Вы находитесь на странице: 1из 303

The Mozi as an Evolving Text

Studies in the History of Chinese Texts

Edited by

Martin Kern, Princeton University Robert E. Hegel, Washington University, St. Louis Ding Xiang Warner, Cornell University

VoLUME 4

The titles published in this series are listed at brill.com/hct

The Mozi as an Evolving Text

Different Voices in Early Chinese Thought

Edited by

Carine Defoort and Nicolas Standaert

Text Different Voices in Early Chinese Thought Edited by Carine Defoort and Nicolas Standaert LEiDEN  •
Text Different Voices in Early Chinese Thought Edited by Carine Defoort and Nicolas Standaert LEiDEN  •

LEiDEN  boSToN

2013

Cover illustration: The Chinese text on the cover is from Tang Yaochen 唐堯臣 (16th century), Mozi 墨子, Ming woodblock edition from 1553 repr. in Mozi daquan 墨子大全, eds. Ren Jiyu 任继愈 and Li Guangxing 李广星, beijing: beijing tushuguan chubanshe, 2004, vol. 3, 131.

Library of Congress Cataloging-in-Publication Data

The Mozi as an evolving text : different voices in early Chinese thought / edited by Carine Defoort and Nicolas Standaert.   pages cm. — (Studies in the history of Chinese texts ; volume 4)  includes bibliographical references and index.  iSbN 978-90-04-23434-5 (hardback : alk. paper) — iSbN 978-90-04-24620-1 (e-book) 1. Mo, Di, fl. 400 b.C. Mozi. i. Defoort, Carine, 1961– author, editor of compilation. ii. Standaert, N., author, editor of compilation.

 b128.M8M627 2013

 181’.115—dc23

2013000783

This publication has been typeset in the multilingual “brill” typeface. With over 5,100 characters covering Latin, iPA, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more information, please see www.brill.com/brill-typeface.

iSSN 1877-9425 iSbN 978-90-04-23434-5 (hardback) iSbN 978-90-04-24620-1 (e-book)

Copyright 2013 by Koninklijke brill NV, Leiden, The Netherlands. Koninklijke brill NV incorporates the imprints brill, Global oriental, Hotei Publishing, iDC Publishers and Martinus Nijhoff Publishers.

All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher.

Authorization to photocopy items for internal or personal use is granted by Koninklijke brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.

This book is printed on acid-free paper.

Dedicated to Watanabe Takashi, Angus Graham, and Roman Malek, inspiring Mozi scholars

CoNTENTS

introduction: Different Voices in the Mozi: Studies of an Evolving Text  Carine Defoort and Nicolas Standaert

1

1. Are the Three “Jian Ai” Chapters about Universal Love?  Carine Defoort

35

2. How to End Wars with Words: Three Argumentative Strategies by Mozi and His Followers  Paul van Els

69

3. Mozi 31: Explaining Ghosts, Again  Roel Sterckx

95

4. Mozi’s Remaking of Ancient Authority  Miranda Brown

143

5. The Ethics of the Mohist Dialogues  Chris Fraser

175

6. From “Elevate the Worthy” to “intimacy with officers” in the MoziHui-chieh Loy

205

7. Heaven as a Standard  Nicolas Standaert

237

bibliography 

271

References to the Mozi

281

Subject index 

287

IntroductIon: dIfferent VoIces In the Mozi:

studIes of an eVolVIng text

carine defoort and nicolas standaert

子墨子曰: 吾非與之並世同時, 親聞其聲, 見其色 也以其所書於竹帛, 鏤於金石, 琢於槃盂, 傳遺 後世子孫者知之.

Master Mozi said: “since I was not alive when the [sages] lived, I have not personally heard their voices or seen their faces. It is because of what they wrote on bamboo and silk, carved in metal and stone, engraved on plates and bowls, and passed on to their descen- dants, that I know it.”

Mozi 16: 28/29–29/1

Mo di 墨翟 (ca. 479–381 Bce) claims to know that the ancient sages were caring and compassionate even though he has not personally heard their voices or seen their faces. fortunately for him, their writings were pre- served on bamboo and silk, metal and stone, or plates and bowls. so he could use their authority to promote his own novel ideas among the ruling elite of his day. he himself, however, was not so lucky: the book named after him was not carved in metal or stone, and it fared less well than the sages’ writings. the Mozi 墨子, a book of seventy-one units,1 was seriously neglected in the course of chinese history partly due to its perceived low literary value and uninteresting content. this agelong neglect has caused such serious textual corruption and interpretive difficulties that even con- temporary scholars are often reluctant to tackle this text. nevertheless, the authors of the current volume have chosen this voluminous source of Mohist thought—or, at least, its best-preserved parts—as their topic. Written over a period of some two hundred years (roughly in the fourth and third centuries Bce) and possibly put into its current shape during the han dynasty, the Mozi appears to have been largely forgotten until its

1 only fifty-three pian (units, chapters) are extant. But the fact that a Mozi version in seventy-one pian was listed in Hanshu 30.1738, has led to the belief that it originally had seventy-one chapters. for the textual history of the text, see Maeder, “some observations on the composition of the ‘core chapters’ of the Mozi,” 29–34.

2

carine defoort and nicolas standaert

inclusion in the Daozang 道藏 (daoist canon) published in 1447.2 despite some emerging attention from the Ming dynasty onward, serious interest in Mozi began only with the textual studies of the Qing dynasty—more specifically, those studies conducted by scholars such as Bi Yuan 畢沅 (1730–1797), Wang niansun 王念孫 (1744–1832), Wang Yinzhi 王引之 (1766–1834), Yu Yue 俞樾 (1821–1907), and sun Yirang 孫詒讓 (1848– 1908).3 Missionary interest emerged with James legge (1861) and ernst faber (1877)4 in the nineteenth century and was followed in the twentieth and twenty-first centuries by Japanese,5 chinese, and Western scholarship and translation. the first Western translation, almost complete, was in german, by alfred forke (1922).6 Important partial english translations were made by Mei Yi-pao 梅貽寶 (1929),7 Burton Watson (1963),8 angus c. graham (1978),9 and Philip Ivanhoe and Bryan Van norden;10 most recently, a complete translation by Ian Johnston (2010) has appeared.11 even though the Mozi is still not a hot topic in academic research, there has been an increasing interest during the last decades: there have been studies on Mohist thought or philosophy, on the social and geographi-

   2 In this edition, which forms the basis of the presently transmitted version, eighteen out of the seventy-one chapters were already missing. for the four earliest extant Ming editions and their supposed song source, see durrant, “an examination of textual and grammatical Problems in Mo Tzu,” 63–68.    3 for an overview of Mohist studies, see Zheng Jiewen, Ershi shiji Moxue yanjiushi. this study does not mention any Japanese or Western Mozi research. li Quanxing, “ershi shiji Mozi yanjiu lunzhu suoyin 20” also includes Japanese scholarship. for a brief overview of different and more recent trends in chinese Mozi research, see the preface to defoort, “Mo Zi research in the People’s republic of china.”    4 see faber, Die Grundgedanken des alten chinesischen Socialismus. legge translates Mozi’s “universal love” writings and discusses their connection to Yang Zhu and Mencius in his The Chinese Classics, vol. 2, 103–126.    5 see hashimoto sumiya, “riben de ‘Mozi’ yanjiu gaiguan”; tan Jiajian, Mozi yanjiu, appendix 3, 623–644.    6 see forke, Mê Ti. forke’s translation of the defense chapters is rather a paraphrase. for more on forke’s study and translation, see Maeder, “some observations on the com- position of the ‘core chapters’ of the Mozi,” 35–37.    7 see Mei, The Ethical and Political Works of Motse. It was republished with the chinese original and a modern chinese translation added under the title The Works of Motze. It contains a complete translation of the opening chapters, core chapters, and dialogues.    8 see Watson, Mo-tzu, Basic Writings.    9 for a study and translation of the dialectical (or logical) chapters (40–45) of the Mozi, see graham, Later Mohist Logic, Ethics, and Science. 10 Ivanhoe and Van norden, Readings, 55–109, is a partial translation of the core chapters. 11   see Johnston, The “Mozi.” for a longer list of chinese and Western translations of the Mozi, see ibid., lxxviii–lxxxi. Johnston is preparing a new translation for the Penguin series.

introduction: different voices in the mozi

3

cal provenance of Mo di and his followers, on the nature and history of the Mohist school or lineage (mojia 墨家), on its division and perceived demise by the end of the Warring states, on its contemporary relevance for and influence (or lack thereof) on chinese culture, and so forth. some research more narrowly concerns the book Mozi, asking questions about its composition, history, textual corruption and reconstruction. the pres- ent volume feeds into this last domain by focusing on the three most read- able parts of the Mozi, namely the opening chapters (chapters 1–7), the core chapters (chapters 8–37), and the dialogues (chapters 46–49/50). With their focus on moral, political, and social matters, these three parts are distinguished from two other parts that are generally identified as being of a more “technical” nature: the dialectical chapters or Mohist canon 墨經 (40–45)12 and the defense or Military chapters (52–71).13 these two somewhat later parts are not discussed in this volume because they are different in style, very technical in content, and bedeviled by tex- tual corruption.14 Versions of the essays collected in this volume were originally presented during workshops and seminars at the university of leuven (Belgium), where the Mozi has been a research topic for a decade. the discussions and reflections during these scholarly meetings shaped the topic of the current volume and, more specifically, its background hypothesis: despite variations in content and approach, all contributors share an awareness of the differences that can be found in the book Mozi, not only between its major parts but also within the parts, the chapters, and the fragments. therefore, generalizing statements about “the” Mohists or Mohist thought in general will often make way for the possibility of different voices in the text and for more specific questions about evolutions or tensions within the three parts of the Mozi identified above. It is not our intention to deny

12 the Western opus magnum about the dialectical chapters is graham, Later Mohist Logic, Ethics, and Science. although impressive in its achievement, this book is not always very easy to consult. for a criticism of graham, see geaney, “a critique of a.c. graham’s reconstruction of the ‘neo-Mohist canons’.” see also Johnston, “choosing the greater and choosing the lesser”; and Johnston, The “Mozi,372–373. 13 robin Yates’s dissertation is probably the most complete source on these chapters:

“the city under siege: technology and organization as seen in the reconstructed text of the Military chapters of the Mo Tzu” (harvard university, 1980). see also Yates, “the Mohists on Warfare”; and Johnston, The “Mozi,732–733. 14 these two parts are tentatively dated around the late fourth and third centuries Bce. see graham, Later Mohist Logic, Ethics, and Science, 337–338; Johnston, The “Mozi,xxxii– xxxiii; fraser, “Mohism,” see “supplement to Mohism: text and authorship.”

4

carine defoort and nicolas standaert

the often noticed and claimed unity of Mohist thought, but rather to com- plement this generalizing reading with a more detailed account. this interest in differences within the Mozi has, in turn, shaped three basic methodological assumptions that are shared by the editors of this volume and have to some extent influenced the contributions. first, we keep a firm focus on the text rather than its context. to put it bluntly:

we think not so much of a master with disciples and opponents bringing about a text, but rather of a text describing (and thereby creating) a mas- ter, disciples, and opponents. When detecting information in the written source about the lives, status, or provenance of its authors or its audience, we refrain from making strong inferences about their historical existence. only when specific issues posed by matters of style, content, rhetoric, or grammar make these matters relevant for our purposes do we occasion- ally reflect on them. a second idea is that we attribute the differences in the core chapters (and some other chapters) mainly to an evolution over time and not to rivalry between opposing Mohist sects. We therefore offer some suggestions about the chronological arrangement of several core chapters. our last leading thought concerns the titles of the core chap- ters, which may have been added at a relatively late stage. We believe that these titles may reflect the shape into which Mohism had evolved by the end of the Warring states period. and we are convinced that they have to a considerable degree influenced interpretations of early Mohism until today. for a fresh interpretation of the text, it may be fruitful to read the chapters while temporarily ignoring their titles. Before presenting the various contributions of this volume, we briefly outline these three guiding ideas that, for want of a better phrase, we call “basic assumptions.”

First Basic Assumption: Focus on the Text

the received Mozi consists of 71 numbered units, conventionally called “chapters” (or “books,” pian ), which were transmitted on fifteen rolls ( juan ). the label “early Mozi” usually refers to the core chapters (8–37), which are believed to date from the early fourth until the early third century Bce and to contain the original ideas of Master Mo and his followers.15 they are often framed as Master Mozi’s responses to sup-

15 graham believed that they go back to the beginnings of the school and are not later than 350 Bce. a. taeko Brooks ranges the core chapters from ca. 390 to ca. 273 Bce. Watanabe takashi postulates a much longer time span for the evolution of the core chap-

introduction: different voices in the mozi

5

posed objections of opponents. although only 23 out of these 30 chapters are extant, the short description of a 71-pian Mozi preserved in the Book of Han suggests that by the han the core of Mozi had already assumed its current shape, consisting of ten sets of three chapters each. conse- quently, these chapters have often been called “triplets” or “triads.” the three chapters in each triplet carry the same title, and the ten titles are believed to reflect the ten “dogmas” or “theses” of early Mohism.16 even though there has been disagreement about the translation of these titles (some are discussed in the contributions to this volume), the general con- tent of the core ideas is relatively clearly reflected by them.17 throughout this volume, we normally transliterate and translate the titles of the trip- lets as follows: “shang xian” 尚賢 (elevate the Worthy; chapters 8–10), “shang tong” 尚同 (conform upward; 11–13), “Jian ai” 兼愛 (Inclusive care; 14–16), “fei gong” 非攻 (against Military aggression; 17–19), “Jie yong” 節用 (Moderation in expenses; 20, 21, with 22 missing), “Jie zang” 節葬 (Moderation in Burials; 25, with 23 and 24 missing), “tian zhi” 天志 (Will of heaven; 26–28), “Ming gui” 明鬼 (explaining ghosts; 31, with 29 and 30 missing), “fei yue” 非樂 (against Music; 32, with 33 and 34 miss- ing), and “fei ming” 非命 (against fatalism; 35–37). the three chapters within each triplet are distinguished as, respectively, shang (upper), zhong (Middle), and xia (lower). “fei ru” 非儒 (against the ru; 39, with 38 missing) is sometimes called a “duplet” or “diad” because it was registered as two chapters and not as a “triplet” or “triad”. the status of the sole extant “fei ru” as a core chapter is questionable, even though its title and position in the corpus both suggest that it could be consid- ered the “eleventh dogma,” containing Mohist criticism of the ru (the classicists, erudites, confucians). not only its exceptionally polemic tone but also its composition, content, and grammar suggest that this duplet postdates the core chapters and fits better with the dialogues.18

ters, from the early fourth century to the end of the third century Bce. for more informa- tion on these views, see below. 16 Hanshu 30.1738. for more details on the textual history of the Mozi, see Maeder, “some observations on the composition of the ‘core chapters’ of the Mozi,” 29–34; and graham, Later Mohist Logic, Ethics, and Science, 65. 17 for a brief discussion of the core chapters and a summary of the discussions about their titles, see Johnston, The “Mozi,xxxiv–lxvi. We discuss the titles in more detail below. 18   for these and other reasons to exclude “fei ru” from the core chapters, see below and desmet, “all good things come in threes,” 224–243; a. taeko Brooks, “the Mician ethical chapters,” 105–106; and ding sixin, “a study on the dating of the Mo zi dialogues and the Mohist View of ghosts and spirits,” 51–52.

6

carine defoort and nicolas standaert

the dialogues (46–49/51) consist of anecdotes, sayings, and conversa- tions between Mozi and his disciples or opponents. slightly postdating or perhaps partly overlapping with the core chapters, they are tentatively attributed to the master’s first generation of disciples and dated to the middle of the fourth century Bce.19 like the core chapters, they discuss a mixture of moral, social, and political matters, but stylistically they are framed as actual dialogues between Master Mozi and historical per- sons. to have such a collection of sayings and wise responses of a master postdating the relatively structured essays of the core chapters seems to reverse the chronological evolution recognized in early chinese texts. the oldest-known texts are often a collation of rather fragmentary nota- tions, such as wise enunciations of or (staged) dialogues with a master, while somewhat structured statements resembling essays postdate them.20 Indeed, these dialogue chapters have been called the “Mohist Analects” and were perhaps modelled on the Lunyu 論語.21 the first chapter, “geng Zhu” 耕柱 (geng Zhu; 46), is named after a disciple of Mozi; “gui yi” 貴 義 (Valuing Morality; 47) mostly contains sayings attributed to Master Mo; “gongmeng” 公孟 (gongmeng; 48) is named after a ru who opposes Mohist views; and “lu wen” 魯問 (lu’s Questions; 49) is a record of con- versations with the ruler of lu. unlike these four chapters, “gongshu” 公輸 (gongshu Pan ; 50) is a long narrative about Mozi convincing the king of chu to call off an attack on song. It is not always counted among the dialogues because of this stylistic difference as well as its military con- tent, which is more in line with the defense chapters (52–71) immediately following the dialogues in the received Mozi. since the content and the title of chapter 51 are lost, we cannot determine its nature. the last group of chapters that we discuss are the seven short mis- cellaneous essays at the beginning of the book, which we call opening chapters. they have also been labeled “appendices,” “digests,” “epitomes,”

19   see fraser, “Mohism,” see “supplement to Mohism: text and authorship.” ding sixin, “a study on the dating of the Mo zi dialogues and the Mohist View of ghosts and spirits,” 73 dates some of them in the Qin and han dynasties. 20 see, e.g., fu sinian, “Zhanguo wenji zhong zhi pianshi shuti—yige duan ji,” 17–21; and Boltz, “the composite nature of early chinese texts.” 21   some scholars believe that the dialogues predate the core chapters and portray the historical Mo di in interaction with his actual disciples and rivals. for a recent recapitu- lation of the arguments, see Zheng Jiewen, zhongguo Mo xue tongshi, 4, 46. for an early refutation of this view, see durrant, “a consideration of differences in the grammar of the Mo Tzu ‘essays’ and ‘dialogues,’ ” 255–256; and, more recently, ding sixin, “a study on the dating of the Mo zi dialogues and the Mohist View of ghosts and spirits,” 57, 73.

introduction: different voices in the mozi

7

or “summaries.”22 they contain a variety of topics, some in line with the core chapters and others remarkably ru in content. their authenticity, affiliation, and textual history have been a topic of debate. chris fraser considers them “probably the latest part of the corpus,” and taeko Brooks tentatively dates them, in reverse order, from 270 to 250 Bce. she calls them “singlets,” as distinguished from the triplets and duplet presented above.23 Various scholars consider the first three opening chapters spuri- ous and un-Mohist: these are “Qin shi” 親士 (Intimacy with officers; 1), “xiu shen” 脩身 (cultivating the self; 2), and “suo ran” 所染 (What has Been dyed; 3). others believe that the last four chapters consist of fragments of otherwise lost material: “fa yi” 法儀 (standards and norms; 4), “Qi huan” 七患 (seven Misfortunes; 5), “ci guo” 辭過 (eschewing faults; 6), and “san bian” 三辯 (three arguments; 7).24 the opening chapters belong to the better-preserved and nontechnical parts of the Mozi, those that that we discuss in this volume. this volume certainly does not aim at providing a complete study of all chapters included in these three parts but rather takes the core chapters, dialogues, and opening chapters as the scope in which all the contri- butions fall. following the chronological order that we attribute to the chapters, the volume begins with three studies that focus on the core chapters: one on the “Jian ai” triplet, by carine defoort; one on the “fei gong” triplet, by Paul van els; and one on the sole remaining chapter of the “Ming gui” triplet, by roel sterckx. the contribution on the author- ity of the ancient sages by Miranda Brown mostly concerns all three parts of the Mozi. the next study, by chris fraser, mainly discusses the dialogues. and the last two contributions, by hui-chieh loy and nicolas standaert, each start from one opening chapter, namely “Qin shi” and “fa yi,” respectively. our focus on the book Mozi rather than on the reality hidden behind it is not meant to deny that the parts, chapters, fragments, or paragraphs

22 these labels are used by many scholars, such as Mei, durrant, graham, Maeder, lowe, and Johnston. 23 fraser, “Mohism,” see “supplement to Mohism: text and authorship”; Brooks, “the Mician ethical chapters,” 107, 117. there is a consensus on the relative lateness of the first seven chapters. see, e.g., Wu Yujiang et al., “Mozi gepian zhen wei kao,” 1025–1026; and Johnston, The “Mozi,xxxii. 24 Mei, The Works of Motze, 2, notes that the first three chapters “are judged to be spuri- ous almost unanimously by competent textual critics.” hu shi, zhongguo zhexueshi dagang, 133, believes that the seven opening chapters are all later forgeries: the first three are not Mohist at all; the last four are constructed out of lost Mohist fragments. for an overview of various chinese views about these chapters, see Yang Yi, Mozi huanyuan, 19–23, 211–12.

