of logic research in Taiwan in this period. There were three important scholars at this time – CHEN Daqi, MOU Zongsan, and YIN Haiguang. CHEN Daqi offered traditional western logic courses in Taiwan while continuously studies Buddhist logic at the same time. MOU Zongsan had already accomplished his study of western logic before he went to Taiwan and, upon arrival, not only did he teach logic but he also applied logic to re-construct Confucian philosophy, the so-called “neo-­ Confucianism.” But the most important person, who really and effectively introduced “genuine,” i.e., formal, logic, to Taiwan, was YIN Haiguang. He devoted himself to studying, teaching, and promoting logic, and trained and cultivated talent students such as LIN Yusheng (林毓生), LIN Zhenghong (林正弘), LIU Fuzeng (劉福增), WANG Xiaobo (王曉波), LI Ao (李敖), CHEN Guying (陳鼓 應), and HUANG Zhanji (黃展驥). His thoughts and theories on logic influenced generations of Taiwanese youths, and played an important role in the development of logic research in Taiwan. Third, the focus of logic research in this period was on traditional logic (Aristotelian logic); meanwhile, the study of Buddhist logic and modern logic also achieved some progress. According to our investigation, nine logic books were published

528

Z. Wu and W.-f. Wang

from 1945 to 1964, written by YIN Haiguang, CHEN Daqi, YU Yu (虞愚), CHANG Tie (張鐵), and HE Xiuhuang (何秀煌). There were also two translated works – one for Introduction to Logic by Max Black and one for Introduction to Logic by Andrew Bachhuber. Traditional logic was the main concern in this period, while modern logic was also taken into considerations. Focuses of logic research in this period include: what logic is, features and functions of logic, logical dichotomy, syllogism, propositional logic and predicate logic, fundamental principles and methods of Buddhist logic, and so on. The Third Stage  This is the development stage for logic research in Taiwan, but it is also the martial law era during the Constitutional Period (1964–1987). The 1960s (especially after 1964) was about Taiwan’s rapid economic development. It was a golden age for private enterprises, which laid a material foundation for the development and reform of the political system in the 1980s. A large number of new small and medium enterprises emerged as a result of the export-oriented economy, encouraging development in science and technology as well as in cultural and educational business. This created a favorable economic condition for the development of logic studies. However, this is also the period in which KMT’s one-party ruling had been insisted until 1987; the enforcement of marshal law caused restriction in creativity and ideologies, which added political overtones to the purely academic study of logic. The development of logic in Taiwan in this period has the following characteristics. First, writings and translated works in logic increased significantly. From 1964 to 1987, nearly 30 books on logic and 22 translated works of logic were published in Taiwan. The authors included MOU Zongsan, CHEN Daqi, LIN Zhenghong, LIU Fuzeng, HE Xiuhuang, HONG Chengwan (洪成完), WU Dingyuan (吳定 遠), ZHANG Shangde (張尚德), SHI Yuanjian (石元健), LIU Shunde (劉順德), YANG Shiyi (楊士毅), XU Naide (許乃紅), CHAI Xi (柴熙) and other scholars. Translators in this field included LIU Fuzeng, YAN Menghui (嚴夢輝), LI Shuren (李樹仁), HE Xiuhuang, LIU Shunde, HE Zhaoqing (何兆清), CHEN Rongbo (陳榮波), WU Dingyuan and more. The research interest and topics in logic were expanded significantly – with a shift from previous studies on traditional logic to studies on modern logic. Research areas included philosophical logic, logical methods, symbolic logic, logic terminology and logic concepts, and so on. Meanwhile, translated works were mainly related to set theory, the history of mathematical logic, symbolic logic, formal logic, etc. Second, the academia saw more and more young researchers and scholars, with higher education and strong English ability. Most of the logic scholars of this period, especially the younger generation, studied abroad. This had two effects on the development of logic research in Taiwan. For one thing, it stimulated and promoted logic studies in Taiwan; for the other, it created a fusion effect: ­Taiwan’s logic studies integrated with western logic studies while absorbing the latter’s strengths. The Fourth Stage  This is the democratization era during the Constitutional Period (1987-present), which is also a blooming period for the development of logic studies in Taiwan. Taiwanese people enjoyed in this period further economical develop-

26  Logic Studies in Taiwan

529

ment and increased level of education, which have expanded their knowledge and changed their ways of obtaining information. 1986 became a turning point for Taiwanese people’s political cognition and evaluation of the political system. In September of 1986, the DPP’s (Democratic Progressive Party) successful “beachhead” operation forced the KMT to recognize the legitimacy of DPP’s activities and ultimately lifted the ban on forming new political parties in July of 1987. Since then, large and small political parties have emerged in Taiwan, disintegrating KMT’s one-­ party ruling history of nearly 60 years. The radical protests became an important political factor of Taiwan’s political transition. The changes in ideologies had a great impact on the development of logic studies in Taiwan: studies of logic became diversified. The development of Taiwan’s logic studies in this period has the following characteristics. First, Taiwan has engaged actively in communication and collaboration with Mainland China, all the while has kept up with contemporary international trends of logic research. Following the heat of such communication, LIN Zhenghong originated the first “Cross-strait Logic Teaching and Academic Conferences,” jointly organized by the philosophy departments of National Chung Cheng University and National Taiwan University, with HONG (Allen) Yuhong (洪裕 宏) as the main organizers. The conference took place in National Taiwan University from June 28th to 30th in 2002. Among the logic scholars that were invited to the event, 10 were from Mainland China, 4 were from Hong Kong, and more than 50 were from Taiwan,” (Zhang 2003: 105) officially marking the first step of cross-­strait academic exchanges on logic research. The second Crossstrait Logic Teaching and Academic Conference was then held in Nanjing University in October, 2006, with a larger scale – 18 logic scholars from Hong Kong and Taiwan, and more than 120 from Mainland China. The third conference was held in May of 2008, organized by the Department of Philosophy at Soochow University in Taiwan, inviting 38 logic scholars from Mainland China. The conference papers were complied into a book, “Logic and Philosophy, which was published by Taiwan’s Xue Fu Culture Com. Limited (學富文化事業有限 公司). Up till the end of 2015, the “Cross-strait Logic Teaching and Academic Conferences” has been held seven times. Additionally, according to Professor WANG Wen-fang’s (王文方) oral report, Taiwan had held 14 other logic conferences that were not of cross-strait nature by 2012. The most significant one was the “Seventh Asian Logic Conference,” from June seventh to tenth in 1999, co-­ hosted by HONG Yuhong, LI Guowei (李國偉), DONG Shiping (董世平), and planned by PENG Mengyao (彭孟堯). Furthermore, the “Experience and Truth Conference” was held four times, respectively in March of 2005, June of 2005, November of 2006, and November of 2007, all founded and organized by PENG Mengyao, WANG Wen-fang and LIANG Yiyu (梁益堉). The “Symposium on Knowledge and Logic” took place in January of 2012, organized by the Department of Philosophy of National Taiwan University, co-chaired by PENG Mengyao and WANG Wen-fang. The “Taiwan Philosophical Logic Seminar” took place on December seventh of 2012, organized by YANG Jinmu (楊金穆).

530

Z. Wu and W.-f. Wang

Second, logic studies in Taiwan in this period have developed rapidly and have a wide interest in different logic systems, including Buddhist logic developed in the long history of logic. The study of Aristotelian logic, however, was overlooked. After 1987, research topics of Taiwanese logicians include the philosophical analysis of the language of logic, the study of philosophical logic, translation of natural language into formal languages, computation theory, applied logic in computer science, studies of mathematical logic, logic research methodology and many other areas. The study of Buddhist logic has also progressed, marked by LIN Chongan’s (林崇安) book A Research on Buddhism logic (因明邏輯研究) and the Chinese Buddhist Journal (中華佛學學報) published by Chinese Institute of Buddhist Studies.

2  I mportant Figures and Academic Achievements of Taiwan in the Fourth Stage The studies of logic in Taiwan at the last stage cover various branches, but the main two areas are mathematical logic and philosophical logic. Mathematical logic includes set theory, proof theory, model theory, and recursive function. Philosophical logic includes modal logic, epistemic and temporal logic, deontic logic, and formal semantics of natural languages. If one looks closely at the various logic works and articles published in Taiwan in this period, one can see that there is no strict boundary between these two categories; most researchers touch on both sides. On the other hand, even though both Chinese logic and Buddhist logic are studied by Taiwanese scholars in this period, their studies have not been the focus of logic researches in general.

