Introduction to Classical Chinese 0198834977, 9780198834977

This textbook provides a comprehensive scholarly introduction to Classical Chinese and its texts. Classical Chinese is t

235 85 21MB

English Pages 592 [587] Year 2021

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover
Introduction to Classical Chinese
Copyright
Preface
Contents
List of figures
List of boxes
List of abbreviations
Abbreviated book titles
Introduction
Defining Classical Chinese
Characteristics of Classical Chinese
Phonology
Writing
Characters vs. words
Syntax
Structure of this book
Part I
Lesson 1: Word classes and constituents
1.1 Introduction
1.2 Word classes
1.2.1 Nouns
1.2.2 Pronouns
1.2.3 Verbs
1.2.4 Prepositions
1.2.5 Adverbs
1.2.6 Conjunctions
1.2.7 Particles
1.2.8 Interjections
1.3 Constituents
1.3.1 Subject
1.3.2 Predicator
1.3.3 Object
1.3.4 Complement
1.3.5 Adjunct
1.4 Clause types
1.5 Exercises
Focus 1: History of the Chinese language
Lesson 2: Noun phrases I
2.1 Introduction
2.2 Asyndetic coordination
2.3 Conjunctional coordination
2.4 Asyndetic subordination
2.4.1 Nouns as modifiers
2.4.2 Pronouns as modifiers
2.4.3 Verbs as modifiers
2.5 Exercises
Focus 2: Reconstruction of Middle Chinese
Lesson 3: Nominal clauses
3.1 Introduction
3.2 Declarative clauses
3.2.1 Identification and classification
3.2.2 Material composition
3.2.3 Reason
3.3 Adjuncts
3.3.1 Negation
3.3.2 Assertion
3.4 Interrogative clauses
3.4.1 Yes-noquestions
3.4.2 Alternative questions
3.4.3 Variable questions
3.5 Exercises
Review 1: Learning vocabulary
Lesson 4: Verbal clauses
4.1 Introduction
4.2 Simple predicators
4.2.1 Single verb
4.2.2 Coordinated verbs
4.3 Simple modification
4.4 Modal auxiliaries
4.4.1 Active voice
4.4.2 Passive voice
4.5 Exercises
Focus 3: Reconstructing Old Chinese
Lesson 5: Objects and questions
5.1 Introduction
5.2 Objects
5.2.1 Single object
5.2.2 Two objects
5.2.3 Negation
5.3 Interrogative sentences
5.3.1 Yes-noquestions and alternative questions
5.3.2 Variable questions
5.4 Exercises
Focus 4: Chinese writing
Lesson 6: Noun phrases II
6.1 Introduction
6.2 Subordination with 之
6.2.1 Nouns as modifiers
6.2.2 Four-syllablephrases
6.3 Nominalization with 者
6.4 Exercises
Focus 5: Personal names
Lesson 7: Adverbial modification
7.1 Introduction
7. 2 Suppositions and questions
7.3 Distributives
7.4 Noun phrases
7.5 然 phrases
7.6 Subordination with 而
7.7 Exercises
Focus 6: Shujing
Lesson 8: Prepositional phrases I
8.1 Introduction
8.2 Phrases with 於, 于, or 乎
8.2.1 General usage
8.2.2 Comparisons
8.2.3 Passive voice
8.2.4 Object and prepositional phrase
8.2.5 Fusion words
8.3 Phrases with 自 or 至
8.4 Exercises
Focus 7: Exemplary rulers
Lesson 9: Prepositional phrases II
9.1 Introduction
9.2 Phrases with 以
9.2.1 General usage
9.2.2 Disposal constructions
9.2.3 Anteposed prepositional ‘objects’
9.2.4 The 以為 construction
9.3 Phrases with 為
9.4 Phrases with 與
9.5 Exercises
Review 2: Analysing clauses
1 Ground rules
2 Using signal words
3 Using parallelism
4 Exercises
Lesson 10: Reference to time
10.1 Introduction
10.2 Past and present
10.3 Duration
10.4 Imperfective and perfective aspect
10.5 Other aspectual distinctions
10.6 Exercises
Focus 8: Telling time
Lesson 11: Complements
11.1 Introduction
11.2 Subject complements
11.3 Object complements
11.4 Comparisons
11.5 Exercises
Focus 9: Two-syllablewords
Lesson 12: Nominalization
12.1 Introduction
12.2 Object nominalization
12.2.1 所 with verbs
12.2.2 所 with prepositions
12.3 Clause nominalization
12.4 Exercises
Review 3: Analysing complex noun phrases
1 者 phrases
2 所 phrases
3 Nominalized clauses
4 Exercises
Lesson 13: Topics
13.1 Introduction
13.2 Subject topicalization
13.3 Object exposure
13.3.1 Regular objects
13.3.2 Prepositional objects
13.4 Absolute topics
13.5 Exercises
Focus 10: Chunqiu
Lesson 14: Anteposition and inversion
14.1 Introduction
14.2 Anteposition of objects
14.3 Inversion of subject and predicate
14.4 Exercises
Review 4: Non-canonicalclauses
1 Adjacent noun phrases
2 Signal words
3 Exercises
Lesson 15: Complex sentences I
15.1 Introduction
15.2 Coordination
15.3 Subordination with 而
15.4 Conditional clauses
15.4.1 Conditional conjunctions
15.4.2 Consecutive conjunctions
15.5 Exercises
Focus 11: Shijing
Lesson 16: Complex sentences II
16.1 Introduction
16.2 Concessive clauses
16.3 Temporal clauses
16.4 Causal clauses
16.5 Final and consecutive clauses
16.6 Exercises
Focus 12: The unification of writing
Part II
Focus 13: Modern editions of classical texts
Lesson 17: Lunyu
17.1 Introduction
17.2 Learning (LY 1.1, 2.4, 16.13)
17.3 Morals (LY 4.15, 12.2, 17.5, 13.19)
17.4 Ritual (LY 12.1, 6.27, 3.15)
17.5 Junzi vs. xiaoren (LY 4.11, 13.23, 12.5)
Focus 14: Commentaries
Lesson 18: Mengzi
18.1 Introduction
18.2 Profit, benevolence, and righteousness (Meng 1A1)
18.3 Benevolent government (Meng 1A5)
18.4 Human nature (Meng 6A2)
Focus 15: The Warring States
Lesson 19: Xunzi
19.1 Introduction
19.2 Encouraging learning (Xun 1)
19.3 On human nature (Xun 23)
19.4 Social order (Xun 9)
Focus 16: Dictionaries
Lesson 20: Zuozhuan
20.1 Introduction
20.2 A family feud in Zheng (Zuo 1.1)
20.3 The story of Zhao Dun (Zuo 7.2)
Focus 17: Composite texts
Lesson 21: Guoyu
21.1 Introduction
21.2 The voice of the people (GY 1.3)
21.3 A femme fatale (GY 7.3, 7.6)
Focus 18: Transmission of classical literature
Lesson 22: Laozi
22.1 Introduction
22.2 The ineffable Dao (Lao 1)
22.3 Coincidentia oppositorum (Lao 2)
22.4 The sage’s order (Lao 3)
22.5 The ideal state (Lao 80)
Focus 19: Manuscripts
Lesson 23: Zhuangzi
23.1 Introduction
23.2 The butterfly dream (Zhuang 2)
23.3 The wheelwright’s wisdom (Zhuang 13)
23.4 The fish’s joy (Zhuang 17)
23.5 Life and death (Zhuang 18)
Focus 20: Text structures
Lesson 24: Mozi
24.1 Introduction
24.2 Against Offensive Warfare (Mo 17)
24.3 Beware of ghosts (Mo 31)
Focus 21: Parallelism
Lesson 25: Guanzi
25.1 Introduction
25.2 Shepherding the People (Guan 1)
Focus 22: Rhymes
Lesson 26: Military texts
26.1 Introduction
26.2 Waging War (Sun 2)
26.3 Planning the Attack (Sun 3)
Focus 23: Schools of thought
Lesson 27: Shangjun shu
27.1 Introduction
27.2 Changing the rules (SJS 1)
Focus 24: Multiple transmissions
Lesson 28: Han Feizi
28.1 Introduction
28.2 Past and present (HFei 49)
28.3 Public and private interest (HFei 49)
Focus 25: Synonyms
Lesson 29: Lüshi chunqiu
29.1 Introduction
29.2 Eliminate Self-Interest(Lü 1.5)
29.3 Promote Farming (Lü 26.3)
Lesson 30: Zhanguo ce
30.1 Introduction
30.2 Feng Xuan trades debt for gratitude (ZGC 11.1)
Focus 26: Dialects
Lesson 31: Shiji
31.1 Introduction
31.2 Biography of Confucius (SJ 47)
Focus 27: Textual criticism
Lesson 32: Yijing
32.1 Introduction
32.2 The Great Tradition I (Yi 7)
32.3 The Great Tradition II (Yi 8)
Lesson 33: Ritual texts
33.1 Introduction
33.2 The Cycle of Ritual (Li 9)
33.3 The Great Learning (Li 42)
Focus 28: Translation
Glossary
Bibliography
Index of linguistic terms
Recommend Papers

Introduction to Classical Chinese
 0198834977, 9780198834977

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Introduction to Classical Chinese

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Introduction to Classical Chinese KAI VO GELSANG

1

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

1 Great Clarendon Street, Oxford, ox2 6dp, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Kai Vogelsang 2021 The moral rights of the author have been asserted First Edition published in 2021 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2020950522 ISBN 978–0–19–883497–7 (hbk.) ISBN 978–0–19–883498–4 (pbk.) Printed and bound by CPI Group (UK) Ltd, Croydon, cr0 4yy Links to third-party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third-party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Preface When Jean-­ Pierre Abel-­ Rémusat, the first European professor ‘de Langue et de Littérature chinoises et tartares’ at the Collège de France published his Élémens de la grammaire chinoise in 1822, he op­ti­mis­tic­ al­ly predicted: The obstacles and prejudices that have up to now prevented the progress of Chinese literature in Europe seem to diminish day by day; and one can foresee the moment when the former will have been completely removed and the latter entirely eradicated: then, the study of Chinese will become as easy as that of all other oriental languages, maybe even as easy as that of certain European languages. (Abel-­Rémusat 1822, p. v)

Yet, despite some great scholarly endeavours, the obstacles in the study of Classical Chinese have by no means disappeared. While lexicography and historical phonology of Old Chinese have made significant progress, the grammar of Classical Chinese remains sadly under-­studied. Georg von der Gabelentz’s Chinesische Grammatik of 1881, a landmark in Sinological scholarship, remains unsurpassed until this day. Although numerous articles devoted to specific phenomena and some short outline grammars have appeared since, there still is no comprehensive state-­of-­the-­art grammar of Classical Chinese. The one work that would have lived up to this promise, Ulrich Unger’s Grammatik des Klassischen Chinesisch in ten volumes, was never completed. A similar situation obtains with regard to textbooks. To be sure, there is no dearth of primers of Classical Chinese in Western languages. As early as 1857, Wilhelm Schott wrote a Chinesische Sprachlehre zum Gebrauche bei Vorlesungen und zur Selbstunterweisung; in 1968, Harold Shadick’s highly successful First Course in Literary Chinese in three volumes as well as Raymond Dawson’s succinct Introduction to Classical Chinese appeared; in 1985, Ulrich Unger published his erudite Einführung in das klassische Chinesisch; in 2005, Robert Gassmann and Wolfgang Behr completed a three-­volume course of Antikchinesisch; and in 2011, Paul van Els published a delightful Dutch lesboek entitled Van orakelbot tot weblog, to name just a few. But there are many more: in European and American universities, many departments use internal compilations of material in class. Indeed, the very abundance of different primers indicates the problem: there is no single scholarly introduction to the

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

p r e fac e

Classical Chinese language that fulfils all needs, providing a comprehensive grammatical outline as well as detailed analyses of Classical texts, reading strategies, fundamentals of ancient Chinese cultural history, methods for the study of ancient texts, and up-­to-­date references. Significantly, Endymion Wilkinson’s authoritative research guide, in the section on studying Classical Chinese, recommends: ‘Try using one of the many textbooks and readers compiled for Chinese university students’ (Wilkinson 2018, 43). Almost 200 years after Abel-­Rémusat, there still is no standard textbook of Classical Chinese for Western students. The present volume is intended to fill this gap. Such an enterprise may seem daunting, and the present author would never have undertaken it were he not standing on the shoulders of giants. The grammatical framework developed by Ulrich Unger in his Grammatik des Klassischen Chinesisch provided a solid basis for my analyses; the verb classes I propose were inspired by the work of John Cikoski; among the inventory of clause constituents, I owe the category of ‘complement’ to Derek Herforth and William  G.  Boltz; the description of nominalization is based on unpublished material by William G. Boltz; and finally, I followed Axel Schuessler for phonological reconstructions of Old Chinese. Moreover, I am profoundly grateful to many people who personally assisted me in writing this book: to the late Hans Stumpfeldt for encouraging and inspiring me in countless ways; to Bill Boltz, ever enthusiastic to discuss grammatical problems, for giving me erudite guidance over many years; to Ruth Cremerius for debating all those thorny details with me over and over; to Ruth Cordes and Roland Kießling for providing me with linguistic guidance; to Joachim Gentz for valuable feedback to an early draft and inspiring the Focus on text structures; to Axel Schuessler and Eric Henry for generously supplying me with their unpublished material; to Wolfgang Behr for loading me with published material; to Yegor Grebnev and Christian Schwermann for extensive comments on parts of the manuscript; to Christoph Harbsmeier for much material and advice; to Derek Herforth for alerting me to many weaknesses in my draft; and to three reviewers for Oxford University Press for their highly stimulating comments. But most of all, I am grateful to my students and assistants at the Asia-­Africa Institute, who offered truly overwhelming support of this project for six years: to Charly Hirsch Morbey, the mistress of the gloss­ ary, who unflinchingly drafted and redrafted the first versions of this vi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

p r e fac e

book; Nils Wieland and Moritz Hesselmann, who tested early drafts in their tutorials; Stefan Christ, Alexandra Rohse, Marie Schierhorn, Valerie Fuhlenbrok, and Charles Schildge, who invested hours, days, and weeks helping to revise every detail of this book; and—last, but certainly not least—my second-­year students, who patiently coped with all the imperfect stages of this book.

vii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Contents List of figures xi List of boxesxii List of abbreviations  xiii Abbreviated book titles  xiv Introduction  xvii

PART I Lesson 1 Word classes and constituents

Focus 1 History of the Chinese language

3 14

Lesson 2 Noun phrases I

17



27

Focus 2 Reconstruction of Middle Chinese 

Lesson 3 Nominal clauses 31

Review 1 Learning vocabulary 

39

Lesson 4 Verbal clauses 41

Focus 3 Reconstructing Old Chinese 

51

Lesson 5 Objects and questions 54

Focus 4 Chinese writing 

63

Lesson 6 Noun phrases II 67

Focus 5 Personal names 

75

Lesson 7 Adverbial modification 78

Focus 6 Shujing 

88

Lesson 8 Prepositional phrases I 91

Focus 7 Exemplary rulers 

100

Lesson 9 Prepositional phrases II 102 Review 2 Analysing clauses 

110

Lesson 10 Reference to time 115

Focus 8 Telling time 

123

viii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

co n t e n t s

Lesson 11 Complements 126

Focus 9 Two-­syllable words 

133

Lesson 12 Nominalization 136 Review 3 Analysing complex noun phrases 

143

Lesson 13 Topics 148

Focus 10 Chunqiu 

155

Lesson 14 Anteposition and inversion 158 Review 4 Non-­canonical clauses 

164

Lesson 15 Complex sentences I 167

Focus 11 Shijing 

177

Lesson 16 Complex sentences II 180

Focus 12 The unification of writing 

188

PART II

Focus 13 Modern editions of classical texts 

193

Lesson 17 Lunyu 196

Focus 14 Commentaries 

208

Lesson 18 Mengzi 213

Focus 15 The Warring States 

221

Lesson 19 Xunzi 223

Focus 16 Dictionaries 

234

Lesson 20 Zuozhuan 238

Focus 17 Composite texts 

251

Lesson 21 Guoyu 254

Focus 18 Transmission of classical literature 

262

Lesson 22 Laozi 264

Focus 19 Manuscripts 

274

Lesson 23 Zhuangzi 278 Focus 20 Text structures 

286

Lesson 24 Mozi 289

Focus 21 Parallelism 

298

ix

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

co n t e n t s

Lesson 25 Guanzi 302

Focus 22 Rhymes 

309

Lesson 26 Military texts 312

Focus 23 Schools of thought 

321

Lesson 27 Shangjun shu 324 Focus 24 Multiple transmissions 

333

Lesson 28 Han Feizi 336

Focus 25 Synonyms 

344

Lesson 29 Lüshi chunqiu 346 Lesson 30 Zhanguo ce 356 Focus 26 Dialects 

363

Lesson 31 Shiji 366

Focus 27 Textual criticism 

375

Lesson 32 Yijing 378 Lesson 33 Ritual texts 389 Focus 28 Translation 

397

Glossary Bibliography Index of linguistic terms

401 531 555

x

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

List of figures Maps 1. The Warring States  2. Jin in the Chunqiu period 

222 261

Figures 1. A page from the Jingdian shiwen  2. Oracle bone inscription  3. Bronze inscription  4. Confucius receives a carp from the Duke  5. Regional characters  6. A page from the Xunzi jijie  7. Three dictionary entries for 莊子 and 莊周  8. A zun tripod  9. Composite texts: Guanzi  10. Bamboo manuscript  11. Shuihudi tomb  12. Mss from Mawangdui and Guodian  13. Stemma codicorum 

xi

28 63 65 76 189 194 236 247 251 274 275 277 376

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

List of boxes 1. Classical Chinese in the twenty-first century xviii 2. Faux amisxxv 3. Numerals 8 4. Euphonic rules 19 5. Personal pronouns 24 6. Literary pronunciations 37 7. Punctuation 49 8. Optional determination 62 9. Types of genitive 71 10. Word class flexibility 86 11. Pseudo-transitivity 97 12. Verb classes 108 13. Numerals and measures 118 14. The verb 為131 15. Titles 142 16. Fusion words 153 17. Mandatory anteposition 162 18. Functions of 其170 19. Combinations of particles 220 20. Taboo characters 232 21. Appositions 249 22. Deletion 271 23. Bracket constructions 285 24. Adverbs of degree 297 25. The verb 有308 26. Weights and measures 319 27. Conceptual history 331 28. Pluralizing words 342 29. Transcriptions of Chinese 354 30. Quantifying objects 362 xii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

List of abbreviations For ease of recognition, syntactic constituents are abbreviated with capital letters, whereas lowercase is used for word classes. A adjunct adv adverb aux modal auxiliary C complement comp comparative conj conjunction cop copula ini initial position interj interjection itr intransitive MC Middle Chinese mod modifier MSC Modern Standard Chinese n noun  nc common noun  nloc local noun  np proper noun  ntemp temporal noun NP noun phrase O object  Od direct object  Oi indirect object  Opr object pronoun  O= coreferential object OC Old Chinese P predicator parta active participle partp passive participle pass passive voice

pr pronoun  prdem demonstrative pronoun  prind indefinite pronoun  print interrogative pronoun  prper personal pronoun  prrefl reflexive pronoun  prrel relative pronoun prep preposition PrP prepositional phrase pt particle  ptint interrogative particle  ptemp emphatic particle S subject temp temporal tr transitive tr2 ditransitive v verb  vi intransitive verb  vn neutral verb  vst stative verb  vtr transitive verb VP verb phrase * reconstructed pronunciation; unattested construction < derived from > developed into ⟨ ⟩ deletion / alternatives → modifies + coordinated with

xiii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Abbreviated book titles Unless noted otherwise, books are quoted by juan number. Scholarly literature is listed in the bibliography. ChuC Chuci zhijie 楚辭直解, ed. by Chen Zizhan 陳子展. Shanghai 1996. CQFL Chunqiu fanlu yizheng 春秋繁露義證, ed. by Su Yu 蘇輿. Beijing 2010, quoted by pian. Erya Erya yishu 爾雅義疏, ed. by Hao Yixing 郝懿行. 2 vols. Shanghai 1983, quoted by pian. FY Fayan yishu 法言義疏, ed. by Wang Rongbao 汪榮寶. 2 vols. Beijing 1996. GLiang Chunqiu Guliang jingzhuan buzhu 春秋榖梁經傳補注, ed. by Zhong Wenzheng 鍾文烝. 2 vols. Beijing 2009, quoted by duke number and year. GLong Gongsun Longzi xuanjie 公孫龍子懸解, ed. by Wang Guan 王琯. Beijing 1996. Guan Guanzi jiaoshi 管子校釋, ed. by Yan Changyao 顏昌嶢. Changsha 1996, quoted by pian. GY Guoyu jijie 國語集解, ed. by Xu Yuangao 徐元誥. Beijing 2002, quoted by juan and paragraph. GYang Chunqiu Gongyang zhuan zhushu 春秋公羊傳注疏, ed. by Xu Yan徐彥 [ed. Shisan jing zhushu]. Shanghai 2009, quoted by duke number and year. HDNJ Huangdi neijing suwen: fu Lingshu jing 黃帝內經素問:附靈樞經. Shanghai 1955, quoted by pian. HFei Han Feizi jijie 韓非子集解, ed. by Wang Xianshen 王先慎. Beijing 1998, quoted by pian. HNan Huainan zi jishi 淮南子集釋, ed. by He Ning 何寧. 3 vols. Beijing 1998. HSWZ Hanshi waizhuan jianshu 韓詩外傳箋疏, ed. by Qu Shouyuan 屈守元. Chengdu 1996. Lao Laozi jiaoshi 老子校釋, ed. by Zhu Qianzhi 朱謙之. Beijing 1984, quoted by zhang. Li Liji zhengyi 禮記正義, ed. by Kong Yingda 孔穎達 [ed. Shisan jing zhushu]. Shanghai 2009. Lie Liezi jishi 列子集釋, ed. by Yang Bojun 楊伯浚. Beijing 1996. LNZ Xinyi Lienü zhuan 新譯列女傳, ed. by Huang Qingquan黃清泉. Taipei 1996, quoted by juan and pian.