8

carine defoort and nicolas standaert

have been written down by and for actual persons. But in the case of the Mozi much has been said and little can be ascertained about these “Mohists.”25 that is one reason why we refrain from speculating on the historical identity of the persons behind the text, whether the authors/ editors or the audience. as for the authors/editors, we alternately ascribe the views expressed in these chapters to “Mozi” (the master after whom the book is named), “the author(s) of this chapter,” or sometimes “the Mohist(s).” We thereby do not insist on the “strong authorship” of any of these persons. on the contrary, like many other early chinese sources, the various chapters may well have been collected and (re)edited at vari- ous times on the basis of older fragments circulating among a group of like-minded people. the authors we have in mind are those “scholar- editors,” or “bricoleurs,” who,26 for their own reasons, gave these various chapters their current composite structure.27 Mozi himself was, of course, not that author, but rather the authority to whom the writers referred. In that sense, “our Master Mozi” 子墨子 was to some extent a creation of the book to which he was expected to lend legitimation and inspira- tion. even though the text presents itself as created by persons (a master and his disciples), these persons were also created by the text. as Mark lewis has argued, “the text, the master, and the disciples were inextricably bound together. Without the text there was no master and no disciples (beyond the lives of the individuals involved); without the master there was no authoritative text or transmitters of the text; without the disciples the text was not written or transmitted.”28 Besides this crucial triangle of text-master-disciples, the audience was also both the cause and the result of the book. We tend to agree with dan robins that the opponents staged in the Mozi defended customs and established privileges of the

25 for a summary of reflections on the identity of Mo di and his followers, see John- ston, The “Mozi,xviii–xxv. for a careful attempt to distinguish between the “Mohists” as a social group, “Mohism” as an intellectual orientation, and “Mohist-inspired thinking,” see Brindley, “ ‘the Perspicuity of ghosts and spirits’ and the Problem of Intellectual affilia- tions in early china,” 230–234. 26 for “scholar-editor,” see Boltz, “the composite nature of early chinese texts,” 59; for “bricoleur,” see Maeder, “some observations on the composition of the ‘core chapters’ of the Mozi,” 81–82. Judging from their novel ideas and relatively exclusive selection of tex- tual fragments, the earliest Mozi authors/editors probably belonged to a relatively closed group, the latest perhaps dating from the han dynasty. 27 for the composite structure of early chinese texts as opposed to individual author- ship of integral, structurally homogeneous texts, see Boltz, “the composite nature of early chinese texts,” 70–71; and Maeder, “some observations on the composition of the ‘core chapters’ of the Mozi,” 28, 82. 28 lewis, Writing and Authority, 58.

introduction: different voices in the mozi

9

ruling elite rather than theories of particular philosophical lineages or schools.29 We also agree with him, against the conventional interpreta- tion, that the opponents in the Mozi do not coincide with the ru, who are seldom mentioned in the book and never in the core chapters.30 the opponents—anonymously staged in the core chapters and presented as specific individuals in the dialogues—play an important role in bringing up objections that are, of course, all convincingly refuted by the master. even though in these two parts of the Mozi the setting still exudes the master’s authority, this is gradually taken over by the increasingly subtle argumentation that one would expect in a philosophical essay. this evo- lution of increasing opposition and refutation can be perceived in these chapters and will be discussed further on.

Second Basic Assumption: Evolution in the Core Chapters

Most studies of Mohist thought tend to consider the core chapters as representative of Master Mo’s original ideas. they attribute to each trip- let one consistent vision, such as the promotion of universal love or the rejection of aggressive warfare, and therefore quote from any of the three chapters to illustrate the relevant thesis or dogma.31 But those who focus on the book Mozi have long been fascinated by the threefold structure of the core chapters. Qing scholars started searching for explanations of the differences within the triplets: did the upper (shang), Middle (zhong), and lower (xia) chapters represent three sets of lecture notes, different branches within Mohism, or stages in its evolution? Yu Yue 俞樾, luan tiaofu 欒調甫, fang shouchu 方授楚, alfred forke, Watanabe takashi 渡邊卓, stephen durrant, angus graham, taeko Brooks, chris fraser, and Karen desmet, among others, have identified consistent differences among the triplets on the basis of such things as particle use, vocabulary, compounds, fixed formulas, rhetoric, style, references to authority, use of

29 see robins, “the Moists and the gentlemen of the World,” 388–389. If anything, the Mohist insistence on using good reasons and objective criteria must have initiated the (philosophical) debate rather than joined it. 30 see ibid., 386. the ru are explicitly attacked in “fei ru” (chapter 39) and to a lesser extent in “gong Meng” (chapter 48). otherwise, they are not explicitly mentioned in the Mozi. 31   Zheng Jiewen considers the core chapters more mature and later than the dia- logues, and traces an evolution between the ten triplets, but he never mentions any evo- lution or difference within the triplets, nor does any chinese author that he discusses in his overview of Mozi research. see Zheng Jiewen, zhongguo Mo xue tongshi, 1–24.

10

carine defoort and nicolas standaert

logic, and the intellectual, political, social, or technological content.32 the following overview outlines the three steps in this debate that have most guided our own reflections: the “three-sects theory” as presented by a. c. graham, the “sequence theory” as presented by taeko Brooks, and the alternative “evolution theory” defended long ago by Watanabe takashi.33

The Three-Sects Theory

Inspired by stephen durrant’s study of the Mozi,34 a. c. graham appor- tioned each triplet (or triad) among three rival sects or factions, which he labeled “Purist,” “compromising,” and “reactionary.” to reach this conclu- sion, he first distinguished grammatical features and vocabulary in the different chapters and then went on to look for differences in content. he thus began by dividing most chapters into three groups, named Y, h, and J after a special grammatical feature: the Y chapters cite Mozi after the opening sentence with the formula zi Mozi yan yue 子墨子言曰 instead

of 子墨子曰 (therefore called Y[an] chapters); the h chapters replace the

postverbal particle yu by hu when possible (therefore called h[u] chapters); and the J chapters use the particle ran ( jan in Wade-giles transcription; therefore called J chapters) after citing an ancient source. on the basis of their content, graham argued that these chapter groups

were written by three competing sects into which Mohism is said to have divided according to Han Feizi “xian xue” 顯學 (eminent learning] chapter 50) and zhuangzi “tianxia” 天下 (the World; chapter 33): the

Y group was seen as defending the purest and most radical Mohist doc-

trine and as residing in the northern part of the realm; the h group, also from the north, was somewhat more accommodating to political reali- ties; and the J group, in the south, was the most accommodating to poli- tics and therefore farthest removed from the original doctrine.35 graham

32 for an overview of various theories concerning the threefold nature of the core chapters, see durrant, “a consideration of differences in the grammar of the Mo Tzu ‘essays’ and ‘dialogues,’ ” 253–255; Yang Yi, Mozi huanyuan, 213–215. 33 these three are not the first nor the only scholars presenting such theories on Mozi, but they represent three steps in our initial acquaintance with the debate. other scholars, mainly chinese and Japanese, are mentioned in the notes. We refer to Karen desmet, “all good things come in threes,” 17–71 for a fuller overview of the topic. 34 see durrant, “an examination of textual and grammatical Problems in Mo Tzu”; and durrant, “a consideration of differences in the grammar of the Mo Tzu ‘essays’ and ‘dialogues.’ ” the three-sects theory was first suggested by Yu Yue in his preface (“Yu xu” 俞序) to sun Yirang’s Mozi jiangu. 35 graham, Divisions in Early Mohism Reflected in the Core Chapters of “Mo-tzu,18–19.

introduction: different voices in the mozi

11

concluded: “We can well understand why the Mohist sects disputed so fiercely. It would seem to the Purist that out of eagerness for political power the true teachings of Mo-tzu had been shamefully diluted by the compromisers and utterly betrayed by the reactionary.”36 although these labels could also be interpreted as reflecting an evolution, graham saw them rather as matching with roughly coexistent and rival sects each using the dialect of their own region.37 since three core chapters did not fit this framework, he considered them later additions: either as “digests” of the Mohist doctrine (14, “Jian ai, shang”; and 20, “Jie yong, shang”) or as a dislocated “manuscripts (fragment)” (17, “fei gong, shang”).38 his divi- sion of the core chapters can be graphically represented as follows: 39

table 1. the division of the core chapters according to graham

triplets

digests and

Y

h

J

fragment

Purist

compromising

reactionary

尚賢

8

9

10

尚同

11

12

13

兼愛

14

15

16

非攻

17

18

19

節用

20

21

(22)

節葬

(23)

(24)

25

天志

26

27

28

明鬼

(29)

(30)

31

非樂

?

?

32

?

非命39

35

36

37

Source: graham, “Mo tzu,” 336–337. Note: the chapters in parentheses are not extant.

36 graham, Disputers of the Tao, 53. 37 Divisions in Early Mohism Reflected in the Core Chapters of “Mo-tzu,28, graham explicitly sets aside questions of dating. 38 graham thought that chapter 17 was mistakenly cut from the end of chapter 26, where, according to him, it belonged. for more details, see ibid., 3–4. see also Maeder, “some observations on the composition of the ‘core chapters’ of the Mozi,” 69–75. 39 this triplet is generally considered very corrupt. graham used parts of chapters 35 and 36 to reconstruct the Y chapter of the “fei ming” triplet, he added a piece of chapter 35 to chapter 37 to form the h chapter, and the J chapter is again a mixture of the original chapters 35 and 36. see graham, Divisions in Early Mohism Reflected in the Core Chapters of “Mo-tzu,12–16.

12

carine defoort and nicolas standaert

although this three-sects theory has been very influential, especially among Western scholars of early chinese thought,40 most Mozi scholars have recently abandoned it. as Qin Yanshi 秦彦士 has pointed out, doc- trinal disputes within the triplets are remarkably absent, which weakens the hypothesis of fierce rivalry.41 Moreover, the opponents mentioned in the core chapters do not seem to be other Mohists, not even other philos- ophers or masters, but rather members of the ruling elite preserving and defending their customs against Mohist attacks. however, the rejection of the hypothesis of fierce disputes between rivals does not necessarily imply the rejection of the possibility of a different regional provenance of some chapters, while leaving open the possibility of temporal progress. taeko Brooks has taken these possibilities—the combination of regional varia- tion and chronological sequence—into account in her sequence theory.

The Sequence Theory

on the basis of graham’s work, taeko Brooks has argued that the dif- ferences within the triplets may bear witness to a political, intellectual, technological, and social evolution rather than to a division into three competing sects. Based on formal features (e.g., initial attribution formu- las, the use of past authority, the elite mentioned in the text, the reference to written sources) and differences in content (e.g., opposition, refer- ences to supernatural sanctions, populism, controversy, self-definition), she argues that the triplets are the result of successive revision and pro- gressive accommodation to political realities within one and the same school, moving from the state of Zheng to Wei and then to song.42 she concludes “that those differences are plausible as developing over time, as the Micians [Mohists] move from outside critics to inside members of the system, philosophize it in their terms, make peace with its intrinsic war aims, and cope with the increasingly harsh conditions which apply to all who serve the state.”43 the layers distinguished by Brooks in the core chapters are presented in the following table and are tentatively dated from 390 to 273 Bce:

40 see, e.g., schwartz, The World of Thought in Ancient China, 137–138; hansen, A Daoist Theory of Chinese Thought, 99; tong shuye, Xian Qin qi zi sixiang yanjiu, 59. 41 Qin Yanshi, Mozi yu Mojia xuepai, 23. see also Johnston, The “Mozi,xxiv. the earli- est rejection of the three-sects theory came from luan tiaofu, “Mozi shu zhi chuanben yuanliu yu pianshi cidi,” 180. 42   for speculation on the location of the Mohists, see a. taeko Brooks, “the Mician ethical chapters,” 116. 43 a. taeko Brooks, “the Mician ethical chapters,” 111.

introduction: different voices in the mozi

13

table 2. tentative dates for the core chapters according to taeko Brooks

尚賢 尚同 兼愛 非攻 節用 節葬 天志 明鬼 非樂 非命 390–375 Bce 17 14 (390)
尚賢
尚同
兼愛
非攻
節用
節葬
天志
明鬼
非樂
非命
390–375 Bce
17
14
(390)
(386)
20
(382)
[23]
(378)
374–345 Bce
11
(372)
21
18
(367)
(362)
[24]
(357)
26
(352)
[29]
(347)
342–324 Bce
15
8
(342)
(338)
[22]
(334)
25
19
(330)
(326)
[30]
(324)
322–317 Bce
12
(322)
32
(320)
35
9
(319)
(317)
310–287 Bce
16
(310)
27
(302)
31
(298)
[33]
(295)
37
(291)
282–273 Bce
28
(282)
36
[34]
(280)
10
(277)
(275)
13
(273)

Source: Brooks, “the Mician ethical chapters,” 117. Note: We exclude the “fei ru” duplet chapter and the opening chapters (or singlets), which taeko Brooks includes in her study of the “ethical chapters” of the Mozi. the chapters between square brackets are not extant. the dates are given in parentheses.

14

carine defoort and nicolas standaert

Without subscribing to every claim and date presented by taeko Brooks, we retain three important conclusions from this table that con- cern our interest in the book Mozi. first, Brooks supports the sequence theory as opposed to graham’s three-sects theory to explain the differ- ences between the three chapters within each triplet. second, she does away with graham’s suggestion that three short chapters were “digests” or a “manuscripts (fragment)” later added to the book. We agree on these two points. the third point is that Brooks mostly44 considers the upper-Middle-lower sequence of the chapters as their actual chronologi- cal order. following the work of Watanabe takashi, we take issue with this point.

An Alternative Evolution Theory

the sequence theory developed by Brooks is in fact more recent than the evolution established by the Japanese scholar Watanabe takashi 渡邊卓, who published his views on the Mohist core chapters as early as 1962.45 after a thorough study of the contents of all the core chapters together and, especially, of the increasing sophistication of ideas and logic, he fitted each chapter into a period and proposed a chronological order. Very roughly, the evolution went as follows: in the early Warring states period, the Mohist movement began with the promotion of jian ai and fei gong; when the movement became more structured and the demands for political advice increased, the Mohists came up with shang xian, jie yong, jie zang, and fei yue. at the end of the Zhou dynasty, when their movement was falling apart, the Mohists promoted shang tong, tian zhi, ming gui, and fei ming. Watanabe takashi does not give exact dates for the chapters but suggests that the last chapter, “shang xian, zhong” was written by the end of the Warring states period or the beginning of the Qin dynasty.46 on the basis of his analysis, Watanabe concludes that in

44 all triplets are chronologically ordered upper-Middle-lower, except the (very cor- rupt) “fei ming” triplet, which is upper-lower-Middle. 45 for the three periods that Watanabe takashi distinguishes, see his “Mozi sixiang,” 4, 26–50. he does not exclude regional differences between the chapters (e.g. the states of Qi, song, chu, and Qin). for other early Japanese scholarship on the core chapters, see hashimoto sumiya, “riben de ‘Mozi’ yanjiu gaiguan,” 259–262, 264–268. 46 Watanabe takashi “Bokushi shohen no chosaku nendai,” part 2, 30–31 identifies a fourth short and overlapping stage ending around 210 Bce, but he places that stage with the third stage in the third period. for a more recent Japanese theory, dating all core chapters between ca. 400 and ca. 250 Bce, see Yoshinaga shinjirō, “Jian ai shi shenme,”

585.

introduction: different voices in the mozi

15

table 3. the chronology and evolution of the core chapters according to Watanabe takashi

Warring Year 尚賢 尚同 兼愛 非攻 節用 節葬 天志 明鬼 非樂 非命 states early 400
Warring
Year
尚賢
尚同
兼愛
非攻
節用
節葬
天志
明鬼
非樂
非命
states
early
400
Bce
14
period
17
Middle
380
Bce
8
period
350
Bce
15
18
20
16
19
21
late
300
Bce
11
period
26
10
13
28
250
Bce
25
32
12
27
35
220
Bce
9
31
36
37

Note: the three triplets with a shang-xia-zhong sequence are underlined.

three triplets—“shang xian,” “shang tong,” and “tian zhi”—the chrono- logical order is upper (shang)–lower (xia)–Middle (zhong) and not the traditional order of upper–Middle–lower.47 table 3 presents an overview of Watanabe’s results. again without subscribing to Watanabe’s actual dating of the various chapters, we believe that his suggested evolution theory is superior to the two previous theories: like taeko Brooks, he considers graham’s “digests” (chapters 14 and 20) and “fragment” (chapter 17) to be early core chap- ters; and he presents an evolutionary picture, although one that differs in the sequence of chapters within the triplets. Moreover, this theory fits well with some data from graham’s own analysis a few decades later: the

47 When forke, Mê Ti, 23, divided the core chapters into “source” (Quelle), followed by “elaboration” (Erweiterung) and finally by “paraphrase” (Paraphrase), he also reversed the order of these three triplets and of “fei ming.”