2.1  Academic Achievements in the Study of Modern Logic In the area of mathematical logic, the main icons in Taiwan in the fourth stage include LIU Shichao (劉世超), HONG Chengwan, LI Guowei, DONG Shiping, and LIAO Chunzhong (廖純中). LIU Shichao was YIN Haiguan’s disciple and had a Ph.D. in mathematical logic from Wisconsin University. He was the first one in Taiwan to obtain a Ph.D. in logic and also the first Taiwanese logic scholar who published his work in renowned international philosophical journals. He was a researcher at the Institute of Mathematics of Academia Sinica in Taiwan and retired in 1991. He specializes in the foundation of logic and mathematics. His articles on recursion theory, proof theory, and nonstandard analysis were published in several prestigious international journals.

26  Logic Studies in Taiwan

531

HONG Chengwan attended Notre Dame University from 1967 to 1969; he studied epistemology of mathematic and mathematical logic, and conducted researches in logic theories espoused by Polish logicians. Afterward, he studied philosophy of mathematics at Wisconsin University. His main specialties are in relevant logic and intuitionistic logic systems, model theory and algebraic semantics, etc. His published papers include “The Philosophical Foundation of Set Theory,” “Entailment, the Paradox of Implication, and Lewy’s Argument,” “The Independence of Connectives and Quantifiers,” “Some Problems on Names and Reference Theory,” “Tracking Thoughts: A Study from First-Order Logic to Philosophical Logics.” LI Guowei received his Ph.D. in mathematics from Duke University in 1976, with specialization in combinatorics, logic, and philosophy of mathematics. His articles are mostly in English and published abroad, earning him international attention; he is also one of the most prolific Taiwanese logicians when it comes to international publications. His only published paper in Taiwan is “The Evolution of Proof: A Comprehensive Look at the Philosophy of Mathematics and the History of Mathematics.” He served as the fifth, sixth and seventh chairs of “National Committee of the R.O.C. for the Division of Logic, Methodology and Philosophy of Science of the International Union of History and Philosophy of Science of the International Council for Science.” He is also the author of many pop-science books in mathematics and logic, contributing immensely to the promotion and popularization of logic in Taiwan. DONG Shiping, who received his Ph.D. in mathematics from Illinois University, is now a professor at the College of Science of Chung Yuan University in Taiwan. He specializes in algebra, computational biology, logic, bio-informatics, and computation theory. He has published more than 20 academic papers and 10 technical reports. He is highly regarded in the field of philosophy of mathematics. WANG Wen-fang commented that, “In the past 60 years, Professor DONG Shiping (and Professor LI Guowei of Academia Sinica) are believed to be the mathematical philosophers in Taiwan who best keep up with international standards.” (Wang 2010: 123). LIAO Chunzhong finished his Ph.D. at National Taiwan University in 1992, and is now a researcher at the Computer Science Institute of Academia Sinica, specializing in reasoning under uncertainty, applied logic, artificial intelligence logic, and so on. Most of his works are published in foreign journals related to artificial intelligence, such as Artificial Intelligence, Journal of Applied Logic, International Journal of General System, and so on. So far, he has published about 40 works. Two more scholars are worth being mentioned. LI Ruilin (李瑞麟), who received a Ph.D. in mathematics from Illinois University, mainly engages his researches in mathematical logic, computational complexity, and calculus. CAI Xingjian (蔡行 健) finished his Ph.D. study at Columbia University in 2005, and is now a professor at the Department of Philosophy of National Chung Cheng University in Taiwan. He mainly studies mathematical logic, philosophy of logic, philosophy of mathematics, metaphysics, and philosophy of language.

532

Z. Wu and W.-f. Wang

When it comes to philosophical logic, the most representative scholars are LIU Fuzeng, LIN Zhenghong, WANG Wen-fang, YANG Jinmu, DENG Dunmin (鄧敦 民), and WANG Yiqi (王一奇). LIU Fuzeng is a Ph.D. candidate at UCLA and had been a long-standing teacher at the Department of Philosophy of National Taiwan University. He specializes in philosophy of language, Wittgenstein, Laozi, mathematical logic, philosophy of mathematics, and general science-based methodology. Liu is the most prolific scholar in this period, both in paper publications and monographs. Nevertheless, “although Liu’s publications in logic are mostly related to philosophy of logic or philosophy of language and present some good viewpoints, they fall short in important discoveries or systematic theories.” (Wang 2010: 116). LIN Zhenghong graduated from UC Berkeley in 1985 with a Ph.D. in philosophy, and had been a professor at the Philosophy Department of Soochow University before he retired in 2015. He also taught at the Department of Philosophy of National Taiwan University for a long period and once served as the president of Taiwan Philosophical Association. His researches are mainly in philosophy of logic, epistemology, and philosophy of science. He is the author of Logic (邏輯) (1970), A White Horse is not a Horse (白馬非馬) (1975), Knowledge, Logic, and Philosophy of Science (知識、邏輯、科學哲學) (1985), Symbolic Logic (符號邏輯) (1986), and several other books. Among them, A White Horse is not a Horse is a collection of 12 papers, mainly related to logic and semantics. He has been awarded the Distinguished Research Award by National Science Committee (now called “Ministry of Science and Technology”) in Taiwan many times for his abundant works. “It is clear that Professor Lin’s works exhibit a higher level of sophistication in comparison to logic books published by his peers or predecessors.” (Wang 2010: 128) More importantly, Lin was the first one to start a communication with logic scholars in Mainland China, achieving outstanding contribution in promoting the exchange, integration, and development of logic studies on both sides of the Taiwan Strait. WANG Wen-fang received his Ph.D. from the University of Iowa in 1996. Upon returning to Taiwan he first served as a professor at the Philosophy Department of Chung Cheng University. Later on, he started teaching at the Institute of Philosophy of Mind and Cognition of National Yang Ming University. His concentrations are in philosophy of logic, especially in the semantic paradox, the sorites paradox, and theories of conditionals. WANG Wen-fang is the backbone of logic studies in Taiwan. He has played an active role in promoting communication between logic researchers on both sides of the Taiwan Strait in recent years. He has published more than 40 papers, including “Three-Valued Pluralsim”, “Adding conditionals to a Fixed Point Theory of Truth,” “Do WE Really Need Possibilia?”, “Truth in PW and the Solution of Hale’s Dilemma for Modal Fictionalism,” “Theories of Abstract Objects without Ad Hoc Restriction” and so on, many of which were published in Synthese, the Philosophical Forum, Journal of Philosophical Research and Erkenntnis. While WANG Wen-fang focuses mainly on philosophy of logic, he sometimes produces more formal works. He is among the best scholars in Taiwan, be it from the standpoint of depth, breadth, innovation, or creativity of his works. Meanwhile, he is also the most active scholar in cross-strait communication in logic

26  Logic Studies in Taiwan

533

studies, making outstanding contributions in promoting collaboration between logic researches across the Taiwan Strait. YANG Jinmu received his Ph.D. from Oxford University in 1993 and is currently a professor at the Department of Philosophy of National Taiwan University. He specializes mainly in formal logic, philosophy of logic, philosophy of language, and contemporary Anglo-American philosophy studies. “Professor Yang’s papers focus on issues in the philosophy of logic. His works present some interesting viewpoints and show a deep understanding of relevant issues.” (Wang 2010: 110). WANG Yiqi graduated from Austin University of Texas with a Ph.D. in philosophy, and primarily engages in philosophical logics, formal semantics, and epistemology. He is a professor at the Department of Philosophy of National Chung Cheng University. He has published around 15 works, many of which were published in international journals such as Synthese, Erkenntnis, and so on. DENG Dunmin obtained his Ph.D. from Cambridge University, with a concentration in philosophical logics and metaphysics. He currently teaches at the Department of Philosophy of National Taiwan University. His research areas include ontology, conceivability, possibility, and more. Through our field study, we have discovered that current researches in logic in Taiwan not only focus mainly on Western logic, but also show a trend of interdisciplinary researches that combine formal semantics, game theory, mathematical logic, philosophical logics, and so on. Representative figures in formal semantics for natural languages include LI Jiaying (李佳穎), an associate researcher at the Academia Sinica (concentrating in neurolinguistic and cognitive neuroscience studies), LIN Ruowang (林若望), a distinguished researcher also at the Academia Sinica (concentrating in formal semantics, syntax, and syntactic-semantic interface), SU Yiwen ( 蘇以文) at the Language Institute of National Taiwan University (specialized in discourse analysis and pragmatics), HE Wanshun (何萬順) (specialized in formals semantics, Chinese linguistics, and computational linguistic), and LIU Zhaolin (劉 昭麟) (specialized in knowledge discovery, information technology and education, model construction and automated reasoning, and computational linguistics) of the Graduate Institute of Linguistics of National Chenchi University, and finally ZHANG Rongxing (張榮興), chair of the Institute of Linguistics of National Chung Cheng University (specialized in cognitive semantics, sign language linguistics, first and secondary language acquisition). Important figures in game theory include YANG Jiancheng (楊建成), a distinguished researcher at the Institute of Economics of Academia Sinica (concentrating in public economic research), ZHENG Xiuling (鄭秀玲), chair of the Department of Economics of National Taiwan University (specialized in microeconomics, innovation and industrial economics, industrial organization), LIN Zujia (林祖嘉) of the Department of Economics of National Chengchi University (specialized in residential economy, cross-strait economic and trading). Icons in mathematical logic include DONG Shiping of Central University, LI Ruilin of Formosa University, CAI Chengzhi (蔡承志) of Mackay University, researcher LI Guowei of Academia Sinica, CAI Xingjian, WANG Yiqi and WANG Renjun (王仁俊) of National Chung Cheng University, WANG Wen-fang and Thomas Benda (卞拓蒙) of National Yang Ming University, MI Jianguo (米建國)

534

Z. Wu and W.-f. Wang

and LIN Zhenghong of Soochow University, QIU Jianshuo (邱建碩) of Fu Jen Catholic University, PENG Mengyao, DENG Dunmin, and YANG Jinmu of National Taiwan University, and finally SU Qinghui (蘇慶輝) of Shandong University in China (originally from Taiwan).