xiv

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

a b b r e v i at e d b o o k t i t l e s

LT Liutao 六韜, ed. by Chen Xi 陳 曦, Beijing 2016. Lü Lüshi chunqiu jiaoshi 呂氏春秋校釋, ed. by Chen Qiyou 陳奇猷. 2 vols. Shanghai 1984, quoted by juan and pian. LY Lunyu zhengyi 論語正義, ed. by Liu Baonan 劉寶楠. 2 vols. Beijing 1998, quoted by pian and zhang. Meng Mengzi zhengyi 孟子正義, ed. by Jiao Xun 焦循. 2 vols. Beijing 1998, quoted by pian and zhang. Mo Mozi xiangu 墨子閒詁, ed. by Sun Yirang 孫詒讓. 2 vols. Beijing 2001, quoted by pian. Shen Shenzi jijiao jizhu 慎子集校集注, ed. by Xu Fuhong 許富宏. Beijing 2013, quoted by pian. Shi Xinyi Shizi duben 新譯尸子讀本, ed. by Shui Weisong 水渭松. Taipei 1997. SHJ Xinyi Shanhai jing 新譯山海經, ed. by Yang Xipeng 楊錫彭. Taipei 2004. SJ Shiji 史記, ed. by Sima Qian 司馬遷. Beijing 1997. SJS Shangjun shu zhuizhi 商君書錐指, ed. by Jiang Lihong 蔣禮鴻. Beijing 1996, quoted by pian. Sun Shiyi jia zhu Sunzi bingfa jiaoli 十一家注孫子校理, ed. by Yang Bing’an 楊丙安. Beijing 1999. SY Shuoyuan jiaozheng 說苑校證, ed. by Xiang Zonglu 向宗魯. Beijing 1987. Wen Wenzi shuyi 文子疏義, ed. by Wang Liqi 王利器. Beijing 2000. WLiao Wei Liaozi 尉繚子, ed. by Xu Yong 徐勇. Zhengzhou 2010. Wu Wuzi 吳子, ed. by Xu Yong 徐勇. Zhengzhou 2010. XJ Xinyi Xiaojing duben 新譯孝經讀本, ed. by Lai Yanyuan 賴炎元 and Huang Junlang 黃俊郎. Taipei 1992, quoted by zhang. Xun Xunzi jijie 荀子集解, ed. by Wang Xianqian 王先謙. 2 vols. Beijing 1996. XS Xinshu jiaozhu 新書校注, by Jia Yi 賈誼, ed. by Yan Zhenyi 閻振益 and Zhong Xia 鐘夏. Beijing 2000, quoted by juan and pian. XX Xinxu jiaoshi 新序校釋, ed. by Shi Guangying 石光瑛 and Chen Xin 陳新. 2 vols. Beijing 2001, quoted by juan and zhang. XY Xinyu jiaozhu 新語校注, ed. by Wang Liqi 王利器. Beijing 1997, quoted by pian. Yan Yanzi chunqiu jishi 晏子春秋集釋, ed. by Wu Zeyu 吳則虞. 2 vols. Beijing 1982, quoted by juan and pian. YDan Xinyi Yan Danzi 新譯燕丹子, ed. by Cao Haidong 曹海東. Taipei 1995. Yi Zhouyi zhengyi 周易正義, ed. by Kong Yingda 孔穎達 [ed. Shisan jing zhushu]. Shanghai 2009. YTL Yantie lun jiaozhu 鹽鐵論校注, ed. by Wang Liqi 王利器. 2 vols. Beijing 1992, quoted by pian. ZGC Zhanguo ce 戰國策. 3 vols. Shanghai 1985, quoted by juan and pian.

xv

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

a b b r e v i at e d b o o k t i t l e s

ZL Zhouli zhengyi 周禮正義, ed. by Sun Yirang 孫詒讓. 14 vols. Beijing 2000. Zhuang Zhuangzi jishi 莊子集釋, ed. by Guo Qingfan 郭慶藩. 4 vols. Beijing 1997, quoted by pian. Zuo Chunqiu Zuozhuan zhu 春秋左傳注, ed. by Yang Bojun 楊伯峻. 4 vols. Beijing 1981, quoted by duke number and year.

xvi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Introduction Classical Chinese is perhaps the most important language in world ­history. Its significance is comparable to that of Latin and English, and the extent of its impact is second to none. Although it was used as a living language only during a relatively short time span of 400 years, it was the foundation of a literary tradition which spanned two millennia, surpassing that of any other civilization. Classical Chinese has shaped thought and literature throughout the history of imperial China and up to the present day. This introduction will give an outline of how Classical Chinese may be defined, what its characteristic features are, and how it is treated in the present book. Defining Classical Chinese Simply put, Classical Chinese is the language of the Chinese classics: the language of the philosophical and historical literature that represented the greatest flowering of Chinese intellectual history. The Classical period—roughly the fourth to the first century bc—saw the ‘Hundred Schools’ (百家) of Chinese philosophy, the beginnings of Chinese historiography, and the earliest attempts at synthesizing the whole of received wisdom. It was the age in which the words of Confucius and his followers were committed to writing, contending with the teachings of Daoism, Mohism, Legalism, and many others that laid the foundations of Chinese thought. Their writings constituted a canon of acknowledged excellence for later Chinese literati who spared no effort to learn the Classical language, which was far removed from the Chinese they spoke.  Just as European writers turned to Homer, Herodotus, Cicero, and Sallust, the Chinese read, emulated, and quoted the Lunyu, Mengzi, Zuozhuan, and Shiji. These and other classical works provided a model that served to embellish the style and strengthen arguments in all respectable Chinese literature well into the twentieth century. Until today, no student of Chinese culture can afford to neglect the Classical Chinese language. Even scholars dealing exclusively with modern China ignore it at their peril: not only scholarly articles, but

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

also literature, political speeches, and newspaper articles employ ­elements of Classical Chinese. In the past decades, there has been a notable revival of Classical Chinese, especially among young scholars, and several internet sites are devoted to promoting the language. Box 1  Classical Chinese in the twenty-first century From an e-mail to the author, September 2014: 尊敬的 Prof. Dr. Kai Vogelsang: 安好 百忙之中,收閱拙文,感激之情,寸言難達。 余姓張名凱*,字銘勗,乃天津市直沽人也。祖居山東平陰,高祖乃民 間中醫,活人無數,至余已第五代。幼好舊學,性喜讀書,曾入梨園 行當之四載有餘,受傳統文化之耳濡目染。少年之季,興至針灸,祖 母慈愛,授予一二。後學塾針灸推拿五年以至學士,再研讀中醫文獻 三年而獲碩士,於二零一一年畢業回津就職至今。...余攻學之外,尚 醉心於三教之學,出入於百家之言。嘗品味佛教唯識禪宗之八識定 慧,玩索道家老子莊子之無為逍遙,世行儒家中庸大學之慎獨修 身。閒暇之餘,吾常修習中華傳統武術,練陳氏太極數年有餘,解棚 捋擠按之意,思采列肘靠之法。時至今日,吾已徜徉於中國古代文化 十年有餘,僅得中醫之初道,望從三教聖賢為師,再修心正義,證天 地大道,則此生之願,足矣. * Name changed. For other examples of CC elements in everyday communication, cf. the illustrations in van Els 2011.

Yet, despite its importance, for two millennia the Classical Chinese language had not been systematically described. The Chinese never developed a grammatical tradition like the ancient Romans did, explaining the structure of their language and the rules that govern it. Countless students, aspiring to official positions in the Chinese empire, learned the language in years of intensive reading, memorization, and recitation of classical texts. Anthologies and annotated editions were meant to inculcate students with the basic tenets of Chinese tradition and teach them the classical language by the intuitive method, as it were. The first systematic accounts of Classical Chinese were given by European scholars like Abel-­Rémusat (Élémens de la grammaire chinoise, 1822) and Georg von der Gabelentz (Chinesische Grammatik, 1881), who laid the groundwork for the academic study of Classical Chinese. Earlier grammars like Martino Martini’s Grammatica sinica (1653) or Francisco Varo’s Arte de la lengua mandarina (1703) dealt with the Early Modern Chinese language. The first

xviii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

grammar written in Chinese, Ma Jianzhong’s 馬建忠 Mashi wentong 馬氏文通, heavily influenced by Latin grammar, was published in 1898 (cf. Ma Jianzhong 2007).

Arguably, the concept of ‘Classical Chinese’ is a scholarly construction of the nineteenth century—and up to the present day there is no consensus as to how it is defined. The term has been used in a variety of different ways. In Western literature it is variously understood as the Chinese language of the fifth to the third century bc (Unger 1985, 1), or ‘the language of the period from the end of the Spring and Autumn period down to the end of the Han dynasty’ (Norman 1988, 83) and beyond that, even including the literary language written until the twentieth century (Ramsey 1989, 4). In Chinese scholarly literature, on the other hand, the concept does not appear at all. The term that usually appears in text book titles, 古代漢語 gudai Hanyu, is a blanket term covering all varieties of written Chinese from the earliest texts down to the thirteenth century. Cf., for example, the classic works by Zhou Fagao 1972 and Wang Li 1981, as well as more recent textbooks such as Zhang and Yan 1990 that include prose and poetry from the Shijing (first millennium bc) to Jiang Kui (c.1155–1221).

The term 文言 wenyan (or 文言文 wenyanwen) is equally broad in scope. It refers to the language employed in all genres of texts written in imitation of classical texts, covering the two millennia from Han times to the twentieth century. Finally, the term 古文 guwen is highly am­bigu­ ous: it may refer to (1) ‘old texts’, roughly equal in meaning to gudai Hanyu; (2) texts written in the ‘old script’ used before the Qin reform as opposed to such written in ‘new script’ of Han times; and (3) to a certain literary style promoted by traditional scholars beginning in the eighth century. None of these terms corresponds to ‘Classical Chinese’: paradoxically, there is no Chinese term for ‘Classical Chinese’. In the present book, ‘Classical Chinese’ will be defined as the language represented in transmitted Chinese literature of the fourth to first centuries bc, which is roughly the time of the Warring States (戰國), Qin (秦), and Western Han (西漢). This delimitation, which differs from others cited above, was chosen for two reasons. Firstly, judging from the present state of scholarship it seems untenable to date any of the classical texts earlier than the fourth century bc. Secondly, the period beginning in the first century ad not only saw a dramatic increase in literary production and a proliferation of literary genres. More importantly, the language of the texts began to differ significantly from the language of earlier texts: the beginning of the first century ad appears to mark the end of the classical period. xix

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

Classical Chinese is thus distinguished from (1) pre-­ Classical Chinese, used in texts before the fourth century bc, and (2) Literary Chinese, being the written scholarly language used from the first century onward. Furthermore, Classical Chinese should be distinguished from (3) the language of texts that were not transmitted by copying but buried and or discarded more than two millennia ago. (1)  Classical Chinese is clearly distinct from pre-­Classical Chinese, that is the language used in inscriptions and literature before the fourth century bc. In particular, pre-­Classical Chinese is the language of Chinese canonical literature (which, unfortunately, is also often called ‘classical’), namely the Five Canonical Texts, 五經, which were written in an archaic—or in many instances: archaizing—language that differed considerably from Classical Chinese. Indeed, the vocabulary and syntax of pre-­Classical Chinese was so different from the subsequent stages of Old Chinese that it apparently was no longer fully understood in the classical period. Significantly, the Five Canonical Texts ‘rarely served as a model for later writers’ (Norman 1988, 83): in this sense, too, they were not classical. (2)  On the other end of the time scale, Classical Chinese is distinguished from Literary Chinese. Whereas in the classical period the written language corresponded more or less closely to spoken language, this correlation became tenuous in the first and second centuries ad. While spoken Chinese evolved, the written language continued to be modelled on the classical language: it turned into a purely literary language, similar to Latin in mediaeval Europe. However, this literary language by no means remained true to Classical Chinese. While preserving archaisms and most syntactical structures, Literary Chinese was constantly infused with new vocabulary. Many elements of Classical Chinese grammar were no longer understood, so that they came to be used in different ways. Moreover, Literary Chinese developed an abundance of genres that each followed specific linguistic conventions. Thus Literary Chinese, while firmly based on Classical Chinese, developed in various ways that set it clearly apart from the latter. (3)  Finally, Classical Chinese should be differentiated from the language of the manuscripts on bamboo, wood, or silk that have recently been found in ancient tombs, wells, or ruins of frontier posts (cf. Focus 19). While written at the same time as the classical texts, these manuscripts are mostly legal or military documents that use administrative xx

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

language and much technical terminology that is not to be found in classical literature. The same is true for medical and mantic texts found in tombs: these are specialized texts that would require a special introduction. Moreover, these texts in tombs, having been buried for two millennia, never became models for later literature: they, too, were not classical. Characteristics of Classical Chinese Phonology Classical Chinese has not been spoken for 2,000 years. We can only guess at how it may have sounded; but it must have been vastly different from Modern Standard Chinese. Linguistic change is observable even within the life-­spans of individuals. Over generations and millennia its effects are tremendous. Consider, for example, the differences between the Beowulf, written in Old English about a thousand years ago, and its modern counterpart: 1 Hwæt! wē Gār-­Dena  in geār-­dagum 2 þēod-­cyninga  þrym gefrūnon, 3 hū ðā æþelingas  ellen fremedon. 4 Oft Scyld Scēfing  sceaþena þrēatum, 5 monegum mǣgþum  meodo-­setla oftēah. 6 egsode Eorle,   syððan ǣrest wearð 7 fēasceaft funden;   hē þæs frōfre gebād: 8 wēox under wolcnum,  weorð-­myndum þāh, 9 oðþæt him ǣghwylc 

þāra ymb-­sittendra

10 ofer hron-­rāde  hȳran scolde, 11 gomban gyldan:   þæt wæs gōd cyning!

So. The Spear-­Danes in days gone by and the kings who ruled them had   courage and greatness. We have heard of those princes’ heroic  campaigns. There was Shield Sheafson, scourge of   many tribes, a wrecker of mead-­benches, rampaging   among foes. This terror of the hall-­troops had come far. A foundling to start with, he would   flourish later on as his powers waxed and his worth was  proved. In the end each clan on the outlying  coasts beyond the whale-­road had to yield to  him and begin to pay tribute. That was one   good king.

Beowulf, tr. by Seamus Heaney. London and New York: Norton, 2000.

Classical Chinese is more than twice as old as the Beowulf. The gulf that separates us from this language is even wider than that between Modern xxi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

English and Old English. The linguist Lü Shuxiang 呂叔湘 once remarked that if the great scholar Zhu Xi 朱熹 (1130–1200), who wrote standard commentaries on the sayings of Confucius, were to travel back in time and witness a conversation between Confucius and his disciples, he would not have understood a single sentence of what they were talking about. Much less a contemporary Chinese, being another 800 years removed from Zhu Xi. We still do not know how Classical Chinese sounded. Only its phonological structure—that is the relationships among the sounds— has been reconstructed with some degree of confidence. The following is a phonological reconstruction of a text from Han Feizi, next to a pinyin transcription in Modern Standard Chinese. It is the famous anecdote about the peasant who sees a rabbit run against a tree stump and break its neck, whereupon he, the peasant, instead of plowing his field decides to wait for further rabbits to do the same. 1 sûŋh nin wəʔ krêŋ lîn taʔ 2 lîn truŋ wəʔ tro 3 lhâh tsôʔ thok tro 4 tet keŋʔ nə siʔ 5 ʔin lhak gə ruiʔ nə hjuʔ tro 6 krəih bukh tə̂k lhâh 7 lhâh pəʔ khâiʔ bukh tə̂k 8 nə lhin wai sûŋh kwə̂k sauh 9 gə̂mʔ lok ləʔ sə̂n waŋ tə teŋh 10 drə tâŋ lhats tə min 11 krî hjuʔ tro tə rus laiʔ

Sòng rén yǒu gēng tián zhě tián zhōng yǒu zhū tù zǒu chù zhū zhé jı ̌ng ér sı ̌ yīn shì qí lěi ér shǒu zhū jì fù dé tù tù bù kě fù dé ér shēn wèi sòng guó xiào jīn yù yı ̌ xiān wáng zhī zhèng chí dāng shì zhī mín jiē shǒu zhū zhī lèi yě

These reconstructions indicate staggering differences between Old and Modern Chinese. Hardly a single syllable in the Old Chinese reconstruction looks like anything we know from Modern Standard Chinese, and a number of features are decidedly alien to it. Initial consonant clusters like in *krêŋ (l. 1) or *truŋ (l. 2) are entirely absent from Modern Standard Chinese, just like the glottal stop *ʔ that occurs in almost every line and the final consonants *-k (as in *thok, l. 3) and *-t (*tet, l. 4). Also the vowels, although we do not know their exact phonetic value, appear to be quite different from those of Modern Standard Chinese. To sum up, Old Chinese and Modern Chinese are entirely different languages. Students of Classical Chinese may rest reassured that they will not be expected to attempt pronunciation of these tongue-­twisters. In fact, they are not meant to be pronounced in the first place (cf. Focus 3). The xxii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

best we can do is to read Classical Chinese words in their modern standard pronunciations. For these, pinyin transliterations are provided in the vocabulary lists in every lesson. The glossary in the appendix includes the phonological reconstructions of Old Chinese for reference purposes. In the body of the work, however, phonology is only treated where it is relevant to the understanding of an utterance.

Writing The differences between Classical and Modern Chinese are largely concealed by the characteristics of Chinese writing. Firstly, Chinese characters do not represent sounds nearly as closely as alphabetic script does, so the complex phonology of Classical Chinese is nowhere apparent in writing. We can easily recite a classical text in Modern Standard Chinese pronunciation without noticing that there is anything amiss about it, which would be well-­nigh impossible with the Beowulf. In fact, Japanese scholars become proficient in reading Classical Chinese in a mixture of Sino-Japanese and native Japanese pronunciations, a practice called Kanbun kundoku (漢文 訓讀). In this tradition, not only the words are pronounced in Japanese (adding gram­mat­ ical suffixes), but entire sentences are read in the syntactical order called for in Japanese, inserting prepositions, conjunctions, and particles as needed. Similar methods are used in Korea for reading Hanmun (漢文) texts.

Secondly, although the character forms used in classical times, specimens of which have been found on manuscripts, differed significantly from modern Chinese writing, all classical texts have been transmitted to us in the still-­current standard script (楷書 kaishu). All standard editions of Classical Chinese literature employ modern orthography, using either full-­form or simplified characters. This makes classical texts look deceptively familiar at first sight. Don’t be deceived. The anecdote from Han Feizi, reproduced in its original version on the left and in Modern Standard Chinese translation on the right, may serve to illustrate the manifold differences that lie behind the similarity of writing. 1 宋人有耕田者。 2 田中有株。 3 兔走,觸株。 4 折頸而死。 5 因釋其耒而守株。 6 冀復得兔。 7 兔不可復得。 8 而身為宋國笑。 9 今欲以先王之政,

宋國有一個種田的人。 田裏有一根樹樁子。 一隻兔子跑過碰在樹樁子上, 折斷了脖子死了。 他因此就放下他的犁耙來守着樹樁子, 希望再得到兔子。 兔子再也得不到, 他自己卻被宋國的人笑話了。 現在假使還要用先王的政治

xxiii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

10 治當世之民。 11 皆守株之類也。

來治理當代的民眾, 那就無疑屬於守株待兔之類的人了。



HFei 49, adapted from Bauer 1991, 18f.

One feature that immediately meets the eye is the difference in length: the Classical Chinese text has fifty-­eight characters, whereas the modern translation has 113, almost twice as many. This is partly due to lexical differences. Note how many words are monosyllabic in Classical but di- or trisyllabic in Modern Standard Chinese: 宋 – 宋國, 兔 – 兔子, 樹 – 樹樁子, 頸 – 脖子, 因 – 因此, 冀 – 希望, 身 – 自己, 笑 – 笑話, etc. Some Classical Chinese words—like 兔 and 笑—have simply been complemented by additional syllables, others have been completely replaced. Further points that contribute to the conciseness of Classical Chinese include the absence of measure words as in 一隻, 一個, 一根, which are mandatory in the modern translation. Although measure words do exist in Classical Chinese, they are far less frequent. Nor do verbal complements such as 跑過, 折斷了, 放下, 得到 feature in the classical text; in every case, the simple verb suffices. Moreover, it may be noted that there seems to have been no equivalent to the third person pronoun 他 in Classical Chinese (ll. 5, 8); only the pronoun 其 (l. 5) appears as a counterpart to the possessive 他的. Furthermore, the conjunction 而 (l.  4) has no equivalent in Modern Standard Chinese; the adverb 皆 (l. 11) is no longer used; and the particle 之 (ll. 9–11) has been replaced by 的. The first thing to be noted about Classical Chinese, then, is that its vocabulary differs significantly from Modern Standard Chinese. Characters vs. words With the above in mind, students of Classical Chinese should be wary of faux amis, i.e. words that look familiar but actually differ in meaning. For example, 走, which appears in line 3, means ‘to walk’ in Modern Standard Chinese; in Classical Chinese, however, it means ‘to run’. The rabbit did not walk against the tree stump and break its neck, it ran. Similarly, the particle 也 (l. 11) is obviously quite different from the adverb meaning ‘also’ in Modern Standard Chinese. The same character may write different words. Students of Classical Chinese should be keenly aware of the problematic relationship between writing and language. Since Chinese characters xxiv

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

do not indicate sounds as precisely as alphabetic script does, it is often difficult to determine which word a given character represents. Many characters in Classical Chinese texts can be read in different ways, which means they may do service for several different words. Take, for example, the clause in line 8: the first character, 而, may represent a conjunction, meaning ‘and, but’ or a personal pronoun, meaning ‘your’; and the character 為 may express, among others, the verb ‘to make’, the copula ‘to be’, or the preposition ‘for’. It is easy to imagine how vastly different the clause may be understood, depending on how one interprets these characters. The first task when reading a Chinese text, then, is to determine which words are represented by its characters. Experienced readers may recognize this fast and accurately. But beginners are well advised to regu­lar­ly look up characters in order to ascertain which words they may stand for. Besides their meaning, it is crucial to determine the lexical properties of the words in question, that is the functions they may perform in a sentence. But this is only the first step. Classical Chinese texts cannot be read merely with the help of a dictionary. In order to answer the question which word is written by a given character in a certain position within a sentence, it is necessary to have a good know­ ledge of Classical Chinese syntax. Box 2  Faux amis Classical Chinese texts are usually edited using modern characters; that is the same writing system is used for different languages. This practice conceals the fact that the words represented by characters are quite different from those associated with them in MSC, not only phonologically but often also se­man­ tic­al­ly. One should beware of such faux amis, ‘false friends’ that seem familiar from MSC but are not. Some examples include: 也

Classical Chinese

Modern Standard Chinese

emphatic particle

‘also’



‘explain’ or ‘persuade’ or ‘be pleased’

‘speak’



‘go away, leave’

‘go to’



‘another’

‘he’



‘commissioner, scribe’

‘history’

Moreover, quite a few CC noun phrases have become lexicalized in later ­periods, and some have significantly changed their meaning over time, for example:

xxv

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

Classical Chinese

Modern Standard Chinese

人物

‘man and beast’

‘person’

東西

‘east and west’

‘thing’

朋友

‘comrades and friends’

‘friend(s)’

大家

‘powerful family’

‘everyone’

妻子

‘wife and children’

‘wife’

Students of Classical Chinese may take these examples as a cautionary reminder of the fact that dictionaries of Modern Standard Chinese may be very misleading when used for reading Classical texts. For some more appropriate dictionaries, cf. Focus 16.