16

carine defoort and nicolas standaert

chapters that Watanabe characterizes as written last in each triplet all happen to coincide with graham’s h chapters. this suggests that what graham regarded as a regional characteristic might have been a (chrono- logical and/or regional) sign of the last group of authors/editors of the triplets. Karen desmet’s recent study of the use of compounds in the core chapters confirms Watanabe’s hypothesis.48

these three theories by graham, Brooks, and Watanabe takashi, which reflect the views of a larger community of Mozi scholars, generated our second basic assumption: focusing on the core chapters, we believe that they contain interesting differences that to some extent suggest an evolu- tion of ideas in the order presented by Watanabe. the various contribu- tors to this volume do not necessarily subscribe to this insight, but they are all aware of it and consider its implications. the four most striking developments that are traced throughout various papers of this volume are the ever-increasing radicalization of ideas, a growing search for a theo- retical foundation and consistency, a move from acts to motivation, and a multiplicity of voices. In the “Jian ai” triplet, for instance, the demand to “inclusively care for everybody” does not diminish its radical nature while Mohists adapt to political realities; on the contrary, it is only slowly con- ceived, and as it undergoes a conceptual evolution, it increasingly gains force. an example of the second development is the ever-more-frequent use of references to heaven and ghosts, which occur relatively late in various core chapters and provide the Mohist proposals with a respected authority. next, the move from actions to motivation is visible in the views on ghosts and the type of behavior that they respond to: not only good acts but also noble intentions. and finally, the dialogues and opening chapters, even more than the core chapters, convince us that there are

48 see desmet, “the growth of compounds in the core chapters of the Mozi,” 111–117. her research also shows that all the h chapters (in graham’s terms) consistently contain a higher ratio of different compounds, which confirms Watanabe takashi’s conviction that they form the latest group. these are always the lower (xia) chapters except in the trip- lets “shang xian,” “shang tong,” and “tian zhi,” where the Middle (zhong) chapters belong to the h group. If one focuses only on those compounds that are unique for one group and hence are not shared throughout the Mozi, one finds that the h chapters contain not only most compounds but also those that begin to occur only in relatively late War- ring states texts. chapters from the Y and J groups contain fewer exclusive compounds, which moreover also occur in some older texts. these conclusions are inevitably tentative because of the controversies on the nature of compounds as well as on the dates of early chinese sources.

introduction: different voices in the mozi

17

different and even inconsistent voices to be heard in this one book, both on the side of the opponents and on that of the defenders. despite reflections on these and other possible differences within the book Mozi, we also share a strong awareness of the fact that the book is both less and more unified than presented in this volume. on the “lesser” side, we believe that not only the book as a whole lacks unity, but so do the various parts in which it has been conveniently divided, among which are the core chapters, dialogues, and opening chapters. Just as we hear a variety of voices speaking in these parts, we also perceive a variety of opponents and addressees: the ruling elite, some ru, Mohist adherents, and critical disciples. the vocabulary and values in the Mozi that also commonly occur in many Warring states texts do not always dis- tinguish the Mohist authors very clearly from other “schools” or “lineages” either, even though they sometimes insist on their own interpretations of treasured values such as ren and yi . But the strongest reminder of the degree of disunity within the Mozi certainly comes from erik Maeder, who has traced differences between the paragraphs (ce ) of the same chapter (pian ). the fact that some characteristics identified by graham with the h group are clustered in only some paragraphs of these chapters, while other characteristics checked by Maeder are equally spread over the chapters, suggests that the former belonged to older documents used by the Mohist authors, while the latter might be from the final hand.49 for such reasons and despite Maeder’s strong support for some of graham’s conclusions, he also argues in favor of a complex temporal evolution of the core chapters.50 Maeder’s study also makes us aware that the Mozi (as probably many other ancient texts) is an evolving text consisting of many layers, fragments, and lacunae, and that its authors were more edi- torial compilers or inventive bricoleurs than the “strong authors” we tend to expect behind a text.51 on the other hand, we also believe that there is more unity in the book Mozi than we have tried to show in this volume. We therefore also

49 Maeder, “some observations on the composition of the ‘core chapters’ of the Mozi,” 44–47. In the same article, Maeder also shows that there are remarkable similari- ties between paragraphs in very different pian. this suggests the use of older texts by the Mohist authors. see ibid., 74–75, 81–82. 50 see ibid., 76. for Maeder’s strong support of graham’s three-sects theory and “digest/ manuscripts (fragment)” theory, see ibid., 39–40, 54–55, 75. 51 Ibid., 81–82. Postmodern theory has shown that even “strong authors” can be seen as bricoleurs of existing quotes. the cut-and-paste habits of contemporary computer use have made this characterization even more apt.

18

carine defoort and nicolas standaert

understand the general tendency to present Mohism as one consistent vision, and each triplet as the expression of one dogma or thesis. first of all, the differences between the chapters of each triplet are sometimes rather small and only implicit; they are not explicitly emphasized as one would expect from rivalling sects holding fierce debates. second, there is a principle of charity that expects the reader or listener, at least to some extent, to make sense of the author or speaker and, hence, to distill a coherent message despite apparent incoherence.52 We do not want to cas- tigate other Mozi scholars for having done exactly that. a third reason to attribute unity to the Mozi is the fact that it was ascribed to one particular master, constructed as one text by the han dynasty, and read as such in the many centuries thereafter, at least when the text was available. even though the Mozi parts that we have focused upon may have been com- posed on the basis of older written or oral sources, they do not appear to be merely “a reservoir of so-called textual building blocks,” in William Boltz’s terms. rather, they attest to “an editorial process,” which “pre- sumes a doctrinal or other similarly purposeful motivation.”53 however multivocal, loose, and corrupt the chapters sometimes are (or appear), there clearly were people who identified with this composed text.54 By preserving the three different versions of the triplets under the labels “upper,” “Middle,” and “lower,” the last of the Mohist authors/editors may have given us an exceptional glimpse into the reworking of perhaps many more early texts and into the efforts that a community put into weaving the tapestry of their intellectual tradition.55 one final reason for attribut- ing unity to the various triplets more specifically is their identical titles,

52   the “principle of charity,” named as such in 1958–1959 by neil Wilson and much dis- cussed by philosophers, requires the reader or listener to interpret an author’s or speaker’s statements as rational, coherent, valid, and interesting. there has also been much discus- sion about the possible disadvantages of such an attitude, especially over the boundaries of times and cultures. see, e.g., feldman, “charity.” 53 Boltz, “the composite nature of early chinese texts,” 59. 54 the perceptible looseness may also differ per chapter. for instance, although the Middle and lower chapters of the “Jian ai” triplet are relatively well structured, they give the impression of having been constructed out of previously existing fragments more than the upper chapter does; the latter is a nicely constructed essay and probably just as much “composed” as is a modern essay. see the essay by carine defoort in this volume. 55 the exceptional threefold nature of the core chapters may result from the han edi- tors’ benign neglect, resulting in their failure to edit the Mozi into a unified text. It is perhaps not irrelevant that in Hanshu 30.1738 the book is listed as the very last of the Mohist writings, analogous with the collections of “daojia yan” 道家言 under taoism, “Za yinyang” 雜陰陽 under the Yin Yang lineage, and “fajia yan” 法家言 under legalism.

introduction: different voices in the mozi

19

with each chapter of a triplet distinguished only by “upper,” “Middle,” and “lower.” But who added those titles to the chapters, and when did they do so? tentative answers to these questions are discussed under the third basic assumption made by the editors of this volume.

Third Basic Assumption: Temporary Suspension of the Titles

We should be cautious with expectations created not only by the modern notions of “book” and “author” but also by “titles.” In his study of one Xunzi chapter, “tian lun” 天論 (about heaven), edward Machle warns that its title “may have created expectations that misled generations of readers as to the real subject matter of the essay.” having removed the title from its privileged position, Machle concludes that “the discussions in the essay are not chiefly about Tian, but about the conditions for suc- cessful government, the full development of human possibilities, the lim- its of human responsibility, moral discipline, the proper attitude toward omens and sacrifices, the necessity of following Li, and the limitations of some prominent philosophers.” he therefore speculates “that xun Qing would be quite surprised to see the title that has been given to the work, and would reject it as a determiner of the essay’s interpretation.”56 In a similar vein, we first reflect on the presence and nature of the core chap- ter titles before speculating about their emergence as slogans or mottos representing Mohist thought.

Titles

unlike Machle, we do not think of one particular person as the real author who would have rejected later added titles. But we do believe that the risk of titles misleading the reader haunts Mozi studies. In the case of the core chapters, the risk is even greater, because early Mohism has been identified almost literally with the titles of these ten triplets. for example, Mozi scholars have taken the “tian zhi” 天志 triplet as containing the master’s original views about the will of heaven. But could it be that when the three chapters were first constructed, they lacked titles? for what it is worth, the expression tian zhi 天志 hardly occurs in the triplet named

56 Machle, Nature and Heaven in the “Xunzi,58. for Machle’s reflections on titles in Xunzi, see ibid., 57–58.

20

carine defoort and nicolas standaert

after it (only once in chapter 26 and once in 28, i.e., “tian zhi, shang” and “tian zhi, xia”), and never in the rest of the book. hence the question:

did the Mohist authors have an inkling of the importance that we now attribute to these titles? for learning about the importance and frequency of chapter titles, we can take advantage of the many manuscripts discovered in the twentieth century. We know from unearthed materials as well as from references in han sources that the current distinction between chapters and books was very fluid in early china.57 lin Qingyuan 林清源 has divided early writings preserved on wood, bamboo, and silk into different categories, one consisting of texts or fragments about thought (sixiang 思想). two conclusions, about the dates and nature of their titles, may be relevant for our research. first, writings about thought only began carrying titles by the mid- or late Warring states period. “Warring states texts about thought often have no titles and the formation of titles does not yet seem to have turned into strict rules; but han texts on thought often have titles and their formation is full of change.”58 thus, unearthed manuscripts suggest that perhaps few philosophical manuscripts from that period carried titles. since no sub- stansive part of the Mozi has hitherto been discovered in a tomb, there is no specific information to be expected from that side.59 But it is possible that the core chapters—the oldest part of the book—at an early (perhaps pre-final) stage did not carry any titles. second, lin’s conclusion about the nature of these writings’ titles is that they often reflect the general content of the text.60 the titles of the Mohist core chapters indeed seem to refer to the general content and have also been perceived as such—hence, for example, the established association

57 han sources discussing writings about thought tend to refer to (what we now know as) chapter titles rather than to (what we now know as) book titles. see Yu Jiaxi, Muluxue fahui, 200–204. 58 lin Qingyuan, Jiandu boshu biaoti geshi yanjiu, 7. the category of “title” is also com- plex: there is discussion of whether an expression is a title or merely a fragment heading. see ibid., 69–105. 59 some (mostly chinese) scholars have identified the shanghai manuscript “guishen zhi ming” 鬼神之明 (title added by contemporary editors) as Mohist. for doubts about this identification, see Brindley, “ ‘the Perspicuity of ghosts and spirits’ and the Problem of Intellectual affiliations in early china,” 218–230; and the essay by roel sterckx in this volume. 60 see lin Qingyuan, Jiandu boshu biaoti geshi yanjiu, 7–9, 48–50. three other types of titles are created by (1) expressing the concrete object discussed in the texts, (2) repeating the first keywords of the text, and (3) quoting the first unit of a series of items.

introduction: different voices in the mozi

21

of the “tian zhi” chapters with Mohist views on the will of heaven. But when was this title chosen and why? neither the expression tian zhi nor tian zhi zhi 天之志 is common in the triplet.61 nor is the expression zun tian 尊天 (revere heaven), a motto that seems to have represented Mohist thought, as we will argue further on. an expression very similar to tian zhi, namely tian yi 天意, appears eleven times in the relevant triplet (ten times in the upper version and once in the Middle version) but was not chosen as its title.62 and on top of all this, we have the strong impression, as did Machle in the case of the Xunzi’s “tian lun,” that these chapters do not mainly discuss heaven’s will but rather “righteousness” (yi ). the three “tian zhi” chapters begin and end with the idea of yi; they plead for a new understanding of righteousness and find a foundation for their novel views in heaven. the absence of the expression tian zhi from the triplet (and even from the whole received text), combined with a content that does not entirely coincide with its title, suggests that the addition of titles may have hap- pened at a relatively late stage, when the argument was already formed. William Boltz has speculated with respect to early chinese writings that “unedited, ‘raw’ source material” at some point was edited into a “text” carrying an authorial, or at least editorial, voice. as an example, he sug- gests that the untitled pre-han guodian fragments (now labeled Laozi 老子) might be an instance of the former, while the two Mawangdui man- uscripts from the han titled “de jing” (classic of Power) and “dao jing” (classic of the Way) might be of the latter type.63 the Mohist core chap- ters, as we now have them, clearly belong to the latter type: they are rela- tively well structured and do make a point, which reflects the presence of some ideological motivation behind the construction of the text. But

61   the combination tian zhi zhi 天之志 never appears in chapters 26 and 27; it appears only three times in chapter 28 and two times in chapter 49. It might be significant for linguistic research that expressions “noun + zhi + noun” (see also tian zhi yi in the following note) appear only in the Middle and lower versions and never in the upper version. 62 Tian zhi yi 天之意 never appears in chapter 26; it appears nine times in chapter 28 and twenty-four times in chapter 27. the Mohists may have invented a new concept (tianyi 天意) to express heaven’s will, since heaven’s Mandate (tianming 天命) was prob- ably too closely associated with fatalism (ming). the occurrences of yi in the Mozi are most concentrated in this triplet (65 occurrences out of a total of 302). 63 Boltz, “the composite nature of early chinese texts,” 58–61. these two titles are based on the first keywords of each part and were added at the end of the manuscripts together with the number of characters. “de jing” and “dao jing” are good examples of titles that prob- ably do not represent the content of the text, but that have almost without exception led to such an interpretation, as if they formed a book about the “Way” and its “Power.”

22

carine defoort and nicolas standaert

the point that they make does not always coincide with their titles. We therefore suspect that the core chapters may have been constructed in various stages, with the addition of titles at a relatively late stage, possibly in the han.64 the addition shang, zhong, and xia to the titles of the trip- lets also suggests the influence of such editorial hands. If one accepts that these individual chapters came into being at different times, as we do in this volume, then some editor or team must have considered them one unity and arranged them in the current redaction. In order to speculate on the insertion of titles in the core chapters, we extend our exploration to determine where and how often these phrases occur in the Mozi as a whole (table 4).

table 4. the occurrence of the core chapter titles in the Mozi

尚賢 shang xian

chapter 8

chapter 9

chapter 10

elsewhere

 

4

15

13

 

1

尚同 shang tong

chapter 11

chapter 12

chapter 13

elsewhere

 

1

13 [+1]

14

1

[+1]

(5 上同)

(2 上同)

(1 上同)

(1 上同)

兼愛 jian ai

chapter 14

chapter 15

chapter 16

elsewhere

 

0

0

1

9

[+1]

非攻 fei gong

chapter 17

chapter 18

chapter 19

elsewhere

 

1

0

1

 

[1]

節用 jie yong

chapter 20

chapter 21

elsewhere

 

0

1

 

1

節葬 jie zang

chapter 25

elsewhere

 

0

 

2

(1 節喪)

 

64 sun Yirang attributed the redaction of the Mozi to the imperial han librarian liu xiang 劉向 (77–6 Bce), who was responsible for the order of the chapters and sections in various works, such as Xunzi, Guanzi, and zhan guo ce. In an edict of 26 Bce emperor cheng di ordered him to collate writings for the imperial library (Hanshu 10.310, 30.1701). there is, however, no proof that liu xiang did this redaction or gave the titles to the chapters of the Mozi. see sun Yirang, Mozi jiangu, 653. Many scholars seem to follow this attribution. see e.g. Zheng Jiewen, zhongguo Mo xue tongshi, 202, 252, 289.

introduction: different voices in the mozi

23

table 4 (cont.)

天志 tian zhi

chapter 26

chapter 27

chapter 28

elsewhere

   

1 0

1

0

明鬼 ming gui

 

— —

chapter 31

elsewhere

 

— —

 

0

0

非樂 fei yue

chapter 32

elsewhere

 

1

1

非命 fei ming

chapter 35

chapter 36

chapter 37

elsewhere

 

0

0

0

3

Note: chapter titles are excluded from the count. numbers between square brackets refer to the occurrences of characters indicated in the Mozi zhuzi suoyin as being reconstructed. alternative characters are added in parentheses.

table 4 reveals three remarkable facts. first, only the “shang xian” and “shang tong” triplets display a clear correlation between the title and the use of the expression in the chapters. In the others there is hardly any correlation at all. second, as we saw with the expression tian zhi, there is remarkably little reference to these so-called “central dogmas” in the rest of the book: most titles occur only once elsewhere in the Mozi, namely in a fragment of chapter 49, “lu wen,” that will be discussed further on.65 a third remark is that the expression jian ai occurs slightly more often in the rest of the Mozi, but hardly in the triplet named after it. on the basis of table 4, we are led to conclude that the authors of the Mozi were simply not aware of the “ten dogmas” of original Mohism or, at least, that they failed to give them the importance that we now attribute to them, except in “shang xian and “shang tong.” for the eight other trip- lets, the chapters themselves contain no strong clue as to why they carry precisely the titles they do.66 hence, we cannot exclude the possibility that the chapters were arranged into their current shape by someone who had

65 the expression jie zang occurs once in chapter 21, “Jie yong,” and the expression fei ming occurs twice in chapter 45, “xiao qu” 小取 (choosing the lesser). 66 Perhaps the fact that the phrases making up the titles do not appear in the text of the triplets can to some extent be explained by the nature of titles of philosophical writings, which are concise mottos as opposed to running text. e.g., while the expression ming gui is totally absent from the received Mozi, the terms gui and ming are discussed in relation to each other.