2.2  A  cademic Achievements in the Study of Ancient Chinese Logic Taiwanese scholars have a rather rare interest and weak understanding toward ancient Chinese logic. This, perhaps, is due to a popular view that is presented in an essay written by LIU Fuzeng, “Is There Logic in Traditional Chinese Culture?” In that article, Liu suggested that logic has never been a part of Chinese culture nor did Chinese people ever value logical thoughts. In spite of this popular view, different conclusions and some achievements still have been made by a limited number of scholars in Taiwan. For instance, according to HUANG Junjie’s (黃俊傑) research, Chinese people have had a very flexible and unique way of thinking, a fact that can be seen clearly by the following reasoning methods that have often been employed by Chinese people: (1) the analogical reasoning to infer to the unknown from the known; (2) the correlative thinking of combining two things or two phenomena of identical nature; (3) the concrete thinking that focuses on concrete objects. (Zhang 2004: 41) Such Chinese ways of thinking also stirred up strong interest in some scholars. For example, A Study of Ancient Chinese Thinking Mode (Taiwan Zhongzheng Book Press (中正書局) 1996), edited by YANG Rubin (楊儒賓) and HUANG Junjie, provided insight to analogical reasoning and concrete thinking. Their research started with the analogical reasoning, placing ancient Chinese thinking in the specific circumstances created, and they argued that the analogical reasoning is the common and the main reasoning type of Chinese classics like The Book of Changes (Yi-Jing 易經), the Zhuangzi (莊子), and the Mobian (墨辨) (a part of the Mozi (墨子)). In the book Mozi (Dong Da Book Co. Ltd. 1996), the author WANG Zanyuan ( 王讚源) made a comparative analysis between the theory of the relation between names and reality proposed by School of Names (名家) in ancient China and the theory of reference proposed by G. Frege. The book also offered new analysis and interpretation of Mohism’s analogical reasoning. LI Xianzhong (李賢中) and SUN Changxiang (孫長祥) also conducted researches on Mohist logic. HUANG Junjie and LI Xianzhong studied analogical reasoning and Confucianism. All of them contributed immensely to the study of the main reasoning methods of ancient Chinese philosophers. LI Xianzhong is currently one of the most distinguished scholars of Chinese philosophy in Taiwan, with fruitful academic research on ancient Chinese logic. He received his Ph.D. from Fu Jen Catholic University in 1991. He is now a professor at the Philosophy Department of National Taiwan University, specializing in

26  Logic Studies in Taiwan

535

Chinese Philosophy, philosophy of School of Names prior to Qin (先秦) Dynasty, Mohism, Chinese logic, etc. He has published more than 40 academic papers, including “the Conception of Name-Reality in Pre-Qin days” (Philosophy and Culture 1991(204)), “Gong Sunlong” (Philosophy and Culture 1995(256)), “Aspects of Chinese Philosophy” (Soochow Journal of Philosophical Studies 1998(3)), etc.

2.3  Academic Achievements in the Study of Buddhist Logic The development of the researches in Buddhist logic in Taiwan can and should be contributed mainly to the research achievements of certain research organizations and scholars such as CHEN Daqi, Master Shuiyue (水月法師), JIANG Longcan ( 江龍燦), LIN Chongan and LIN Zhenguo (林鎮國). CHEN Daqi published his Commentaries and Analyses of Buddhist Logic (mimeograph edition) in Mainland China in 1938, and the stereotype edition was re-published in Chongqing in 1945. In this book, CHEN Daqi made proficient use of Aristotelian logic in his research of Buddhist logic, constantly demonstrating his keenness and depth. (Chen 1945: 1) The book has become a significant work in the development of Buddhist logic on account of its important academic value. He continued his research upon his arrival in Taiwan. In 1970, he published A Preliminary Analysis of Buddhist Logic (Taiwan Zhong Hua Book Press(中華書局)), which included a deeper analysis of the concepts and content of Buddhist logic. Many original points were raised in this book, including the discovery that “Similar yu-body (同喻體) is provable by the conjunction of the second and the third principles of Tri-rūpa-hetu (or “yin-shan-xiang” (因 三相) in Chinese) rather than by the second principle of Tri-rūpa-hetu alone”. (Chen 1970: 104). Master Shuiyue is also known as Shenghe (聖禾) or, by his secular name, WANG Junling (王俊嶺). He was born in Changli County, Hebei Province, in 1928. He spent his younger years following Master Huifeng (慧豐法師) of Zhan Ran Temple (湛然寺) in southern Taiwan. He became the master of Zhan Ran Temple after the death of Master Huifeng. Master Shuiyue is the most distinguished Buddhist follower in the popularization and exploration of Buddhist logic in modern times. He spent 10  years writing and editing The Journal of Buddhist Logic. He published three books: An Anthology of Buddhist Logic, A Study of Ancient Buddhist Logic, and The Origin of Buddhist Logic. He also led the reprinting project of two books: A Study of Buddhist Logic and Commentaries and Analyses of Buddhist Logic by CHEN Daqi and several monographs written by ZHOU Shujia (周叔伽) and Huiyuan (慧圓). As for academic research, Master Shuiyue concerned himself with the Buddhist logic developed in ancient China; his A Study of Ancient Buddhist Logic is the only monograph that contains a comprehensive analysis and ­interpretation of the Buddhist logic developed in ancient China since XU Dishan ( 許地山) published his Buddhist Logic of Zhong Guan School and Yoga School in Pre-Chenna (Pre-Dignaga) Periods in 1931. (Valeng 2013).

536

Z. Wu and W.-f. Wang

LIN Zhenguo is a distinguished scholar at the Research Center on Religion of National Chengchi University. He received a Ph.D. in religious study from Temple University. He specializes in comparative study between Mahayana Buddhism and Chinese Philosophy. With inspirations from modern western philosophy and Hermeneutics, he explored Buddhist traditions such as idealistic empiricism, Meso, Buddhist logic, Pramānavāda, Tian Tai and Zen. He also conducted researches on modern New Confucianism, Kyoto School of Japan, Neo-confucianism, and Taoism under the approach of cross-cultural comparison. His works include Sunyata and Modernity: Buddhist Hermeneutics from Kyoto School, New Confucianism, to Multivocality (1991), On Dialectics (2002), and Sunyata and Methodology: Fourteen Discussions on Cross-Cultural Buddhist Philosophy (2012). His research papers include “Worries of Sunyata: East-Asian Confucianism and Modernity” as well as many others published in English and Japanese. Another achievement worth mentioning is the publication of Buddhist Logic Studies, written by HUO Taohui (霍韜晦) and published by Taiwan’s Foguang Publishing House in 1978 and subsequently by Hong Kong’s Fazhu Academic Society in 1984. Among the papers in the book, “The Development of Buddhist Logic after Dignaga” was a follow-up work to XU Dishan’s Buddhist Logic of Zhong Guan School and Yoga School in Pre-Chenna (Pre-Dignaga) Periods. LIN Chongan’s An Investigation on Buddhist Logic (Beiju Press 1991, reprinted in 1992) became college textbook material. Finally, the Research Institute of Chinese Buddhism also published Journal of Chinese Buddhism, which is also worth mentioning.