Syntax Syntax, that is the way words combine to form sentences, is the primary concern of this book. Together with the study of lexical properties, it constitutes the core of Classical Chinese grammar. Whereas students of the language can make do with modern pronunciations and modern characters, they cannot do without a thorough understanding of Classical Chinese syntax. By contrast, morphology—the internal structure of words—only plays a marginal role in structuring Classical Chinese utterances: there are hardly any structural features (and none which are apparent in writing) that indicate whether a Classical Chinese word is a verb, noun, or adverb, there are no affixes that show us whether it is used as a subject or object, nor are tense and number indicated by the forms of words. In the absence of such morphological distinctions, syntax acquires all the more importance for marking grammatical ­relationships. This is quite the opposite from the grammar of Hebrew, Greek, and Latin, where the greatest emphasis is placed on morphology. Indeed, the absence of inflectional morphology has led students of the language to claim that (Classical) Chinese has no grammar—which, of course, is patently absurd. If Classical Chinese had no grammar, it would cease to be a l­anguage.

The lack of inflectional morphology not only entails that words remain constant in their form but also that syntactic rules apply more strictly than in inflectional languages. In fact, Classical Chinese grammar displays some very regular features that may be explained in much the same terms used for Indo-­European languages. We speak about ‘words’, ‘phrases’, and ‘sentences’ in Classical Chinese as a matter of xxvi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

course; and we can further analyse and classify these linguistic units by the same categories that apply in Western grammar: word classes like ‘noun’, ‘verb’, ‘conjunction’, etc., and constituents like ‘subject’, ‘predicator’, and ‘object’. This is not to say that these categories are unproblematic or unanimously accepted; they are not. All grammars are artificial sets of rules imposed on a given language. Such rules can never fully describe the infinite variety of linguistic expression; much less if it has been developed for a different language than the one it is applied to. Since the grammatical categories used for describing CC are transferred from Latin, they should be taken as metaphoric. In the absence of an indigenous grammatical tradition, these metaphors are the best we have.

These categories allow us to formulate some very simple ground rules that apply to Classical Chinese syntax. Thus the order of the basic sentence parts, with few exceptions, is subject—predicator—object, just like in the English language. Furthermore, modifiers always precede that which is modified. Consider the phrases 宋人, ‘people of Song’, in line 1 of the anecdote above, where the modifier precedes the head, or 復得 in line 7, where the adverb precedes the verb. This is a principle that should inform any analysis of syntactical structures in Classical Chinese: the core of any utterance, be it a phrase, a clause, or complex sentence, is likely to be found in the final position. This kind of analysis, assuming rules that may be couched in Western terminology, goes a long way in explaining Classical Chinese utterances. Nevertheless, a word of caution is called for. Grammatical rules are not laws. They apply in most but certainly not in all cases. Whereas the first part of this book introduces basic rules, the texts in the second part will show how many of these rules were being ignored or became obsolete over time. Students of Classical Chinese should be prepared for this experience, which may be frustrating at times. It is well to remember that Classical Chinese was once a living language with as many peculiarities, redundancies, and illogicalities as other languages. It is complicated, even enigmatic at times, but certainly not incomprehensible. An understanding of Classical Chinese does not come intuitively, in a double sense: it is neither reached immediately nor without resort to conscious reasoning. But if attained, through patience and rational analysis, it is hugely rewarding. Structure of this book This Introduction to Classical Chinese is conceived as a textbook for university-­level courses in Sinology or Chinese studies as well as for xxvii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

autodidactic study. Its object is to impart the knowledge necessary to read, understand, and analyse Classical Chinese texts, namely: • a basic core vocabulary of Classical Chinese, including word classes; • the most important syntactic structures and the rules that govern them; • acquaintance with historical and cultural background as well as important texts of the classical corpus; • the ability to use dictionaries and other reference works for the study of ancient China; • philological methods for the critical analysis of Classical Chinese texts. In all this, a basic knowledge of Modern Standard Chinese is presupposed. The book does not provide detailed information on modern standard pronunciations, pinyin transliteration, the stroke order of characters, or other points treated in beginners’ courses of Chinese. The one language dealt with in this textbook is Classical Chinese. The book is divided into two parts. Part I introduces the basic grammar of Classical Chinese; its lessons are organized according to gram­ mat­ ical features, beginning with words and phrases, progressing through simple sentences (or clauses), and ending with complex sentences. Concise vocabulary lists are provided to accompany the example texts, and a detailed glossary in the appendix contains all Classical Chinese words that appear in the book. This is where students should regularly turn for all lexical information not given in the vocabulary lists or the pertinent comments. Part II of the book is devised as an introduction to Classical Chinese literature, arranged by texts, not by syntactical features. It provides information about the texts’ structure, date, contents, and historical importance, as well as selected readings. Annotations deal not only with grammatical points but also with literary or historical issues. Whereas in Part I, every lesson is essential and should be studied in proper sequence, Part II is more loosely organized: readers may digress from the suggested order or even skip single chapters, depending on their time and interests. Finally, this book is not only about grammar. In order to deal with Classical Chinese texts in a scholarly manner, knowledge of the language is not enough. Readers will encounter a host of names, dates, xxviii

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

introduc tion

­ istorical allusions, and references to other texts that they have to make h sense of; they will need to use dictionaries, concordances, commentaries, and other philological tools; and they will need to understand the genesis and transmission of texts in order to analyse them in a methodological way. These issues are dealt with in a number of digressions on the historical background, chronology, names, bibliographic matters, reference works, critical philology, etc. These ‘Foci’ are meant to provide basic knowledge of Classical Chinese studies in a broad sense. In sum, this book is meant to introduce students of Classical Chinese to all aspects relevant for working in the field.

xxix

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Part I

The lessons in the first part of this book are meant to lay the groundwork for the understanding of Classical Chinese texts. Each one of them is centred around a specific grammatical issue, proceeding from the smallest grammatical units—words and phrases—to more complex structures: clauses and complex sen­tences. Thus Part I encapsulates the very basics of Classical Chinese grammar: this is what every student of the language must know in order to deal with texts such as those given in Part II. Every lesson provides: 1. A short introduction into the larger topic as well as preliminary questions. 2. Explanations of specific grammatical points and ex­amples to illustrate these points. All these example phrases or sentences are corpus-­based, that is derived from classical texts and reproduced as they appear there: in full-­form characters and arranged in ­traditional fashion from top to bottom and from right to left. No attempt has been made to invent ‘purer’ examples in order to highlight certain phenomena. Extensive comments serve to explain the examples. The punctuation of these example sentences is simplified. It is limited to commas and circles, ex­cluding question marks and other punc­tu­ation marks; occasional omission of parts is indicated by ellipses; following Chinese editorial custom, proper names are underlined.

3. Vocabulary lists accompanying the examples. It should be noted that these lists provide only the meaning of a given character in the particular example. For fuller entries, the reader should turn to the glossary in the appen­dix. Once a character has been listed, it will not be repeated in subsequent vocabulary lists unless it represents a different word; such entries are marked with superscript numerals. 4. Exercises at the end of each lesson. No vocabulary lists are provided for these; stu­dents are expected to look up unknown words in the glossary.

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

pa r t i

5. Abundant cross-­references for review purposes as well as references to other grammars and scholarly literature. 6. Interspersed boxes providing summary information of grammatical topics that are spread over several lessons. Throughout the book, additional information and references for more inquisitive ­readers are provided in small script para­graphs like this one. Students who want to ­concentrate on the basics may simply skip these.

Occasional reviews serve to summarize important points and provide additional examples for repetition. Additionally, a series of foci introduce basic knowledge, resources, and methods for dealing with Classical Chinese texts. These are no mere diversions but important elem­ents in the study of Ancient China: they are just as fundamental as the grammar lessons and should be studied with equal attention.

2

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 1

Word classes and constituents

1.1 Introduction 1.2 Word classes 1.2.1 Nouns 1.2.2 Pronouns 1.2.3 Verbs 1.2.4 Prepositions 1.2.5 Adverbs 1.2.6 Conjunctions 1.2.7 Particles

1.2.8 Interjections 1.3 Constituents 1.3.1 Subject  1.3.2 Predicator  1.3.3 Object  1.3.4 Complement  1.3.5 Adjunct  1.4 Clause types 1.5 Exercises

3 5 5 6 6 8 8 9 9

9 9 10 11 11 11 12 12 13

1.1 Introduction This lesson introduces the two basic grammatical elements of Classical Chinese that will be dealt with in this book: words and constituents. Words may be defined for the present pur­poses as the smallest independent meaningful ele­ments of speech; accord­ing to their lexical properties, they may be assigned to certain word classes such as noun, verb, pre­position, etc. One could add that words consist of one or more morphemes. However, morphemes, being the smallest semantic units of language, will only occasionally be dealt with in this book, for two reasons. (1) Deriva­tional morphology, which describes the way morphemes contribute to word-­formation, properly belongs to the field of lexicology; and (2) in OC, inflectional morpho­logy is scarce and plays no im­port­ant role in grammatical analysis. Therefore, morphological distinctions will only be discussed where they have a bearing on grammar.

Constituents (also called sentence parts), on the other hand, are the syntactical elements that make up a sentence: subject, predicator, object, etc. The former classification refers to the form of a gram­matical unit, the latter to its function within a sentence. Within a simple hierarchy of grammatical units, words and con­stituents may be placed as follows: 3

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 1 : w o r d c l a s s e s a n d co n s t i t u e n t s

Function Form sentences, which consist of   one or more clauses, which consist of one   or more constituents, represented by phrases, which consist of one or  more words Adapted from Quirk et al. 1985, 42f.

It is important to appreciate the categorical difference between words and constit­uents. Words belong to fixed word classes, but they may assume the function of different constituents; conversely, constituents have a clearly defined slot within a sentence, but this slot may be filled by different words. Take, for example, the English sentence ‘The sheriff rode a horse’. The word ‘sheriff ’, which belongs to the class of nouns, here functions as subject; the verb ‘ride’ functions as predicator; and ‘horse’, which is also a noun, functions as object. Conversely, ‘sheriff ’ may also function as object—‘I shot the sheriff ’—but not as predicator; and the participle of ‘ride’ may function as subject or object, as in ‘I like riding’. The interplay between both categories is what the grammar of Classical Chinese is all about. On the one hand, the lexical properties of a word limit the syntactic functions it may fulfil; on the other hand, constituents can only be arranged in a limited number of ways: seven canonical clause types will be introduced in this lesson. Thus the ana­ lysis of Classical Chinese sentences—an exercise we will rehearse over and over again in the following lessons—consists in correlating the lexical properties of the words it contains with the positions they occupy in the sentence. The precondition for such analysis is acquaintance with the word classes and constituents of Classical Chinese. Preliminary questions: What word classes can you distinguish in your native language? What constituents do you know? How do the two categories relate to each other: in other words, what functions within a sentence may members of a given word class perform?

4

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

1.2 w o r d c l a s s e s

1.2  Word classes The words of Classical Chinese, just like those of Indo-­ European ­languages, may be arranged in different word classes. Whereas in inflectional languages morpho­logy is an important aspect for such classification, Classical Chinese in most cases lacks formal indicators of word class. Words must therefore be classified almost exclusively by their intrinsic meaning and their syntactic behaviour in sentences. Chinese scholarship has traditionally distinguished between 實字 and 虛字, ‘full’ and ‘empty words’. The former correspond to ‘content words’ such as nouns and verbs that carry semantic substance; the latter correspond to ‘function words,’ including preposi­tions, conjunct­ions, and particles that mainly serve to structure sentences. This twofold scheme may be much refined by applying categories from Greco-­Latin grammar. In what follows, we will distinguish between: • nominal word classes, including nouns and pronouns; • verbal word classes, including verbs and prepositions; • others, including adverbs, conjunctions, particles, and interjections. The following paragraphs will briefly introduce these word classes and their secondary word classes, deferring more detailed discussion of their syntactic behaviour to the subsequent lessons. It is worth emphasizing that there is no agreement among scholars on the classification of words in Classical Chinese. For various views on the topic, cf. Norman 1988, 87–94; Vochala and Vochalová 1990, 32–116; and Cikoski 1970.

1.2.1 Nouns These are words that denote people (e.g. 王, ‘king’), places (e.g. 城, ‘city’), or things (e.g. 木, ‘tree’, 春, ‘spring’), including names (e.g. 周, ‘Zhou’). According to their semantic value and syntactic behaviour, nouns may be further classified as: • common nouns, which designate classes of persons (e.g. 王, ‘king’), places (e.g. 國, ‘state’), things (e.g. 木, ‘tree’), or abstract concepts (e.g. 德, ‘virtue’, 禮, ‘ritual’); 5

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 1 : w o r d c l a s s e s a n d co n s t i t u e n t s

• local nouns, which express spatial or temporal posi­tions (e.g. 上, ‘top’, 前, ‘front’); • temporal nouns, which express points in time (e.g. 昔, ‘ancient times’, 夏, ‘summer’). • proper nouns, which are names of people, places, or things (e.g. 堯, ‘Yao’, or 河, ‘He’, the ‘Yellow River’). Syntactically, nouns typically function as subject, object, complement, or predicator. They may be modified by other nouns, pro­nouns, and certain verbs. OC nouns are not morphologically marked as such; there are no nominal suffixes as MSC 子 and 頭 (in 椅子, 石頭, etc.). But in many cases there is a morpho­logical differ­entiation between nouns and the verbs derived from the same root, e.g. 王 wáng < *waŋ, ‘king’, vs. wàng < *waŋh, ‘to be king’. Compare the pho­nological difference between certain English nouns and the related verbs: belief vs. believe, advice vs. advise, c̍ onvict vs. con v̍ ict, d̍ ecrease vs. dec̍ rease, etc.

1.2.2 Pronouns Pronouns are words with nominal function that are used to refer either to parti­cipants or things mentioned in the discourse. They include: • personal pronouns (e.g. 我, ‘I, we’); • demonstrative pronouns (e.g. 此, ‘this’); • interrogative pronouns (e.g. 誰, ‘who?’); • indefinite pronouns (e.g. 某, ‘some’); • reflexive pronouns (e.g. 己, ‘him-/her-/itself ’); • a relative pronoun (所, ‘that which’). As the name implies, pronouns ‘replace’ nouns. Syntactically, they function in much the same ways as nouns; but there are certain restrictions. Some pronouns, for example, may only be used as modifiers. However, none of them may be modified themselves. 1.2.3 Verbs Verbs are words that describe an action, motion, state, quality, or occurrence. They typically function as predicators, adjuncts, or as modifiers. Verbs can be further cate­gorized by the kind of complementation they

6

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

1.2 w o r d c l a s s e s

permit—e.g. one or two objects, a comple­ment, or none—and how this affects their meaning. In Classical Chinese, four subclasses of verbs may be distinguished: • transitive verbs usually govern an object and, if used without an object, may be understood in the passive voice (compare ‘I broke the window’ and ‘The window broke’), e.g. 分, ‘divide’: X分Y X分

X divides Y. X is divided.

• intransitive verbs usually do not take an object and, if followed by an object, are to be understood in a causative sense (compare ‘He walked’ and ‘He walked the dog’), e.g. 休, ‘rest’: X休 X休Y

X rests. X makes Y rest.

• stative verbs express states or qualities, corresponding se­man­ tic­al­ly to English ad­jectives (e.g. 小, ‘small’, 老, ‘old’); while syn­tac­ tic­al­ly similar to intransitive verbs, they may occur in comparative and superlative constructions; • neutral verbs may be used with or without an object without changing their rela­tionship to the subject (compare ‘He eats’ and ‘He eats spaghetti’); e.g. 從, ‘fol­low’: X從 X從Y

X follows. X follows Y.

I owe the category of neutral verbs to John Cikoski (who calls them ‘direct’ verbs). The first three of the above categories correspond to what Cikoski has called ergative verbs (Cikoski 1970, 54, 1978b, 79). Cikoski’s categories have been adopted and expounded by Ōnishi 2004 and Boltz 2011, 2015a.

Transitive, intransitive, and stative verbs may serve to modify nouns. This use is es­pecially frequent for stative verbs and for numerals, which belong to the subclass of intransitive verbs. On the other hand, neutral verbs, as a rule, may not serve as modifiers: there is no *從子, ‘obedient child’, in Classical Chinese.

7

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 1 : w o r d c l a s s e s a n d co n s t i t u e n t s

Box 3  Numerals In Classical Chinese, numerals are intransitive verbs. The numerals from one to ten, 一, 二, 三, 四, 五, 六, 七, 八, 九, 十, are the same as in MSC, and so are the multiples of ten: 百, 千, 萬. However, there is no distinction between ­ordinals and cardinals. Com­pound numer­als are formed through determination (or, mathematically speaking, multiplication) with a smaller number, e.g. 三十, 百萬. Different powers of ten are coordinated asyndetically: 十萬九千 二百一十二 (CQFL 28), ‘109,212’. (Note that a word for ‘zero’ did not exist in Classical Chinese.) Sometimes, a lower power of ten may be joined by 有, as in 二十有五 (Zuo 3.25), ‘twenty-­five’ (compare old English ‘five and twenty’), or 百有二十 (Li 5), ‘hundred and twenty’. If a numeral is preceded by a smaller or equal numeral, the two are usually to be multiplied, as in 三三而九 (CQFL 28), ‘three times three is nine’, or 二八十六 (SY 18.12), ‘two times eight is sixteen’. However, approximate numbers, too, are expressed by simple juxtaposition: 六七 (Meng 2A1), ‘six or seven’, 再三 (HFei 32), ‘two or three (times)’. Fractions may be expressed like in MSC: 三分之一 (Guan 76), or simply 十九 (‘nine tenths’, Zhuang 25). Finally, it should be pointed out that numerals in Classical Chinese often do not indicate exact quantities: 萬物 simply means ‘all (kinds of) beings’, 百官 means ‘all offices’, i.e. the ‘admini­stration’, and so on. References: Harbsmeier 2002; Martzloff 2006.

1.2.4 Prepositions Prepositions express spatial, temporal, or other relations between two entities, one being a verb (which usually precedes the preposition), the other being a noun phrase (which usually follows the preposition), as in ‘She walks in the park’. All Classical Chinese pre­positions derive from verbs (e.g. 於, ‘in, at, on’ < ‘be [in a place]’). According­ly, pre­positional phrases occur only in verbal clauses, where they function as adjuncts. There are only about a dozen prepositions in Classical Chinese; among the most common are 於, ‘in, at, on’, and 以, ‘with, by means of ’. Some scholars call prep ‘coverbs’ or ‘prepositional verbs’, others do not use the category at all but simply treat prep as verbs. However, since by the classical period many of them are used almost exclusively as prep, they are classified as such in this book.

1.2.5 Adverbs While in Classical Chinese many nouns and verbs may function adverbi­ally, there are some words that primarily serve to modify verbs: 8

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

1.3 co n s t i t u e n t s

these may be classified as ad­verbs. Syntactically, ad­verbs exclusively function as adjuncts, preceding the predicator. They modify predicators by expressing degree (e.g. 最, ‘most’), nega­tion (e.g. 不, ‘not’), time (e.g. 已, ‘already’), manner (e.g. 忽, ‘suddenly’), or scope (e.g. 或, ‘some’, or 各, ‘each’). 1.2.6 Conjunctions These are words that connect words, phrases, or clauses, establishing different kinds of logical relationships between them, e.g. 而, ‘and, but’, 則,‘then’. Conjunctions may express coordination (‘apples and bananas’), dis­junction (‘him or me’), conditions (‘If I understand correctly. . .’), as well as temporal (‘Before I saw her. . .’), adversative (‘You can run, but you can’t hide’) and other relations. Syntac­tically, they function as connect­ ive adjuncts. 1.2.7 Particles This class includes a variety of function words that add modal, as­pect­ual, accentu­ating, interrogative, or other meanings to an utterance. Frequent Classical Chinese parti­cles include emphatic 也, interrogative 乎, perfective 矣, and others. Syntacti­cally, they always function as adjuncts. Many of these particles do not correspond to English words, since in English their function is typically ful­filled simply by intonation (e.g. high pitch at the end of questions) or, in ­writing, by punctuation such as question or exclamation marks.

1.2.8 Interjections Interjections are emotive words, such as English ‘ah!’, ‘oh!’, ‘dear!’— expressing sur­prise, joy, anger, and so forth—that do not enter into syntactic relations. They are typically placed at the beginning of sentences. As the following lessons will show, CC words often transcend these neatly defined word classes: nouns may be used as verbs or verbs as nouns (cf. Box 10). This does not invali­date the basic classifica­tions given above.

1.3 Constituents We now turn to constituents, being the functional elements of ­sentences. Simple sentences, or clauses, always consist of two major 9

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 1 : w o r d c l a s s e s a n d co n s t i t u e n t s

functional categories, a subject and a predicate. The subject almost always presents the ‘topic’ of the sentence or its ‘subject matter’, expres­ sing what the sentence is about. The predicate is what is being stated about the subject, the ‘comment’ about the ‘topic’, as it were. As a rule, the predicate follows the subject. However, in this book we will only rarely refer to this level of analysis. Rather than focussing on the predicate as a sep­ar­ate structural unit, we will mainly analyse the elements that con­ stitute the predicate. Whereas in English the predicate must always contain a verb, in Classical Chinese it may contain either a verb or a noun phrase as predi­ca­tor. While subject and predicator are sufficient for a complete clause, in many cases they are joined by other constituents: a predicator may take one or two objects, and it may be supplemented by a complement; additionally, various adjuncts may add to the meaning of the utterance. The basic clause structure may be visualized in the following formula, which shows pos­sible clause constituents and their syntactic order (constituents which may be ab­sent are put in par­en­theses): Subject  Predicate Constituent Position

S initial

(A) P medial

(Oi) (Od) (C) (A) final

The focus of this book will be on the constituents highlighted in grey. In what follows, these will be briefly introduced. 1.3.1 Subject Being the ‘topic’ of the sentence, the subject (S) can be the agent who does (or the patient who suffers) the action expressed in the sentence; or it can be a person, thing, or action that is defined or characterized; or it can designate the place where something happens, in which case one can speak of a locative subject. In regular constituent order, the subject occupies the initial position in a clause. The subject is typically represented by a noun phrase; but it may also be a verb phrase (in which case we speak of an infinitive subject), or even an entire ‘demoted’ clause. When understood from the context, the subject is regularly deleted. This does not mean that the clause has no sub­ject, it does; it  is  merely not explicitly stated. When translating Classical Chinese 10

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

1.3 co n s t i t u e n t s

sentences, the subject must often be gleaned from the context—and needs to be supplied in an English translation. 1.3.2 Predicator The pre­dicator (P) is the central and indispensable part of the clause; its normal position is after the subject and before objects or complements, if there are any. Be sure to distinguish the predicator from the predicate, mentioned in 1.3: the former is only a part of the predicate. Where­as in English and MSC, the pre­dicator is always represented by a verb, Classical Chinese clauses may have verbal or nominal pre­dicators. Accord­ingly, one can distinguish verbal and nominal clauses. While the sub­ject may be deleted, the predicator must be ex­pressed. The minimal clause in Classical Chinese thus formally consists of a predicator only, as in 反, ‘⟨He⟩ returned’, 善, ‘Good’, or imperatives like 坐, ‘sit down!’. However, most clauses are more complex, containing one or several of the following constitu­ents. 1.3.3 Object The object (O) represents a person or thing to which an action is directed. Since objects are governed by the verb that expresses the action, they only occur in verbal clauses. Like the sub­ject, objects are usually represented by a noun phrase; but verb phrases (infinitive ob­jects) or ‘demoted’ clauses also occur in this function. Most Classical Chinese verbs license an object, neutral verbs may even take two. In the latter case, the first is the indirect object (Oi) and the second the direct object (Od). Usually, the object occurs in final position, directly following the predicator; but in certain cases, it may be anteposed, that is placed before the predicator, or even exposed before the sub­ject. 1.3.4 Complement Com­plements (C) are constituents which are not affected by the verbal action—hence are not objects—but which complement the sub­ject or the object by applying some attribute or definition to it. Compare English sen­tences like ‘Peter is a student’, or ‘He called me a fool’. In Classical Chinese, complements may be realized by noun phrases or verb phrases. Their normal place in the sentence is the final position. 11

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 1 : w o r d c l a s s e s a n d co n s t i t u e n t s

1.3.5 Adjunct We will call adjuncts (A) all other parts of the predicate, including adverbial expressions of place, time, and manner, particles, as well as connectives linking clauses or consti­tuents. In contrast to the more central constituents—subject, predicator, object—which are usually only realized once, it is common to find several adjuncts in a sentence. Con­ versely, they are usually optional and hence peripheral to the syntactic structure: as a rule, adjuncts can be omitted without affecting the grammaticality of the sentence. They may occur in various positions: before the subject (initial), between subject and predi­cator (medial), and after the predicator (final). In principle, members of all word classes, except pronouns and interjections, may function as adjuncts. 1.4  Clause types Having introduced the different constituents, we may now take the formula given in 1.3 one step further and fill in the distribution of constituents that usually occur within the clause. Leaving aside all optional adjuncts, seven canonical clause types of Classical Chinese emerge: ① ② ③ ④ ⑤ ⑥ ⑦

S 舜 三人 聖人 父 子 箕子 王

P Oi 人 行。 作 教 之 為 為 之 使

舜 S hùn, np, a mythical ruler 人 rén, nc, ‘person, man’ 也 yě, ptemp, marks nom P 行 xíng, vi, ‘go, travel’ 聖人 shèngrén, nc, ‘sage’ 作 zuò, vi, caus ‘make’

Od C

A 也。 Shun was a man. Three men walk. 樂。 The sages made music. 忠。 The father teaches him loyalty. 誰。 Who are you? 奴。 Jizi was a slave to him. 人 來。 The king made somebody come. 樂 yuè, nc, ‘music’ 父 fù, nc, ‘father’

誰 shéi, print, ‘who?’ 箕子 Jīzı,̌ np

教 jiào, vn, ‘teach’

奴 nú, nc, ‘slave’

之 zhī, prper, ‘him, her, it’

王 wáng, nc, ‘king’

忠 zhōng, nc, ‘loyalty’

使 shı ̌, vn, ‘make s.o. do sth.’