24

carine defoort and nicolas standaert

no idea that these expressions would come to represent Mozi’s thought and that they would be added as titles. for that reason, we believe it is worth trying to read these core chapters without the dominant influence of their titles. some advantages of this approach are that the reader carries weaker expectations concerning the unity of the chapter, that there is also more attention to the differences between the three chapters of a triplet, that there is no (possibly misleading) presupposition about the content of the chapter, and that one is more on the alert to the importance and recurrence of other expressions in the chapter.

Mohist Mottos

But the existence of the titles shows that there must have been a moment in the history of Mohism when the master’s ideas were thought of in terms of these relatively fixed expressions or mottos. one indication that the extant titles functioned as such short slogans is perhaps the occurrence of several “fei x” (against x) constructions: against military aggression ( fei gong), against music ( fei yue), against fatalism ( fei ming), and against the ru ( fei Ru ) .67 the use of fei as a transitive verb “be against . . .” cor- responds to the characterization of the debates among masters in the late Warring states as “shi fei是非 (pro and contra), but it occurs more in the Mozi titles than in the running text.68 this seems to confirm their status as mottos or slogans summarizing the content of the chapters. there is also one piece of textual evidence that Mohist thought at some point came to be closely identified with the ten expressions that now function as the titles of the core chapters: it occurs in “lu wen” 魯問, which records a conversation between Mozi and a disciple about what to expound first when meeting the lords of the four quarters.

子墨子游魏越, :「既得見四方之君子則將先語?」子墨子曰:「凡 入國, 必擇務而從事焉國家昏亂, 則語之尚賢尚同; 國家貧, 則語之 節用節葬, 國家說音湛湎, 則語之非樂非命; 國家淫辟無禮, 則語之尊 天事鬼; 國家務奪侵凌, 則語之兼愛非攻69 曰擇務而從事焉

67 three of these titles are from chapters that are usually thought to be among the lat- est core chapters (“fei yue” and “fei ming”) or even to postdate them (“fei ru”). they may represent the increasing severity of the Mohist ideology. “fei gong,” however, is generally considered an early triplet. 68 about “pro and contra” or “right versus wrong” debates, see graham, Disputers of the Tao, 36, 167, 176–177; and the essay by chris fraser in this volume. 69 Gong and do not occur in the daozang edition but are indicated as being restored in the Mozi zhuzi suoyin. hence, the oldest extant Mozi edition contains nine (and not ten) dogmas.

introduction: different voices in the mozi

25

When our Master Mozi was traveling, Wei Yue asked: “having been granted an audience with the lords of the four quarters, what would you expound first?” our Master Mozi said, “Whenever you enter a state, you must select a task and work on it. If the state is in disorder, expound to them ‘elevating the worthy’ and ‘conforming upward’; if the state is impoverished, expound ‘moderation in expenditure’ and ‘moderation in burial’; if the state over- indulges in musical entertainment, expound ‘against music’ and ‘against fatalism’; if the state is dissolute and indecorous, expound ‘revering heaven’ and ‘serving ghosts’; if the state is devoted to aggression and intimidation, expound ‘inclusive care’ and ‘against military aggression.’ therefore, I say:

select a task and work on it.” (49: 114/7–10)

this fragment tells us at least three things. first, in what sounds like a summary of Mozi’s political doctrine, fei Ru (非儒) does not occur; this provides further support for its rejection as a core chapter. the second piece of information concerns the possible dates of the core chapter titles. one cannot fail to notice that the political remedies ascribed here to Mozi correspond almost exactly to the titles of the ten core chapters except for the expressions zun tian 尊天 (revere heaven) and shi gui 事鬼 (serve the ghosts), which appear as tian zhi and ming gui in the titles. since this is the only fragment in the extant corpus of pre-Qin texts listing these ten slogans, it may have had some relation—as cause or result—to the creation of the titles. chris fraser dates the dialogues around the middle of the fourth century Bce, and taeko Brooks dates this particular chapter to 262 Bce.70 a third reflection inspired by this fragment is the question whether the expressions zun tian and shi gui would have provided more appropriate titles for the “tian zhi” and “Ming gui” triplets. the slogan zun tian occurs eight times in the Mozi but only once in the “Will of heaven” triplet (26: 43/11). It is always paired with shi gui (shen ), an expression that occurs alone in three more instances but never in the extant “Ming gui” chapter.71 thus, from their occasional appearance in the book Mozi, we might postulate that the alternative expressions zun tian and shi gui

70 fraser, “Mohism,” see “supplement to Mohism: text and authorship”; a. taeko Brooks, “the Mician ethical chapters,” 115. 71   In chapter 26, it occurs in the threefold statement about the ancient sage-kings: “In their work, they upward revered heaven, in the middle served the ghosts, and downward took care of men” 其事上尊天, 中事鬼神, 下愛人. In the other cases, the expression 尊天事鬼 is used (4: 5/1, 9: 12/22, 35: 59/7–8, 48: 107/27, 48: 111/7, 49: 111/23, and 49: 114/9).

26

carine defoort and nicolas standaert

were more widespread slogans in early Mohism than the actual chapter titles, but they hardly occur in the two relevant triplets either.72 as several contributions to this volume show, many topics from the core chapters are taken up in other parts of the Mozi. But, aside from this fragment from “lu wen,” there is little evidence of the emergence of the core chapter titles from mottos or fixed expressions representing Mohist thought.73 table 4 shows that these expressions do not often occur in the Mozi. their appearance together in clusters is even rarer, except for the unique fragment quoted above. there are three short and rather similar clusters to be found in the dialogues, which could attest to the emerging association of Mohism with some key ideas or mottos: revering heaven, serving ghosts, and caring for others. In “gongmeng,” for instance, gong- mengzi 公孟子 asks Mozi why confucius was never made son of heaven despite his broad knowledge of the classical heritage. Mozi explains that knowledge is not enough: “a wise person must revere heaven and serve the ghosts, care for others, and moderate expenditures. the combination of these makes one wise” 夫知者必尊天事鬼, 愛人節用。合焉為知矣 (48: 107/27). this answer clusters three slogans (zun tian and shi gui from the list in chapter 49, not the actual titles), one title “Jie yong,”74 and per- haps an echo of jian ai, namely ai ren 愛人. In the same chapter Mozi explains that he does not mind being accused of a lack of humanity (bu ren 不仁) as long as he is acknowledged as “revering heaven, serving the ghosts, and caring for others” 尊天事鬼愛人 (48: 111/7). here the master explicitly endorses the two slogans mentioned above, again followed by a possible echo of jian ai. a last fragment occurs in “lu wen,” where the lord of lu asks for assistance against the attacks by Qi. referring to the exem- plary rulers of the past, Mozi’s advice is as follows: “I wish that the lord would, upward, revere heaven and serve the ghosts and, downward, care for and benefit the people” 吾願主君之上者尊天事鬼,下者愛利百姓. he then continues with a longer list of concrete suggestions (49: 111/23). again, we have no more than the two slogans from the list in chapter 49, followed by a general instruction to care for the people.

72 It is striking that the ten mottos in this fragment, including zun tian and shi gui, can be understood as verb + object constructions. this is not the case for the actual title “tian zhi.” 73 for more about these Mohist mottos, see defoort, “do the ten Mohist theses rep- resent Mozi’s thought?” 74 as the Ics edition of Mozi zhuzi suoyin indicates, the original edition (= Daozang) had yong jie. apparently, Bi Yuan corrected this ‘mistake’ without leaving any comment.

introduction: different voices in the mozi

27

even though the exact motto jian ai does not occur in these short clus- ters, there is always a mention of caring. taken together with the occa- sional appearance of jian ai in the whole book (see table 4), and including slightly variant expressions such as ai ren, jian xiang ai 兼相愛 (care for each other inclusively), and jian er ai 兼而愛 (caring inclusively), it some- what stands out among the ten early dogmas.75 even in the core chap- ters postdating the “Jian ai” triplet, the expression already occurs, usually paired with the two slogans found in the dialogues: “revere heaven” and “serve the ghosts.” In “shang xian, zhong,” for instance, the author claims that the sage-kings, “when ordering everybody in the world, inclusively cared for them all and hence benefited them, and also led all the peo- ple of the world to elevate and revere heaven and to serve the ghosts” 其為政乎天下也, 兼而愛之, 從而利之, 又率天下之萬民以尚尊天事 鬼 (9: 12/21). therefore, they were rewarded and made sons of heaven. In one “tian zhi” chapter, the expression jian ai occurs no fewer than four times (28: 48/4 [twice], 28: 48/8, 28: 48/15) and in another “downward, loving others” 下愛人 complements the instruction to “revere heaven above and serve the ghosts in the middle” 上尊天中事鬼神 (26: 43/11). finally, in “fei ming, shang,” Mozi talks about the sage-kings “caring for all the people mutually and benefiting each other in interaction” (與其 百姓 兼相愛交相利, followed by the claim that they “led these people to thereby, revere heaven and serve the ghosts above” 率其百姓, 以上 尊天事鬼 (35: 59/9).76 Moving from the core chapters postdating the “Jian ai” triplet toward the dialogues, jian ai becomes even more explicitly identified as a spe- cific Mohist ideal or slogan. Wumazi 巫馬子, for instance, confronts Mozi with the following claim:

子兼愛天下, 未云利也; 我不愛天下, 未云賊也功皆未至, 子何獨自 是而非我哉?

75 兼相愛 occurs 13 [+1] times in the book (three times in chapter 14, five times in chapter 15, twice [+1] in chapter 16, once in chapter 26, and twice in chapter 35); 兼而愛 occurs five times (twice in chapter 4, once in chapter 9, once in chapter 26, and once in chapter 28); 兼天下而愛 occurs twice (in chapter 27). see also sato, “the Idea to rule the World,” 38–40. 76 also, in “fa yi” it is said that the sage-kings “inclusively [cared for] all the people in the world and led them to revere heaven and serve the ghosts” 兼[愛]天下之百姓, 率 以尊天事鬼 (4: 5/1), with the character ai restored by Bi Yuan “on the basis of the meaning.”

28

carine defoort and nicolas standaert

You inclusively care for everyone in the world but cannot quite be said to benefit them; I do not care about everyone but cannot quite be said to hurt them. since neither of us has had any effect, why do you consider yourself alone right and me wrong? (46: 100/20–21)

In the same chapter Wumazi also tells Mozi that in one respect he “dif- fers from the master, [since] he is not able to inclusively care for others” 我與子異, 我不能兼愛, explaining that he cares more for those who are located more in his own vicinity (46: 102/24). Mozi, in one of the dia- logues, speaks about his sense of justice ( yi ) in terms of caring, which constitutes his “hooks and clamps”; at the end of the day, they are far superior to “the hooks and clamps used in naval battles” 舟戰之鉤強 (49: 115/18–19). We tentatively conclude that the titles of the core chapters postdate the earliest construction of the chapters themselves and even most of the received Mozi. In only two triplets, “shang xian” and “shang tong,” did the authors probably finalize the text in full knowledge of their titles. In the other core chapters it is difficult to understand why they would have constructed a text (possibly from older fragments) or rewritten it without making any reference to the slogan or motto that was chosen as title. It is also striking that the rest of the Mozi hardly shows awareness of these mottos, even of the expressions shang xian and shang tong. there is, however, an emerging identification in the book of a Master Mozi’s thought with fixed expressions or mottos. the oldest seems to have been “care (for all),” joined by two expressions—“revering heaven” and “serv- ing ghosts”—which for some reason were not chosen as chapter titles. the unique fragment in “lu wen” is the only testimony in Warring states texts identifying Mozi’s core ideas with ten (or nine) mottos that are very close to the current core chapter titles. But uncertainty about its date leaves many hypotheses open.77 other questions that remain unresolved concern when these titles were added, why they were chosen (especially in the case of “tian zhi” and “Ming gui,” since alternatives were circulat- ing), and how they relate to the threefold structure of the triplets. as was the case with the two previous basic assumptions, we do not claim to have firmly proven an alternative view (there is not enough evidence to do this), but we believe in the methodological value of approaching titles

77 a. taeko Brooks, “the Mician ethical chapters,” 115, believes that “the triplet chap- ters had been rounded off, and the Mician doctrines officially fixed at ten” somewhat before 262 Bce.

introduction: different voices in the mozi

29

critically. the point we want to get across is that disregarding the titles of the core chapters (and probably of many other chapters of early chinese texts) is not only reasonable but also fruitful for a fresh interpretation of the text. a title both leads and misleads the reader; it should not remain a shackle by which all past and future interpretations are held captive.

Parts and Characteristics of this Volume

Most of the essays in this volume originated in presentations at the work- shop “the Many faces of Mozi: a synchronic and diachronic study of Mohist thought” (leuven, 2009);78 others were originally presented at the workshop “argument and Persuasion in ancient chinese texts” (leuven, 2005);79 and several of them were developed during monthly seminars with colleagues from the netherlands on the textual nature of the Mozi (leuven, 2002–2005).80 the three basic methodological assumptions—the focus on the text itself, understanding the differences among the chapters as reflecting evolution over time rather than Mohist sectarian differences, and the temporary disregard for the chapter titles—informed the basis of our selection. We have also asked the authors to rewrite their contri- butions with these assumptions in mind. the seven studies are mostly arranged in what we believe is the chronological order of the Mozi chap- ters that they discuss. each of the first three essays focuses on one triplet. the first essay con- centrates on the “Jian ai” (Inclusive care) triplet (chapters 14–16). carine defoort asks the basic question: do these three chapters really concern the topic of “universal love” or “inclusive care”? she discerns an ever- increasing radicalization of moral demands in the triplet: first, “caring for oneself ” is rejected in favor of “caring for each other” in relationships of

78 three papers focusing on Mohist influence in late Zhou and han were published in oriens Extremus 48 (2009): nylan, “Kongzi and Mozi, the classicists (ru ) and the Mohists (Mo ) in classical-era thinking”; sato, “the Idea to rule the World”; and gentz, “Mohist traces in the Chunqiu fanlu.” two papers were published in chinese: guo, “ru Mo liang jia zhi ‘xiao,’ ‘san nian zhi sang’ yu ‘ai’ de qubie yu zhenglun”; and ding sixin, “Moyu chengpian shidai kaozheng ji qi Mojia guishen guan yanjiu.” they were both translated in Contemporary Chinese Thought 42.4 (2011). 79 three papers on Mozi were published in oriens Extremus 45 (2005–2006): defoort, “the growing scope of Jian ”; loy, “on a Gedankenexperiment in the Mozi core chap- ters”; and desmet, “the growth of compounds in the core chapters of the Mozi.” 80 the participants in these seminars were carine defoort, Karen desmet, dirk Meyer, nicolas standaert, Karel van der leeuw, and Paul van els.

30

carine defoort and nicolas standaert

reciprocity (chapter 14); then the scope is gradually broadened from one’s familial or political in-group to include others, by “inclusively caring for each other and mutually benefiting each other” jian xiang ai, jiao xiang li 兼相愛, 交相利 (chapter 15); finally, with the advice to “replace exclusive- ness with inclusiveness” (chapter 16), the ideal of jian ai slowly emerges, but its full-fledged promotion takes shape only later. defoort then turns to the somewhat later triplet the “Will of heaven” (chapters 26–28), in which the expression jian ai occurs more often than in the chapters that have the motto in their title. In this triplet the emergence of the model of heaven leads to an unconditional type of caring for everybody else, ultimately without any expectation of reward. defoort thus shows that one of the core concepts attributed to Mohist thought underwent an evolution of which only the beginning can be observed in the “Jian ai” triplet. another way of observing variation can be found in the “fei gong” (against Military aggression) triplet (chapters 17–19), which is the topic of the second essay in this volume. Instead of treating these three chap- ters as one homogeneous unit, Paul van els asks another basic question:

what are the differences between them? his starting point is that each “fei gong” chapter displays a remarkable conceptual coherence and clear argumentation, which indicates that it forms a closed textual unit in the eyes of whoever created its received version, be it Mozi, his followers, or later editors. then, van els searches for conceptual differences between the chapters and discerns three types of arguments against aggressive war- fare. chapter 17 approaches warfare from the angle of morality, which van els calls the “moral argument.” chapter 18 is all about counting and cal- culating, as it quantifies the costs and benefits of a military campaign: the “economic argument.” chapter 19 speaks of ghosts and spirits and repeat- edly claims that warfare harms the interests of heaven, which van els calls the “religious argument.” In sum, the analysis of these core chapters shows that the Mohists did not uphold just one argument against military aggression but instead actively pursued different lines of argumentation, possibly to persuade different audiences. the third study concentrates on Mozi 31, which is the only extant chapter of the “Ming gui” (explaining ghosts) triplet and which is sup- posed to represent the Mohist view on spirits. the basic question asked by roel sterckx is whether there is one clear Mohist view on the issue. sterckx shows that the absence of two out of the possibly three original chapters forming the “Ming gui” triplet does not prevent us from pictur- ing a more polyphonic Mohist view of the spirit world. first, he under- takes a close reading of units at the subchapter level, then he compares

introduction: different voices in the mozi

31

them with passages in the other core chapters and the dialogues, and finally he introduces a chu bamboo-slip manuscript on ghosts and spir- its that is now housed at the shanghai Museum. sterckx’s analysis sug- gests that Mohist views on spirits evolved or, at least, diversified across the received Mozi text. some Mohists, for instance, were skeptical about the prescience of the spirit world, although that was a firm belief of Mozi. the philosophical issue of the “existence” of ghosts and spirits, on the one hand, and the more pragmatically inspired question of whether they were capable of punishing and rewarding, on the other hand, were seen as separate issues. the fourth essay in this volume functions as a transition between the discussion of the core chapters in the first three studies and the dia- logues and opening chapters in the last three. Its focus is the notion of the exemplary past in the core chapters, a notion that was central to the development of chinese traditions of thought. Miranda Brown won- ders what role the Mohists played in shaping it. her starting point is the observation that there are numerous appeals to ancient sages (sheng ) and sage-kings (sheng wang 聖王) in the Mozi, while there is a paucity of references to the compound “sage-king” in what she identifies as the pre-Mohist corpus. she compares the vocabulary and rhetorical strate- gies of the Mohist core with other early chinese texts, while also pay- ing attention to the differences between the various strata of the Mozi. she concludes that the Mohist view of the ancient rulers differed from that found in earlier works: while the early Mohists were not the first to make appeals to past rulers, they nevertheless played a role in creating the image of the three dynasties as a golden age with reference to a full set of exemplary kings: Yao , shun , Yu , tang , KingWen 文王, and King Wu 武王. such a vision, Brown argues, was motivated largely by rhetorical necessity: invocations of the sage-kings bolstered, rather than undermined, the Mohist attack on aristocratic traditions. the investiga- tion of this idea of wise rulers within and without the Mozi reveals how this central notion in the chinese tradition was shaped by the Mohists. While the next contribution concentrates on the dialogues, it does not leave the core chapters behind. chris fraser’s essay argues that the ethics of the dialogues is largely consistent with the middle and late stratum of the core chapters, but that the dialogues also developed new ethical ideas. tracing evolutions in Mohist thought, fraser presents four important exten- sions of older Mohist ethical ideas. first, the dialogues further develop the Mohist conception of morality as norms that can be promulgated through statements or teachings and that lead to beneficial, self-consistent