3  A Reflection on the Development of Logic in Taiwan Since the Japanese Occupation, the development of logic in Taiwan in the past 100 years has undergone four distinctive periods, including the Japanese Occupation Period before 1945, the Retrocession Period of Taiwan to the R.O.C from 1945 to 1964, the Marshal Law Period between 1964 and 1987, and the Democratization Period from 1987 until now. Through detailed analysis, we have discovered three problems: The first problem is that the Ministry of Education and the Ministry of Science and Technology of Taiwan have long ignored and not placed enough emphasis on the promotion and research of logic. The educational system in Taiwan has undergone reformation, transition and modernization and has finally become more developed and well rounded. As for logic, however, the Ministry of Educational and the Ministry of Science and Technology of Taiwan have not laid enough emphasis on its research and promotion, which can be seen from the following two aspects. First of all, logic has not been integrated well enough into college curriculum. During the Japanese Occupation, although there was only one university, i.e., TIU, and a few higher normal universities in Taiwan, they nevertheless already

26  Logic Studies in Taiwan

537

offered courses in logic. Nowadays, the number of colleges in Taiwan has increased significantly. There are 171 colleges in total, which include 32 public universities (including normal universities, art colleges and sport universities), 28 private universities, 9 public technology universities, 5 public vocational colleges, 20 private technology universities, 46 private vocational colleges, 17 professional colleges, and 9 police or military academies. And yet, within the abundance, none of these colleges have an independent department or institute dedicated to logic. Aside from the “National Committee of R.O.C. for the Division of Logic, Methodology and Philosophy of Science of the International Union of History and Philosophy of Science of the International Council for Science” established in 1979, there has since not been any other organization or institute devoted specifically to logic. And even this committee could not help the development of logic in Taiwan because its nature is irrelevant to promoting or researching logic. This phenomenon is attributable to the poor integration between the development of logic in Taiwan and Taiwan’s educational system. Second of all, political ideologies have interfered with and affected the development of logic in Taiwan. WANG Wen-fang pointed out in “Works on Logic Published in Taiwan in the Past 60 Years (1949-2009)” that such political interference might be traced back to an incident between logician YIN Haiguang and the political authorities of the time. He wrote, “Professor YIN Haiguang was the most influential philosopher and logician in Taiwan in the 1950s, yet he often wrote articles that publicly criticized KMT’s party-centric education policy and its “Strike Back at Mainland China” military policy. He also advocated for the right to establish new political parties. All of these publications and speeches offended the ruling authorities. He was forced to leave his post at National Taiwan University in 1966; soon afterwards, he was diagnosed with gastric cancer in 1967 and passed away in 1969 at the age of 49. Ever since Professor YIN Haiguang, for a long time, the educational authorities have been reluctant towards the emergence of any logic-related institute, committee, or academic journal.” (Wang 2010: 110) Since YIN Haiguang criticized the government so openly and publicly, the authorities held a grudge against him and the logic he promoted so fiercely; subsequently, since most Taiwanese logicians of the time were either Yin’s students or his disciples, the authorities held the same grudge against them as well, which consequently led to its reluctance in helping the promotion and development of logic. The second problem is the lack of research in ancient Chinese logic; a gap exists between the research of modern logic and that of the traditional Chinese logic because the latter has not been passed down. Taiwanese scholars only started studying ancient Chinese logic after Mainland Chinese scholars brought it to this side of the Taiwan Strait after 1946. During this period, Taiwanese scholars approached ancient Chinese logic from two aspects. Either they employed Chinese culture as their framework and explored the traditional Chinese logic thoughts or they tried to construct archetypes and

538

Z. Wu and W.-f. Wang

reasoning methods of traditional Chinese logic. Significant works of this period related to ancient Chinese logic studies include A White Horse Is Not a Horse (LIN Zhenghong 1975), Logic (WU Kunru (鄔昆如), et al. 1983), A New Theory on Logic (Logic Teaching Committee of National Taiwan University 1989), A Logical Interpretation of “I Ching” (ZHANG Banglian (張邦廉) 1990), A Study on the “Name-Reality” of Pre-Quin School of Name (ZHANG Banglian 1992), and more. Nonetheless, Taiwanese scholars’ attention and devotion towards ancient Chinese logic was still pale in comparison to their efforts in modern logic. This can be attributed to two reasons. The first reason is historical: Taiwan and Mainland China have isolated from each other due to historical circumstances, cutting off any communication in the academic research of logic. The second reason pertains to the direction in which Taiwan’s development of logic was heading. Most Taiwanese logicians were, at some point, trained in the Western world, so the development of logic in Taiwan was greatly influenced by Western logic, which naturally leads to overlooking or avoiding researches in traditional Chinese logic. The third problem is rooted in the deficiency of research in Buddhist logic, also known as Hetuvidya. When Buddhist logic spread around China, it took on two paths and was divided into Tibetan Hetuvidya and Han Hetuvidya. We have seen two times in history in which Han Hetuvidya was studied feverishly. The first time was when Master Xuanzang returned from his pilgrimage to India; studying Buddhism became a trend in the Tang Dynasty. The second time was during the 30 years after China’s May Fourth Movement in 1919; the liberal scholars were once passionate about Buddhist logic. Such scholars include LV Zhi (呂瀓), YU Yu, and CHEN Daqi. In 1948, CHEN Daqi came to Taiwan and became one of the earliest Buddhist logic scholars in Taiwan. He published Commentaries and Analyses of Buddhist Logic in 1938 (mimeograph) and 1945 (stereoptype) in Mainland China. He then published Introduction to Nyaya-pravesa in a New Context in Taiwan in 1970, and the book focused on comparative study between Aristotelian logic and Hetuvidya. Following these came two works: A Study on Buddhist Logic (HUO Taohui 1979) and An Investigation on Buddhist Logic (LIN Chongan 1991); the latter became college textbook material and was soon reprinted in 1992. Aside from these, there was hardly any publication on Buddhist logic. All in all, Taiwan should consider expanding its researches in Buddhist logic in order to supplement its development of logic research. Let us conclude by making a brief summary of what we said in the above passages. Logic studies in Taiwan have a history over 100 years, with each period possessing unique characteristics. From the early stage’s exploration of concepts in logic to today’s rather complete systems of modern logic, Taiwanese logicians have contributed greatly to the field. By systematically studying the history of development and current circumstance of logic research in Taiwan, Mainland Chinese logicians can benefit and gain a comprehensive understanding of the research of logic in Taiwan. Meanwhile, such a study is significant in promoting logic in both Taiwan

26  Logic Studies in Taiwan

539

and Mainland China. The research of logic in Taiwan has presently reached a new phase; the field has exceptional professors and the finest scholars, who bring themselves into line with logicians of the world. These exceptional scholars include LIN Zhenghong, WANG Wen-fang, CAI Xingjian, WANG Renjun, LIN Zhenguo, WANG Yiqi, DENG Dunmin, YANG Jinmu, DONG Shiping, ZHUANG Zhengliang, LI Guowei, LI Rulin, LI Xianzhong and more. As logic studies progress and scholars from the Greater China area communicate, the research of logic in Taiwan is bound to head towards a blooming period. Meanwhile, we must also recognize certain worrisome phenomena. After undergoing the initial stage, the secondary stage, and the developing stage, logic research in Taiwan is still heavily focused on Western logic, shaping a purely formal path for itself, which calls for concern. On the other hand, most Taiwanese logicians trained for their Ph.D. in the Western world; it can be anticipated that after the older generation of logicians and ancient Chinese philosophy scholars retire from the academic field, there will be lack of scholars with expertise in ancient Chinese logic, which also calls for attention. Moreover, many Taiwanese logicians use English or other Western languages when writing their academic works. This is perhaps a good way to demonstrate their academic ability, but it also means that Chinese does not have its own voice in academia. The disadvantage of mother-tongue-oriented thinking had made it difficult for logic, especially Chinese logic and Buddhist logic, to earn a place in the development of logic in Taiwan, which is also a shared concern.

References Chen, Daqi 陳大齊. 1945. Commentaries and Analyses of Buddhist Logic 因明大疏蠡測. Chongqing: Mimeograph Edition. Chen, Daqi 陳大齊. 1970. A Preliminary Analysis of Buddhist Logic 因明入正理論悟他門淺釋. Taipei: Zhong Hua Book Press. Fan, Shoukang 范壽康. 1931. “Review of Taiwan Education during Janpanese Occupation” 日治 時期台灣教育綜述. In The Education Magazine 教育雜誌. 1931, 23(9). Fan, Shoukang 范壽康. 1948. “Review of Taiwan Post-war Education after Two Years” 戰後兩 年來的台灣教育綜述. In Zhonghua Educational Review 中華教育界, 1948(2) 1948年復刊 第2卷, 29–56. Valeng 翰林修. 2013. “Master Shuiyue” 水月法師. http://baike.haosou.com/doc/7314954.html. Qi, Hongshen 齊紅深. 2004. Educational History of the Japanese War of Aggression against China 日本侵華教育史. Beijing: People’s Education Press. Wang, Wen-fang 王文方. 2010. “Works on Logic Published in Taiwan in the Past 60 Years (1949– 2009)” 台灣六十年 (1949–2009). In Studies in Logic 邏輯學研究, 2010(3), 110–138. Zhang, Binfeng 張斌峰. 2004. “Review and Prospects of Contemporary Logic Development in Taiwan” 當代台灣邏輯學發展的回顧與展望. In Journal of Hunan University of Science &Technology (Social Science Edition) 湖南科技大學學報(社會科學版), 2004(9), 39–44. Zhang, Jianjun 張建軍. 2003. “Exchange, Review and Prospect, −- Report of the first ‘Cross-­ Strait Logic Teaching and Academic Conferences’.” 交流、回顧與前瞻—記首次”兩岸邏輯 教學學術會議”. In Philosophical Researches 哲學研究, 2003年增刊, 105–108. Zheng, Wenhui 鄭文輝. 1994. History of Logic in Europe and America 歐美邏輯學史. Guangzhou: Zhongshan University Press.