為 wéi, vn, as cop, ‘be’

來 lái, vi, ‘come’

子 zı ̌, nc, ‘you’

12

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

1.5 e x e r c i s e s

All other clauses may be understood as variations of these seven types. The follow­ing lessons will offer detailed explanations of these canonical clauses, of their variation by deletion of constituents or addition of adjuncts, of their transformation into non-­canonical clauses by changes in syntactic order, and finally of their combination into complex sentences. In doing so, they will constantly rehearse the analysis of these con­structions with reference to the lexical properties and syntactic structures outlined in this chapter. 1.5 Exercises Look up the characters 也, 行, 作, 樂, 教, 為, 誰, and 王 in the glossary and note which words they represent. What do you notice? References  ① Meng 4B28 ② LY 7.22 ③ Li 19 ④ Zuo 5.23 ⑤ LY 18.6 ⑥ LY 18.1  ⑦ Meng 2A2

13

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 1

History of the Chinese language With a history of over 3,000 years, Chinese is one of the oldest recorded languages in the world. In the course of this history, the Chinese language has undergone such profound phonological, lexical, and grammatical changes that calling it a single language is a bold abstraction. Rather, it seems fair to say that Chinese has become a different language several times over. Linguists roughly divide the history of the Chinese language into four distinct periods, Old, Middle, Early Modern, and Modern Chinese: (1) Old Chinese (上古漢語): thirteenth century bc to third century ad (2) Middle Chinese (中古漢語): fourth to twelfth centuries (3) Early Modern Chinese (近代漢語): twelfth to nineteenth centuries (4) Modern Chinese (現代漢語): twentieth to twenty-­first centuries For this periodization, cf. Wang Li 1980, 35. Wang assumes ‘transitory periods’ between each of these periods, namely in the third and fourth centuries, twelfth and thirteenth centuries, and from 1840 to 1919.

Note that Classical Chinese does not appear as a distinct period in this scheme. Rather, it is a stage within the long period of Old Chinese which may be further subdivided as follows: (1a) pre-­Classical Chinese: thirteenth to fifth centuries bc (1b) Classical Chinese: fourth to first centuries bc (1c) post-­Classical Chinese: first to third centuries ad (1a) Old Chinese was a largely monosyllabic language with a productive derivational morphology, diverse affixes, and much more complex syllables than Middle or Modern Chinese. The earliest evidence of Old Chinese is found in inscriptions: oracle records of the Shang dynasty from c. thirteenth to eleventh centuries bc, incised on bones, and commemorative inscriptions in sacrificial bronze vessels from the Zhou dynasty. Composed in a terse, specialized language and written in 14

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 1 : h i s to ry o f t h e c h i n e s e l a n g uag e

ancient script, both bone and bronze inscriptions are very limited in their contents. Apart from recurring standard formulae, much of their vocabulary is still unknown, and the syntax of these inscriptions differs markedly from later stages of Chinese. The same is true of a few transmitted texts that may be (partly) dated to pre-­classical times: the Canon of Odes (詩經, cf. Focus 11); the Canon of Documents (書經, cf. Focus 6); the Canon of Changes (易經, cf. Lesson 32), and the Springs and Autumns (春秋, cf. Focus 10). Together with another text, the Canon of Rites (禮經, cf. Lesson 33), this corpus was revered as the Five Canonical Texts (五經) in later times. However, already in the Classical period, their vocabulary and syntax appeared so archaic that they were no longer understood without extensive commentary. Just like the ancient inscriptions, they were remnants of a bygone age which, even to Confucius and his contemporaries, was clouded by myths. (1b) In the classical period, beginning in the fourth century bc, a momentous change took place: now, texts were increasingly written on bamboo or wooden slips, sometimes on silk. The spread of these writing materials ushered in the age of Chinese manuscript culture (cf. Focus 19). Dozens of book-­length texts were written—the longest reaching half a million characters—whose scope and diversity far surpassed all textual production of the preceding period. It was the classical age of Chinese philosophy in which ‘a hundred schools of thought’ flourished. They not only introduced a host of new terms, but also used a diction that was much closer to spoken language than earlier texts. It included many new function words—prepositions, conjunctions, particles—various syntactic options and a nuanced vocabulary (including dialectal variants) that allowed for an unprecedented richness of written expression. This is the Classical Chinese language that will be described in detail in this book. (1c) By the first century ad, Classical Chinese had undergone many subtle changes. On the one hand, some older constructions fell out of use and others were no longer clearly understood; on the other hand, the vocabulary grew as many compound words were generated, new ways of phrase creation were exploited and syntactic constructions introduced that had been unknown in the classical period. In this period, which may be called post-­Classical Chinese, written language again began to separate from the vernacular. Moreover, the former increasingly diversified into genre-­specific styles: historical, poetic, religious, literary, philosophical, and technical texts began to develop along distinctly different lines. 15

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 1 : h i s to ry o f t h e c h i n e s e l a n g uag e

This was the beginning of Literary Chinese (cf. p. xx), a purely written language based on Classical Chinese which was nevertheless continually influenced by the developments of the spoken language. Especially during periods of disunion, under the influence of migration and the influx of foreign peoples, this spoken language would change dramatically. (2) Middle Chinese, which developed after the fall of the Han dynasty, is well attested in a great corpus of literature, some of which written in the vernacular, and, notably, in dictionaries arranged according to rhymes. It was phonologically quite different from Old Chinese: now, affixes and other morphological features developed into eight tones, making Chinese a tonal language. The influence of Buddhist scriptures as well as the languages of Central and Middle Asian peoples that came into China contributed to changes in syntax and vocabulary. For example, the use of measure words greatly increased, nominal pre- and suffixes appeared, the pronoun 是 developed into a copular verb, postverbal aspect markers like 了 or 著 appeared, and measures began to be used for verb phrases. (3) With the flight of the Song court to South China in the twelfth century and the Mongol occupation began the period of Early Modern Chinese. This period saw the development of the modern Chinese dialects. Whereas the southern dialects remained phonologically closer to Middle Chinese, the northern dialects further reduced phonological distinctions, leading to the four tones of modern Mandarin. The latter also received distinct influences from the Altaic languages, including the development of ‘disposal’ constructions with 把, perhaps also the prevalence of absolute topics at the beginning of sentences. A sizable corpus of colloquial literature, including the classic Chinese novels, was written in a language fairly close to Modern Chinese. (4) Modern Chinese was characterized by further phonological changes and, most conspicuously, the influx of new vocabulary from Japanese and Western languages in the twentieth century. In fact, the entire inventory of modern concepts—社會, ‘society’, 民族, ‘nation’, 自由, ‘freedom’, and hundreds more—was imported into Chinese via Japan. Moreover, for the first time, a standardized ‘national language’ (國語) or ‘common language’ (普通話) was propagated in order to supersede the various Chinese dialects: yet another new Chinese language that was entirely different from its predecessors. For native speakers of Modern Chinese, too, Classical Chinese is a foreign language. References:  Tai and Chan 1999 give a succinct overview of periodization; Norman 1988 and Ramsey 1989 both contain historical sketches; Wang Li 1980 is a standard account in Chinese; Wilkinson 2018, 17–27, is a good summary with many references.

16

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 2

Noun phrases I

2.1 Introduction17 2.2 Asyndetic coordination n1 + n218 2.3 Conjunctional coordination n1 與 n220 2.4 Asyndetic subordination 21 2.4.1 Nouns as modifiers n1 → n222 2.4.2 Pronouns as modifiers pr → n23 2.4.3 Verbs as modifiers v → n25 2.5 Exercises26

2.1 Introduction Having discussed words and constituents, we now turn to the link between these two categories, phrases. Phrases are single words or groups of words that serve as con­stituents. They are named after the class of that word which has the primary function in it. Thus a noun phrase centres around one or several nouns, and its overall value is nom­inal. Syntactically, it typically functions as subject, object, predicator, or comple­ment. This lesson will discuss the structure of noun phrases that consist of more than one word. In such constructions independent elements are linked either by coordination or by subordination. Coordination means the juxta­position of two or more linguistic units of the same status, placing both on equal rank, as in English ‘pots and pans’: one noun does not modify the other, but they stand side by side. Subordination places two nouns on unequal rank, the second one being the head of the phrase, while the first one serves as a modifier. Preliminary questions: What is meant by ‘coordination’ and ‘sub­or­ din­ation’? Find examples of noun phrases constructed according to 17

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 2: n o u n p h r a s e s i

these principles in your mother tongue and in Modern Standard Chinese. 2.2  Asyndetic coordination In Classical Chinese, nouns may be coordinated by simply juxtaposing them with­out the aid of a linking element. The form is simply: ① 日 月

sun and moon

n + n2 1  NP

日 rì, nc, ‘sun’ 月 yuè, nc, ‘moon’

This usually means ‘n1 and n2’ or, in some cases, ‘n1 or n2’. We shall call this un­marked form of coordination, which may consist of two or more words, asyndetic coordination. Not all nouns can be coordinated in this way. The preconditions for asyndetic coordination are that, firstly, both nouns belong to the same secondary word class: they must both be common, proper, local, or temporal nouns. Secondly, there must be semantic corre­spondence between them: both nouns must belong to the same semantic category. ‘Sun and moon’ are both celestial bodies; and in the following examples, ‘spring and autumn’ are both seasons; ‘father and mother’ are both parents, etc. Apart from enumerations and poetry, asyndetic coordination is not possible in English. To translate such noun phrases, we have to insert ‘and’ in between their constituent elements.

Such noun phrases often serve as expressions for the superordinate semantic cate­gory of their constituent elements (just like in English, one may say ‘pots and pans’, meaning ‘cooking utensils’). As such, many noun phrases became lexicalized in later times. However, in Classical Chinese their individual elements mostly retained their specific mean­ ing. Students of Classical Chinese should take the individual com­ ponents of such noun phrases seriously.

天地  . . . 

春秋冬夏

父母

上下

子女

東西南北

堯 舜  . . .  湯武 父兄子弟

父子  . . .  君臣  . . .  長幼  . . .  朋友  . . . 

夫婦

⑨ ⑧ ⑦ ⑥ ⑤ ④ ③ ②

春 chūn, ntemp, ‘spring’ 秋 qiū, ntemp, ‘autumn’ 冬 dōng, ntemp, ‘winter’ 夏 xià, ntemp, ‘summer’ 母 mǔ, nc, ‘mother’ 上 shàng, nloc, ‘top’ 下 xià, nloc, ‘bottom’

18

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

2.2 a s yn d e t i c co o r d i n at i o n

Comments ② The seasons—all of them 天 tiān, nc, ‘heaven’ ntemp—are normally enumerated in this order 地 dì, nc, ‘earth’ for euphonic reasons (cf. Box 4). ③ Does 父 子2 zı̌, nc, ‘son’ 母 mean ‘parents’ or ‘father and mother’? 女 nǚ, nc, ‘daughter’ When in doubt, it is advisable to choose the 東 dōng, nloc, ‘east’ latter translation, taking each word seriously. 西 xī, nloc, ‘west’ ④ Local nouns, too, may be coordinated 南 nán, nloc, ‘south’ asyndetically. ‘Heaven and earth’ also means 北 běi, nloc, ‘north’ ‘the world’. ⑤ In CC, 子 means ‘child’, which 堯 Yáo, np, a mythical ruler could also be a daughter. In the combination 湯 Tāng, np, Shang ruler 子 女 , it obviously means ‘son’. On the other 武 Wǔ, np, Zhou ruler hand, 女 can also mean ‘girl’ or ‘woman’, as in 兄 xiōng, nc, ‘older broth­er’ 男 女 , ‘men and women’. ⑦ If two np are 弟 dì, nc, ‘younger brother’ coordi­nated asyn­deti­cally, semantic corre­s­ 君 jūn, nc, ‘ruler’ pondence is also neces­sary: Yao and Shun 臣 chén, nc, ‘servant’ were both sage ­rulers of antiquity; and Tang 夫 fū, nc, ‘husband’ and Wu were the virtuous founders of the 婦 fù, nc, ‘married woman’ Shang and Zhou dynasties, respect­ively. ⑧ A 長 zhǎng, vst, as nc, ‘senior’ co­ord­in­ation of two coordina­tions, designat­ 幼 yòu, vst, as nc, ‘younger ing elder and younger relatives. ⑨ These are one’ the well-known ‘five relations’ (五 倫 ) that 朋 péng, nc, ‘comrade’ order society according to the ‘Confucian’ 友 yǒu, nc, ‘friend’ tradition. Note that 朋 友 , too, is a noun ­ phrase, not a word, and that it is a plural: ‘comrades and friends’ (cf. Box 2). As this example shows, asyndetic co­ord­in­ations of near-synonymous words often function to express the plural which cannot be distinguished morphologically. There is never just one 朋 友 in Classical Chinese. Box 4  Euphonic rules The order of the individual nouns in asyndetic coordinations is not arbitrary. Just like in English, where we always say ‘pots and pans’ but not ‘pans and pots’, ‘pomp and circumstance’ but not ‘circumstance and pomp’, asyndetic ­coordinations in Classical Chinese seem to follow certain euphonic rules. Ulrich Unger has specified five such rules with reference to the sound system of Middle Chinese (cf. Focus 2): 1.  Even tone precedes oblique tones, as in 身體 or 天地. 2.  Rising tone precedes departing tone, as in 土地.

19

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 2: n o u n p h r a s e s i

3.  Entering tone succeeds all others, as in 酒肉. 4.  High tone precedes low tone, as in 子女. 5.  Nasal final precedes vowel final, as in 兄弟 or 朋友.

Occasionally, these euphonic rules may be overruled for semantic reasons, e.g. in 父子 and 國家, where the higher-ranking term takes pride of place. In some cases, such as 山川, both words are phonologically equal; their order may be purely conventional (*川山 is not attested)—unless the phrase is to be understood as a sub­ordination, in which case the euphonic rules do not apply. In the case of 國家 *kwə̂k-krâ, the consonance of final *-k and the following initial *k- may have been an independent euphonic reason for the order of words. To sum up, while the euphonic rules proposed by Unger do not cover all asyndetic coordinations, they are a highly useful tool for understanding the structure of these noun phrases; moreover, in cases of doubt, they may serve as an indication of whether a given noun phrase is to be construed as a coordination (in which case the euphonic rules should apply) or a sub­or­din­ ation (in which case they do not apply). References: Unger 2019, 44–82.

2.3  Conjunctional coordination As mentioned above, only nouns belonging to the same semantic cat­ egory may enter into asyndetic coordination. Words belonging to different semantic categories must be coordinated with the conjunction 與, which means ‘and’ or, in some cases, ‘or’. ⑩ 君  與  子

the ruler and you

⑪ 君   與  父

ruler or father

n conj n2 1     NP n conj n2 1     NP

與 yǔ, conj, ‘and, or’

Note that simple ‘and’ often does not convey the force of 與, which emphatically means ‘as well as’ or ‘together with’; conjunctional coordina­ tion emphasizes both elements individually. While their conjunction is stressed, the relation between the two nouns is not as close as in asyn­ detic coordinations; nor do conjunctional coordina­tions designate a super­ordinate term. On the emphatic force of 與, cf. Boltz 1983.

20

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

2.4 a s yn d e t i c s u b o r d i n at i o n

Moreover, the conjunction 與 is mandatory for coordi­nations of two-syllable nouns or noun phrases with one-syllable nouns, as in: ⑫ 天地  與  我   Heaven, earth, and I

conj  n NP   NP

父與夫

萬物與我

人與人

利與命與仁

此與彼

我與爾

吾與汝

⑲ ⑱ ⑰ ⑯ ⑮ ⑭ ⑬

物 wù, nc, ‘creature’ 我 wǒ, prper, ‘I’, ‘we’ 利 lì, nc, ‘profit’ 命 mìng, nc, ‘fate’ 仁 rén, nc, ‘benevolence’ 此 cı ,̌ prdem, ‘this’

彼 bı̌, prdem, ‘that’ 爾 ěr, prper, ‘you’ 吾 wú, prper, ‘I’, ‘we’ 汝 rǔ, prper, ‘you’

Comments ⑬ The context makes clear that this means ‘father or hus­ band’. ⑮ 與 is also used when identical words are joined. Asyndetically coordinated, 人 人 often means ‘man by man’ in the sense of ‘one by one’. ⑯ Here, 與 means ‘together with’: the phrase states what C ­ on­fucius allegedly rarely talked about. ⑰ 與 is also necessary to c­oordinate ­pronouns with one another; in this case, the two most common ­demonstrative pronouns. 此 refers to something close at hand or just mentioned, whereas 彼 refers to something more distant. ⑱–⑲ These are the most common personal pronouns of the first and second persons in Classical Chinese (cf. Box  5); they cannot be ­ ­coordinated asyndetically. 2.4  Asyndetic subordination In noun phrases constructed by subordination, the first element is sub­ordinated to the second, it serves as a modifier to the latter, which is called the head. This follows the general rule that in Classical Chinese the modifier precedes the modified. Just like coordinations,

21

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 2: n o u n p h r a s e s i

subordinations may be achieved by simple juxtaposition of two words without an inter­vening conjunction, in the form: ⑳ 東  海

eastern sea

n 1 ® n2     NP

海 hǎi, nc, ‘sea’

In this construction, n2 is the head, that is the word that governs all the other words in the phrase. The head alone can represent the whole noun phrase: the ‘eastern sea’ is first and foremost a sea. This being said, n1 serves as modifier to the head: it specifies which or what kind of sea is meant. While the head of a noun phrase must always be a noun, the modifier may be a noun, pronoun, or verb. 2.4.1  Nouns as modifiers The precondition for asyndetic subordination of two nouns is that they belong to different semantic categories (or else they would have to be construed as coordinated; cf. 2.2). Noun phrases with a noun as modifier may be understood in analogy to genitive constructions in English, more specifically, they may be construed as descriptive genitives: the modifier describes a quality of the head. Often such noun phrases would become lexicalized, as it were, turning into compound nouns; and they may often conveniently be translated into English as such. 聖人

中國

孔子

金人

山林

山下

堂上

前王

王前

㉙ ㉘ ㉗ ㉖ ㉕ ㉔ ㉓ ㉒ ㉑

中 zhōng, nloc, ‘middle’ 國 guó, nc, ‘state’ 孔 Kǒng, np, family name 金 jīn, nc, ‘bronze’ 山 shān, nc, ‘mountain’ 林 lín, nc, ‘forest’ 堂 táng, nc, ‘hall’ 前 qián, nloc, ‘ front’

Comments ㉑ ‘Wise man’, or simply ‘sage’, means a ‘teacher for a hundred genera­tions’ (Meng 7B15) like Yao, Shun, the Duke of Zhou, and Confucius, or an ideal ruler. ㉒ As seen in ⑳, local nouns, too, may enter into asyndetic subordinations, either as the head or as the modifier of a NP designating spatial or temporal relations. 中國 does not yet mean ‘China’ but rather 22

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

2.4 a s yn d e t i c s u b o r d i n at i o n

‘the middle states’ located in the North China plain: it is a plural. ㉓ 孔子, ‘Master Kong’, was the usual appellation of Confucius by his disciples. Names are a special case of asyndetic subordination: in this case the head is a honorific title, specified by a family name (cf. Focus 5). ㉔ 金人 means a statue made of bronze. Note that the logical relation between modifier and head is not indi­cated by the construction itself: for ­example, a 玉人 is not a man made of jade but a man working with jade, a ‘jade-smith’. Compare English ‘carrot cake’ vs. ‘cupcake’ vs. ‘birth­ day cake’. ㉕ Some cases may seem ambiguous: is 山林 to be understood as a coordina­tion or a subordina­tion? ㉖–㉗ The exact pos­ition expressed by local nouns must be gleaned from the context. Thus 山下 is not ‘underneath the hill’ but ‘at the bottom of the hill’; and 堂上 is not ‘on top of the hall’ but ‘up in the hall’: temple and palace halls were usually built on hills or platforms. ㉘–㉙ Note the crucial differ­ence in these two NP: the first one means a ‘king’, 王 being the head, more specifically a ‘former king’ (the local noun here expres­ses a temporal relation); the second one, by contrast, means a place ‘in front of ’, 前 being the head, more specifically ‘in front of the king’. 2.4.2  Pronouns as modifiers Pronouns in general may serve as modifiers, but never as heads of noun phrases: they cannot themselves be modified. Pronouns do not modify the head in terms of a quality but determine it with reference to its situ­ ational context: they answer to the question ‘which?’ or ‘whose?’ When personal pronouns serve as modifiers, we may compare this to an English possessive genitive (cf. Box 9); in fact, this is the main function of the pronouns 而 and 其, which cannot be used in­de­pend­ent­ly. ㉚

此人



我國



吾王



而母

其妻子



而 ér, prper, ‘your’ 其 qí, prper, ‘his’ 妻 qī, nc, ‘wife’

Comments ㉚ A prdem as modifier. ㉜ 吾, being a less emphatic form of the first person pronoun, is much more common as a modifier than the stressed form 我. ㉝–㉞ The pronouns 而 and 其 are only used as modi­ fiers. Note that 妻子 is itself a NP which is being further modified. 23

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 2: n o u n p h r a s e s i

Box 5  Personal pronouns The inventory of personal pronouns in Classical Chinese differs considerably from that of MSC. It comprises the following: 我 wǒ and 吾 wú, ‘I, we’, 予 (var. 余) yú, ‘I’, 爾 ěr and 若 ruò ‘you’, 汝 rǔ (also, deceptively, written 女),‘you’; 而 ér,‘your’, 其 qí,‘his, her, its’, and 之 zhī ,‘him, her, it’. Of these, 我 survives in Modern Chinese, and so does 爾 *neɁ, which developed into colloquial nı̌. Even the character 爾, simplified as 尔, entered into the character 你. The pronouns 乃, ‘your’, and 厥, ‘his’, were archaisms already in the Classical period.