32

carine defoort and nicolas standaert

consequences if “constantly” followed by all people. second, they pres- ent a series of views on moral worth that tie it to the agents’ character and intentions. third, they develop the Mohist view of moral motivation and indicate how the Mohists might approach issues related to weakness of will. and finally, they also set forth a more radical ideal of personal sagehood. on the whole, fraser concludes, the dialogues present a more demanding conception of the moral life than the triplets do. later evolutions in Mohist thought can also be found in the opening chapters, which are the subject of the last two studies in this volume, again in connection with the core chapters. hui-chieh loy analyzes “Qin shi” (Intimacy with officers), which is the first chapter in the received Mozi. It has often been considered either a mere appendix to the core chapters, specifically, the “shang xian” (elevate the Worthy) triplet, or even a non-Mohist essay. and although the ideas in this chapter are akin to counsels found in “shang xian,” questions remain as to how these chap- ters relate to each other. does “Qin shi” simply repeat—briefly or with elaboration—the points made in one or more of the “shang xian chap- ters? or does it contradict, go beyond, or qualify the latter? By pointing at various differences, hui-chieh loy argues that “Qin shi” improves upon “shang xian” in the sense that it provides a more sophisticated construc- tion of the motivations that worthies have to enter government service. If in “shang xian” worthies are portrayed as motivated by a somewhat mercenary pursuit of wealth, honor, and power, in “Qin shi” they are pre- sented as acting from apparently nobler motivations. the final essay in this book takes “fa yi” (standards and norms) as its starting point. commonly presented as a “summary” of the “tian zhi” chapters, it portrays heaven in close relation to the idea of a “standard.” Which stage in Mohist thought does this chapter represent? What evolu- tion is there in the Mozi concerning the relationship between tian and fa? By analyzing the similarities and differences with the core chapters, nicolas standaert argues that “fa yi” may have been one point in an evolu- tion of the Mohist doctrine, not necessarily the final one. In the course of the book Mozi there is a growing need for certainty and for a foundation of the core ideas. the analogy with artisan tools, such as the compass and square, is an expression of this need. consequently, the instruments of the artisan are taken as a metaphor for the use of standards in human behavior: the rather abstract fa (such as “inclusive care of each other and mutual benefit to each other”) and the more concrete standards (such as the “ancient sage-kings”) are all supposed to function in the same way:

introduction: different voices in the mozi

33

objectively, measurably, and infallibly. the ultimate standard is heaven, an idea that appears in “fa yi” as well as in the late core chapters. despite the methodological connections and shared interest of these seven contributions, they can also be read separately. this explains why each essay contains some basic information that may appear repetitious but allows the reader to freely determine the order of reading the contri- butions. as for the Mozi text, though various editions have been consulted, such as those by sun Yirang and by Wu Yujiang 吳毓江, all fragments are quoted using the chinese text in d. c. lau’s Mozi zhuzi suoyin 墨子逐字 索引 (A Concordance to the “Mozi” ).81 all references to Mozi fragments are to this edition; the chapter number is given first, followed by a colon and then the page number and line number separated by a slash—for example, 16: 29/2. We have followed the editors’ reconstructions except where indicated otherwise. When we count the number of times terms appear in the Mozi, we explicitly indicate reconstructions marked by the Ics editors. although we have consulted one or more existing translations, such as those by Mei Yi-pao (from 1929), Burton Watson (from 1963), and Ian Johnston (from 2010), the translations are by the authors themselves except where indicated otherwise. for other primary sources, the original chinese text is not quoted and we only refer to the chapter, except when specific fragments are commonly recognized by a number (e.g., Lunyu, 1.2, Laozi, 24, and Mengzi 3a9).

Acknowledgements

like the chapters in the book Mozi, the studies in this volume went through at least three different versions. at each stage different audiences and “opponents” shaped the content and sharpened the arguments. the first versions of most papers were submitted to the workshop “the Many faces of Mozi: a synchronic and diachronic study of Mohist thought.” We are grateful to the chiang ching-kuo foundation for International scholarly exchange for supporting this workshop as well as the publi- cation of the papers in the present volume, and to the participants for engaging in the debate and thereby forcing us to rethink our arguments. aside from the contributors to this volume, these participants included

81 Mozi zhuzi suoyin 墨子逐字索引, edited by d. c. lau, Ics ancient chinese texts concordance series 41 (hong Kong: commercial Press, 2001).

34

carine defoort and nicolas standaert

ding sixin, Joachim gentz, guo Qiyong, roman Malek, Michael nylan, Michael Puett, sato Masayuki, shun Kwong-loi, hashimoto sumiya, Mar- tin svensson, and Zeng hantang. the second versions of the essays and the introduction were discussed in two seminars that took place in leuven in the spring of 2010 and 2011. We thank Karen desmet, annick gijsbers, Burchard Mansvelt Beck, Paul van els, griet Vankeerberghen, and sara Vantournhout for their critical reading. the third version of the chapters was submitted to the careful copyediting of Pamela J. Bruton. In addition we thank els ameloot, lee ting-mien, and lise Merken for their multiple contributions in efficaciously finalizing the manuscript. We also thank Martin Kern for his critical comments on the almost final manuscript. unlike the editors of Mozi’s core chapters, we decided not to publish the three different versions of the various papers, a decision for which we are sure the readers will thank us. support for the Mozi project at the university of leuven was gener- ously provided by fWo-Vlaanderen. We want to thank the contributors of this volume for their stimulating papers and patient response to vari- ous comments. and finally, we express our gratitude to all the old and new friends we encountered through the study of the Mozi, above all Watanabe takashi, angus graham, and roman Malek. they shared our enthusiasm, shaped our arguments, caused evolution in our own think- ing, and encouraged us to pursue the investigation of this important but often neglected ancient master. We hope that this volume may further stimulate the discovery of other voices and faces of master Mo.

Are the three “JiAn Ai” ChApters About universAl love?*

Carine Defoort

Jian ai 兼爱, which has been variously translated as “universal love,” “impartial caring,” “concern for everyone,” “inclusive care,” “co-love,” and “allumfassende liebe,”1 is more than just one among the ten central dogmas of early Mohism: it is generally considered the heart of Mozi’s thought, the very core of the so-called Core Chapters. Almost every expo- sition of Mohist philosophy begins with a claim to the effect that jian ai is “the corner-stone of the system,” “Mozi’s core and quintessence,” its “unifying principle of morality,” the “center of Mohism,” the “heart of their ethics,” a “startling, original, and even revolutionary concept,” generally opposed to the ru view of “graded love.”2 hence the three consecutive chapters titled “Jian ai” in the Mozi (chapters 14, 15, and 16) are considered crucially important. the very simple question of this essay may therefore be somewhat surprising: do these three chapters really concern the topic of “universal love”? this question can be conveniently divided into two stages: first, do they really discuss jian ai, as their titles suggest? And sec- ond, does the Chinese expression mean “universal love”? by focusing on the former question, this contribution reshapes the formulation and rel- evance of the latter one. since the Mohist view on jian ai has always been illustrated with quota- tions from the “Jian ai” triplet, the former question may come as a surprise. in this essay i argue that these three Core Chapters do not yet discuss jian ai but provide steps toward its formulation. this conviction makes the second

* this chapter is a further elaboration of the ideas presented in Defoort, “the Growing scope of Jian .” 1 For a detailed discussion of the meaning of jian, see schumacher, “An outline of the evolution of the Concept of Jian in Mohism,” 3–12. see also Graham, Disputers of the Tao, 41; and Johnston, The “Mozi,xliii–xliv. sato, “the idea to rule the World,” 39, has recently suggested translating jian ai as “Kingly love for all.” even though jian ai is indeed first of all (but not exclusively) meant for rulers, that does not warrant such an overly specific translation of the expression. 2 these quotes are, respectively, from Forke, Mê Ti, 82, about “den eckpfeiler des sys- tems”; tan Jiajian, Mozi yanjiu, 35; Graham, Disputers of the Tao, 41; Ding Weixiang, “Mojia jian ai guan de yanbian,” 70; Fraser, “Mohism”; and Watson, Mo-tzu, Basic Writings, 10. see also Zheng Jiewen, Zhongguo Mo xue tongshi, 14.

36

carine defoort

question somewhat irrelevant, at least in relation to this particular triplet, since the expression jian ai occurs only once in the whole triplet and thus needs no consistent translation there.3 but the terms jian and ai do occur separately in all three chapters: ai in itself remains uncontested by oppo- nents in the triplet and will be translated as “care” or “caring,” thereby referring to actions (take care of) as well as feelings (care for).4 the term jian, however, is new and contested almost as soon as it enters the argu- ment. Considering its novelty and sensitivity, i shall translate it flexibly, moving from “inclusive” toward “impartial” as jian gradually gains shape throughout these three and some other Core Chapters. the evolution that i will trace throughout these chapters could be the reflection of changing responses to the criticism of opponents or the result of a radicalization in the internal dynamics of early Mohist thought. While not denying an important degree of similarity among chapters 14, 15, and 16 of the Mozi, i shall highlight the differences and thus concen- trate on those views and arguments that are usually overlooked. i argue that, strictly speaking, the three chapters do not focus on the concept of jian ai and that the titles, which were added somewhat later, have steered our reading in a particular direction. this essay thus starts by temporarily ignoring the titles of the triplet (in the first two sections); then it works its way through the three chapters separately (in the following three sec- tions) and concludes with a tentative understanding of the concept of jian ai (in the last section) as it appears in the Core Chapters of the Mozi. to clarify the argument, a chart of the structure of the three chapters is appended to this essay.

Beyond the Title “Jian ai”

one of the main reasons prompting almost all scholars to treat the three “Jian ai” chapters as equal in philosophical content is their identical titles.5

3 the earliest triplet where the expression jian ai occurs more often is the “Will of heaven” (chapters 26–28). see below. 4 Johnston, The “Mozi,xliii–xliv, agrees with James legge, who, in 1861, did “not know how to render it better than by ‘universal love.’ ” see legge, The Chinese Classics, vol. 2, 104 n. 4. see also lowe, Mo Tzu’s Religious Blueprint for a Chinese Utopia, 93. 5 see, e.g., legge, The Chinese Classics, vol. 2, 103–126; liang Qichao, Mozi xue’an, 15–26; tan Jiajian, Mozi yanjiu, 35–58; Xue bocheng, Mojia sixiang xintan, 11–28; Qi Wen and li Guangxing, Mozi shi jiang, 278–285; Fung, A History of Chinese Philosophy, 76–105; schwartz, The World of Thought in Ancient China, 135–172; and han lianqi, Xian Qin liang Han shi

are the three jian ai chapters about universal love?

37

We know, however, that titles in Warring states sources were usually added later and not invented by their earliest author(s), let alone by the master to whom the ideas were ascribed.6 the titles of the Core Chapters of the Mozi are not derived from their first important words or from the main concepts in the chapters; in fact, the mottos used as chapter titles often do not appear in the chapters at all.7 in the whole “Jian ai” triplet, the expres- sion jian ai occurs only once: in the last chapter (16: 29/2), in a reflection on King Wen’s attitude as described in a quotation from the “Grand oath” 泰誓.8 While it is not totally clear when and by whom the chapter titles were chosen, or to what extent the editor determined the shape of the current book, the titles seem designed to indicate the central tenet of the different chapters of each triplet.9 they may have been added after or around the formation of chapter 49, “lu wen” 魯問 (lu’s Questions), in which Mozi’s doctrines are described on the basis of ten expressions that largely coincide with the titles of the Core Chapters.10 in an attempt to read the Mozi without being distracted by a later editor’s choice of titles, i have provided the three “Jian ai” chapters with different working titles on the basis of their content: “Caring for each other” (xiang ai 相愛) for chapter 14, “inclusively Caring for each other, Mutually ben- efiting each other” ( jian xiang ai, jiao xiang li 兼相愛, 交相利) for chap- ter 15, and “With inclusiveness replace exclusiveness” ( jian yi yi bie 兼以 易別) for chapter 16. only after the completion of these three chapters, i argue, did the motto jian ai come to represent their thought. in this essay, the three chapters are temporarily freed from their anachronistic title and

luncong, 298–321. some scholars, such as lowe, Mo Tzu’s Religious Blueprint for a Chinese Utopia, 92, 94, notice the differences but dismiss their relevance. the scholars mentioned below in this essay are counterexamples to this general current.   6 since i do not know who these authors were, i will alternatingly attribute the words to “Mozi,” “the author(s),” or “the Mohist(s),” without thereby trying to pinpoint the person(s) behind the text.   7 For more about the titles, see Yu Jiaxi, Muluxue fahui, 200–204; and lin Qingyuan, Jiandu boshu biaoti geshi yanjiu, 7–9, 48–50, discussed in the introduction of this volume.   8 For an interpretation of this passage, see below.   9 Graham presumes that the editors were liu Xiang 劉向 and liu Xin 劉歆 of the han. see Graham, Divisions in Early Mohism, 17. this idea was proposed by sun Yirang, Mozi jiangu, 653. 10 in a conversation between Mozi and Wei Yue (49: 114/8–10), Mozi advises shang xian 尚賢 and shang tong 尚同 to order a state, jie yong 節用 and jie zang 節葬 to enrich the state, fei yue 非樂 and fei ming 非命 to prevent debauchery, zun tian 尊天 and shi gui 事鬼 (neither is the literal title of the respective triplet) to prevent wantonness, and jian ai 兼愛 and fei gong 非攻 to prevent military invasions. see also the introduction to the present volume.

38

carine defoort

allowed to make their point separately, a point that does not totally coin- cide with “universal care” but rather marks steps in that direction.11 i shall show that an evolution of increasing radicalization in moral demands can be traced from chapter 14, through chapters 15 and 16, and continuing in the triplet “tian zhi” (Will of heaven; chapters 26, 27, and 28).12 Ding Weixiang 丁为祥 is somewhat unique among Chinese schol- ars for having highlighted the differences between the three chapters and traced an evolution from chapters 14 to 16. however, his views are radically opposite to what i shall try to establish. he believes that the most original Mohist stance contained the challenging idea of “universal love” ( jian ai), which amounted to a radical self-sacrifice for the benefit of everyone and a strong opposition to other trends of thought such as those of the ru and Yang Zhu 楊朱, known as the defender of individual- ism or egoism (wei wo 為我). later this ideal was tempered, according to Ding, as Mohism became a school that adapted to current values—hence the addition of reciprocity ( jian xiang ai 兼相爱) and utilitarian motives (li ). other scholars who have noticed the differences between the three “Jian ai” chapters also tend to interpret them in terms of compromise or adaptation.13 i shall argue, however, that Mohism did not start off in a con- frontational manner—although some contemporaries may have found it quite challenging—but became more specific and demanding over time. As we shall see, the idea of self-sacrifice and unidirectional concern for the weak and poor is absent from chapter 14 but most clearly present in chapter 16, which is why, for this triplet at least, the evolution cannot be characterized as one of compromise or dilution.

11   For similar and other evolutions traced in the three “Jian ai” chapters, see Yoshinaga shinjirō, Sengoku shisōshi kenkyū, 77–78, 106–108; Fraser, “Doctrinal Developments in MZ 14–16;” schumacher, “An outline of the evolution of the Concept of Jian in Mohism,” 12–19; and A. taeko brooks, “Mwòdž 14–16 兼愛 ‘universal love.’ ” 12 Japanese scholars have indicated the link between the “inclusive Care” and “Will of heaven” triplets. see e.g. Yoshinaga shinjirō, Sengoku shisōshi kenkyū, 96, 106. sakai Kazu- taka, “Makki boku no ken’ai shisō,” 101–105 further includes the “Fa yi” chapter. 13 see Ding Weixiang, “Mojia jian ai guan de yanbian.” A. taeko brooks, “the Mician ethical Chapters,” 111, sees the Mohists as moving “from outside critics to inside mem- bers of the system . . ., mak[ing] peace with its intrinsic war aims, and cop[ing] with the increasingly harsh conditions which apply to all who serve the state.” see also Graham, Divisions in Early Mohism, 24; Graham, Disputers of the Tao, 35–36. Yoshinaga shinjirō, Sengoku shisōshi kenkyū, 77–78, 106 sees Mozi’s own moral view translated in an ever more political and utilitarian political strategy. sakai Kazutaka, “Makki boku no ken’ai shisō,” 101–105 traces an evolution from “care” toward “benefit.” Japanese scholarship has been very attentive to differences and evolutions within the triplets.

are the three jian ai chapters about universal love?

39

to present the three “Jian ai” chapters as different and consecutive steps within one line of thought, i shall first summarize previous research on the structure of the triplet chapters, with an emphasis on the nature of this particular triplet.

The Structure of the “Jian ai” Triplet

the threefold structure of the Mozi’s Core Chapters has long been a source of speculation. From the Qing dynasty onward, scholars have identified consistent differences among the triplet chapters—not necessarily along the divisions shang, zhong, and xia—on the basis of variation in particle use, vocabulary, fixed formulas, rhetoric, style, references to authority, use of logic, or intellectual content. two major lines of interpretation are the “three-sects theory” versus the “sequence/evolution theory.”14 Angus C. Graham believed that the differences are best explained by reference to three regionally distinguished, relatively contemporary, and competing sects. on the basis of linguistic and philosophical differences, he identified them as (1) the “purists” in north China, who defended the doctrine against rival thinkers, also called the Y group because of their use of the particle yu ; (2) the “Compromising,” who were also from the north but adapted the doctrine to the ideology of the state, also called the h group because of their use of hu as postverbal particle instead of yu ; (3) and, finally, the “reactionary” in the south, who adapted even more to the political situation, identified as the J group because of their use of the particle ran (in Wade Giles jan) following the title of a quoted source. When applied to the “Jian ai” triplet, Graham’s hypothesis of regional diversity and doctrinal opposition boils down to the following picture:

chapter 15, or “Jian ai, zhong,” belongs to Y and defends the philoso- phy’s theoretical purity against rival thinkers; chapter 16, or “Jian ai, xia,” belongs to h and proposes a “watered-down” version that was acceptable to the politicians in power.15 because chapter 14, “Jian ai, shang,” does not contain characteristics of J, nor does it further adapt the doctrine to

14 see the introduction of the present volume. For an overview of various views, see e.g. Desmet, “the Growth of Compounds in the Core Chapters of the Mozi,” 99–104; and Ding sixin, “A study on the Dating of the Mo Zi Dialogues and the Mohist view of Ghosts and spirits,” 39–51. 15 see Graham, Divisions in Early Mohism, 24.