Chinese-English Glossary

Chinese Terms bai 白 (white) bai-ma 白馬 (white horse) bai ma fei ma 白馬非馬 (white horse is not horse) bai-shi 白石 (white stone) bai-wu 白物 (white things, white objects) bi liang 比量 (anumāna in Sanskirt) (inference) bian 遍 (quantificational adverbial) bian 辯 (argumentation, disputation, debate, distinction-drawing, to distinguish, to discriminate, to argue) bian 辨 (to distinguish, to make distinctions) bian 變 (to change, to make changes) biangu 變故 (event, incident) bianshì 辯士 (debater, dialectician) bianzhe 辯者 (debater, dialectician) bie ming 別名 (particular name) bu 不 (not) bu cheng 不成 (asiddha in Sanskirt) (not established) bu ding suo bai 不定所白 (not fixing white) bu-he 不合 (difference in terms of two things not joining together) bu-le 不類 (difference in terms of two things not being of the same class/kind) bu-ti 不體 (difference in terms of two things not in one body) buzu 不足 (insufficiency) cha lei 察類 (examining classes) cheng 誠 (really) cheng jian 成見 (a phenomenon of seeing) chong-tong 重同 (sameness of duplication in terms of identity or equivalence) © Springer Nature Switzerland AG 2020 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7

541

542

Chinese-English Glossary

ci 辭 (sentence, utterance, expression, proposition, phrasing) dang 當 (fit, coincide) dao 道 (way, principle) de 德 (virtue, property) dagu 大故 (death) dang 當 (fit, should be) dangli 當理 (fit the principle, in accordance with the principle) daoli 道理 (theoretical principle or principle of dao) duan 端 (a point) er 而 (if) er ming yi shi 二名一實 (two names of one thing) er wu zuo 二無左 (two does not have left) er wu yi 二無一 (two does not have one) er wu you 二無右 (two does not have right) fa 法 (model, standard, rule, model) fayi 法儀 (standard, rule or model) faze 法則 (standard, rule or model) fei 非 (unless, not) fei yi zhi lei 非義之類 (not of the class/kind of righteousness) fu 弗 (not) gong攻 (attack) gong míng 共名 (classifying names, general names) gou 苟 (really) gu 故 (cause, reason) guyue 故曰 (so I say that) he-tong 合同 (sameness of commonality in terms of two names referring to two things joining together) he tong yi 合同異 (the unity of similarity and difference) ji 即 (unless, then) ji cheng 極成 (prasiddha in Sanskirt) (well established) jia 假 (supposing) jian 見 (seeing) jian 兼 (compound, total, collection whole) jian bai 堅白 (hard and white) jian xiang ai 兼相愛 (mutual love) jiao xiang li 交相利 (mutual benefit) jiashi 假使 (supposing) jiazhi 假之 (let us assume this as an arbitrary hypothesis) jie 藉 (then) jin 今 (now) jiu 舊 (old, past)

Chinese-English Glossary

543

jiu ju yin 九句因 (nine forms of reason) ju 句 (phrase, sentence) ju 舉 (to refer) judou 句讀 (pausing at the end of a phrase or sentence which is used as sentence, clause and phrase markers) ke 可 (assertible, permissible, acceptable) lei 類 (class, kind, sort, category) lei-tong 類同 (sameness of commonality in terms of things belonging to a class/ kind) li 理 (principle, reason, reasonableness) li 禮 (rite, ritual) li 利 (benefit, utility) li jian bai 離堅白 (the separation of hardness and whiteness) liang 量 (pramāṇa in Sanskirt) (means of valid cognition) ling 令 (supposing) liu 牛 (ox, cow) liu-ma 牛馬 (ox-horse) lun 論 (reasoning, theory) ma 馬 (horse) ming 名 (names, terms, labels, or reputation) ming lei 明類 (understanding classes) ming ming 命名 (to name) mou 侔 (parallelism, parallelizing) mou 某 (some, someone) neng li 能立 (sādhana n Sanskirt) (demonstration) pi 譬 (analogy, illustration) pian 篇 (chapter) qi 氣 (vital force) qing 情 (facts, conditions, affective states) ran 然 (so, like this) ren. 仁 (humanity, benevolence) ru 如 (if) ruo 若 (if) ruoshi 若使 (supposing) san zhi zuo fa 三支作法 (three-membered argument) san xiang 三相 (trairūpya in Sanskirt) or yin san xiang 因三相 (three characteristics) she 設 (supposing) shen 身 (body) shu 屬 (sub-class, sub-kind)

544

Chinese-English Glossary

shuì 說 (persuasion) shuo 說 (explanation, discourse, theory) si 私 (private, private name, individual name) shi 實 (stuff, objects, features, events, or situations, reality, actuality) shi 是 (this, that, right) shi 事 (affair, event) shi 使 (supposing) shigu 事故 (event, incident) shili 事理 (principle of affairs) siwang 死亡 (death) suo li 所立 (sādhya in Sanskirt) (the inferable) suo li fa 所立法 (sādhyadharma in Sanskirt) (the property to be proved or the inferable property) tang 儻 (in case) tianli 天理 (natural principle) tian ren he yi 天人合一 (heaven and man are unified) ti 體 (unit, individual, part) ti-tong 體同 (sameness of commonality in terms of two names referring to two unifying parts) tong 同 (sameness, similarity or unity) tong pin 同品 (sapakṣa in Sanskirt) (similar instance) tong-yi 同異 (sameness and difference) tui 推 (generalization, inferring, extending to an isomorphic example) tuilei 推類 (the generalization of class/kind) wei 微 (unless, if it had not been for but for) wei 位 (position) wei 偽 (intentional activity) wei shi bi liang 唯識比量 (inference of consciousness-only) wu 無 (no) wu 物 (objects, things) wu hou 無厚 (dimensionless) wu nei 無內 (having nothing inside) wu wai 無外 (having nothing outside) wuli 物理 (physical principle, property of physical object) wu zhi li 物之理 (physical principle, property of physical object) xi 昔 (past) xian 咸 (quantificational adverbial) xian liang 現量 (pratyakṣa in Sanskirt) (perception) xiang fu 相副 (mapping) xin 信 (trustworthy, sincere, reliable) xin 信 (real, really, trustworthy) xing 性 (nature) xing 形 (form, shape)

Chinese-English Glossary

545

xing 行 (behavior, action, practice) xu hua 虛化 (de-substantiation) yan 言 (language, sentence, statement) yi 異 (difference, dissimilarity or disunity) yi 議 (disputation, exposition, judgment) yi 意 (meaning, idea, intention) yi 義 (righteousness, appropriateness) yi pin 異品 (vipakṣa in Sanskirt) (dissimilar instance) yili 義理 (principle of appropriateness, principle of righteousness) yin 因 (the criterion) yin 因 (hetu in Sanskirt) (true reason) yin fa 因法 (reason-property) yin-ming 因明 (hetuvidyā in Sanskirt) (demonstration with reasons or science of reasons) yin-yang 陰陽 (two opposing and complementary forces in nature) yìzhe 議者 (debater, dialectician) yong 用 (use) you 有 (have, there is) you fa 有法 (dharmin in Sanskirt) (property-possessor) you yi tong 有以同 (two things have something in common) youyu 有餘 (excess) yu 喻 (dṛṣṭānta in Sanskirt) (true example) yu ti 喻體 (example-itself) yu yi 喻依 (example-base) yu ci 於此 (here, this) yuan 援 (adducing, citing a similar example) yuzhu 語助 (words of assistance or particles) ze 則 (then) zhang 章 (passage) zhe quan 遮詮 (paryudāsa) (negation with affirmative implication) zheng 正 (rectify) zheng 正 (associative properness) zhengmíng 正名 (rectification of names) zhì 志 (intention) zhi 指 (pointing, to point, to refer, what is pointed out, referent, meaning, idea) zhì 治 (to govern, to put in order, to regulate, achieving justice and order) zhi 之 (that, this) zhi 志 (intention, will) zhi lan 止濫 (prasajyapratiṣedha) (negation with no more implication) zhi lei 知類 (knowing classes, knowing kinds) zhiyan 卮言 (goblet words) zhongli 中理 (fit the principle, in accordance with the principle) zhou 周 (quantificational adverbial)