These pronouns differ in their range of meaning and functions. Of the first person pronouns, 我 is the most important: it can be singular or plural, and it can function as S, O, and P in nominal sentences. 吾 is a less stressed form of 我; while sharing the latter’s range of meaning, it cannot occur in final ­position. Finally, 予/余 only expresses the singular. Of the second person pro­ nouns, 爾 and 若 may be singular or plural, while 汝/女 is only singular; all three can function as S or O. On the other hand, 而 and 其 can only function as modifiers. Remarkably, Clas­sical Chinese lacked a third person pronoun used as S or P: 之 is only used as O, and 其 is only used as modifier (他 tā, in Classical Chinese, is a demonstrative pronoun meaning ‘other’). It should be noted that the status of the latter two is contested among grammarians; for example, Schuessler 2007 takes 之 as a prdem, and Unger 1985 takes both as ‘anaphoric pronouns’.

1st pers.

2nd pers.

3rd pers.

sing.

plur.

mod.

S

O

P



yes

yes

yes

yes

yes

yes



yes

yes

yes

yes

yes*

no

予/余

yes

no

yes

yes

yes

yes



yes

yes

yes

yes

yes

no

汝/女

yes

no

yes

yes

yes

yes



yes

yes

yes

yes

yes

no



yes

no

yes

no

no

no



yes

yes

no

no

yes

no



yes

yes

yes

no

no

no

*As an O, 吾 may not occur in final position but only, in negated P, before the P (cf. 5.2.3).

Personal pronouns were often avoided in Classical Chinese texts and substituted by other forms of address: humble self-references like 臣, ‘your humble servant’, or 寡人, ‘my humble self’ (used by the ruler) replace first person pronouns; 夫子 or 子, ‘master’, or 君, ‘ruler’ are used instead of second person pro­nouns, when talking to superiors; the non-existing third person pronoun in the nominative case is replaced by a noun or a demonstrative pronoun, or the S is simply omitted. References: Graham 1969; Kennedy 1956. Cikoski 1970; 127–30; Pulleyblank 1996, 76–85; Vochala and Vochalová 1990, 40–3; Harbsmeier 1997; Chen Cuizhu 2013.

24

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

2.4 a s yn d e t i c s u b o r d i n at i o n

2.4.3  Verbs as modifiers Nouns may also be modified by certain verbs. Of the basic subclasses of verbs (cf. 1.2.3), transitive, intransitive, and stative verbs can serve as modifiers, but neutral verbs—with very few exceptions—cannot. Stative verbs, in particular, lend themselves to usage as modifiers, since they are semantically equi­valent to English adjectives: they describe a quality of the head. While stative verbs as modifiers are translatable as adjectives in English, the case is different for transitive and intransitive verbs. Intransitive verbs describe an action that the head is per­forming, they should be translated as active participles (in English, -ing parti­ ciples), as in ‘the crying baby’. Transitive verbs, on the other hand, describe an action that the head has undergone, they must be construed as passive participles (in English, -ed/-en participles), as in ‘the destroyed building’. To sum up: vst → n ~ adjective vi → n ~ parta vtr → n ~ partp

高 gāo, vst, ‘be high, tall’ 大 dà, vst, ‘be big’

大家

小家  . . . 

高山

弱燕

彊 齊  . . . 百 姓  . . .

四民

飛鳥

補衣

生物

愛子

㊷ ㊶ ㊵ ㊴ ㊳ ㊲ ㊱ ㉟

小 xiǎo, vst, ‘be small’ 家 jiā, nc, ‘family’ 弱 ruò, vst, ‘be weak’ 燕 Yān, np, a state 彊/強 qiáng, vst, ‘be strong’ 齊 Qí, np, a state 民 mín, nc, ‘people’ 姓 xìng, nc, ‘clan’ 飛 fēi, vi, ‘fly’ 鳥 niǎo, nc, ‘bird’ 補 bǔ, vtr, ‘to mend’ 衣 yī, nc, ‘clothes’ 生 shēng, nc as vi, ‘live’ 愛 ài, vtr, ‘love’

Comments ㊱ Of course, 大家 has nothing to do with the modern word; it means a ‘great’, that is powerful aristocratic family. ㊲ Some proper names may be modified by vst. ㊳ Two phrases with numerals as modi­fiers: 四民 means the idealized ‘four kinds of people’, i.e. the four estates in traditional 25

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 2: n o u n p h r a s e s i

Chinese society, 士農工商, ‘scholars, peasants, artisans, and merchants’ (in descending order of esteem); 百姓 first meant ‘the aristoc­racy’ and then broadened to mean ‘the people’. ㊴ This designates a class of animals, to be translated as parta. ㊵–㊷ Consider the verb classes to decide whether the modifiers are to be construed as parta or partp. 2.5 Exercises Analyse and translate the following noun phrases. Use the glossary to look up words you do not know. 陰陽

死生

富貴  . . . 

齊魯  . . . 

士農工商

吳越

百川  . . . 

禮與食

秋水

老子

予手

㊿ ㊾ ㊽ ㊼ ㊻ ㊺ ㊹ ㊸

References  ① LY 19.24  ② Zhuang 18  ③ LY 4.19 ④ Meng 7A13  ⑤ XS 8.1  ⑥ ZGC 4.4  ⑦ Meng 7A30 ⑧ Zuo 9.14 ⑨ Meng 3A4 ⑩ Meng 3A3 ⑪ SY 19.23 ⑫ Zhuang 2  ⑬ Zuo 2.2  ⑭ Zhuang 2  ⑮ Mo 15  ⑯ LY 9.1  ⑰ Mo 45  ⑱ LY 7.11  ⑲ Zhuang 6  ⑳ SY 5  ㉑ Meng 2A2  ㉒ Meng 1B11  ㉓ LY 2.19  ㉔ SY 10  ㉕ Xun 14  ㉖  Xun 21  ㉗ Meng 1A7  ㉘ CQFL 1.1  ㉙ Meng 2B2  ㉚ SJ 31  ㉛ Mo 19  ㉜ Meng 1B1  ㉝ ZGC 20.13  ㉞ Meng 7A22  ㉟ Zhuang 21 ㊱ Mo 16 ㊲ SY 1 ㊳ SY 7 ㊴ Meng 2A2 ㊵ Lü 15.5 ㊶ Li 19  ㊷ Zuo 7.2  ㊸ SY 19  ㊹ LY 12.5  ㊺ SY 7  ㊻ Meng 6B1  ㊼ Zhuang 29  ㊽ Zhuang 17  ㊾ Zhuang 23  ㊿ LY 8.3

26

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 2

Reconstruction of Middle Chinese As discussed in Focus 1, the Chinese language has changed significantly throughout its history. Chinese scholars have been aware of this fact for about half a millennium, and they have recognized the importance of historical phonology, that is the recon­struction of historical sound systems: ‘If you know the ancient rhymes, then you can grasp the ancient pronunciations; and if you know the ancient pronunciations, then you can realize the ancient meanings’ (Zhang Zhidong). However, since Chinese characters re­present entire syllables without specifying their internal phono­logical structure, the recon­struction of ‘ancient pronunciations’ is a complicated task. This and the following Focus will give a rough outline of how it is done. We have no systematic account of how Old Chinese was pronounced. The earliest available descriptions of sounds that we possess refer to Middle Chinese. Starting in the third century, Chinese philologists had begun to employ a system of describing the pro­nunciation of words through the help of two others (反切 fanqie), the first representing its initial, the second its final and the tone. Thus 寄 jì was described as 居 jū plus 義 yì; in Middle Chinese: *kje = *k( jwo) + *(ŋ)je Such descriptions of ancient pronunciations are given in many commentaries on classical texts. An especially valuable collection of such glosses is the Jingdian shiwen 經典釋文 by Lu Deming 陸德明 (c.550–630), which provides the pronunciation of a great number of characters in fourteen classical and canonical texts: it is a prime source for the reconstruction of Middle Chinese. A sample page with glosses on Zhuangzi is shown in Figure 1: note, for example, how the char­acter 好 (3rd line from the left) is glossed as 呼報 反, that is *x(uo) + *(p)âu = *xâu. Just like 報, which defines the tone, the character 好 must be read in the fourth tone in modern pronunciation. Moreover, in 601, the first dictionary, the Qie­yun 切韻 by Lu Fayan 27

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 2 : r e co n s t r u c t i o n o f m i d d l e c h i n e s e

Figure 1  A page from the Jingdian shiwen Source: Iriya Yoshitaka (ed.). Kobun Shōsho, Sōshi ongi. Tōkyō: Tenri Daigaku Shuppanbu, 1982, 252.

陸法言, was written that arranged words by rhymes, distinguishing thirty-­ six initials and 193 rhymes. This system was refined in later centuries, and a vocabulary for the description of articulation—including words for ‘labials’ (唇), ‘dentals’ (齒), and so on—was devel­oped. With these tools, it was possible to describe the phonological relations of Middle Chinese ­syllables to one another—but not yet their actual phonemic structure. It was not until Euro­pean scholars of the twentieth century combined the methods of Western linguistics with these specifically Chinese techniques that such reconstructions became possible. In a pioneering work, the Swedish Sinologist Bernhard Karlgren (1889–1978) undertook a systematic comparison of twenty-­two Chinese dialects as well as the Sino-­ Japanese, Sino-­Korean, and Sino-­Vietnamese pronunciation of Chinese characters. Assuming that the Chinese dialects had developed genetically from Middle Chinese like the Romance languages from Latin, Karlgren traced the changes in pronunciation to a hypothetical Middle Chinese which he correlated to the phonological categories of rhyme diction­aries. 28

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 2 : r e co n s t r u c t i o n o f m i d d l e c h i n e s e

Karlgren’s picture was, of course, highly simplified, since it neglected the existence of dialects (which, in fact, had existed already in Old Chinese, cf. Focus 26). His reconstruction—an intricate system including fifteen distinct vowels—represented a composite of many regional languages which was never actually spoken. For Karlgren’s life and work, cf. Malmqvist 2011.

Karlgren’s system has been improved by later generations of philologists, so that we now have a rather precise picture of Middle Chinese phonology. It is the picture of a language quite unlike the Chinese we know. For example, Middle Chinese had a vowel system very different from Modern Standard Chinese, and final consonants *-p, *-t, *-k, and *-m that no longer exist in Modern Standard Chinese. Moreover, Middle Chinese had eight tones: four contour tones (平, ‘even’, 上, ‘rising’, 去, ‘departing’, and 入, ‘entering’) in two registers, high (陰) and low (陽). The following table shows the eight categories in­clud­ing examples with Middle Chinese reconstructions for every one: tone

平聲

上聲

去聲

入聲



天 *thien

犬 *khiwen

半 *puân

足 *tsjwok



牛 *ŋjəu

馬 *ma

大 *dâi

目 *mjuk

register

As the examples indicate, the difference in register correlates with the difference be­tween voiceless initials (Yin register) and voiced initials (Yang register); moreover, the ‘entering’ tone correlates with the finals *-p, *-t, *-k. These tonal categories developed into the four tones of Modern Standard Chinese in the following way: 陰平 > first tone, 陽平 > second tone, 上聲 > mostly third tone, 去聲 > fourth tone, 入聲 > diverse tones. In this way, the separation into registers disappeared from Modern Standard Chinese, just like the final *-p, *-t, *-k, and *-m and several other features of Middle Chinese phonology. They do survive, however, in some southern Chinese dialects and in transliterations of some Chinese words that may seem puzzling to students of Modern Standard Chinese. Consider the following: 北京 青島 新加坡 揚子江 雜碎

Peking Tsingtao Singapore Yangtsekiang Chop suey

MC *pək-kjɐŋ MC *tshieŋ-­tâu MC *sjen-­ka-­phuâ MC *jiaŋ-­tsɨ-­kåŋ MC *dzəp-­swậi 29

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 2 : r e co n s t r u c t i o n o f m i d d l e c h i n e s e

These words testify to the enduring influence of Middle Chinese ­phonology; some of its features lasted throughout the period of Early Modern Chinese, only to disappear in the nineteenth century. Knowledge of Middle Chinese is useful for the understanding of southern Chinese dialects and the explanation of seemingly exotic transliterations; and, as we shall see, it is pivotal for the reconstruction of Old Chinese. References: Karlgren 1915, 1954, 1957; Pulleyblank 1984, 1991; Unger 1985, 14–22; Norman 1988, 23–42; Ramsey 1989, 116–34; Li Xinkui 1991; Pan Wuyun 2000.

30

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 3

Nominal clauses

3.1 Introduction31 3.2 Declarative clauses 32 3.2.1 Identification and classification S P 也 33 3.2.2 Material composition S P 也 33 3.2.3 Reason S P 也 34 3.3 Adjuncts 34 3.3.1 Negation S 非 P (也) 34 3.3.2 Assertion S 固 P 也 35 3.4 Interrogative clauses 36 3.4.1 Yes-no questions S P 與 36 3.4.2 Alternative questions S P 與 P 與 36 3.4.3 Variable questions S 何 也 37 3.5 Exercises 38

3.1 Introduction Having introduced noun phrases in isolation, we now turn to the ways these units are used in context. Noun phrases usually occur embedded in larger constructions which we shall call simple sentences or clauses (cf. 1.3). In Classical Chinese, two basic types of clauses may be distinguished according to their predicator: clauses with a verbal predicator shall be called verbal clauses, and such with a nominal predicator, nominal clauses. Nominal clauses, which are treated in this lesson, express declarative state­ments or questions; they do not narrate events (hence they occur more often in treatises than in narrative texts). Instead, they simply place the subject and the predicator side by side (hence they have also been called ‘appositional sentences’). The basic statement of a nominal clause is S = P, ‘the subject is the predicator’. It is important to 31

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 3 : n o m i n a l c l au s e s

note that the copula ‘is’ only appears in the English translation. In Classical Chinese nominal clauses, there is no copula: they are verbless, hence timeless. The statements they make are categoric in the double sense of being absolute and unconditional and of typically involving a category. This lesson will discuss the syntax of declarative and interrogative nominal clauses as well as some common adjuncts that modify them. The distinction between verbal and nominal clauses does not exist in English grammar because strictly nominal predicates do not occur. Even clauses like ‘Peter is a student,’ which some call ‘verbless clauses’, still require a copula, that is a verb linking S and P.

Preliminary questions: Review the canonical clause types listed in 1.4: which type will be discussed in this lesson? Do you know examples of purely nominal predicates in Modern Standard Chinese? 3.2  Declarative clauses A nominal clause has the basic form ①

宋 小國 也。 Song is a small polity. S P A

宋 Sòng, np, a polity

The subject is directly followed by the predicator, without a copula or any other word linking the two. Instead, the characteristic sign of the nominal clause is the emphatic particle 也, which marks nominal ­predicates. It bears emphasis that 也 is not a copula: it is neither a verb nor does it function to link subject and predicator. Syntac­tically, it is an adjunct to the predicator, and its function is to empha­size the latter. It is hardly ever omitted in positive predicates. In pre-­Classical Chinese, nominal clauses had the basic structure S 唯 P, the word 唯 wéi being an emphatic particle. When 唯 ceased to be used, 也 appeared as a functionally equivalent element in nominal clauses.

Recall that the subject may be deleted if known from the context (cf.  1.3.1). Hence the minimal nominal clause formally consists of the predicate only: ②

⟨S⟩

善人 P

也。 ⟨He⟩ is a good man. A

善 shàn, vst, ‘be good’

Declarative nominal clauses express either an identification or classification of the subject, its material composition, or the reason for the subject’s occurrence. The follow­ing sections will discuss these cases separately. 32

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

3.2 d e c l a r at i v e c l au s e s

3.2.1  Identification and classification Nominal clauses may serve to identify the subject, as in ‘This is Peter’, or, more frequently, they may classify the subject, as in ‘Peter is a graduate student.’ ③

舜人也 。



余而祖也 。



此酒也 。

是良史也 。



余 yú, prper, ‘I’ 祖 zǔ, nc, ‘ancestor’ 酒 jiǔ, nc, ‘ale’ 是 shì, prdem, ‘this’ 良 liáng, vst, ‘be excellent’ 史 shı̌, nc, ‘scribe’

Comments ③ For an English translation, we need to choose a tense; however, the original statement is atemporal. ‘Shun is a man’: he always has been and he always will be. ④ In the absence of a copula between S and P, the line between the two is not made explicit: why is it nonetheless clear that it must be between 余 and 而? ⑤ Here, the prdem is S; but what if it was a modifier? ⑥ The prdem 是 refers to something that has just been mentioned: ‘this (thing in question)’, ‘the aforementioned’. 3.2.2  Material composition Apart from identifying or classifying, nominal clauses have yet another function. Just like in English statements, such as ‘This dress is cotton’ or ‘The building is brick’, they can also express the material composition of the subject in the sense of ‘the subject is made of or consists of the predicator’. ⑦

車木也

船木也 。

 . . .

子肉也 。

我肉也 。



車 chē, nc, ‘cart’ 木 mù, nc, ‘wood’ 船 chuán, nc, ‘boat’ 肉 ròu, nc, ‘flesh’

Comments ⑧ 子 functions as a substitute for the second person pronoun (cf. Box 5).

33

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 3 : n o m i n a l c l au s e s

3.2.3 Reason Finally, nominal clauses can express reasons in the sense that ‘the subject is due to the predicator’. Compare English sentences like ‘That was the wind’, or ‘That was my little brother’, meaning that the wind or the brother did it. ⑨

此天也 。

治亂非天也 。



治 zhì, nc, ‘order’ 亂 luàn, nc, ‘disorder’ 非 fēi, pt, ‘is not’

Comments ⑨ The text continues: ‘. . . what can there be done about it?’ ⑩ The S is a NP. For a discussion of 非, cf. the next section. 3.3 Adjuncts The statement of a nominal clause may be modified by adjuncts that express certainty, doubt and other nuances of meaning, or negation. Such adjuncts always appear in medial position, between subject and predicator. This section will discuss the function of the particle 非 as a negatory adjunct, and it will give some examples for words that express assertions. 3.3.1 Negation Nominal clauses are negated by the particle 非, which is placed between subject and predicator. In negated clauses, the particle 也 is usually omitted. ⑪

子 非 魚。 You are not a fish. S A P

魚 yú, nc, ‘fish’

非 fēi is a contraction of the pre-­ classical ptemp 唯 wéi and a negation: 不唯 > 非. Hence, the ptemp 也, which replaces the obsolete 唯 (cf. note at 3.2), is not needed in clauses with 非 (cf. Pulleyblank 1959). Grammarians do not agree as to what word class 非 belongs to (it certainly should not be called ‘copula’, since it is not a verb). In view of the etymology described above and its syntactic use, which is similar to other particles in nominal clauses, we will treat 非 as a particle.

34

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

3.3 a d j u n c t s



爾非吾君也 。



吾非瞽史 。



我非其主也 。

非酒也 。



瞽 gǔ, nc, ‘musician’ 主 zhǔ, nc, ‘lord’

Comments ⑬ Note the asyndetic coordination meaning ‘n1 or n2’ (cf. 2.2). ⑮ What do you notice about this clause (compare ⑤)? 3.3.2 Assertion Apart from negation, most adjuncts that appear in nominal clauses serve to add confidence and force to the utterance: they express assertions. This corresponds with the general function of nominal clauses to encode categoric statements. The most important of these ‘assertive’ words are 固, 必, and 誠. ⑯

子固非魚也 。



固周制也 。



士誠小人也 。

必非誠士 。



固 gù, vst, as adv, ‘certainly’ 周 Zhōu, np, a dynasty 制 zhì, vtr, as nc, ‘institu­tion’ 士 shì, nc, as np 誠 chéng, vi, as adv, ‘truly’ 小人 xiǎorén, nc, ‘petty man’ 必 bì, vi, as adv, ‘definitely’

Comments ⑯ This reinforces the statement cited in ⑪. 固 expresses certainty, implying that the truth of the statement is evident or commonly known. Note how the statement is further underlined by the use of 也, which is not obligatory. ⑰ The ‘institutions of Zhou’ were the (imaginary) model that political advisers in the classical period often invoked to justify their policies. ⑱ 士, here, is the name of the speaker; it is used instead of a prper. 誠, ‘truly, really’, is probably related to the vtr 成, ‘finish, com­plete’, so we may translate: ‘I am truly / a complete bumpkin’. A 小 人, more formally rendered as ‘petty man’, is an unsophisticated person with a parochial mentality. ⑲ 必, ‘definite­ly, certainly’, expresses that something is always and inevitably the case. Since there is an assertive A before 非 (as in ⑯), 誠 cannot be an assertive A in this case but must be 35

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 3 : n o m i n a l c l au s e s

construed as a mod. Compare the difference between English ‘He is really not a sportsman’ and ‘He is not a real sportsman’. 3.4  Interrogative clauses Apart from making declarative statements, nominal clauses can also express questions. Three basic types of interrogative clauses may be distinguished by the set of logically possible answers they define: yes­no questions that call for affirmation or negation, alternative ­questions that expect one of two proposed answers, and variable ques­ tions, which are open questions expressed by interrogative pronouns. In contrast to the English language, interrogative clauses in Classical Chinese do not require an inversion of constituents: the word order remains the same as in declarative clauses. 3.4.1  Yes-­no questions Yes-­no questions are marked by the interrogative particles 與/歟 yú, or 耶/邪 yé, which substitute for the final 也. These ‘vocalized ­question marks’ are the only characteristic that differentiates yes-­no ques­ tions from declarative clauses. 與/歟 yú is a fusion of 也 *laɁ and the ptint 乎 *ɦâ: *laɁ + *ɦâ > *la. It is not the same word as the conjunction 與 yǔ: the morphological difference still exists in the modern pronunciation. 邪/耶 only occurs in certain texts, especially in Zhuangzi, whereas it is entirely absent in Lunyu, Mengzi and almost entirely absent in Zuozhuan and Guoyu.