40

carine defoort

(southern) political demands, it does not fit into the scheme. its brevity, its simplicity, and the absence of certain formulas suggest to Graham that this chapter is a later digest of the Mohist doctrine of “concern for every- one,” added by the editor in the place of an older, lost J chapter.16 the explanation of the Core Chapters in terms of the “three-sects theory” has been widely accepted, whether or not in the form defended by Graham.17 other scholars before and after Angus Graham, such as Watanabe takashi 渡邊卓, helwig schmidt-Glintzer, A. taeko brooks, Chris Fraser, Yoshinaga shinjirō 吉永真二郎, and Ding Weixiang, explain the differ- ences within the triplets by postulating a chronological evolution rather than synchronic alternative stances, but not necessarily along the origi- nal order of the shang, zhong, and xia chapters. this evolution, which, according to A. taeko brooks, took place from ca. 390 to ca. 273 bCe, may later have been followed by a division of the Mohist school, but that is irrelevant to the structure of the triplets.18 like brooks, Fraser believes that the three “Jian ai” chapters, in their traditional order, represent the evolution of a relatively early and central idea in Mohism.19 this outline of the two current major views on the structure of the Mohist Core Chapters in general and of the “Jian ai” triplet in particular, suffices as a background for the interpretation of these three chapters pre- sented below. Although the hypotheses of regional diversity and chrono- logical evolution do not necessarily exclude each other, my reading of this triplet in general supports the evolutionary interpretation and more spe- cifically rejects Graham’s two major claims concerning this triplet. First i believe that chapter 14, “Jian ai, shang,” is probably not a later digest but

16 see ibid., 20–27; and Graham, Disputers of the Tao, 36. Graham, Divisions in Early Mohism, 28, explicitly sets aside questions of dating; he focuses on particles, first quota- tion formulas, introductory and concluding formulas, and only then looks at differences in thought. 17 see Graham, Disputers of the Tao, 51. A convinced follower of Graham’s three-sects theory is Maeder, “some observations on the Composition of the ‘Core Chapters’ of the Mozi,” 44, 47, 54–55, 75–76, 82. see also schwartz, The World of Thought in Ancient China, 137–138; hansen, A Daoist Theory of Chinese Thought, 99; tong shuye, Xian Qin qi zi sixiang yanjiu, 59; lewis, Writing and Authority, 59; Xinyi Mozi duben, 5. 18 see A. taeko brooks, “the Mician ethical Chapters,” 117. she dates chapter 14 to ca. 386 bCe, chapter 15 to ca. 342 bCe, and chapter 16 to ca. 310 bCe. Watanabe takashi sees the evolution of the Core Chapters as taking place over a much longer time span, from the early fourth century bCe till the Qin dynasty. see Watanabe takashi, “Bokushi shohen no chosaku nendai,” part 2. 19 see Fraser, “Mohism,” see supplement. i agree with Watanabe takashi in not respect- ing the chronological order of the shang, zhong, and xia chapters in some Core Chapters. see the introduction to the present volume.

are the three jian ai chapters about universal love?

41

an early step in the reasoning toward “universal care.” second, the politi- cally more compromising stance of chapter 16 as compared to chapter 15 is not the major difference between them. For this triplet at least, Gra- ham’s labels of “purist” and “Compromising” are not warranted. As for the first point, others have argued that chapter 14 could very well be the first, or “original,” of the three chapters, rather than a later summary.20 there are indications of a gradual evolution from shang via zhong to xia, at least in this triplet. throughout the three chapters, the argumentation becomes richer, the objections become more varied and specific, and the answers become better supported by narratives, quota- tions, theories, or analysis of technical concepts such as “inclusive, uni- versal” ( jian ), “exclusive, partial” (bie ), “distinguishing terms” ( fen ming 分名), “right versus wrong” (shi fei 是非), and “category” (lei ). Although many of these and other characteristics are closely intercon- nected, one could tentatively unravel them as different aspects of a pos- sible evolution. First, comparing the atmosphere in the consecutive chapters, one notices diminishing optimism and growing impatience: chapter still expresses confidence that order can be brought about by the sage who follows the Mohist advice; chapter 15 voices concern about the current situation; while chapter 16 complains about more and worse disasters in the world. the growing impatience with opponents is reflected in the increasing use of bi indicating the “necessary” or “inevitable” steps of sound reasoning that, according to the author, one should make.21 A sec- ond aspect of the evolution, analogous with the first, concerns the critics who are mentioned: none in chapter 14; the scholarly gentlemen (once 天 下之士, 15: 25/19, and once 天下之士君子, 15: 26/14) in chapter 15; and in chapter 16 the stubborn “critics of jian in the world” 天下之士非兼者 (16: 27/28), those “who have all heard of jian but reject it” 皆聞兼而非之 (16: 29/15) and whose objections “still don’t stop” 猶未止也 (16: 28/12). third, and not surprisingly, the objections of these critics also increase in number and seriousness. While none are reported in chapter 14,

20 see, e.g., Fraser, “Mohism,” see supplement; and A. taeko brooks, “the Fragment theory of MZ 14, 17 and 20,” 120. 21   there is no occurrence of bi (must, necessarily) in chapter 14, thirteen occur- rences in chapter 15, and twenty-seven in chapter 16. it occurs, not in the narratives or in the quotations, but in the arguments, in which the point is made either that something inevitably results from a certain cause or that a certain answer to a (didactic) question is inevitably right (bi yue 必曰, in chapter 16). this increasing use of bi in the argumentation may indicate development toward the later Mohist focus on argumentation.

42

carine defoort

those in chapter 15 question the applicability of jian. so do four out of the five objections in chapter 16 (e.g., “how could it be used?” 豈可用

, 16: 27/28). Mozi is said to reply impatiently that “if it were not appli- cable, even i would object to it” 用而不可, 雖我亦將非之 (16: 27/29). one objection in chapter 16 goes so far as to question the very value of “inclusiveness,” not just its applicability, by opposing it to the duties of

a filial (xiao ) son. Along with this growing opposition comes a fourth

and more complex evolution, namely in the increasing types of defense or argumentation. the plain argument of chapter 14 is supported in chapter 15 by well-known stories and model figures (with one quotation) indicat- ing the feasibility of the Mohist project; in chapter 16 the views are further supported by five quotations from authoritative sources and explicitly attributed to the exemplary figures mentioned in them. For example:

雖子墨子之所謂兼者, 於湯取法焉.

even what our Master Mozi calls “inclusive” is derived from the model of tang. (16: 29/10–11)

A

fifth aspect of the evolution could be called rhetorical: while chapter 14

is

one piece of monologue, chapters 15 and 16 make use of didactic ques-

tions and answers, like a catechism in which purely hypothetical alterna- tives are posed, reflections on good reasoning are presented, and technical vocabulary is stipulated. Admittedly, none of these characteristics, even when combined, is totally conclusive in the rejection of Graham’s theory, since the absence in chapter 14 of seemingly later characteristics could be determined by the style or decision of later authors. but this possibility, we will see, is much further from being proven convincingly. three more arguments in favor of the evolution theory relate to the similarities and differences between the three chapters. First, it is generally known that the three “Jian ai” chapters share ideas, vocabulary, and sentences. unique similarities exist between chapters 14 and 15, on the one hand, and between chapters 15 and 16, on the other, but there are almost none between chapters 14 and 16.22 this strongly suggests that chapter 15 made use of some written or remem- bered version of chapter 14, adding arguments and narratives to support the central idea; and that chapter 16 did the same with chapter 15, again

22 the sole exception is the use of the verb luan (to disrupt) in a parallel passage in chapters 14 and 16, where chapter 15 uses cuan (to usurp). the latter also occurs in a parallel passage in “tianzhi, shang,” which in various other ways parallels chapter 15. see Mozi 26: 43/26, 26: 43/28.

are the three jian ai chapters about universal love?

43

supplementing reflections, arguments, objections, and responses as the authors found necessary in the context of their time. thus, while further elaborating upon each other, the chapters increased in length, each dou- bling the previous one in number of characters.23 in Graham’s alterna- tive the later writer of chapter 14 would have made a summary of only chapter 15 without any indication of knowing the content of the best- developed arguments in the last chapter. A second indication of the chronological priority of chapter 14 could be the fact that, while increas- ing tension may have caused objections and responses to increase, some arguments seem to have disappeared, perhaps because they were consid- ered generally accepted and not disputed by opponents. For instance, the first chapter argues at length that political chaos should be handled by the sage just as a disease is treated by a doctor, namely through diagnosis and remedy: like a doctor, the sage has to find the cause of chaos and suggest a solution. this argument does not occur in chapter 15 or chapter 16, but the medical analogy is taken for granted: it is simply used without any explicit legitimation. if chapter 14 were a later summary of chapter 15 (and perhaps of chapter 16), it would be difficult to explain why it starts out by arguing a point that the other authors briefly apply and that nobody, as far as we can tell, ever calls into question. And a third indication of the prior- ity of chapter 14 is that the four cases of chaos mentioned in the first half of this chapter appear in a reversed order not only at the end of the same chapter but also in the whole of chapters 15 and 16. As will be illustrated in the next section when discussing the argument of the chapter, these differences suggest that a reversal, made for good reasons in chapter 14, was retained throughout the two following chapters.24 the final argument in favor of the chronological priority of chapter 14 concerns the second point of disagreement with Graham, namely on the characterization of the different chapters of the triplet. Connected with the changing atmosphere, the critics and their criticism, the arguments and

23 According to A. taeko brooks, “Mwòdž 14–16 兼愛 ‘universal love,’ ” 129–130, chap- ter 14 has 585 words, chapter 15 has 1,312 words, and chapter 16 has 2,716 words. this means a progression at the ratio of 1:2:4. 24 For more proof of this evolution, see Desmet, “the Growth of Compounds in the Core Chapters of the Mozi.” erik Maeder, in “some observations on the Composition of the ‘Core Chapters’ of the Mozi,” considers Graham’s identification of some later “Digests” as one “of the most important results of Graham’s work” (39), and he also indicates: “Atten- tion to pattern is dominant in all three ‘Digest’ chapters . . ., reaching its seeming perfec- tion in chapter 14” (55). but he does not give any extra argument in relation to chapter 14, neither for its status as a later digest nor for his more general claim that the basic unit of the chapter is the “paragraph.”

44

carine defoort

rhetoric, there is a subtle evolution in the content of the three chapters, which is not well captured in the labels “later digest” (chapter 14), “purist” (Y, chapter 15), and “Compromising” (h, chapter 16). Graham’s arguments in relation to this particular triplet are rather thin. As for chapter 14, he believes that the “central importance of universal love in Mohism makes it incredible that ch. 14 . . . can be one of the authoritative statements of the doctrine,” because it is short, lacks quotations as well as answers to objections, has few parallels with chapters in the same triplet, and lacks attributions to Mozi (except for one mention in the conclusion).25 As for Graham’s characterization of the two other chapters, his only and indi- rect argument is that chapter 15 addresses “rival thinkers who question the doctrine of universal love,” presumably the officers (shi ) or officer- gentlemen (shi junzi 士君子),26 while chapter 16 addresses princes and men of state and hence is more political in nature. Graham does not quote any example to support his case, but there is exactly one occur- rence of “kings, dukes, great men” 王公大人 in the conclusion of chapter 16 (30/7). the more common opponents in chapter 16 are “those among the officers of the world who reject it [ jian],” and their criticism differs from that in chapter 15 only in amplitude and philosophical subtlety, not in political demands for conformity. Graham, moreover, admits that, at least in the “Jian ai” triplet, there is no major difference in content: “in the next triplet, Chien ai, there is no evidence of compromise on the central Mohist doctrine of universal love. the J chapter is missing, but Y and h, as well as the digest ch. 14, all say explicitly that each should regard the family of another as though it were his own.”27 hence, the content of the “Jian ai” triplet does not support his labels. i have similar reservations about characterizations of the triplet as revealing an increasing political

25 Graham, Divisions in Early Mohism, 4. stephen Durrant concluded, on the basis of the use of grammatical particles, that either of the hypotheses (the three-sects theory and chronological evolution) could be correct. see Durrant, “An examination of textual and Grammatical problems in Mo-tzu,” 172. For arguments in favor of the chronological sequence 14–15–16, see, e.g., Watanabe takashi, “Bokushi shohen no chosaku nendai,” part 1, 4–9; Ding Weixiang, “Mojia jian ai guan de yanbian,” 71–72; and Fraser, “Mohism,” see supplement. 26 Graham, Divisions in Early Mohism, 20. he must have overlooked the shi junzi at 15: 26/14 when he argues that “thinkers who oppose Mohism on the issues of universal love . . . are never called officer gentlemen” (19). 27 ibid., 24.

are the three jian ai chapters about universal love?

45

conformism28 and a gradual acceptance of war.29 All this indicates that, at least for the “Jian ai” triplet, Graham’s labels—“later digest,” “purist,” and “Compromising”—are not warranted. My alternative characterization of the three chapters highlights their views concerning the nature of “caring.” nobody seems to deny that it is good to love or care (ai). but the crucial question is: to whom should this love or care be directed? oneself, each other, specific others, or everyone? And how specific should it be? the answer differs in the three “Jian ai” chapters, which, i argue, can be seen as stages in the growing scope and specificity of caring, reaching the ideal of “inclusive” or “impartial care” only in the last chapter, where the expression jian ai occurs for the first time. in this respect, what distinguishes chapter 16 from chapter 15 is not an increased willingness to compromise but rather a further radicalization of the moral stance. the evolution that will now be traced throughout the triplet starts off with relatively vague reciprocal love within familiar rela- tionships and moves toward specific and unidirectional concern of the rich and strong for the poor and weak. More striking than the tendency toward an ever-growing scope of ai, there is an increasing specification of the moral stance: the attitude of caring in chapter 14 is specified in chap- ter 15 as a double duty: feelings of concern (ai ) as well as beneficial acts (li ). thus, it is only in chapter 15 that the concept of benefit or profit enters the scene in a positive sense.30 What exactly is counted as “caring” becomes ever more specific in chapter 16.

“Caring for Each Other” 相愛 (Chapter 14)

unlike the two following chapters, chapter 14 lacks short statements attrib- uted to Master Mozi in response to various critiques but instead consists of one long reflection on political chaos and order. As suggested by its

28 As argued by, e.g., Yoshinaga shinjirō, “Jian ai shi shenme”; and A. taeko brooks, “the Fragment theory of MZ 14, 17 and 20,” 120. there certainly is an increasing political interest in the triplet, but nothing indicates that Mohists occupied positions at the court. 29 As argued by, e.g., Graham, Divisions in Early Mohism, 19–20, and, with more nuances, by A. taeko brooks, “Mwòdž 14–16 兼愛 ‘universal love,’ ” 131. 30 in chapter 17, li is also related to egoism and harming others in terms similar to those in chapter 14 (以虧人自利也, 17: 30/18). see also Fraser, “Doctrinal Developments in MZ 14–16;” and the essay by paul van els in this volume. According to A. taeko brooks, “the Mician ethical Chapters,” 117, chapter 17 is the earliest Core Chapter, slightly predating chapter 14.

46

carine defoort

conclusion, the whole chapter is someone’s specific interpretation of what Master Mozi may have meant by his insistent entreaty to care for others:

故子墨子曰不可以不勸愛人者, 此也。

thus, the fact that our Master Mozi says that we must encourage people to care for others is because of this. (14: 24/22)

hence, the encouragement to “care for others” (ai ren 愛人) may have been the only original echo of the master’s thought, if we consider the sayings attributed to him in the two following chapters as didactic tools rather than instances of actual speech. As Yoshinaga shinjirō has sug- gested, the encouragement to simply “care” may very well represent the earliest Mohist concern: if a master had explicitly pleaded for “inclusive care,” and if that motto was already current, the author of the first “Jian ai” chapter would not have failed to mention it.31 the Mohist starting point of jian ai is probably “care for others,” a view that was also attributed to Confucius in Lunyu (12.22):

樊遲問仁子曰: 愛人

Fan Chi asked about goodness. the master said: care for others.32

Why did Master Mozi, according to the author of this chapter, urge his audience to care for others? in other words, what does the “this” of the conclusion refer to? it refers to his own interpretation of this moral imperative to care for others, arguing that someone who dedicates his life to the noble cause of ordering the world has to diagnose the political disease and subsequently suggest a remedy. Furthermore, the diagnosis, which consists of the first half of his argument (from 14: 24/4 to 24/12), is the failure to care for each other (xiang ai 相愛) because people care only for themselves (zi ai 自愛); hence, they benefit themselves to the detriment of others. the remedy, in the second half of the argument (from 14: 24/14 to 24/19), is that people are made to33 “care for each other inclu- sively” ( jian xiang ai 兼相愛), so that the causes of disorder are removed.

31   see, e.g., Yoshinaga shinjirō, Sengoku shisōshi kenkyū, 75, 78; and Yoshinaga shinjirō, “Jian ai shi shenme,” 31. 32 this may be a relatively late saying. i refrain from speculating about cause and effect between Mohist and Confucian ideas. see ibid., on ren as a non-familial duty. 33 Shi 使 in the sense of “order them to,” “make them,” is a political initiative, which is further developed in chapters 15 and 16. if we translate 使 as “suppose that,” then what follows is hypothetical thinking. A third possibility is that a mere hypothesis in chapter 14 was later interpreted as a political initiative.

are the three jian ai chapters about universal love?