546

Chinese-English Glossary

zhu 誅 (punish, execute) zong 宗 (pakṣa in Sanskirt) (thesis) zong fa 宗法 (pakṣadharma in Sanskirt) (a property of the subject)

Chinese Names Chan (Zen) Buddhism 禪宗佛教 Confucianism 儒家 (Ru Jia) Confucius 孔子 (Kong Zi) Daoism 道家 (Dao Jia) Deng Xi 鄧析 Fa Jia 法家 (Legalist School or School of Law) Gongsun Long 公孫龍 Hanfeizi 韓非子 (Han Fei Zi) Hui Shi 惠施 Kong Yingda 孔穎達 Laozi 老子 (Lao Zi) Later Mohism 後期墨家 (Later Moism) Liu Xie 刘勰 Mencius 孟子 (Meng Zi) Ming Jia 名家 (School of Names) Mohism or Moism墨家 (Moh Jia) Mozi 墨子 (Mo Zi) Song Xing 宋鈃 Xu Shen 許慎 Xunzi 荀子 (Xun Zi) Yin Wen 尹文 Yin-Yang Jia 陰陽家 (School of Yin-Yang) Zheng Xuan 鄭玄 Zhu Xi 朱熹 Zhuangzi 莊子 (Zhuang Zi)

Chinese Classics Baima Lun 白馬論 (Disquisition on White and Horse) Chunqiu Guliang Zhuan 春秋穀梁傳 (Guliang’s Commentary of the Spring-­ Autumn Annals) Chunqiu Gongyang Zhuan 春秋公羊傳 (Gongyang’s Commentary of the Spring-­ Autumn Annals) Chunqiu Zuo Zhuan 春秋左傳 (Zuo’s Commentary of the Spring-Autumn Annals)

Chinese-English Glossary

547

Daqu 大取 (Greater Pick or Choosing the Greater) Gongsun Longzi 公孫龍子 (Gongsun Long Zi) Hanfeizi 韓非子 (Han Fei Zi) Huainanzi 淮南子 (Huai Nan Zi) Huangdi Neijing 黃帝內經 (Inner Canon of the Yellow Emperor) Jianbai Lun 堅白論 (Disquisition on Hard and White) Jing Shang 經上 (Canon A) Jing Xia 經下 (Canon B) Jingshuo Shang 經說上 (Explanation A) Jingshuo Xia 經說下 (Explanation B) Laozi 老子 (Lao Zi or Dao-De-Jing 道德經) Liji 禮記 (Li Ji or the Book of Rites) Lijizhu 禮記注 (Li Ji Zhu or Commentary of the Book of Rites) Lunyu 論語 (Analects) Lun Heng 論衡 (Balanced Discourse, Balanced Enquiries or Critical Essays) Lushi Chunqiu 呂氏春秋 (Spring and Autumn Annals of Mr. Lu) Ming Shi Lun 名實論 (On Name and Actuality) Mao Shi 毛詩 (Mao’s Commentary of the Book of Poetry) Maoshi Zhengyi 毛詩正義 (Explanatory Notes to Mao’s Commentary of the Book of Poetry) Mencius 孟子 (Meng Zi or Meng Zi) Mojing墨經 (Mohist Canons) Mozi 墨子 (Mo Zi) Shi Jing 詩經 (Book of Poetry) Shuo Wen Jie Zi 說文解字 (Shuowen Jiezì) (Describing single-component graphs and explaining compound characters) Tongbian Lun 通變論 (Disquisition on Continuity and Change) Wai Chushuo Zuo Shang 外儲說左上 (Outer Congeries of Sayings, Upper Left Series) Wenxin Diaolong 文心雕龍 (The Literary Mind and the Carving of Dragons) Xiaoqu 小取 (Lesser Pick or Choosing the Lesser) Xunzi 荀子 (Xun Zi) Yi Zhuan 易傳 (Commentary on the Book of Change) Yinwenzi 尹文子 Zhi Wu Lun 指物論 (On Referring to Things) Zhuangzi 莊子 (Zhuang Zi)

Author Index

A Ames, Roger, 2, 59, 78, 115, 116, 118, 438 Aristotle, 129, 130, 136–139, 162, 167, 254, 313, 377, 439, 441, 445, 452, 453, 489, 527 Au, Terry Kit-fong, 99, 101

B Baizhang, Huaihai, 391–393, 395 Bao, Zhiming, 336 Bloom, Alfred, 97–101 Brandom, Robert, 185

C Carnap, Rudolf, 226, 265, 497, 500 Chan, Wing-tsit, 42, 55, 58, 207, 360 Cheng, Chung-ying, 3, 37, 38, 59, 130, 131, 148, 300, 304, 402 Chmielewski, Janusz, 2, 3, 38, 148, 149, 209, 317–322, 377 Chong, Kim-Chong, 9, 351–366 Copi, Irving, 351–353, 415, 423, 451, 458 Creel, Herrlee, 55, 330, 333

D Davidson, Donald, 5, 80, 86, 87, 100, 182, 183, 400 Dewey, John, 144, 466, 474–476 Du, Guoping, 10, 511–519 Duyvendak, J.J.L., 181

E Elster, Jon, 108 Engelfriet, Peter, 452, 458–460 F Fang, Wan-chuan, 86, 87 Fodor, J.A., 185 Forke, Alfred, 102, 145, 147, 148, 173 Fraser, Chris, 6, 8, 44, 59, 113–125, 144, 154, 210, 240, 253, 264, 271, 285–305, 343, 345 Frege, Gottlob, 16, 37, 279, 312, 324, 534 Fung, Yiu-Ming, 1–10, 37, 59, 71–93, 97–110, 156, 181–201, 210, 213–230, 237, 275, 287, 296, 300, 304, 309–326, 403, 417, 426 Fung Yulan (Feng Yu-lan, Feng Youlan), 37, 131, 132, 135, 136, 138, 140, 240, 314, 315, 325, 333 G Gabbay, Dov M., 220 Gamut, L.T.F., 62 Geach, P.T., 6, 46, 62 Geaney, Jane, 5, 6, 15–30 Gernet, Jacques, 443, 444 Gongsun Long, 3, 6, 7, 9, 36–43, 46, 49, 51, 61, 62, 65, 68, 82, 86, 105, 106, 124, 164, 173, 175, 184, 193, 209, 210, 213, 218–222, 227, 239, 253, 267, 274, 286–289, 291, 293, 297, 298, 300–305, 309–326, 470, 489, 491, 492

© Springer Nature Switzerland AG 2019 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7

549

550 Graham, A.C., 2, 3, 44, 77–83, 97, 105, 113, 115, 118–124, 144, 148, 168, 174, 175, 186, 190, 192, 193, 196, 197, 209, 213, 218, 219, 224, 242, 253, 254, 259, 270, 280, 281, 287, 292, 295, 296, 300, 301, 304, 319, 320, 342, 356, 361, 377, 379, 381, 382, 438, 441, 442, 490 Granet, Marcel, 2, 97 Guo, Xiliang, 72 H Hall, David, 2, 59, 78, 115, 116, 118, 444 Hansen, Chad, 2, 6, 16, 37, 59, 75, 78–83, 85, 86, 113–116, 118, 120, 153, 154, 174, 192–194, 199, 210, 211, 219, 224, 227, 228, 295, 296, 300, 304, 311, 323, 324, 372 Harbsmeier, Christoph, 3, 59, 78, 83, 101, 102, 116, 119, 131, 148, 209, 210, 240, 241, 286, 295, 300, 301, 303, 372, 381 Hearne, James, 304, 320, 321 Ho, Chien-hsing, 10, 389–403 Hodges, Wilfrid, 197 Hu Qiguang, 74 Hu Shih, 4, 144, 309–315, 329 Huangbo, Xiyun, 394–396, 403 Hung, Bo, 88 I Im, Manyul, 300 Indraccolo, Lisa, 7, 171–179 J Jacquette, Dale, 1 Jiang, Xinyan, 6, 129–140 Jin, Yuelin, 466, 474, 481, 485, 493–495, 497, 499, 511, 512 Johnston, Ian, 161–163, 165–168, 174, 175, 200, 204, 213, 217, 218, 242, 377 K Kao, Kung-yi, 304 Knoblock, John, 117, 121, 183, 228, 238, 239, 247, 248, 275, 288, 330, 358, 359, 362–366 Kripke, Saul, 37, 57, 275, 325