治亂天邪 。



是魯 孔丘與 。

子真是邪 。



邪/耶 yé, ptint, 魯 Lǔ, np, a polity 丘 Qiū, nc as np 與2 /歟 yú, ptint 真 zhēn, vi, as adv, ‘truly’

Comments ⑳ For the answer, cf. ⑩. ㉒ Here, 真 functions as assertive A. 3.4.2  Alternative questions Alternative questions call for answers from a set of alternatives given in the question itself. They are posed by simply repeating the question with a different nominal predicator; the interrogative particle must be repeated. 36

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

3.4 i n t e r r o g at i v e c l au s e s



天與…

人與 。

松耶

柏耶 。



松 sōng, nc, ‘pine tree’ 柏 bǎi, nc, ‘cypress’

Comments ㉓ The answer to this question is 天也,非人也。What do you think the question asks about: an identification, a material com­ position, or a reason? 3.4.3  Variable questions

㉕ 是誰也 。

㉖ 何也 。

㉗ 周公何人也 。

 . . .

㉘ 是何鳥也 。

誰非君臣 。

㉙ 何非君土

Variable questions are open propositions that contain a variable, realized by an interrogative pronoun. When asking for people, the pronoun 誰 is used, when asking for things (including animals), 何 is used. The interrogative pronoun may function as predicator by itself or as a modifier to a noun. By contrast, questions in which the interrogative pronoun is used not as predicator but as subject are to be understood as rhetorical questions. 何 hé, print, ‘what?’ 公 gōng, nc, ‘duke’ 土 tǔ, nc, ‘land’

Comments ㉖ An elliptical clause, which may simply be translated as ‘Why?’ ㉗ The print 何, used as modifier, means ‘what kind of?’ ㉙ These are rhetorical questions. Box 6  Literary pronunciations Conservative scholars would raise their eyebrows at some of the colloquial pronunciations given in this book. For 誰, they would not read shéi but shuí; instead of 柏 bǎi, they would insist on bó; and they would prefer to pronounce 車 as jū rather than chē. Such literary pronunciations (讀音) go back to the information given in Middle Chinese rhyme dictionaries (cf. Focus 2), where, for example, the MC pronunciation of 車 is described as 九魚切, *k(wəʔ) + *(ŋ)a = *ka, which should have developed to MSC jū (in analogy to 居 *kaʔ >

37

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 3 : n o m i n a l c l au s e s

jū, 據 *kah > jù, and others). So despite the fact that the character 車 is col­ loquially pronounced chē, it ‘should be’ read jū. By the same rationale, purists read 給 jı ̌, not gěi, 六 lù rather than liù, 他 tuō instead of tā, etc. And of course, the name of the great poet 李白 (701–62) should be pronounced Lı ̌ Bó, not Lı ̌ Bái. These literary readings are still common in Taiwanese editions of Classical texts, and they are given in some dictionaries. For the purposes of this book, we will confine ourselves to the colloquial pronunciations, indicating literary readings only in the glossary. Reference: Demiéville 1950.

3.5 Exercises Analyse the following nominal sentences, identifying their subject and predicator, and translate them with the help of the glossary. ㉚

白馬非馬 。



子誠齊人也 。



夫子聖人也 。



非禮也 。



此何故也 。



何言也 。



子非我 。

非我也 ,兵也 。



References  ① Meng 3B5  ② Meng 7B25  ③ Meng 4B28  ④ Zuo 7.3  ⑤ HFei 19  ⑥ Zuo 10.12  ⑦ Mo 45 ⑧ Lü 11.4 ⑨ XS 3.9 ⑩ Xun 17 ⑪ Zhuang 17 ⑫ SY 9 ⑬ GY 3.1  ⑭ Lü 10.5  ⑮ HFei 10  ⑯ Zhuang 17  ⑰ Zuo 10.23  ⑱ Meng 2B12  ⑲ Xun 5  ⑳ Xun 17  ㉑ LY 18.6 ㉒ SJ 32 ㉓ Zhuang 3 ㉔ SJ 46 ㉕ Yan 1.18  ㉖ Meng 1A3 ㉗ Meng 2B9  ㉘ Lü 18.2  ㉙ Zuo 10.7  ㉚ GLong 2  ㉛ Meng 2A1  ㉜ Zhuang 5  ㉝ Zuo 7.2  ㉞ Mo 17  ㉟ Zhuang 13  ㊱ Zhuang 17  ㊲ Meng 1A3

38

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Review 1

Learning vocabulary So far, we have not encountered a great amount of vocabulary, and many of the words that occurred in the first three lessons are familiar from Modern Standard Chinese. This will change. The amount of genu­ inely new words will grow substantially within the next lessons, and keeping abreast of vocabulary will become an issue. Although this book places greater emphasis on syntax than on lexicon, this does not mean that learning vocabulary is not important; it is. This will become very apparent in examinations, when working on term papers or doing any kind of research based on Classical Chinese sources: there is never enough time to look up all or even most of the words.

Ideally, students should learn all words that occur in this Introduction in order to build up a basic vocabulary of Classical Chinese. However, if time constraints make it impossible to do so, choices must be made. Therefore, it is useful to know which words should be given priority, and how to learn them efficiently. The following are some rules of thumb. Function words—prepositions, conjunctions, particles—are more important than content words. Since they serve to structure sentences, they are much more frequent than content words. Moreover, they can show us at a glance what kind of utterance we are dealing with: a nom­ in­al or a verbal clause, a question, an exclamation, a simple or a complex sentence, and so on. Without knowing them, it would be impossible to make any sense of most sentences. Thus function words should be given highest priority: they should all be memorized, with no exception. It is no coincidence that quite a few lexicographic works are devoted to function words (虛詞): notably the Qing dynasty model by Wang Yinzhi 王引之, Jingzhuan shici 經傳釋詞 (Wang Yinzhi 1985b), and in the same tradition Yang Shuda 1986, Yang Bojun 1981, He Leshi et al. 1985, as well as many others. Western examples of this genre are Brandt 1929 and Jonker 2011.

The importance of content words, of course, depends on their fre­ quency. Here, the vocabulary columns given in the lessons may serve as a useful indicator, since they only list words on their first occurrence. So if a word from an example sentence is not in the vocabulary column, this is at least the second time it appears, and it should be memorized 39

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 1 : l e a r n i n g v o c a b u l a ry

then, at the latest. Moreover, do not neglect names or key concepts, even though they may not appear very frequently. The names of states like 齊 or 魯, or people like 周公, 堯, or 舜 are so much part of Chinese lore that no student of Chinese culture can afford to ignore them; and concepts like 仁, 義, 禮, and others have shaped Chinese tradition over millennia: they belong to China’s cultural core vocabulary. When learning vocabulary, do not just learn bare words. Always learn the word class as well: this is the key to understanding a word’s usage. Do not just rely on the vocabulary lists in the lessons, but make use of the glossary and read the example sentences given there. Even if you do not understand much of these in the beginning, you will profit from this reading experience very soon. It is advisable to memorize at least one paradigmatic sentence for every word to illustrate the way it is employed in context. In addition, it is highly recommended to learn the semantic classifier of the characters you learn: this will make it much easier to look them up again in the glossary or in another dictionary. Finally, repetition is key. Flash cards—physical or as computer applications—are very useful tools for learning vocabulary since they encourage regular repetition. Make sure to include the word class and example sentences on the cards. Always bear in mind Confucius’ adage: 學而時習之, 不亦說乎!

40

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 4

Verbal clauses

4.1 Introduction41 4.2 Simple predicators S  P42 4.2.1 Single verb42 4.2.2 Coordinated verbs v1 + v243 4.3 Simple modification 45 4.4 Modal auxiliaries S A P46 4.4.1 Active voice46 4.4.2 Passive voice47 4.5 Exercises49

4.1 Introduction Having discussed nominal clauses, we now turn to the second Classical Chinese clause type, verbal clauses. These offer far more sophisticated and wide-­ranging possibil­ities of expression than nominal clauses. Verbal clauses are narrative. In contrast to nominal clauses, they encode temp­oral statements such as actions, events, processes, or (temporary) conditions. They may serve as conditional clauses, whereas nominal clauses are in every sense unconditional. By their very nature, verbal clauses occur much more frequently—perhaps accounting for 90% of all clauses in the classical corpus—and they may be much more elab­or­ate than nominal clauses. Whereas the latter rarely go beyond the two major constituents— subject and predicator—and a simple adjunct, verbal clauses often contain several other constituents that expand the statement. Despite these differences, verbal clauses may be analysed in close analogy to nominal clauses. They centre around a verb phrase, which consists of the predicator and elements that add to its meaning: adjuncts and any direct or indirect object, preposi­tional phrase, or complement. 41

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 4 : v e r b a l c l au s e s

This lesson will discuss the most basic verbal clauses, i.e. type 2 as introduced in 1.4. They consist of either a predicator only, a predicator with an adverbial modifier, or a predicator with a modal auxiliary. Building on this, the subsequent lessons will intro­duce more complex verbal predicates. Preliminary questions: What are adverbial modifiers and modal aux­il­ iar­ies? Can you come up with examples from your mother tongue and Modern Standard Chinese? 4.2  Simple predicators The most elementary verbal predicate consists of a predicator only. This predicator, in turn, may consist of a single verb or two coordinated verbs. 4.2.1  Single verb The basic form of verbal clause is: ① 道

S

遠。 P

The way is long.

道 dào, nc, ‘way’ 遠 yuǎn, vst, ‘be long’

In order to analyse even such simple clauses correctly, it is essential to consider the verb class of the verb that functions as predicator (cf. 1.2.3). In ①, 遠 is a stative verb, used intransi­tively, i.e. without an object. The clause is in the active voice. The same analysis holds for intransitive verbs (recall 三人行 in 1.4) and neutral verbs. But contrast this with the following example: ② 簡公 弒。

S

P

Duke Jian was assassinated.

簡 Jiǎn, np, posth. name 弒 shì, vtr, ‘assassinate’

Here, the predicator is a transitive verb, which normally expresses an action directed towards an object. When such verbs are used without an object, as in ②, they express an action that the subject undergoes. The subject of such clauses is not the agent, but the patient: it is affected by the action expressed. Hence, the clause is to be construed in the passive voice. By their very nature, only transitive verbs may be passivized in this way. 42

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

4.2 s i m p l e p r e d i c ato r s



道行  . . . 

教 尊  . . .

人死。



名定

德成

國治。 . . . 



死 sı ̌, nc as vn, ‘die’ 名 míng, nc, ‘name’ 定 dìng, vtr, ‘establish’ 行2 xíng, vtr, ‘carry out’

志通。  . . .

官正  . . . 

志 zhì, nc, ‘will’ 通 tōng, vi, ‘be accessible’ 德 dé, nc, ‘virtue’ 成 chéng, vtr, ‘achieve’ 尊 zūn, vst, ‘be respected’ 官 guān, nc, ‘officer’ 正 zhèng, vst, ‘be straight’

Comments ③ A neutral verb, used intransitively: this is clearly active voice. ④ Pay attention to 道行: which word is represented by the character 行 in this case? ⑤ An example for transitive and stative verbs, passive and active intransitive usage in parallel constructions. Quite a few CC verbs show morphological distinctions between transitive and (passive >) intransitive forms, e.g. 治 chí < *drə, ‘order’, vs. 治 zhì < *drəh, ‘be ordered’, 敗 bài < *prâts, ‘defeat’, vs. 敗 bài < *brâts, ‘be defeated’, 盡 jı̌n < *tsinʔ, ‘exhaust’ vs. 盡 jìn < *dzinʔ, ‘be exhausted’. Linguists agree that these are not forms of inflectional but of derivational morph­ology, i.e. the variant forms represent differ­ent words. References: Downer 1959; Unger 1983, 1993; Sun Yuwen 2015.

4.2.2  Coordinated verbs Just like the elements of noun phrases, verbs may be coordinated to form a verb phrase by simply juxtaposing them without the aid of a conjunction. Such asyndetic coordinations mean ‘do v1 and v2’. ⑥ 聞 見

v1 + v 2  VP

hear and see

聞 wén, vn, ‘hear’ 見 jiàn, vn, ‘see’

Verbs joined by asyndetic coordination usually belong to the same subclass, and they must share a semantic category; often, they are either antonyms or synonyms. Just as asyndetic coordinations of nouns may express a superior category, so asyndetic coordi­nations of verbs often stand for superordinate expressions, e.g. ‘hear and see’ > ‘perceive’. This close

43

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 4 : v e r b a l c l au s e s

a­ ssociation between coordinated verbs, which often became lex­ical­ized in later times, justifies treating them as one predicator. Such two-­syllable predicators often correspond to two-­syllable noun phrases as subject, creating a rhythmically balanced four-­syllable clause: ⑦ 三軍 行止。

S

P

Three armies move around.

軍 jūn, nc, ‘army’ 止 zhı̌, vi, ‘stop’

Asyndetically coordinated stative verbs, too, often express abstract concepts for which there is no other word: 長短, ‘long and short’ > ‘length’, 大小, ‘big and small’ > ‘size’, 多少, ‘many and few’ > ‘amount’. As such, these expressions may be used as NP. The order of verbs in asyndetic coordination is governed by the same euphonic rules that apply to NP (cf. Box 4).



往來



學習



陰陽長短。

君臣恐懼。



往 wǎng, vn, ‘go (to)’ 學 xué, vn, ‘learn’ 習 xí, vn, ‘review’ 陰 yīn, nc, ‘Yin’ 陽 yáng, nc, ‘Yang’ 長2 cháng, vst, ‘be long’ 短 duǎn, vst, ‘be short’ 恐 kǒng, vn, ‘fear’ 懼 jù, vi, ‘be afraid’

Comments ⑧ The translation ‘come and go’ turns the phrase 往來 around. If English usage requires a different word order than CC, such an inverted translation is entirely admissible. ⑨ In CC, 學習 is not yet a compound verb but a VP; both verbs may be used independently, as in 學而時習之,不亦說乎 (LY 1.1). ⑩ Yin and Yang are the two ­complementary forces that were thought to make up all things. The stative verbs 長短 may best be translated as inchoative: Yin and Yang become longer and shorter, they ‘wax and wane’. ⑪ Clauses with two coordinated nouns as S and two coordinated verbs as P may be ­elegantly translated in a crisscross pattern: ‘the ruler was afraid, and the ministers scared’ (rather than ‘the ruler and the ministers were afraid and scared’).

44

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

4.3 s i m p l e m o d i f i c at i o n

4.3  Simple modification Verbal predicates may be amplified by those optional constituents we call adjuncts (cf. 1.3.5). Adjuncts supplement the predicate by providing information about circum­ stances: place, time, manner, degree, means, and so on. The most flexible of all constit­uents, adjuncts may occur in initial, medial, or final position and may be realized by a var­ iety of words and phrases. This section deals with adjuncts that serve as adverbial modifiers. As the name indicates, they typically occur in adverbial, i.e. medial position, before the predicator. Among the kinds of information provided by adverbial modifiers, the most common is negation. Like in MSC, verbal predicates are negated by the adverb 不 or the adverb 毋/無/无, which usually immediately precedes the predicator. There is a crucial difference between the two: Whereas 不 simply negates the predicator, 毋/無/无 expresses a prohibition, a request, or a desire not to do something. ⑫ 日月 不 高。 Sun and moon are ⑬

S

⟨S⟩

A

P

not high.

不 bù, adv, ‘not’ 毋/無/无 wú, adv, ‘do not’

毋 死。 Do not die! A P

On negation in general, cf. Harbsmeier 1981, 17–48. On the properties of 毋/無/无, cf. Van Auken 2004. Van Auken argues that wu is always ‘associated with the speaker’s perception that the agent is able to exert control over his actions’ (209) and therefore treats it as a ‘modal’ negative.

Besides adverbs (which are the quintessential adjuncts insofar as they can fulfil no other function in a clause), most verbs, especially stative verbs, may serve as adverbial modifiers. ⑭ 吾樂 甚 多。

S

A

P

My joys are very numerous.

樂2 lè, nc, ‘joy’ 甚 shèn, vst, as adv, ‘very’ 多 duō, vst, ‘be numerous’

45

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 4 : v e r b a l c l au s e s



火益熱。  . . . 

棺不買。  . . . 

 . . .

水益深



君不果行。



無訟



蔡必來。

人不死



水 shuı ̌, nc, ‘water’ 益 yì, vi, as adv, ‘increas­ ingly’ 深 shēn, vst, ‘be deep’ 火 huǒ, nc, ‘fire’ 熱 rè, vst, ‘be hot’ 果 guǒ, nc, as adv, ‘really’ 訟 sòng, vn, ‘litigate’ 蔡 Cài, np, a polity 棺 guān, nc, ‘coffin’ 買 mǎi, vtr, ‘buy’

Comments ⑮ 益, as a verb, means ‘to increase’; used adverbially, it serves to express the comparative of stative verbs. ⑯ 果 indicates that an event occurred as expected; in negated sentences, it means ‘not actually’ (contrary to expectations or plans). ⑰ The character 無 is used more frequent­ly than 毋. ⑱ For 必, cf. 3.3.2–⑲. ⑲ Observe the parallel between the intransitive 不死 and the passive 不買: both constructions are formally identical, but since 死 is a vn, it cannot be passive. 4.4  Modal auxiliaries Certain verbs may be used as modal auxiliaries, which syntactically function as adjunct, usually immediately preceding the predicator. Unlike adverbial modifiers, they express necessity, potentiality, volition, or other notions concerning the action encoded by the main verb (compare English, ‘can’, ‘must’, ‘should’, etc.). The passive voice, too, may be formed with the aid of modal auxiliaries. This section will introduce some of the main modal auxiliaries and their function within verb phrases. 4.4.1  Active voice The most important auxiliaries expressing the ability and wish to perform an action are 能, ‘be able to’, 可以, ‘can’, 敢, ‘dare to’, 肯, ‘be willing to’, 知, ‘know how to’, 得, ‘get to’, 務, ‘strive to’, 欲, ‘want to’, 願, ‘wish to’, 好, ‘love to’, and 宜, ‘should’. As adjunct, they occur in medial position: 46

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

4.4 m o da l au x i l i a r i e s

⑳ 夫子 可以 行 . . .

S

A

You can leave, master.

P

夫子 fūzı ̌, nc, ‘master’ 可以 kěyı̌, vn, ‘can’

Quite a few of these verbs may also be used with an O, as in 好德 . . . 好色 (LY 9.18), or 願車馬 (LY 5.27), 得罪 (Xun 2), and perhaps also 能水 (Xun 1).



敢問



 . . .

 . . .

 . . .

弟子不能學  . . .



予欲無言。



民不肯止。



知止

不得免



敢 gǎn, vn, ‘dare to’ 問 wèn, vn, ‘ask’ 弟子 dìzı̌, nc, ‘disciple’ 能 néng, vn, ‘be able to’ 欲 yù, vn, ‘want to’ 言 yán, vn, ‘speak’ 肯 kěn, vn, ‘be willing to’ 知 zhī, vn, ‘know how to’ 得 dé, vn, ‘get (to)’ 免 miǎn, vi, ‘escape’

Comments ㉑ A stock phrase: ‘May I dare to ask. . .’ A similar expression is 請問, ‘pray tell/may I ask?’ ㉒ 能 denotes the capability to do something, whereas 可以 expresses a possibility or occasion to act. Note that this clause has two adjuncts: the negation and the aux. What if the order of the two were reversed? ㉓ After aux that express a wish or desire the usual negative is 無, since they involve the subject’s will (cf. note at 4.3). ㉕ ‘Know ⟨when⟩ to cease’; like French ‘savoir’, 知 expresses the ability to do something. 4.4.2  Passive voice The passive voice, too, may be expressed with the help of modal aux­il­ iar­ies, namely 可 and 足 as well as, in later classical texts, 見 and 被. While syntactically equivalent, these words are somewhat different in their meaning: whereas 見 and 被 simply passivize the following verb, 可 and 足 additionally have a semantic value, meaning ‘can be’ (followed by the passive verb). Moreover, the stative verbs 易 and 難, preceding the predicator, serve to passivize the latter in the sense of ‘easy to be’ or ‘difficult to be’.

47

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 4 : v e r b a l c l au s e s

Whereas only transitive verbs can be passivized by simply omitting an object (cf. 4.2.1), all classes of verbs may be passivized with the help of modal auxiliaries. ㉗

跖見害 S A P

Zhi gets hurt . . .

跖 Zhí, np 見 jiàn, vn as aux for pass. 害 hài, nc as vn, ‘hurt’

The passive meaning of a phrase like 可v may be derived from ‘allows action’ > ‘can be done’; and 足v ‘adequate for action’ > ‘is worth being done’; 見v ‘meets with action’ > ‘is done’; and perhaps 被v ‘under­goes/suffers action’ > ‘is done’. For studies of the passive voice, cf. Cikoski 1978c, Wu Yuemei 2007, and Li Yin 2015.



可欺也,  . . . 

括見殺。



實足食也。



君子

難通。  . . . 

不可罔也。



其勢易見也。





括 Kuò, np 殺 shā, vtr, ‘kill’ 實 shí, nc, ‘fruit’ 足 zú, vi, ‘be worth’ 食 shí, vn, ‘eat’

君子 jūnzı,̌ nc, ‘gentleman’ 可 kě, vst, ‘can (be)’ 欺 qī, vn, ‘lie to’ 罔 wǎng, vi, ‘be con­fused’ 勢 shì, nc, ‘circumstances’ 易 yì, vst, ‘be easy (to)’ 難 nán, vst, ‘be diffi­cult (to)’

Comments ㉙ Verbal predicates with 可 and 足 are often marked with 也, since like nominal clauses they may be understood as categoric statements that claim to be unconditionally valid: ‘fruits are edible’, not just at one time but generally. ㉚ 君子, ‘gentleman’ (originally the ‘lord’s son’), means a man involved in public life. The model of ‘Confucian’ education, he embodies refined, principled behaviour. Be sure to dis­tinguish 君子 from 君, the ‘lord’. ㉛ Again, 也 marks the passive predicate.

48

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

4.5 e x e r c i s e s

Box 7  Punctuation The ancient Chinese, just like the ancient Greeks and Romans, basically wrote their texts in scriptio continua, thatiswithoutspacesbetweenwordsorothergrap hicdevi­cestostructuresentences. Although some excavated manuscripts attest to the use of dots to distinguish paragraphs and, occasionally, squares, slashes or hooks to mark other textual units, there was no systematic punctuation separating sentences or their elements. Only in post-­classical paper manucripts small circles—and later, comma-­like dots—were used to indicate the internal structure of paragraphs and sentences. However, unlike English punctuation, this practice was not primarily governed by grammatical considerations: the dots and circles did not necessarily separate syntactic units but rather cola (sing.: colon), that is rhythmical units of utterances. They were often inserted not by the authors themselves but by the readers: simply put, wherever the reader paused to take a breath, he would put a circle. This practice was called judou 句讀, ju meaning syntactical units—clauses or complex sentences—and dou meaning shorter ones, typically constituents, but also clauses.