47

According to the author, this is what the master meant when he encour- aged his audience to care for others. the first step that chapter 14 makes toward what will later be known as “universal caring” ( jian ai) is to widen the scope of one’s care for oneself by including others in relations of reci- procity (xiang ). Jian enters only in the second half of the chapter: it is neither stressed nor explained but broadens the scope from one’s familial or political in-group to include others in relations of mutual care. this broadening of scope begins implicitly through the presentation of four analogous cases of the diagnosis of disorder in the first part. At the beginning the author seeks not to antagonize contemporaries, but rather to convince them by showing that their rejection of certain types of self- ish behavior logically ought to lead to a rejection of analogous cases of “care for oneself ” (zi ai 自愛). he thus first describes a situation in which there is a lack of respect on the part of the lower actors in dyadic and hierarchical relations: of a son versus his father, a younger versus an older brother, a minister versus his lord. everyone, including the conservative elite, would call these attitudes (which i label “case 1”) “disorder” (luan 亂) and hence would reject them. in a second step, the audience is invited to also reject three other instances of egoism, namely of the father, the older brother, and the ruler in relation to, respectively, the son, younger brother, and minister (case 2). thus far, most members of the elite would have no problem sharing the Mohist concern. the third analogy concerns the indisputably despicable behavior of thieves and murderers because they, respectively, care only for their own houses and their own per- sons (case 3). the fourth and last analogy condemns the top ministers and lords, who, respectively, disrupt and attack each other’s families and states out of concern for their own families and states (case 4). At this point, we have reached the problem of political disorder that the sage is eager to solve. the reader is tempted to follow the author in his consecutive rejection of these seemingly analogous cases of “care for oneself,” up to the rejec- tion of top ministers (dafu 大夫) and lords (zhuhou 諸侯) who fight for their own families and states. i strongly suspect that this last point is what the author wants to bring home: in the political remedy presented in the second part of chapter 14 as well as in the two following chapters, the order of these four types of disorder (luan ) is reversed and most atten- tion is directed toward this fourth case of the diagnosis: the chaos gener- ated by ministers and feudal lords. unfilial sons (case 1), unloving fathers (case 2), and thieves and murderers (case 3) were mainly brought into the picture as commonly rejected egoists, just like, at least according to these

48

carine defoort

Mohists, these powerful aristocrats. but this last analogy was probably a step that some contemporaries were not willing to take. not that they particularly favored war in itself, but they generally admired knights as filial sons or loyal ministers for the honor and wealth that they acquired for their ancestors, parents, and lords. A failure to do so was considered a threat to the cornerstone of morality, namely the family or clan. in this argument, however, the Mohist challenge forces conservative contempo- raries to show where the analogy breaks down: since they reject the selfish son, the unloving father, and the thief as “caring for himself ” and failing to “care for each other,” why not also repudiate the aristocratic ministers and feudal lords who engage in war? if unable to counter the last step of this reasoning, they are compelled to join the rejection of these aristocrats on the basis of their rejection of the three analogous cases. in the second part of chapter 14, the remedy to the chaos caused by ego- ism is that we “inclusively care for each other” ( jian xiang ai), not only in hierarchic familial and political relationships (cases 1 and 2) but also more broadly in relation to strangers (case 3) and, most importantly, to other families and states (case 4). it is worth noticing that in the final summary of this remedy, the four cases are repeated in opposite order, with most attention given to the two “egoists” of case 4: feudal lords and top minis- ters attacking and disordering each other’s states and families, respectively (14: 24/18–19). As pointed out above, this reversal is preserved through- out the two following chapters and reflects well the Mohist concern with political chaos. thus, in his argument for reciprocity in these four cases, the author also builds in an explicit plea for broadening one’s scope of care and concern: jian. precisely this aspect will be criticized by opponents and defended by the Mohists in the two following chapters. opponents could have indicated a flaw in the argument, since the scope of caring is inherently ambiguous: very often, egoism or “care for oneself” coincides with altruism or “care for others,” such as when it benefits more than just oneself, as the third case shows (a thief steals for the people of his own house, not just for himself). While obviously rejecting the behavior of a thief, opponents risk also rejecting as egoism an attitude that they would under other circumstances (those of case 4) consider altruism: top minis- ters and feudal lords who promote the benefit of their own states or fami- lies are analogous not only with thieves and selfish sons but also with their opposites, namely respectful sons in relation to their family. the ambiguity of the scope of caring allows one to stress either one of the viewpoints. in the Mozi there is obviously no space for this line of defense. the content of chapter 15 shows, on the contrary, that Mohism evolved away

are the three jian ai chapters about universal love?

49

from conventional morality by further widening its scope of concern and by increasingly insisting on concrete beneficial acts. As stated above, the double fact that, first, only chapter 14 really takes pains to incite the audi- ence in its diagnosis of political problems by initially giving priority to family virtues and, second, that the focus on top ministers and feudal lords in its second part is taken up by the two following chapters seems to further support the chronological priority of chapter 14.

“Inclusively Caring for Each Other, Mutually Benefiting Each Other” 兼相愛、交相利 (Chapter 15)

Chapter 15 is double the length of chapter 14: about one-third consists of a summary of the argument of chapter 14, including both the diagnosis and the remedy; the remaining two-thirds formulate a defense of the Mohist view in the face of two very similar criticisms. only in that part of chapter 15 does the idea of jian enter the debate as an independent concept for Mohists to defend in response to specific objections to the practicability of “inclusiveness.” the expression jian ai has not yet appeared. the introductory summary not only briefly contains the views of chap- ter 14 but also includes some differences. the two main differences are the addition of “benefit” and the increasing specificity of morality. First, the moral duty of “caring” in chapter 14 is now divided into “caring and ben- efiting.” the author of chapter 14 may have considered ai (caring) a matter of both feeling (“care for”) and consequently acting (“take care of ”), the latter apparently amounting to a rather passive attitude of not disturbing or attacking others. the explicit splitting of the moral duty in chapter 15 somewhat moves ai to the realm of emotions or attitudes and separately stresses the importance of beneficial acts. “benefit” (li ), which was exclusively related to egoism in the diagnosis of chapter 14,34 now posi- tively joins the duty of “caring” in a moral tandem—hence the insistent promotion throughout chapter 15 of the method ( fa ) of “inclusively caring for each other, mutually benefiting each other.” the second major difference in the introductory summary of chapter 15 is the increased specificity about the scope of one’s care and beneficial

34 see also Watanabe takashi, “Bokushi shohen no chosaku nendai,” part 1, 8–9, on this new social dimension. i do not suggest that the authors of this chapter (and chapter 17) thought exclusively negatively about benefit, but it was clearly not yet an “umbrella term covering all their core values,” as Dan robins puts it in robins, “Mohist Care,” 61.

50

carine defoort

acts, although perhaps less in the case of benefiting than of caring.35 the author’s concern is more specific and goes far beyond hierarchical rela- tions in which reciprocity conventionally dominates:

天下之人皆不相愛, 強必執弱, 眾必劫寡, 富必侮貧, 貴必敖賤, 詐必 欺愚

if people in the world all fail to care for each other, the strong will inevi- tably have power over the weak, the many will inevitably force the few, the rich will inevitably insult the poor, the noble will inevitably be arrogant toward the vulgar people, and the cunning will inevitably cheat those who are simple of mind. (15: 25/6–7)

Despite the presence of jian in the expression “inclusively care for each other,” the argument still largely runs in terms of “caring for each other,” but “inclusively” represents the widening scope that explicitly encom- passes the weak, the poor, the vulgar, and the simple of mind. these peo- ple stand for those with whom the elite usually did not have a relationship of reciprocity (xiang). Following this introduction, two objections in chapter 15 specifically attack the scope of morality, namely “inclusiveness.” now, for the first time, this term is discussed separately as a concept (inclusiveness) and not just mentioned as an adverb within a longer expression (inclusively). the first objection is:

然,乃若兼則善矣。雖然,天下之難物于故36 也。

Admitted, “inclusiveness” is good indeed. but it is, nevertheless, the most difficult thing in the world. (15: 25/21)

the second objection is very similar: it also admits that inclusiveness is a worthy ideal and equally finds it “something that cannot be put into prac- tice” 不可行之物也 (15: 26/14). Criticism of the Mohist view is remark- ably mild in this chapter. i find none of the resentment and outrage that James legge reads in the opponents’ remarks.37 no critic rejects caring

35 the double slogan of this chapter “inclusively care for each other and mutually ben- efit each other” could suggest a difference between a very broad scope of ai ( jian xiang ai 兼相愛), as in the second part of chapter 14, and the mere reciprocity promoted for li ( jiao xiang li 交相利). 36 sun Yirang interprets 于故 as 迂故, “an unrealistic task” 迂遠難行之事. see sun Yirang, Mozi jiangu, 103–104. 37 legge, The Chinese Classics, vol. 2, 120: “the essay [legge treats the three chapters as one essay] shows that it was resented as an outrage on the system of orthodox belief during all the life-time of Mih and his immediate disciples.” the rather mild objection to jian also occurs in four of the five objections expressed in chapter 16.

are the three jian ai chapters about universal love?

51

or reciprocity, not even the value of benefit. What they do object to is only “inclusiveness,” more specifically because of the apparent difficulties it entails in practice. According to the first response in chapter 15, the problem of these crit- ics lies in a failure to understand:

天下之士君子, 特不識其利辯其故也

the officer-gentlemen of the world really don’t understand their benefit, nor do they distinguish their motivation [or: the causes of their actions]. (15: 25/22)

such a failure can be undone by explaining, arguing, and teaching, a duty that the author takes to heart. the misapprehension of the critics con- sists of two major aspects: shortsightedness and self-contradiction in their motivation, at which the Mohist responses are consequently aimed. As for the shortsightedness, the author sets out to show that reciprocity also exists on a wider scope than only within conventional and hierarchi- cal relations, as there obviously is between fathers and sons, elder and younger brothers, ministers and lords:

夫愛人者, 人必從而愛之; 利人者, 人必從而利之惡人者人必從 而惡之害人者人必從而害之此何難之有?特上弗以為政,士 不以為行。故38 也。

Well, one who cares for others will inevitably as a consequence be cared for by them; one who benefits others will inevitably as a consequence be benefited by them. one who hates others will inevitably as a consequence be hated by them; one who harms others will inevitably as a consequence be harmed by them. What is difficult about this? it is only that the superiors don’t make it their policy and that officers don’t make it their lifestyle. that’s what it is. (15: 25/24–25, see also 15: 26/11–12)

the author thus assumes that a failure to see this broader reciprocity makes people reluctant or even incapable of doing good to those with whom they have no specific relation. hence, these officer-gentlemen find it very difficult to include strangers in their scope of caring. but if only they recognized the benefit to be reaped from treating others well, they would certainly incorporate inclusiveness in their policies and their behavior. second, the implicit contradiction in understanding their own motiva- tion lies in the fact that these gentlemen reject “inclusiveness” as being too difficult or impracticable while they are willing and able to do things that are much more difficult than that:

52

carine defoort

今若夫攻城野戰殺身為名此天下百姓之所皆難也苟君說

則士眾能為之。況於兼相愛交相利則與此異

As for attacking a city and fighting in the fields, offering one’s life for a repu- tation, these are things that people in the world consider difficult. but if the lord finds pleasure in them, then masses of knights/officers are able to do them. how much easier is it, compared to this, to care for each other inclu- sively and to benefit each other mutually. (15: 25/22–24)

if capable of these military feats, how much more should they be able to “inclusively care for each other and mutually benefit each other,” an imperative that is both easier (at least not as life threatening) and more beneficial (because of the inevitable reciprocity of one’s acts)? the author refers to three stories that show how easy inclusiveness would be for sub- jects if only they understood the great benefits of this moral imperative. the stories describe cases in which ministers starve themselves, dress shabbily, or give their lives for the sake of a reputation, simply because this is what pleases their lord. these examples of fights and attacks do not necessarily indicate a growing acceptance of warfare by Mohists but illustrate a contradiction that they perceive in the elite’s code of behavior:

in arguments, the officer-gentlemen reject “inclusiveness” because of the difficulties in practicing it; but in reality, they are very willing to under- take acts that are much more difficult, life threatening, and harmful. the second Mohist response of chapter 15 mainly refers to the feats of three ancient model rulers, Yu , King Wen 文王, and King Wu 武王, and their ability to “practice inclusiveness” or “impartiality” 行兼 (15: 26/21, 15: 26/24, 15: 26/26).39

“With Inclusiveness Replace Exclusiveness” 兼以易別 (Chapter 16)

Chapter 16, “Jian ai, xia,” again doubles the previous chapter in length. the introductory summary of the argument, also consisting of a diagnosis and remedy, now only occupies one-fifth of the chapter, while the objec-

39 While the first response highlights the behavior of subjects, the second refers to the lords and contains references to three sagely kings from antiquity who practiced jian: Yu benefited people from all regions, including barbarians; King Wen helped the old, child- less, and widowed; and King Wu selected workers impartially, taking the blame upon himself whenever something went wrong. their acts were supposed to help people learn to practice jian themselves (15: 26/21, 15: 26/24, 15: 26/25). it is a striking indication of the power of expectations that translators often add the word “love” to their translation after ( jian), sometimes between brackets, where there is no (ai) in the original text. see, e.g., Johnston, The “Mozi,139, 143, 145; and Mei, The Works of Motze, 166, 170, 172.

are the three jian ai chapters about universal love?

53

tions and their responses occupy almost four-fifths. the two differences pointed out about the summary in chapter 15—the positive reference to benefit and the wider scope—seem to be further developed here: the social concern has increased, and “inclusiveness” is promoted from the beginning of chapter 16 as an independent value. in chapter 16, the moral value of “caring” is taken for granted, reciprocity ( jiao or xiang ) has become somewhat less dominant, and the focus lies on the scope of care, namely “inclusiveness” ( jian ). hence, my temporary title for this chapter is jian yi yi bie 兼以易別 (16: 27/14). A total stranger to the debate, who reads only this chapter, might initially wonder in what respect one has to be inclusive. Certainly not in hate and harm! the argumentation seems to be based on the previous chapter(s) and simply assumes that we are talking about caring. the diagnosis has become more serious: chaos caused by attacking states and disorderly families (case 4 in chapter 14) is the major cause of harm in the world, as it was in chapter 15. this assessment is immediately followed by examples of situations in which the strong, numerous, cun- ning, and noble maltreat the weak, few, simple, and vulgar. only then are ungenerous rulers, disloyal ministers, unloving fathers, and disrespect- ful sons mentioned.40 And finally, instances are added where common people harm and hurt each other in various ways. the cause of all this misery is the opposite of “inclusiveness,” namely bie , “exclusion” or “exclusiveness,”41 which is said “to be wrong” 別非也 (16: 27/13). the rem- edy is its replacement by jian, the central topic of this chapter. in the argu- ment below, “caring” and “benefiting” are not even explicitly mentioned:

是故子墨子曰兼以易別然即兼之可以易別之故何也?曰藉為 人之國若為其國夫誰獨舉其國以攻人之國者哉為彼者由42 為己 也

thus, our Master Mozi says: replace “exclusiveness” with “inclusiveness.” but what is the basis for replacing “exclusiveness” with “inclusiveness”? Well, assume that people treat someone else’s state as they treat their own state.

40 A selection of cases 1 and 2 from chapter 14 is presented, but in a different order: in chapter 14, the causes of disorder (case 1) mentioned first are ministers, sons, and younger brothers, whereas in the chaos described in chapter 16, rulers are mentioned before min- isters, and fathers before sons (16: 27/8–9). 41   it is used as an adverb in the expression “exclusively/partially hate each other” 別 相惡 in chapter 26. 42 read as you as in the two following parallel lines.

54

carine defoort

Who would then mobilize his own state to attack someone else’s state? they would treat the others as they treat themselves. (16: 27/14–16)43

if, thanks to the ideal of “inclusiveness,” people cared for others as much as they did for themselves, “one would have to conclude that this would amount to benefit for the world” 即必曰天下之利也 (16: 27/18). More specifically, elderly widowers without sons would be supported in their old age, and orphans would be provided for so they could grow up (16:

27/24–25).44 it seems that in a gradual evolution from chapters 14 to 16, the Mohists expect people to show a moral concern for others, not just in traditional relations, nor just in more and larger reciprocal connections, but expanding to encompass all those who need help—an extremely “inclusive” care, gradually tending toward “impartiality” or “universality.” the major part of this chapter (almost four-fifths) consists of five objec- tions and responses. the content of the objections and the order of their appearance strongly suggest that they do not reflect a continuous reason- ing but were collected from existing textual fragments.45 Four objections (1, 2, 3, and 5) largely resemble those of the previous chapter. they are remarkably positive toward the Mohist project: they find it good (shan ), humane (ren ), and right (yi ; this is the first appearance of the important Mohist concept of “right” or “righteous” in the whole “Jian ai” triplet!). but they too mainly doubt its practicability. the first two Mohist responses are new: they consist of thought experiments, respectively about officers (shi ) and lords ( jun ) who, in words and deeds, stand for bie or for jian. the author argues that everybody without exception would, in fact, prefer to deal with an “inclusive” officer or lord, even if, in theory, one rejects the value of “inclusiveness”:

雖非兼之人必寄託之於兼

之有是也此言而非兼擇即取兼即此言行費46不識天下之

我以為當其於此也天下無愚夫愚婦:

所以皆聞兼而非之者其故何也

43 this passage is followed in 16: 27/16–17 by the same reasoning but with respect to one’s city du and then one’s family jia . 44 this is also said about the second model ruler in the second response of chapter 15

(26/22).

45 the first and second responses go together, respectively arguing about the “officer” (shi ) and the “lord” ( jun ). this separation was implicitly present in chapter 15 in the sense that its first response refers to narratives about officers and its second response to exemplary rulers. the third and fifth responses in chapter 16 are inspired by, respectively, the second and first responses of chapter 15. only the fourth objection makes a further point. 46 read in the sense of fu “go against”. see Xinyi Mozi duben, 104 n. 8; and Mozi jiaozhu, 185 n. 43.

are the three jian ai chapters about universal love?