Author Index Kurtz, Joachim, 144, 145, 465, 467–470, 475, 489, 491, 492 L Lai, Whalen, 304 Lange, Marc, 296 Lao, Sze-kwong, 215, 333 Lau, D.C., 55, 152–154, 174, 300, 343, 354, 356, 357 Legge, James, 19, 83, 84, 87, 91, 92, 105, 107, 150, 181, 189, 195, 236, 333 Leong, Wai Chun, 116, 119 Li, Kuangwu, 77, 493 Li, Zhaoguo, 194, 195 Li, Zhizhao, 453 Liao, W.K., 82, 87, 109, 185 Linski, Leonard, 312 Liu, Fenrong, 163, 258, 260, 261, 268, 275, 278, 519 Liu, Peiyu, 44 Loy, Hui-Chieh, 9, 114, 329–347 Lucas, Thierry, 8, 209, 233–248, 253–282, 296, 300 M Makeham, John, 335, 470, 489 McLeod, Alexus, 53, 113, 116, 118, 334 McNamara, Paul, 256 Mei, Y.P. (Mei, Yi-Pao), 106, 146, 183, 187, 196, 255, 260 Mill, John Stuart, 37 Møllgaard, Eske, 369–386 Mou, Bo, 33–68, 300 Mou, Zongsan, 473, 484, 485, 494, 527, 528 Munro, Donald, 113, 114, 118 N Needham, Joseph, 2, 97, 131, 132, 139, 317, 333 O Obenchai, D.B., 304 P Palmer, Frank, 83 Parsons, Terence, 92 Priest, Graham, 7, 62, 389, 390, 392, 402 Pulleyblank, Edwin, 101, 122

Author Index Q Quine, W.V., 1, 226–228 R Reding, Jean-Paul, 148–151, 154, 156, 304, 330, 333 Rieman, Fred, 304 Ricci, Matteo, 437–462 Robins, Dan, 116, 119, 120, 122 Roetz, Heiner, 59, 85, 116 Rorty, Richard, 437, 443 Ross, John, 82 Russell, Bertrand, 1, 36, 312, 467, 469, 484, 493, 495–497, 511, 518 Ryle, Gilbert, 265 S Salmon, Nathan, 313 Saunders, Frank, 116 Schleichert, Hubert, 304 Searle, John, 6, 79, 82, 83, 237, 312 Seligman, Jeremy, 258, 260, 261, 268, 278 Sellars, Wilfrid, 185 Shen, Yuting, 499, 511, 513, 514, 518, 519 Shi, Ningzhong, 239 Sider, Theodore, 62 Solomon, Bernard, 296, 300 Stemmer, Nathan, 226 Stevenson, Frank, 296, 304 Strawson, P.F., 36, 474 Suter, Rafael, 10, 465–500 T Tan, Jiefu, 490 Tang, Mingjun, 407–434 Thompson, Kirill Ole, 300 Trauzettel, Rolf, 300

551 V Van Benthem, Johan, 258, 260, 261, 268, 278 Van Norden, Bryan, 116, 330 Volkov, Alexeï K, 7, 143–157 W Wang, Dianji, 466, 471, 493, 494, 511 Wang, Hong, 311 Wang, Hongguang, 10, 511–519 Wang, Rongpei, 200, 311 Wang, Wen-Fang, 525–539 Wardy, Robert, 99, 448, 453, 470 Whitehead, Alfred North, 139, 496 Wiggins, David, 62 Williams, Joseph M., 87 Williamson, Timothy, 108 Wilson, Jessica, 226 Wittgenstein, Ludwig, 495, 500 Woods, John, 220 Wu Feibai, 217, 491 Wu, Zeqiang, 10, 525–539 X Xu, Guangqi, 437, 439, 442, 453, 454, 461 Xu, Keqian, 296 Y Yan, Fu, 144, 468, 470, 471, 489, 511 Yang, Wujin, 7, 8, 144, 147, 161–169, 203–211 Yu, Jiyuan, 445 Yuan, Jinmei, 10, 437–462 Z Zhang, Binglin, 144, 145, 147, 317, 466, 471, 483, 486, 487, 489, 490 Zhang, Jialong, 131, 519 Zhang, Wanqian, 8, 203–211

Subject Index

A Analogical arguments, 9, 145, 153, 155–156, 272, 291, 351, 352, 354, 366 Analogical reasoning, 44, 144–151, 154, 156, 210, 272, 454, 534 Analogy, 7, 44, 109, 143–157, 161, 176, 197, 271, 347, 351–357, 359–361, 363–366, 395, 488, 490, 514 Anaphora, 88, 90 Argumentation, 2, 3, 5, 7, 8, 10, 37, 43, 44, 98, 99, 125, 148, 151, 156, 157, 165, 167, 168, 171–179, 206, 208–210, 218, 242, 254–258, 271, 272, 277, 278, 282, 375, 489, 491 Arguments, 25, 38, 78, 99, 115, 134, 145, 163, 172, 184, 209, 218, 253, 287, 314, 331, 351, 378, 402, 407, 439, 488, 531 Aristotelian logics, 10, 129, 132, 138, 139, 144, 438–443, 445, 446, 449, 450, 453, 457, 458, 461, 526, 527, 530, 535, 538 Associate properness, 242, 340, 370, 372, 374, 437–462 Awakening, 384, 390, 391, 396–398, 400, 401 B Buddhist logic, 145, 411–414, 419, 420, 422, 431–434, 467, 471, 472, 485–488, 490, 492, 527, 528, 530, 535–536, 538, 539

C Cataphora, 88 Categorical syllogisms, 439, 441, 445, 451, 456, 457 Changes, 45, 46, 76, 91, 133, 136, 137, 140, 172, 175, 181, 197, 218, 229, 247, 288, 298, 301, 313, 330, 333, 341, 344, 359, 361, 373, 391, 439, 441, 445, 446, 466, 480, 525, 529 Chinese Logic, 15, 143, 203, 272, 438, 466, 513, 530 Class, 48, 76, 145, 169, 183, 203, 214, 233, 259, 287, 311, 332, 439, 472, 512 Collective designation, 45 Collective name, 45, 60, 61, 68, 204, 207 Comparative logic, 4, 496 Complementizer, 87–89 Context, 16, 34, 80, 99, 114, 130, 151, 168, 172, 183, 223, 235, 255, 287, 325, 330, 351, 396, 408, 438, 466, 538 Contradiction, 6, 9, 10, 49, 104, 124, 129–140, 153, 176, 178, 262, 280, 285, 305, 372, 389, 392–395, 400, 402, 403, 442, 476, 479, 518 Contradictory, 6, 52, 105, 106, 124, 130, 131, 167, 174, 280, 300, 304, 377, 389, 390, 392, 393, 395, 396, 399–402, 497 Contrary, 28, 52, 97, 102, 130, 131, 136, 139, 179, 191, 266, 274, 285, 292, 324, 393, 443, 445 Correcting names, 4, 9, 329–347

© Springer Nature Switzerland AG 2019 Y.-m. Fung (ed.), Dao Companion to Chinese Philosophy of Logic, Dao Companions to Chinese Philosophy 12, https://doi.org/10.1007/978-3-030-29033-7

553

Subject Index

554 Correlative thinking, 2, 97, 444, 534 Counterfactual, 97–110, 123, 183 Counterfactual conditionals, 6, 97–110 D Definiendum, 233, 235, 243, 245, 281 Definiens, 233–235, 243, 281 Definition, 15, 63, 84, 104, 147, 173–174, 184, 205, 226, 233, 258, 342, 353, 423, 440 Demonstrative (pronoun), 86–90, 93, 105, 216, 220, 325, 452, 457, 458, 460, 461 Descriptions, 1, 23, 29, 34, 36, 57, 62, 66, 74, 75, 79, 113, 117, 143, 144, 153, 196, 215, 236, 247, 270, 312, 313, 325, 352, 353, 356, 360, 366, 375, 377–379, 398, 454 Designation, 45, 57, 199, 246, 472, 492 Designators, 57, 58, 259, 325, 326 Dialectical logic, 132, 476–483 Dialectical materialism, 479, 482, 483 Dialectical relations, 7, 131, 139 Dialectical thinking, 131, 139, 140 Dialectics, 140, 174, 286, 289, 466, 473, 474, 476–483, 490, 512, 536 Difference, 17, 39, 76, 97, 115, 134, 167, 175, 196, 205, 213, 236, 262, 288, 309, 332, 353, 370, 395, 412, 444, 475 Direct reference, 57 Double reference, 6, 39, 49, 62 E Emptiness, 16, 22, 52, 360, 362, 395, 396, 398, 400, 403 Encounter dialogues, 10, 389, 390, 396–400, 402 Equivalence, 151, 214, 216, 223, 237, 241, 255, 257, 262, 263, 266, 269, 310, 366 Experimental logic, 474–476, 482 F Formal logic, 131, 139, 147, 234, 282, 410, 420, 466, 469–483, 485, 488, 499, 512, 527, 528, 533 G Grammaticalization, 87–89