Even when Western-­style punctuation was introduced in the twentieth century, this practice was retained: thus a comma may separate the subject from the predicate, or the predicator from a prepositional phrase, which may be irritating to Western eyes. Readers of the present book should bear these differences in mind, since it follows the Chinese practice of punctuation throughout. Whereas only circles and commas are used in the example sentences of the first part, the second part introduces all punc­tu­ation marks found in the original Chinese editions. References: Bodde 1991; Guan Xihua 2002; Galambos 2014; Richter 2015.

4.5 Exercises Try to analyse the following clauses, marking subject, adjuncts (if ap­plic­ able), and predicator, and translate them with the help of the ­glossary. ㉝

 . . .

君臣憂懼。



兄弟被侵

 . . . 49



其餘不足觀也。



楚不敢伐。



直木先伐,

甘井先竭。

吾言足用



OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 4 : v e r b a l c l au s e s

References  ① LY 8.7  ② HFei 7  ③ Meng 1A3  ④ Xun 22  ⑤ Li 8  ⑥ SY 13  ⑦ LT 4.5  ⑧ CQFL 16  ⑨ SJ 6  ⑩ Li 24  ⑪ ZGC 11.1  ⑫ Xun 17  ⑬ Zuo 6.10  ⑭ SY 17  ⑮ Meng 1B17  ⑯ Zuo 10.7  ⑰ LY 12.13  ⑱ SJ 35  ⑲ HFei 17  ⑳ SJ 47  ㉑ Meng 3B9  ㉒ LY 7.34  ㉓ LY 17.19  ㉔ SY 7  ㉕ Li 42  ㉖ Meng 1B15  ㉗ HFei 26  ㉘ Meng 7B29  ㉙ HFei 49  ㉚ LY 6.26  ㉛ SJ 91  ㉜ GY 8.2  ㉝ ZGC 33.10  ㉞ HFei 49  ㉟ LY 8.11  ㊱ Zuo 2.6  ㊲ Zhuang 20  ㊳ Mo 47

50

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 3

Reconstructing Old Chinese Reconstructed Middle Chinese serves as the starting point for the reconstruction of Old Chinese. This reconstruction based on a reconstruction is a highly intricate project, and the few specialists in the field do not agree on its details. There are no living languages to supply the sound values from which Old Chinese may be deduced, like in the case of Middle Chinese. Furthermore, there are no contemporary phonological descriptions like fanqie glosses or rhyme dictionaries for Old Chinese. Fortunately, there is other evidence which helps to extrapolate from Middle Chinese to Old Chinese. The first set of evidence comes from the rhymes in the Canon of Odes (Shijing, cf. Focus 11). Assuming that the Odes were written in (more or less pure) rhyme, systematic information may be gleaned about correspondences of sounds in Old Chinese. Linguists have classified the rhyme words of the Shijing in thirty-­one ‘rhyme groups’, which differ significantly from those of Middle Chinese rhyme dictionaries. The second set of evidence is provided by Chinese writing. While Chinese characters do not represent sounds as closely as alphabetic scripts, the vast majority consist of a semantic as well as a phonetic component that indicates their pronunciation. By assembling such phonetic compound characters (諧聲字 xieshengzi or 形聲字 xing­ shengzi) written with the same ‘phonetic’ element, one obtains ­phonetic series like the following: 余 除 茶 途 徐

MSC yú chú chá tú xú

MC *jiwo *ḍjwo *ḍa *duo *zjwo

OC *la *dra < r-­la *dra < r-­lâ *lâ *s-­la

Cf. Karlgren 1957, No. 82.

As these examples show, the Modern Standard Chinese readings of characters within a ‘phonetic series’ may differ widely. However, at the time these characters were created, which was the period of Old Chinese, 51

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 3 : r e co n s t r u c t i n g o l d c h i n e s e

their pronunciations must have been fairly close. Firstly, characters within the same phonetic series all belonged to the same rhyme group. This observation has enabled scholars to correlate not only the rhyme words of the Shijing but virtually all Old Chinese words according to their finals. Secondly, the initials of characters within the same phonetic series all share the same point of articulation: they are all dental, palatal, velar, or the like. If the point of articulation differs in Middle Chinese, so the hypothesis goes, they must go back to initial consonant clusters in Old Chinese. Take the following example: MSC 監 jiān 藍 lán

MC OC *kam *krâm *lâm *râm

Since both words are from the same series, but their initials differ even in Middle Chinese—*k- vs. *l-—it has been concluded that in Old Chinese there must have existed an initial consonant cluster *kr (*l and *r were both medial consonants). Such consonant clusters, in fact, are characteristic features of Old Chinese that set it apart from later stages of the Chinese language. They are in many cases confirmed by a third set of evidence for the reconstruction of Old Chinese: cognate words from Tibetic, Burmic, Baric, Karenic, and other Tibeto-­Burmese languages. Since Old Chinese is a member of the Sino-­Tibetan language family, the ancient words of these languages, reconstructions themselves, add further aspects to reconstructed Old Chinese. For example, the reconstructed word for ‘tea’, *dra < r-­lâ, given above finds parallels in Loloish *la, Proto-­Min *da, and others; and the Tai word for ‘indigo’ (藍) has initial *gr-. It should be borne in mind that these reconstructions are abstractions which reduce the infinity of actual physical sounds produced in spoken language to a few dozen phonemes, that is the smallest units that can make a semantic difference. While they provide us with a detailed picture of the phonological (or phonemic) structure of Old Chinese words, they cannot give any information about their phonetic value. Simply put, they do not record sounds, but only differences between sounds. We do not know how Classical Chinese was actually pronounced.

One feature Old Chinese has in common with most Sino-­Tibetan languages is a rich derivational morphology with diverse affixes. Among others, it possessed a prefix *s- for causative or exoactive verbs, deverbal nouns, and others; a suffix *-s for denominal verbs and deverbal nouns; a suffix *-k for distributives; a suffix *-t for increased valence; an infix *-r- for repetitive actions; and a suffix *-n with nominalizing 52

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 3 : r e co n s t r u c t i n g o l d c h i n e s e

function. Such reconstructions disprove the generalized notion that Chinese is an ‘isolating language’: Old Chinese still possessed a ­morphology that only became obsolete in later stages of the Chinese language. The pioneering study is Downer 1959, who pointed out that more than 200 Old Chinese words had derivatives resulting in MC ‘departing’ tone (i.e., they had an OC suffix *-s/-h). Unger 1983 argued that another 200 verbs had morphologically distinct perfective forms.

Besides the complex syllable structure, there is another striking difference between Old Chinese and Middle Chinese. Analysing the evolution of Middle Chinese phonology, it becomes apparent that its tonality (cf. Focus 1) derived entirely from non-­tonal features of Old Chinese. The ‘rising tone’ goes back to a glottal stop *-Ɂ in Old Chinese; the ‘departing tone’ derives from an OC suffix *-s; the ‘entering tone’ is not really tonal anyway, since it is associated with the MC finals *-p, *-t, and *-k; and the ‘even tone’ simply derived from unmarked syllables. If this theory holds true, Old Chinese did not have any tones at all. If Middle and Modern Chinese are tonal languages, then Old Chinese indeed was an entirely different language. References: Li Fanggui 1971; Norman 1988, 42–8; Baxter 1992; Pulleyblank 1992; Sagart 1999; Pan Wuyun 2000; Gassmann and Behr 2005, vol. 3, 387–445; Schuessler 2007, 2009; Baxter and Sagart 2014; Harbsmeier 2016. Jeffrey Tharsen’s Digital Etymological Dictionary of Old Chinese is the most convenient online resource for OC phonology.

53

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 5

Objects and questions

5.1 Introduction54 5.2 Objects55 5.2.1 Single object S  P  Od55 5.2.2 Two objects S  P  Oi    Od57 5.2.3 Negation S  不 Opr  P58 5.3 Interrogative sentences 60 5.3.1 Yes-no questions and S  P  乎/與 alternative questions S  P    與    P 與60 5.3.2 Variable questions S  何/誰  P61 5.4 Exercises62

5.1 Introduction This lesson continues the discussion of verbal clauses, introducing predicates with one or two objects as well as interrogative clauses. After subject and predicator, objects are the most important constituents of verbal clauses: they represent a person or thing (or even a place) to which an action is directed. Objects are typically realized by nouns or pronouns. The most important object pronoun is 之, ‘him, her, it, them’, which frequently replaces a previously men­tioned noun in this function; 之 is the object par excellence in Classical Chinese. In principle, all Classical Chinese verbs license an object, that is they may be used transitively. However, the way the object relates to the action differs according to the word class of the predica­tor. With transitive and (usually) with neutral verbs as predicator, the object is affected by the action. But when stative or intransitive verbs are used transitively, they are to be understood in a causative sense: that is, the object is caused to behave in the 54

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

5.2 o b j e c t s

way expressed by the predicator. Neutral verbs also lend themselves to ditransitive usage, i.e. they license an indirect and a direct object. The canonical word order is S P (Oi) Od, the object(s) following directly after the predicator with no other constituent intervening. But in certain cases this word order is changed, namely when interrogative pronouns or the indefinite pronouns 有 or 無 serve as object, or when a clause with a personal pronoun as object is negated. Preliminary questions: Which of the canonical clause types listed in 1.4 are discussed in this lesson? Review the subclasses of verbs introduced in 1.2.3: have you clearly understood their lexical properties? Can you think of causative constructions in your mother tongue? 5.2 Objects The way clauses including objects are to be construed varies significantly depending on the word class of the predicator. Moreover, if such clauses involve a negation, certain objects are transposed from final to medial position. 5.2.1  Single object The basic structure of a verbal clause involving an object is S P O, the object usually being the direct object. Such sentences typically serve to express an action by the subject that directly affects the object: ①

臣 S

弒 P

其君。A minister assassinates his ruler. Od

However, there are certain cases in which this canonical word order does not apply. If the indefinite pronouns 有, ‘something’, or 無, ‘nothing’, function as object, they are always placed in medial position, before the predicator. ②

耳 S

無 Od

聞。 His ears heard nothing. P

耳 ěr, nc, ‘ear’ 無2 wú, prind, ‘nothing’

There are different explanations for this construction. For example, Cikoski 1970, 42, asserts that medial 有 and 無 simply ‘obviate an object’. Other scholars assume that the Prrel 所 has been deleted between 有/無 and the verb. The above interpretation follows Unger 2019, 585–91.

55

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 5: o b j e c t s a n d q u e s t i o n s

Whereas 弒 and 聞 in the preceding examples—being transitive and neutral verbs, respectively—may be expected to take an object, and, indeed, usually do, there are other verbs that usually do not take an object: namely stative and intransitive verbs. But in quite a few instances, such verbs are followed by an object, in which case they are to be construed as causative. In such clauses, the subject makes the object do something, as in: ‘The grand old Duke of York, / He had ten thousand men; / He marched them up to the top of the hill, / And he marched them down again’, or: ‘there’s the little problem of where to sleep you’ (James Purdy). In Classical Chinese, such causative constructions are quite common. Compare the clause 是人來, ‘This man came’ (GYang 1.6), with the following: ③

⟨S⟩

來 P

之。 Od

Make them come.

Causation may be effected factually, as in this case, or mentally, in which case the subject believes the object to be or behave a certain way. We will call the former usage factitive and the latter putative. Whereas intransitive verbs, followed by an object, are to be construed as factitive, stative verbs may also be used in a putative sense. Consider the following case: ④

武王 S

之。 King Wu believed they were good/approved of them. Od ⑤

君使臣,臣事君。



史為書,瞽為詩。



有見



無為



休卒徒。



不遠千里  . . .



愚其民。

聖人不高山,不廣河。



善 P

事 shì, nc, as vn, ‘serve’ 書 shū, nc, ‘document’ 詩 shī, nc, ‘song’ 有 yǒu, prind, ‘some­thing’ 休 xiū, vi, ‘rest’ 卒 zú, nc, ‘soldier’ 徒 tú, nc, ‘footsoldier’ 遠2 yuàn, vtr, ‘consider far’ 里 lı̌, nc, ‘mile’ 愚 yú, vst, ‘be stupid’ 廣 guǎng, vst, ‘be wide’ 河 Hé, np, ‘Yellow River’

56

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

5.2 o b j e c t s

Comments ⑤ 使 means ‘to employ’ subordinates, 事 means ‘to serve’ superiors. ⑥ For the uses of 為, cf. Box 14. ⑧ This is probably the most mystified of Daoist doctrines; yet its literal translation is quite exoteric. ⑩ A 里 is usually rendered as ‘(Chinese) mile’; in actual fact, in the ­classical age its length was about one quarter of a mile. ⑪ Whereas 愚民 may be understood as a NP, 愚其民 is unambiguous: since the pr 其 cannot be modified (cf. 2.4.2), 其民 must be construed as the O to 愚, which, in turn, must be causative.

Note that the verb 遠 in ⑩ is read yuàn, not yuǎn. The causative or exoactive (i.e., directed outward) form of several OC verbs was morphologically distinguished by a suffix *-s/-h, which developed into the MSC fourth tone. Compare the following pairs from the same root: 飲 yı ̌n < *ɁəmɁ, ‘drink’, and 飲 yìn < *Ɂəms, ‘make (s.o.) drink’; 食 shí < *m-­lək, ‘eat’, and 食 sì < *s-­ləkh, ‘feed’; 好 hǎo < *hûɁ, ‘be good’, and 好 hào < *hûh, ‘consider as good, like’; 惡 è < *Ɂâk, ‘be bad, evil’, and 惡 wù < *Ɂâkh, ‘treat as bad, hate’; 視 shì < *giɁ, ‘look at’, and 示 shì < *gih, ‘show’; In the last case, the causative/exoactive form is written with a distinct character. In most cases, how­ever, the script gives us no clue to the morphology of Old Chinese words. References: Downer 1959; Schuessler 2007, 42–5; Sun Yuwen 2015.

5.2.2  Two objects Whereas, in principle, all Classical Chinese verbs may take an object, some verbs license two objects. Typically, these verbs express some kind of transfer. Consider, for example, English sentences like ‘I gave her the book’, or ‘She taught me a lesson’. In both sentences, an indirect object (‘her’, ‘me’) is followed by a direct object. We will call this usage ditransitive. In Classical Chinese, such ditransitive constructions are possible only with neutral verbs. ⑬

父 S

教 P

之 Oi

忠。 Od

The father teaches him loyalty.

Here, too, the object immediately following the predicator is the indirect object, the direct object coming in second place. In such constructions, only the indirect object may be realized by the pronoun 之, which ­carries little emphasis; the direct object, occupy­ing the prominently stressed final position, may not.

57

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 5: o b j e c t s a n d q u e s t i o n s



予之法制,

 . . .

告之訓典。



與犬肉

飲小臣酒。



賜之千金。



賜金錢。



彼奪其民時。

為之椁。



予2/與3 yǔ, vn ‘give’ 法 fǎ, nc, ‘rule’ 告 gào, vn, ‘announce’ 訓 xùn, nc, ‘teaching’ 典 diǎn, nc, ‘codex’ 犬 quǎn, nc, ‘dog’ 飲 yìn, vn, ‘give to drink’ 賜 cì, vn, ‘present’ 錢 qián, nc, ‘coin’ 奪 duó, vn ‘steal’ 時 shí, ntemp, ‘time’ 椁 guǒ, nc, ‘(outer) coffin’

Comments ⑭ 之 may serve as Opr, because the noun it replaces is known from the context. Can 予 represent a prper in this clause? ⑮ 犬肉 does not mean ‘dog meat’ in this case; rather, 與 takes two O: 犬 (Oi) and 肉 (Od). Ob­serve that 飲 has the fourth tone, it is morphologically distinct from the vn 飲, ‘drink’ (cf. the note above). ⑯–⑰ Examples of the same verb being used with two and with only one O; the latter is usually the Od. ⑱ Not only verbs that express giving may accommo­date two O, but also such that express taking. Note that the prdem 彼 is used here as a substitute for a third person prper. ⑲ 為椁 means ‘make a coffin’; the additional Oi may simply be rendered as a dative, ‘him’, in English. On these verbs, cf. Herforth 2008, who distinguishes between ‘donative’ (e.g. 與), ‘privative’ (e.g. 奪), and ‘benefactive’ (e.g. 為) verbs.

5.2.3 Negation As we have seen in 4.3, verbal predicators are usually negated by 不 or 毋/無/无 without a change in word order. But if the object (or one of the objects) is a first or second person personal pronoun, it is moved into medial position, between the negation and the predicator (compare French ‘J’aime les livres’ vs. ‘Je ne les aime pas’). The resulting structure is: ⑳ 君 不 吾 聽 The ruler did not listen to me.

S

A

O

P

58

聽 tīng, vn, ‘listen to’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

5.2 o b j e c t s

The same rule also applies to the prrefl 己, as in 不患人之不己知 (LY 1.16). Interestingly, it also occurs with 子, which substitutes for the second person pronoun (cf. Box 5), as in 不子 欺 (Lü 15.7).

The pronoun 之 also moves into medial position in negated sentences. However, due to reasons of intonation, 之 is usually not realized as an independent word following the negation. Since it carries little stress (similar to the pronoun ‘le’ in French ‘Je ne l’aime pas.’), it often merges with the preceding negation, yielding the fusion words 弗 (< 不 and 之) or 勿 (< 毋, ‘don’t’, and 之). We thus have the curious case that one  character represents two words and must be analysed as two ­constituents: ㉑ 民

弗 去。The people won’t leave it. S AO P ㉒ 人 勿 聞。The people should not hear it. S AO P

弗 fú, fus, ‘not . . . it’ 去 qù, vn, ‘leave’ 勿 wù, fus, ‘do not . . . it’

The very presence of 弗 or 勿, then, indicates that the following verb is used transi­tively. More deceptively, 之 is simply deleted after 不. ㉓

孔子 S

不 A

⟨O⟩

見。 P

Master Kong did not meet ⟨him⟩.

Recall that 不 may also negate intran­sitive or passive verbs: the construction, in both cases, is S 不 P. Read in context, however, the two are easily distinguishable. Instances of 之 after 不, such as 寡人不之疑矣 (SJS 1), are rare. However, awareness of the fact that 之 is inherent in 弗 and 勿 seems to have receded over time; cf. sentences like 亡則 弗之忘矣 (Li 3), 弗失之矣 (Li 31), or 王欲行王政,則勿毀之矣 (Meng 1B5). Cf. Graham 1952 and Pulleyblank 1978.

59



人不我順。



余不女忍殺。



愛之…弗仁。



勿奪其時。

我無爾詐,

爾無我虞。



順 shùn, vn, ‘follow’ 忍 rěn, vn, ‘endure to’ 詐 zhà, vn, ‘cheat’ 虞 yú, vn, ‘take precautions (against)’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 5: o b j e c t s a n d q u e s t i o n s

Comments ㉕ If the negated predicate includes an aux (cf. 4.4), the Opr is placed be­fore it. ㉖ Here, the stative verb 仁, following 弗, must be construed as causative. ㉗ A P with two O (cf. ⑱): the Oi is subsumed in 勿, while the Od remains in final position. Note that 勿 does not simply negate the P but expresses a negative injunction: ‘do not . . . it’. 5.3  Interrogative sentences Just like nominal clauses, verbal sentences may express yes-­no questions and alternative questions, marked by interrogative particles, as well as variable questions, expressed by interrogative pronouns. 5.3.1  Yes-­no questions and alternative questions The word order in interrogative clauses is the same as in declarative verbal clauses, the only distinguishing feature being an interrogative particle in final position. The interrogative particles used in verbal sentences are 與/歟 and 邪, which also occur with nominal predicates, and, most frequently, 乎. If 乎 follows the object pronoun 之, the two words are usually contracted into the fusion word 諸 *ta < *tə + *ɦâ (cf. Box 16). Strictly speaking, 乎 is not a purely interrogative pt, since it also marks affirmative sentences like 吳其亡乎 (SY 9), but its primary function is certainly interrogative. For 與, cf. Graham 1957.

㉙ 魯 可 取 乎。 Can Lu be conquered?

S A

P

A

事齊乎,事楚乎。  . . . 

與之與。  . . .

60



汝知之乎。



湯放桀,武王伐紂,有諸。



滕小國也



求之與

女弗能救與。



取 qǔ, vn, ‘conquer’ 乎 hū, ptint 放 fàng, vtr, ‘get rid of ’ 桀 Jié, np, the last Xia king 伐 fá, vn, ‘attack’ 紂 Zhòu, np, Shang king 有2 yǒu, vn, ‘occur’ 諸 zhū, fus < 之乎 滕 Téng, np, a polity 楚 Chǔ, np, a polity 求 qiú, vn, ‘demand’ 救 jiù, vn, ‘save’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

5.3 i n t e r r o g at i v e s e n t e n c e s

Comments ㉚–㉛ Observe the difference between 之乎 in the first clause and the contraction 諸 in the second. While the meaning remains the same, the second clause may have sounded more colloquial, as in English ‘what is’ compared to ‘what’s’. ㉜ For alternative questions, just like in nominal clauses (cf. 3.4.2), the interrogative particle is repeated. ㉝ An alternative question with 與. Note that the second of the three 與 is not a pt but a verb—why? ㉞ There are six constituents in this clause: can you pinpoint them? 5.3.2  Variable questions In variable questions, an interrogative pronoun such as 何 or 誰 is used either as subject or as object. The first construction, analogous to nominal interrogative clauses (cf. 3.4.3), serves to express rhetorical questions. ㉟ 誰 敢 興

S

A

P

之。Who would dare to raise it? O

興 xīng, vi, ‘rise’

With the second—more frequent—construction we encounter another exception to the ground rule of Classical Chinese word order introduced at pp. xxvi–xxvii. When a print serves as object, it is always placed before the predicator; the resulting clause structure is: ㊱

國 何 S O

患。 What does the state fear? P ㊲

天…何欲何惡。



吾誰欺。欺天乎。

誰能止之。



患 huàn, vtr, ‘fear’

惡 wù, vn, ‘loathe’

Comments ㊴ This is a rhetorical question. For a real question, a ­ ominal clause with 誰 as P would be called for. n 61

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 5: o b j e c t s a n d q u e s t i o n s

Box 8  Optional determination A characteristic of Classical Chinese that deserves to be pointed out is the principle of optional determination. Whereas in European languages, as a rule, gram­mati­cal features like number, case, tense, etc. are expressed—indeed, must be expres­sed—morpho­logically (through ablaut or affixes), this is not the case in Classical Chinese. Instead, these features can be indicated by free morphemes (e.g. particles), but they need not. Hence temporal adverbs may be employed to signify past, present, or future tense, con­junctions may mark conditional clauses, counters may indicate number; but in all these cases, determination is optional. In fact, economy of determi­native elements was a rhetorical principle in Classical Chinese. Overly determined texts were held in low esteem. For modern readers, this is one of the major factors that make Classical Chinese texts difficult to understand. Deletion of the S may be treated as a special case of optional determination insofar as the S can be re­garded as a modifier of the predicate. References: Unger 1985, 104f; Harbsmeier 1998, 143–50.