55

i believe that facing this [choice], there are no stupid men or women in the world: even those who reject inclusiveness would certainly depend on the one who considers inclusiveness right. this is to reject inclusiveness in words but to select it in one’s choices, which is a contradiction between words and deeds. i really don’t understand why the officers of the world reject inclusiveness once they have all learned about it. (16: 28/8–10; see also 16: 28/21–23)

this is how the two first responses in chapter 16 end, concerning the best officer (response 1) and the best lord (response 2). the focus is “inclu- siveness” or “impartiality.” the concept of “care” is absent but replaced in both hypothetical cases by acts that are considered to be caring: feeding the hungry, clothing the cold, supporting the sick, and burying the dead (16: 27/31–28/1, 16: 28/3–4, 16: 28/15–16, and 16: 28/18). the argument, more- over, illustrates well the further intellectual evolution of the Mohists: while self-contradiction was implicitly criticized in the responses of chapter 15, it has now become the major argument. by indicating the contradiction between words and deeds, the author considers the objection refuted and proceeds to the next objection. the slogan from chapter 15, “inclusively caring for each other, mutually benefiting each other,” appears in the response to the third objection. this response, together with the fifth, closely resembles the two of the previ- ous chapter, although they are now argued more profusely.47 the third response (like the second in chapter 15) promotes the Mohist interpreta- tion of “inclusiveness” through reference to ancient sages and kings, but now explicitly supported by quotations from respectable sources: the “tai shi” 泰誓 (Grand oath),48 “Yu shi” 禹誓 (oath of Yu),49 “tang shuo” 湯說 (Declaration of tang),50 and “Zhou shi” 周詩 (odes of Zhou).51 it is here that the sole use of the expression jian ai in the whole triplet appears; it will be discussed in the following section. the fifth response contains the three narratives (of the first response in chapter 15) in which subjects prove to be able and willing to do much more difficult feats than what the

47 the third response in chapter 16 further elaborates (using more quotations from clas- sic sources) on the second response in chapter 15. the fifth response in chapter 16 largely copies the three narratives (in different order) of the first response in chapter 15. 48 this passage does not occur in the current “tai shi” in the Shangshu. 49 this title does not occur in the current Shangshu, but a similar passage occurs in “Da Yu mo.” see legge, The Chinese Classics, vol. 3, 64–65. 50 this passage does not occur in the current Shangshu, but the last lines contain simi- larities with “tang gao”; see ibid., 189–190. 51   this passage does not occur in the current Shijing, but the first two lines occur in the Shangshu, “hongfan.” see ibid., 331.

56

carine defoort

Mohists ask, as long as this pleases their lord. As in chapter 15, the contra- diction between their willingness to offer their lives, on the one hand, and their rejection of inclusiveness as something too difficult, on the other, is implicitly present but not further elaborated. the most interesting objection of chapter 16 is the fourth, because it takes the argument a step further by showing the complexities of an ever- increasing scope of caring: it is not mere egoism combined with a failure to recognize reciprocity on a large scale that prevents the critics from treating strangers well; it is the care that they owe above all to their own parents. only the critics of the fourth objection explicitly worry about the practice of filial love:

意不忠52 親之利而害為孝乎

should we perhaps abandon our parents’ benefit and harm the practice of filial love? (16: 29/17)

their question resonates with more conventional or familial concerns and the fear that the larger scope may endanger the smaller one.53 Chapter 14 started from a need for reciprocal caring within the conventional types of hierarchy and only then broadened the scope, mainly asking people to refrain from harmful acts; but chapter 16 clearly demands active care for widows and orphans, the hungry, the cold, the sick, and even the dead. but how can one take care of one’s parents as a filial son when all these other—sometimes even opposite—duties claim moral priority? to offer one’s life for a reputation is not just a failure to understand one’s own benefit but also an act of respect and care for one’s parents and ances- tors. reference to the reputation to be won in battle is one way in which the elite defends the drive of their knights to engage in war.54 thus, what looks like “harm to others” (strangers) from one viewpoint can be “benefit for others” (within one’s family) from another viewpoint. to fight another family is thus not unambiguously a harmful act, as the Mohists want us to believe. or, to put it in terms of the diagnosis in chapter 14, the knight who fights another family for the reputation of his own family is analogous, not with the unfilial son in case 1 but, on the contrary, with a filial son.

52 replacing the character zhong by , meaning “to fit, accord with.” see Mozi jiao- zhu, 194 n. 104. 53 the same concern seems to have motivated ban Gu’s criticism of Mohism: “they extend the idea of inclusive care to the point that they don’t know how to distinguish between kin and stranger” 推兼愛之意而不知别親疏. see Hanshu 30.1738. 54 see, e.g., Mozi 19: 33/26–27.

are the three jian ai chapters about universal love?

57

the response of the Mohists to this objection is that moral acts in the largest scope will inevitably be rewarded in the smaller scope, and not just in relation to oneself (as was the case in chapter 15). since a filial son obviously also wants other people to be good to his parents, he must be good to their parents and not harm them. thus, serving one’s own parents ultimately amounts to caring for all other parents inclusively. this is how truly filial sons reason:

若我先從事乎愛利人之親然後人報我以愛利吾親乎意我先從事

乎惡賊人之親然後人報我以愛利吾親乎即必吾先從事乎愛利人

之親然後人報我以愛利吾親也

Will others reward me by taking care of and benefiting my parents if i first work hard at taking care of and benefiting their parents? or will they reward me by taking care of and benefiting my parents if i first work hard at hating and hurting their parents? Certainly, they will reward me by taking care of and benefiting my parents if i first work hard at taking care of and benefiting their parents. (16: 29/19–21)

this reasoning not only is a clear instance of common sense, according to the author, but can also be founded on ancient sources that celebrate the functioning of rewards or compensation (bao ).55 An ode of the “Da ya” 大雅 (Great odes) makes exactly this point:

『無言而不讎無德而不報

投我以桃報之以李

即此言愛人者必見愛也而惡人者必見惡也

“every word gets its answer. every good deed has its recompense. You throw a peach to me, i’ll reward you with a plum.”56 this is saying that the one who cares for others will inevitably be cared for, and who hates others will inevitably be hated. (16: 29/23–24)

With this response to the concerns expressed by conservative opponents, i conclude the analysis of the “Jian ai” triplet, leaving the single occur- rence of the expression jian ai in the triplet to be discussed in the context of its fuller elaboration in other Core Chapters. in the whole triplet, “care” in itself is never an object of controversy, but only its scope and specific content: the value of reciprocity (xiang) that was prominent in chapter 14 was slowly replaced by inclusiveness ( jian) from the middle of chapter 15

55 on this concept, see Yang lien-sheng, “the Concept of ‘pao’ as a basis for social relations in China.” 56 these are two separate lines in “Yi” , ode 256. see legge, The Chinese Classics, vol. 4,

514–515.

58

carine defoort

onward, and most explicitly so in chapter 16. the deep-rooted idea of reci- procity has not disappeared but has become a part of the explicit argu- ment in favor of impartiality: those who are not shortsighted realize that being good to others will involve compensation for themselves (chapter 15) and for their loved ones (chapter 16). but this is clearly not the final stage of jian ai: its occurrence in chapter 16 as well as in other Core Chapters illustrates the inherent dynamics of the Mohist idea: the new demand for inclusive caring moves further on, almost leaving behind all reflections in terms of reciprocity.

The Birth of Jian Ai

What was the steering force behind the above-traced evolution of jian in these three and other Core Chapters of the Mozi? one could imagine that Mohist thought evolved as a consequence of various types of criticism, by gradually accommodating the critics. As noted above, Ding Weixiang reads in the three “Jian ai” chapters an evolution of accommodation and watering down of an originally quite radical doctrine. Yoshinaga shinjirō traces an evolution from a moral vision in chapter 14 toward an increas- ingly political and utilitarian stance in chapter 16.57 if we focus on those objections that were quoted in the triplet—admittedly perhaps only an unrepresentative portion of the actual critical voices—we can detect a minor and a major trend. the minor and most critical trend is represented only by the fourth objection in chapter 16, which concerns the duty of “filial love” or “care for one’s kin” (xiao ). since a filial son must reciprocate the care that his par- ents gave him, he cannot treat them merely like others to whom he owes nothing. even a theoretical understanding of the reciprocity that rules the larger scope of one’s relations cannot undo this priority. the growing scope of jian that i have tried to describe is certainly not an accommoda- tion to this objection. there are other records of attempts to counter the criticism of traditionalists by nuancing the Mohist view.58 such compro- mises between the broad scope of jian and the conventional priority of

57 Yoshinaga shinjirō, “Jian ai shi shenme.” 58 Mozi 44 and Mengzi 3A5 are often understood as compromises of later Mohists stressing the division ( fen ) of labor between one’s feeling of caring for everybody and the concrete implementation that starts with one’s own relatives. For the Mohist canon, see Graham, Later Mohist Logic, Ethics, and Science, 249–250. For Mencius, see nivison, “two roots or one?,” 740–747; and shun, “Mencius’ Criticism of Mohism.”

are the three jian ai chapters about universal love?

59

the familial scope, are more subtle and mature than the response in chap-

ter 16, which promises certain benefit for the parents of one who benefits the parents of others. there are no indications of accommodation to fam- ily values as a response to this or other objections, although the author

is remarkably respectful of conventional values and hierarchical roles

throughout the three chapters: not only in the beginning of chapter 14,

as a possible captatio benevolentiae of the audience, but also at the end of

chapter 16, where “inclusiveness” is ascribed to the gentleman ( junzi 君 子) who works hard at being a wise ruler, a loyal minister, a loving father,

a caring son, a friendly elder brother, or a brotherly younger brother

(16: 30/7–9).59 perhaps the Mohist departure from conventional morality only began in this early triplet and is mirrored in the sequence of its three chapters. its thought has not yet reached the point of accommodation but instead follows its inherent logic of demanding more equality and social concern. the major and milder trend of all other recorded objections in chap- ters 15 and 16 is a combination of approval and doubt concerning the prac-

ticability of something as difficult as “inclusiveness.” this mixed criticism of Mohism is also echoed in other early sources.60 Again, the response

is not a bit accommodating: first, the author points out that aristocrats

are capable of much more difficult feats if they are really motivated; and second, he quotes authoritative sources concerning model rulers who were all examples of jian and inspirations for Mozi. it is in such a context that the expression jian ai occurs for the very first time—in chapter 16’s response to the third objection.

《泰誓》曰:『文王若日若月乍照光于四方于西土。』

即此言文王之兼愛天下之博大也譬之日月兼照天下之無有私也

即此文王兼也雖子墨子之所謂兼者於文王取法焉

the “Great oath” says: “King Wen was like sun and moon, spreading and shining his light over the four quarters and the Western region.” this says that King Wen’s impartial caring for the world was broad and great. he is being compared to how sun and moon are impartial in their shining over the world. this is the inclusiveness of King Wen. even what

59 Dan robins, “Mohist Care,” 63–67 shows how role-governed relationships and parti- ality for one’s own family is compatible with the Mohist idea of inclusive care. 60 see, e.g., Zhuangzi 33; and Shiji 130.

60

carine defoort

our Master Mozi called “inclusive” was taken from the example of King Wen. (16: 29/1–3)61

the model of jian ai is an exemplary ruler who cares for the whole world equally. he is compared to heavenly bodies, such as the sun and the moon. the emergence of this model seems to mark a further growth in the scope of “inclusiveness,” which does not result from a compromise with objections but, on the contrary, follows an inherent and radicalizing urge within Mohist thought. the ideal of jian ai in this passage and in some other Core Chapters is one in which reciprocity is further stretched and adapted, perhaps to the point of being abandoned: like heaven, the sage is able to care for others “impartially” without starting from a particular center of concern and without expecting any reward in return, either for himself or for his loved ones. the appearance of exemplary kings ruling over All under heaven (tianxia 天下) and their description in heavenly terms indicate this further growth of the concept of jian. With the advent of the model of heaven, implicit in chapter 15 and for the first time explic- itly related to jian ai in chapter 16, the ideal is no longer described as an ever-growing “inclusive” extent of caring beginning with one’s own per- sonal and familial center of concern; rather, the ideal becomes a matter of fundamental “impartiality.” to trace this evolution further in the Core Chapters,62 we have to turn to the somewhat later triplet “Will of heaven” 天志.63 Chapters 26, 27, and 28 share ideas about the scope of caring and benefiting with the “Jian ai” triplet, using expressions such as 兼相愛交相利 (the slogan of chap- ter 15), “inclusively care for everybody in the world” 兼天下而愛之, the independent concepts jian and bie , and even the expression jian ai,

61   the same description of King Wen occurs in chapter 15 (26/21), but without quoting any source and without any interpretation in terms of jian ai. 62 it is remarkable that the occurrences of jian (er ) ai ()in other Mozi chapters are related to heaven. see, e.g., 4: 4/20–24 and 9: 12/17–22. Chapters 4 and 35 speak of 兼相愛, 交相利 in similar terms. For more occurrences, see sato, “the idea to rule the World,” 38–39. 63 A. taeko brooks, “the Mician ethical Chapters,” 117, dates chapter 26 to ca. 352 bCe, which is slightly earlier than chapter 15 (ca. 342 bCe); chapters 27 (ca. 302 bCe) and 28 (ca. 282 bCe) are slightly later. According to Watanabe takashi, “Bokushi shohen no chosaku nendai,” part 1, 27–31, the whole triplet (26, 28, 27) is much later (end of third century). there is disagreement on the order of the two last chapters of the triplet but agreement on the priority of chapter 26. For the evolution of jian ai in the “tian zhi” triplet, see also sakai Kazutaka, “Makki boku no ken’ai shisō,” 101, 103–105.

are the three jian ai chapters about universal love?

61

which occurs no fewer than four times in chapter 28.64 the fragments in this triplet about caring and benefiting further illustrate the growing scope of jian in two ways, both related to the idea of reciprocity. First, they establish a reciprocity between heaven and all human beings, as a new way to motivate the inclusion of strangers in one’s scope of caring. sec- ond, they further radicalize their moral stance to the extent that reciproc- ity becomes a duty toward heaven rather than something to be expected from others. here the obligation of “inclusive caring” is for the first time explicitly identified as the will of heaven (tian zhi yi 天之意):

今天下之士君子之欲為義者則不可不順天之意矣

順天之意何若

兼愛天下之人

As for the officer-gentlemen of the world who want to do what is right, they must follow the will of heaven. but what is the will of heaven? it is to impartially care for everybody in the world. (28: 48/1–4)65

We know that heaven cares for everyone in the world because it has in all times and all regions accepted offerings from all peoples, including the bar- barians. the offerings made by men to heaven seem to initiate a general relation of reciprocity and cause heaven’s positive response to everybody:

苟兼而食焉必兼而愛之譬之若楚越之君今是66楚王食於楚之

四境之內故愛楚之人越王食於越之四境之內故愛越之人

天兼天下而食焉我以此知其兼愛天下之人也

if it impartially accepts food from them,67 it must impartially care for them. Compare it to the lords of Chu and Yue. Well, since the king of Chu is fed by all those in the territory of Chu, he takes care of the people of Chu; since the king of Yue is fed by all those in the territory of Yue, he takes care of the

64 For other passages in common with the “Jian ai” triplet, see also the essay by nicolas standaert in this volume. 65 see also Mozi 28: 48/23–24: “to follow the will of heaven is ‘inclusiveness’; to go against the will of heaven is ‘exclusiveness.’ if ‘inclusiveness’ determines the Way, it is gov- ernment through justice. if ‘exclusiveness’ determines the Way, it is government by force” 順天之意者,兼也; 反天之意者,別也。兼之為道也,義正。別之為道也,力正. 66 Following Wang Yinzhi in reading jin shi 今是 as jin fu 今夫. see sun Yirang, Mozi jiangu, 211. 67 here as well as in a similar fragment in chapter 4, Johnston, The “Mozi,27, 267 trans- lates shi as heaven “feeding them” or “providing food for them”. i agree with Mei, The Works of Motze, 30, 306, that it should be that heaven “accepts offering from them” or “is fed by them.”

62

carine defoort

people of Yue. Well, since heaven is fed by everybody impartially, i know that it takes care of everyone in the world impartially. (28: 48/6–8)68

Another indication of heaven’s caring for everybody is the fact that it punishes anyone who hurts others and it rewards those who are good to others. the best proof is:

故昔也三代之聖王堯武之兼愛天下也從而利

.

the impartial caring for and consequently benefiting of the world by the sage-kings of antiquity, Yao, shun, Yu, tang, Wen, and Wu. (28: 48/14–15)

since they cared for those whom heaven cared for and also benefited them, they were rewarded by heaven and became known as “sagely kings.” but punishments were inflicted upon violent kings such as Jie, Zhou, You, and li, who hated the world and harmed it against the will of heaven (28: 48/10–21). heaven is thus promoted as the foundation of jian ai: it meaningfully relates to all humans, whether by accepting offerings or by compensating human actions. the promise of a broad type of reciprocity among humans that was gradually formulated as a response in the “Jian ai” triplet is now strengthened or replaced by the promise of a reward to be expected from a communicative heaven,69 an idea that is almost totally absent from the “Jian ai” triplet.70 the idea of a responding heaven as it was inspired by

68 because jian xiang ai, jiao xiang li is now being paired with bie xiang wu, jiao xiang zei, with both jian and bie used adverbally, schumacher, “An outline of the evolu- tion of the Concept of Jian in Mohism,” 19, 21, believes that chapter 26 precedes chapter 16, where both characters stand for independent and central concepts, namely “inclusive- ness” (which is being promoted) and “exclusiveness” (which is being rejected). it is also possible that the two opposite concepts were already established when chapter 26 was written. otherwise, it is difficult to understand why the author wrote “exclusively hate each other” (bie ( xiang) wu ()) instead of “inclusively hate each other” ( jian ( xiang) wu ()), which grammatically makes more sense (as it occurs in Mozi 4: 5/2 and 9:

12/26). it seems that the concept of jian was considered invariably good in chapter 28, as it was in chapter 16 (see also Mozi 26: 43/8 and 26: 43/15), and hence was not considered usable as an adverb of the verb “hate.” since bie was rejected as the opposite of jian, it is here used in connection with hate. 69 in chapter 26, there is a unique passage in which the will of heaven is twice literally quoted (故天意曰) about those sagely kings who also cared for those people that heaven cared for, and about those violent kings who did not. see the essay by nicolas standaert in this volume. 70 the sole reference in the jian ai triplet to heaven is in chapter 15 in relation to King Wen, who is compared to the sun and moon (15: 26/22), as he is in the ode in chapter 16 quoted above.

are the three jian ai chapters about universal love?

63

ancient textual references71 and expanded to all human beings, may have seemed a more reliable and respected foundation for the novel moral principle. the same argument is developed in chapters 26 and 27,72 but in the latter with explicit reference to “repay” or “compensate” (bao ). heaven is portrayed as such an endless source of bounty and welfare that human beings should try to repay at least a minor fraction of its generos- ity: it orders sun, moon, and stars, regulates the seasons, sends frost, rain, and dew, grows grain and silk, provides us with hills and rivers, gathers metal and wood, bestows on us birds and beasts, and so on. hence, the motivating force of reciprocity lies not only in heaven’s promise of pun- ishment or reward but also in our duty to respond appropriately to such an enormous gift:

今夫天兼天下而愛之撽遂萬物以利之若豪之末末非天之所為