H Hypothetical syllogism, 451 I Identity, 28, 33, 135, 175, 199, 214, 263, 291, 310, 361, 391, 456, 492, 517 Implication, 1–3, 15, 17, 18, 21, 22, 27, 28, 35, 38, 41, 42, 49, 59, 98, 108, 114, 122, 145, 147, 223, 255, 279, 310, 318, 329, 343, 344, 361, 386, 424, 426, 429, 444, 478, 481, 494, 495, 499, 513, 518, 531 Implicit logic, 2, 143, 253, 258, 281, 282, 492 Incommensurability, 3, 97, 100 Indexicality, 294 Indexicals, 295 Indication, 8, 260, 281, 357, 379, 399, 400, 421 Indirect discourse, 86, 92 Indirect speech, 87–89 Individual, 39, 87, 106, 133, 145, 172, 192, 203, 214, 233, 259, 287, 309, 336, 362, 392, 408, 457, 491 Induction, 144, 166, 352, 420, 441, 446, 454, 479, 500 Intensional, 168, 199, 200, 223, 275, 311 Intensional context, 199, 200, 223, 311 Intentions, 26, 35, 79, 82, 118, 151, 157, 173, 184, 199, 236, 243, 281, 290, 311, 343, 345, 397, 401, 483 K Kinds, 15, 39, 72, 101, 117, 135, 148, 161, 176, 184, 203, 213, 238, 256, 287, 311, 347, 353, 371, 389, 419, 438, 485, 511 L Law of non-contradiction, 6, 9, 129–130, 132, 136–139, 377, 379, 380, 382, 385, 456 Leibniz’s law of indiscernibility of identicals, 105, 302 Liar paradox, 277 Linguistic relativity hypothesis, 97 Logic in China, 414, 469–474, 511, 513, 519 Logical validity, 147, 149, 437–462

Subject Index M Mass-noun hypothesis, 36, 37, 59, 75, 193, 210, 211 Mass nouns, 37, 59, 60, 210, 227, 228 Mass-stuffs, 192–195, 224, 227, 228, 230, 324 Material inference, 185 Mathematical logic, 1, 3, 209, 465, 467, 469, 472, 474, 475, 481, 482, 484, 493–500, 511, 512, 516, 517, 528, 530–533 Meaning, 2, 15, 39, 75, 100, 151, 169, 172, 185, 203, 216, 234, 259, 312, 338, 370, 399, 409, 443, 472, 526 Mereological, 37, 59, 60, 192, 224, 227 Metaphors, 2, 28, 145, 150, 154–156, 163, 176, 236, 321, 352, 353, 357, 362, 363, 365, 400 N Names, 3, 15, 36, 74, 114, 132, 152, 162, 177, 203, 214, 239, 258, 285, 314, 329, 352, 371, 393, 414, 438, 469, 531 Naming, 2, 4–6, 19, 33–68, 75, 80–82, 85, 192, 206, 211, 215, 216, 237, 259, 260, 293, 301–303, 338, 344, 347, 369–372, 380–382, 386, 450, 492 Nominalism, 36, 38, 226, 315 No-mind, 395, 396, 398, 400 Non-classical logic, 62, 518 Non-monotonic reasoning, 519 Nouns, 36, 71, 102, 119, 181, 210, 224, 342, 370, 438 O Original mind, 391, 394–396, 399–400, 403 Ostensive definition, 84, 234, 440, 458, 459 P Paradox, 8, 132, 176, 280, 288, 290–292, 297–299, 301, 304, 305, 310, 314, 319, 363, 389, 390, 394, 395, 399, 402, 495, 499, 518, 519, 531, 532 Paradoxical expressions, 9, 390, 393–397, 399, 401 Paradoxical language, 9, 10, 389–403 Parallel inference, 37, 43–49, 196, 197, 200, 201, 218, 222–224 Parallelism, 6, 77, 145, 147, 148, 157, 196, 201, 218, 222, 242, 255–257, 273, 275, 310, 311, 339

555 Paratactic, 6, 86–90, 93, 101, 103, 107 Performative contradiction, 377–379, 383 Perspectivalism, 296 Philosophical logics, 1, 2, 4, 483, 517–519, 527–533 Philosophy of logic, 1–10, 33–35, 208–210, 244, 488, 494, 519, 531–533 Pivotal sentence, 92, 93 Pragmatist logic, 466, 476 Predicates, 9, 47, 79, 101, 122, 130, 144, 168, 197, 219, 233, 258, 291, 321, 392, 416, 439, 494, 511, 527 Principle of charity, 5, 40, 47, 100 Principles, 30, 33, 76, 100, 124, 132, 145, 175, 183, 209, 218, 238, 271, 294, 324, 333, 359, 371, 402, 408, 438, 469, 518, 526 Propositional attitudes, 6, 71–93 Propositions, 77, 115, 129, 144, 167, 172, 186, 220, 258, 318, 356, 376, 408, 439, 470, 511, 528 Q Quantifiers, 41, 63, 67, 108, 228, 258, 262, 268, 273, 280, 304, 324, 451, 531 R Realism, 36, 38, 48, 192, 213, 316, 317 Reality, 2, 4, 6, 16–18, 20, 25, 28, 30, 33, 34, 36, 40, 41, 50, 52, 54, 59, 82, 85, 98, 100, 103, 113, 131, 132, 139, 140, 176–179, 189, 214, 217, 240, 242, 244, 285, 298, 301, 312, 315, 333, 361, 371, 372, 378, 382, 390, 400–402, 418, 423, 478, 484, 534 Reasons, 18, 38, 81, 100, 115, 132, 143, 161, 173, 187, 210, 216, 239, 253, 285, 318, 330, 358, 383, 396, 408, 438, 467, 517, 538 Reductio ad absurdum, 7, 104, 176, 178, 179, 238, 355 Reference, 16, 33, 80, 116, 145, 192, 203, 224, 256, 292, 312, 333, 354, 378, 413, 451, 474, 531 Referents, 17, 35–37, 40, 46, 48, 50, 51, 57, 60, 64, 66, 122, 124, 194, 196, 216, 218, 223, 229, 295, 299, 301, 304, 312, 315, 316, 326, 342, 391–393, 395, 396, 398–401, 474 Relative identity, 6, 46, 49, 61, 62, 65, 68 Rigid designators, 57, 325, 326

Subject Index

556 S Sameness, 5, 8, 41, 43, 47, 48, 134, 135, 164, 168, 175, 196, 204–206, 209, 213–230 Sapir-Whorf hypothesis, 97, 99 Semantic sensitivity, 36, 43, 46, 47, 49, 61, 68 Sentence-belief, 75, 78 Sentential contexts, 39, 40, 48, 80 Serial verb, 92, 93 Sets, 43, 83, 100, 117, 136, 150, 178, 204, 227, 246, 257, 290, 317, 341, 356, 373, 389, 421, 438, 516, 528 Skepticism, 100 Small clause, 93 Sophists, 76, 150, 156, 163, 173, 301–304, 309, 310, 374, 382, 489, 491 Sorts, 17, 24, 64–66, 102, 139, 163, 175, 197, 209, 227, 262–267, 276, 288, 289, 301, 341, 343 Syllogisms, 144, 147, 167, 176, 318, 418, 437–462, 526–528 Syllogistics, 143–145, 167, 254, 317, 318, 418, 438–442, 448–455, 458, 467, 469, 472, 474, 475, 485, 488, 490

T Term-beliefs, 80 Terms, 15, 33, 72, 103, 116, 131, 144, 162, 177, 182, 205, 213, 243, 253, 285, 310, 332, 351, 371, 392, 408, 439, 472, 514 That-clauses, 6, 80, 86–93, 223 Thing-kind, 193–195, 224, 227 Three standards, 114 Transcendental logic, 466, 478, 483–485 Truth, 34, 78, 100, 113, 136, 167, 190, 223, 261, 314, 338, 361, 379, 390, 408, 440, 499, 532 U Unity of opposites, 7, 131, 133, 134, 138, 140 Universal, 1, 63, 108, 136, 189, 216, 239, 253, 298, 309, 362, 418, 439, 473, 516 Unspeakable, 36 Z Zeno’s arrow paradox, 299 Zero form of that, 88–90, 93