5.4 Exercises Only four sentences, but they are likely to be challenging: try to analyse and translate them as best you can. ㊵

吾得見王。



寡人弗敢忘。



爾無我叛。

三軍可奪帥也,

匹夫不可奪志也。



References  ① Meng 1B8  ② Meng 3B10  ③ LY 16.1  ④ Xun 25  ⑤ LY 3.19  ⑥ Zuo 9.14  ⑦ Meng 3B10  ⑧ Lao 48  ⑨ Zhuang 29  ⑩ Meng 1A1  ⑪ Yan 8.1  ⑫ HNan 19  ⑬ Zuo 5.23  ⑭ Zuo 5.23  ⑮ GY 8.1  ⑯ SY 8  ⑰ SJ 118  ⑱ Meng 1A5  ⑲ LY 11.8  ⑳ SY 1  ㉑ Meng 1B13  ㉒ SY 9  ㉓ LY 17.1  ㉔ Zuo 9.30  ㉕ Zuo 10.1  ㉖ Meng 7A45  ㉗ Meng 1A3  ㉘ Zuo 7.15  ㉙ Zuo 4.1  ㉚ Zhuang 2  ㉛ Meng 1B15  ㉜ Meng 1B13  ㉝ LY 1.10  ㉞ LY 3.6  ㉟ SY 5  ㊱ SY 7  ㊲ Mo 7  ㊳ LY 9.12  ㊴ SY 9  ㊵ Meng 2B14  ㊶ Zuo 1.3  ㊷ Zuo 10.16  ㊸ LY 9.26

62

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 4

Chinese writing The earliest evidence of Chinese writing dates from late Shang oracular inscriptions on turtle plastrons and cattle scapulae. Although it is likely that the Shang also wrote on perishable material—bamboo slips, prob­ ably—that has not survived the millennia, it does not seem likely that the invention of writing significantly predates the thirteenth century bc. Occasional marks on neolithic pottery fragments are tantalizing evidence of earlier recording systems, but they cannot qualify as proper writing. Shang script (cf. Figure 2), by contrast, already displays all the charac­ teristics of later Chinese writing. Although many characters still betray their picto­graphic origin, they clearly stand for words (or, more pre­ cisely, morphemes, being the smallest seman­tic units of the language), not pictures or concepts: by conven­tion, they had a fixed pronunciation and meaning, hence they may rightly be called writing.

Figure 2  Oracle bone inscription Source: David N. Keightley. Sources of Shang History: The Oracle-­Bone Inscriptions of Bronze Age China. Berkeley: University of California Press, 1978, fig. 14.

63

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 4: c h i n e s e w r i t i n g

Chinese writing shares its pictorial origin with the other original writing systems of the world, Mesopota­mian, Egyptian, and Maya; and it developed in the same way as these, expanding its inventory by the principles of rebus, poly­phony, and secondary determination. Naturally, an inventory of signs that only includes pictographs ­cannot serve as a writing system, it must be significantly expanded in  order to be functional. The Shang used three basic methods to achieve this: 1. The first was that of rebus, by which a character invented for a certain word was also used for another, homophonous word. For ex­ample, the graph (modern: 禾), originally a pictograph repre­ senting a stalk of grain, was used for the word *(g)wâi, ‘grow grain’. But additionally, it was also used to write the homo­phonous (but semantically unrelated) word *(g)wâi, ‘harmony’, an abstract word for which there was no pictograph. By the same method, the graph (modern: 其), originally depicting a winnowing fan and used to write that word (*gə), was also employed for the third per­ son pronoun, which was pronounced exactly the same way. By this rebus principle, characters became multivalent, expanding the scope of writing. 2. The Shang apparently used characters not only to write hom­oph­ onous, although semanti­cally unrelated words, but also to write semantically related but phonetically entirely different words. Hence the above-­mentioned character for ‘grow grain’, , was also used for the word ‘harvest’, *nîn (modern: nián). In this way, char­ acters acquired multiple readings, they became polyphonic. 3. While adding greatly to the flexibility of the Chinese writing sys­ tem, rebus and poly­phonic usage of graphs also increased their ambiguity. In order to resolve the ambiguity, another element was added to respecify the meaning of polyvalent graphs. In the case of polyphonic usage, a phonetic determinative was added for dis­ ambiguation, e.g. *nin was added to write the word *nîn, ‘year’ ( > 年). In the case of rebus usage, a semantic determinative was added: for example, was attached to to dis­tinguish the word ‘har­mony’ (resulting in modern 和); and the determinative for ‘bam­boo’ was added to in order to write the word ‘winnowing fan’ (modern: 箕), while 其 con­tinued to be used for the pronoun. As a result, most simple pictographs were eventually used for a 64

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 4: c h i n e s e w r i t i n g

differ­ent word than the one they had been invented for, while the latter was written with a compound character. These three methods of rebus usage, polyphonic usage, and second­ ary determina­tion most likely did not succeed one another but were applied simultaneously. They resulted in the mature, fully operative writing system of the Shang. But the development of Chinese writing continued. As the Shang script was adopted by the Zhou conquerors (cf. Figure 3), new charac­ ters were developed by using the same principles. Especially the rebus principle and the consequent use of semantic determinatives proved to be highly fruitful. It could be applied to new characters, and it could be re-­applied to compound charac­ters, adding a third, fourth, or fifth determinative element. This resulted in a large number of characters

Figure 3  Bronze inscription Source: Edward L. Shaughnessy. Sources of Western Zhou History: Inscribed Bronze Vessels. Berkeley: University of California Press, 1991, fig. 1.

65

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 4: c h i n e s e w r i t i n g

with the same phonetic element and differ­ent semantic determinatives, so-­called ‘phonetic series’ (cf. Focus 3). Semantic determina­tives are also—quite deceptively—called ‘radicals’, although they were by no means at the root of character forma­tion. The root of every character, no matter how many elements it consists of, is always a character asso­ ciated with a sound, that is with a word. This analysis may serve to debunk the tenacious myth of ‘ideo­graphic’ writing that allegedly expresses not language but ‘ideas’. This myth has been sustained by the explanation of certain characters as being ‘seman­ tic com­posites’ (會意, literally ‘combined meanings’). Thus the charac­ ter 好 is interpreted as expressing the union of a woman (女) with her child (子), hence ‘love, good’; or 明 is analysed as sun (日) and moon (月), hence ‘bright’; 安 is taken to convey the idea of a woman under­ neath a roof, that is at home, hence ‘peaceful’. While useful as mne­ monic aids, such analyses are in most cases wrong. Most—perhaps all—alleged ‘semantic com­ posites’ are, like more than 90% of the Chinese characters, actually semantic-­phonetic composites (諧聲 or 形 聲): they repre­sent not ideas but words. References: Hu Pu’an 1984; Bauer 1991; Boltz 1994, 2000; Qiu Xigui 2000; Bagley 2004.

66

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 6

Noun phrases II

6.1 Introduction67 6.2 Subordination with 之68 6.2.1 Nouns as modifiers NP 之 NP68 6.2.2 Four-syllable phrases VP 之 NP69 6.3 Nominalization with 者  VP 者72 6.4 Exercises73

6.1 Introduction Continuing the discussion of noun phrases, we now turn to more complex con­structions, which make use of the particles 之 and 者. Two of the most important func­tion words in Classical Chinese, these particles greatly increase the expressive potential of noun phrases. With the use of 之, simple nouns or noun phrases may be subordinated, expressing relations that can be construed as possessive, descriptive, or other ­genitives. What is more, 之 may also subordinate verb phrases or entire verbal clauses. Such com­plex modifications may often be translated as relative clauses in English. In a similar way, 者 nominalizes verb phrases, yielding complex and sometimes very long noun phrases. Preliminary questions: What kind of genitive constructions do you know in your mother tongue? What relation between the modifier (the genitive) and the head do they express? How are relative clauses expressed in Modern Standard Chinese?

67

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 6: n o u n p h r a s e s i i

6.2  Subordination with 之 For grammatical and stylistic reasons, asyndetic subordination of nouns (cf. 2.4) is often not feasible. In such cases, subordination is marked by the particle 之 between modifier and head. This is necessary in particular for possessive genitives, expressing ownership or other kinds of belonging, or for stylistic reasons, to create four-­syllable phrases. The modifier in such four-­syllable phrases may well be a verb phrase, ex­press­ing a descriptive genitive. Generally, in noun phrases with 之 the connection be­tween modifier and head is not as close as in asyndetic subordination; neither do they express categories nor do they tend to become lexicalized. Note that 之 is a particle, not a conjunction like 與. It is functionally similar to modern 的 (which first appears as a genitive marker in Song times) albeit not equivalent. For the English translation, it is useful to consider 之 as the equivalent to the genitive -’s.

6.2.1  Nouns as modifiers Whereas personal pronouns are directly subordinated to form a possessive genitive, this is not possible for nouns. Since nouns as modifiers in asyndetic subordination express a descriptive genitive (cf. 2.4.1), they must be subordinated with the particle 之 (not the pronoun 之) to form a possessive genitive. ① 子之車

n1 ® n2    NP

the master’s cart

之2 zhī, pt, subordinates NP

To be sure, the difference between the two types of genitive is not always sharply defined, and some expressions allow an alternative phrasing. Compare, for example, 君命 (Zuo 2.6) vs. 君之命 (Zuo 7.2), or 禮經 (Zuo 1.7) vs. 禮之經 (Zuo 1.11).



 . . .

臣,王之臣也。



父之兄妻

父之弟妻



堯 舜之道

君之倉廩實。



倉 cāng, nc, ‘storage’ 廩 lı̌n, nc, ‘granary’ 實2 shí, vi, ‘be full’

Comments ② Observe the difference to direct subordination here: 王之臣 expresses an attachment to the king, whereas 王臣 means ‘royal 68

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

6.2 s u b o r d i n at i o n w i t h



minister’, denoting a quality of the minister and a category (similar to 公臣, ‘ducal minister’). ③ The position of 之 is crucial here; 父兄之妻 means ‘wives of elder relatives’. Always begin analysing such phrases from the back, since that is where the head is. ⑤ This is an entire verbal clause. 6.2.2  Four-­syllable phrases Another common function of the particle 之 is to subordinate two-­ syllable modifiers and one-­syllable heads or vice versa in order to create a four-­syllable phrase (cf. ③–⑤). This is strictly speaking a stylistic convention; but it is so consistently followed in Classi­cal texts as to actually be considered a grammatical rule. Consider the difference be­tween 門外 and: ⑥ 大門 之 外

NP ® n  NP

outside the great gate 門 mén, nc, ‘gate’ 外 wài, nloc, ‘(the) outside’

This may be loosely compared to the MSC usage of 的 in 很好的朋友 vs. 好朋友.

We have already discussed how simple verbs serve as m ­ odifiers in noun phrases (2.4.3); with 之, verb phrases and, indeed, entire verbal clauses— including an object and, sometimes, a subject—may be subordinated to compose a noun phrase. Such verb phrases do not express a possessive genitive, but typically a descriptive genitive: ⑦ 有德 之 君

VP n ®  NP

a virtuous ruler

Such descriptive genitives are parallel in their function to English relative clauses (‘a ruler that possesses virtue’), which do not exist in Classical Chinese. Another kind of genitive expressed in this way is the subjective genitive, as in: ⑧ 父母 之 愛

NP ® n  NP

the love of father and mother

If one con­verts this NP into a clause, the parents would be the subject: *父母愛⟨O⟩. Note that in this example the verb 愛 is used nominally as 69

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 6: n o u n p h r a s e s i i

the head of a noun phrase. This may seem perplexing. But compare English participles like ‘Gimme all your lovin’ (ZZ Top), which are also used like nouns. In Classical Chinese, such nominal use of verbs is rather com­mon. As the following examples will show, quite a few other types of genitive relations may be expressed by subordination with 之. 四海之內

不善之行

用兵之道

殺人之罪

臣生之歲

無子之時

社稷之利

得道之人

不貴難得之貨。

外無敵國之患,內無亂臣之憂。

⑱ ⑰ ⑯ ⑮ ⑭ ⑬ ⑫ ⑪ ⑩ ⑨

內 nèi, nloc, ‘(the) inside’ 行3 xìng, nc, ‘behaviour’ 用 yòng, vtr, ‘employ’ 兵 bīng, nc, ‘soldier’ 罪 zuì, nc, ‘crime’ 歲 suì, nc, ‘year’ 無3 wú, vn, ‘have not’ 社稷 shèjí, nc, ‘altars of soil and millet’ 貴 guì, vst, ‘be valuable’ 貨 huò, nc, ‘goods’ 敵 dí, nc, ‘enemy’ 憂 yōu, vn, as nc, ‘worry’

Comments ⑨ This means the extent of the realm. Of course, 海 cannot literally mean ‘sea’ in this context, since North China, even two millennia ago, was not sur­rounded by four seas. Rather, it refers to wild, ­inaccessible ‘borderlands’. Compare this construc­tion with 海內: either two- or four-­syllable phrases were preferred in Classical Chi­nese. ⑩ A descriptive genitive with a negated verb as modi­fier. ⑪–⑫ In both cases, a verb and its O—i.e. a VP—serves as modifier. ⑬ Here, a S and a P—an entire ‘demoted’ clause—modify the head. ⑮ This is an objective genitive; turned into a clause, the altars would be the O: *⟨S⟩利社稷. ⑰ This is a complete clause; start your analysis by considering the functions of 不 and 之. Note that subordination with 之 makes it possible to use vn as mod. ⑱ Two parallel clauses with objective genitives: the enemy states and unruly ministers are the object of ­anxieties. Also note that the two S have a locative role (cf. 1.3.1): they are not agents but they designate the place of the action (compare French ‘il y a’, or colloquial German ‘da hat’s’). Such constructions ­usually have the form S 有/無 O: ‘in/at/among the S there is/is not the O’.

70

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

6.2 s u b o r d i n at i o n w i t h



Box 9  Types of genitive While all genitive constructions in CC are syntactically alike, the logical relation­ship between modifier and head that they express may be quite different. The most important types of genitive in CC are the following. Possessive genitive, in which the head belongs to the modifier, as in: 吾王 子之車    

my king (Meng 1B1) the master’s cart (LY 11.8)

Descriptive genitive, in which the modifier describes a quality of the head, as in: 金人 有德之君

a bronze figure (SY 10) a virtuous ruler (Zuo 10.20)

Subjective genitive, in which the modifier logically is the subject of the action, as in: 父之命     君上之欲

the father’s orders (Zuo 2.16) the wishes of the lord and the superiors (HFei 36)

Objective genitive, in which the modifier logically is the object of the action ex­pressed by the head, as in: 禮之用      the use of rites (LY 1.12) 天下之禍 a calamity for all under Heaven (Guan 64)

Identifying genitive, in which the modifier is identical to the head, as in: 春秋冬夏四時 the four seasons spring, autumn, winter, and summer (Zhuang 18) 耳目之官 the organs ear and eye (Meng 6A15)

Quantitative genitive, in which the modifier specifies a quantity relating to the head, as in: 一車薪 三年之喪

one cartload of firewood (Meng 6A18) three years’ mourning (Zuo 10.15)

Partitive genitive, in which the head is a subset of the modifier, as in: 十九 玄之又玄

nine out of ten (SY 21) the obscurest of obscure (Lao 1)

71

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 6: n o u n p h r a s e s i i

6.3  Nominalization with 者 In the preceding constructions, the particle 之 may be interpreted as having a nomi­nalizing effect, turning verbal elements into parts of noun phrases. The particle 者—perhaps etymologically related to 之—performs a similar function: verb phrases are nominalized when followed by 者. The resulting construction is a noun phrase, typically functioning as subject, object, or predicator. In order to appreciate the function of 者 phrases, recall the phrase 得道之人 in ⑯; this phrase may also be expressed as follows: ⑲ 得道者

VP pt  NP

one who has found the right way

者 zhě, nominalizing pt

Quite a few of these constructions still exist in Modern Standard Chinese: for example, 學者, ‘scholar’ (< ‘one who learns’), 記者, ‘journalist’ (< ‘one who records’), 作者, ‘author’ (< ‘one who composes’), 患者, ‘patient’ (< ‘one who suffers’), 愛國者, ‘patriot’ (< ‘one who loves the country’).

The particle 者, then, is functionally equivalent to 之人: it nominalizes verbal expres­sions in the sense of a person meaning ‘the one/someone who does [something]’. Logi­ cally, it stands for the subject of the ­corresponding sentence. As such, it can nominalize very complex verb phrases, as for example: ⑳ 不好犯上而好作亂 者 Someone who

VP

NP

pt

好 hào, vn, ‘love to’

does not like 而2 ér, conj, ‘and, but’ 犯 fàn, vn, ‘offend against’ to offend against superiors but likes to cause trouble

While the majority of such constructions are to be understood ad personam, some must be con­strued ad rem, meaning not the person who does something but the thing that does something or even the fact that something is done:

72

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

6.4 e x e r c i s e s



㉒ 為善者得道, 為惡者失道。

㉓ 賢者在位,能者在職。

㉔ 宋人有得玉者。

㉕ 殺人者,臣之父也。

㉖ 為此詩者︙知道乎。

㉗ 如是者再三。

㉘ 殺人者死,傷人者刑。

㉙ 明主…安天下之危者也。

知我 者 ⟨people’s⟩ understanding me VP pt    NP

惡2 è, nc, ‘evil’ 失 shī, vtr, ‘lose’ 賢 xián, vst, ‘be worthy’ 在 zài, vn, ‘be (in, on, at) ’ 位 wèi, nc, ‘position’ 職 zhí, nc, ‘office’ 玉 yù, nc, ‘jade’ 如 rú, vn, ‘be like’ 再 zài, vi, ‘be two’ 傷 shāng, vtr, ‘hurt’ 刑 xíng, nc, as vtr, ‘punish’ 明 míng, vst, ‘be clearsighted’ 安 ān, vst, ‘be peaceful’ 天下 tiānxià, nc, ‘all under heaven’ 危 wēi, vst, as nc, ‘danger’

Comments ㉒–㉓ The structure of these clauses is clearly indicated by the position of 者: start from there with your analysis. ㉔ A clause with a locative S. The formula 有. . . 者, ‘there was someone who. . .’ is rather frequent and very helpful for analysis: everything between 有 and 者 is nominalized. ㉗ For 再三, cf. Box 3. ㉙ With long 者 phrases, the problem is to determine exactly how far back the phrase reaches. Remember that there needs to be at least one verb included in it, since 者 nominalizes verbal expressions. 6.4 Exercises Translate the following expressions into Classical Chinese noun phrases, using 之 and 者 as appropriate.

73

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 6: n o u n p h r a s e s i i

㉚ The ruler’s mother and wife ㉛ A king who hurts his people ㉜ Someone who does not like to kill people ㉝ The fish of the eastern sea ㉞ Someone who can be cheated ㉟ A benefit for all under Heaven ㊱ Someone who is good at teaching rituals References  ① LY 11.8  ② ZGC 19.5  ③ Erya 4  ④ Meng 6B2  ⑤ Meng 1B12  ⑥ Meng 5B6  ⑦ Zuo 10.20  ⑧ HFei 49  ⑨ LY 12.5 Zhuang 18  ⑪ SJS 6  ⑫ Meng 1B2  ⑬ Zuo 9.30  ⑭ ZGC 5.15  ⑮ HFei 49  ⑯ Lü 2.2  ⑰ Lao 3  ⑱ HFei 14  ⑲ Meng 2B1  ⑳ LY 1.2  ㉑ Meng 3B9  ㉒ SY 16  ㉓ Meng 2A4  ㉔ XX 7  ㉕ HSWZ 2  ㉖ Meng 2A4  ㉗ HFei 32  ㉘ Xun 18  ㉙ Guan 64

74

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 5

Personal names The variety of personal names encountered in Classical Chinese ­literature may seem perplex­ing to students of the language. Even for experienced readers, names may present consider­able problems, which is why they are usually marked by lines in modern edi­tions. But even with this reading aid, analysing personal names remains a challenge be­cause of the complicated system underlying them. The following para­graphs will introduce the most common types of personal names and their usage. Given names (名): Every child received a given name three months after birth (at least in theory, cf. Li 28: 三月之末 . . . 父執子之右手, 咳而名之。 ). There was no circumscribed canon of given names; in ­theory, any word or phrase could serve this function. In practice, the given name would often relate to some event surrounding the birth of a child. For example, since Confucius’ mother had prayed for a child on Ni hill, 尼丘, Confucius was given the name Qiu 丘 (‘Hill’), after his birth; and Confucius himself, having received a carp as a present from the duke of Lu upon the birth of his son (cf. Figure 4), called the latter Li 鯉 (‘Carp’). Some of these given names, like Chong’er 重耳 (‘Swol­len Ear’), Wu­sheng 寤生 (‘Breech Birth’), or Hei­tun 黑臀 (‘Black Ass’), sound rather unflattering: these were apotropaic names, meant to avert evil influences (besides referring to physical character­istics the babies may have ­displayed at birth). Unless their semantic value is specifically at issue, Chinese personal names should simply be tran­scribed. They should not be translated: this would often give them an awkward sound which they prob­ably did not have in the ears of the contemporaries. Exceptions may be made for posthumous names (cf. below), since they were specifically meant as epithets.

Courtesy names (字): Upon coming of age, youths were given a courtesy name (cf. Li 1: 男子二十冠而字): this is the name they are usually referred to by people outside their family. These courtesy names typically include either a honorific element—Zi 子, as in Zilu 子路, 75

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 5: p e r s o n a l n a m e s

Figure 4  Confucius receives a carp from the Duke Source: Hans Stumpfeldt. Das Leben des Konfuzius: Bilder zu den Taten des Weisen. Zürich: Manesse Verlag, 1991, fig. 13.

being the most frequent—or an element indicating the rank of birth order (later called paihang 排行): thus 伯 means the first-­born, 仲, the second-­born, 叔 the third-­born, and 季 the young­est; moreover, the first-­born of another than the main wife was called 孟. Significantly, these birth rank elements were given not at birth but only when it was clear that the child had survived (cf. Li 3: 幼名,冠字,五十以伯仲,死謚,周道也, which, however, seems greatly exaggerated). Thus they do not indicate an order of all children born by a mother, but only of those who lived to adulthood. Presumably for the same reason, the given name was only chosen after the baby had survived the first three months.

Another characteristic of courtesy names was that they bore a semantic relationship to the given name. Thus Confucius’ courtesy name was 仲尼 (‘Second-­born Ni’), Ni being the name of the ‘Hill’ that was his given name; and his somewhat unfortunate son received the courtesy name Boyu 伯魚 (‘Eldest Fish’). Family names (姓): Family names only became widespread in the late Chunqiu and Zhanguo periods, when former clan or lineage names (cf. below) as well as toponyms (esp. polity names like Zhou 周, Zheng 鄭, etc.), titles (e.g. Gongsun 公孫, ‘duke’s grand­son’, or Sima 司馬, ­‘marshal’), elements of cour­tesy names (like Meng 孟, or Kong 孔), or occupational names (e.g. Bu 卜, ‘diviner’, or Qidiao 漆雕, ‘lacquer carver’) turned into family names. Altogether, there were not more than 76

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 5: p e r s o n a l n a m e s

a few hundred family names in com­mon use. Like in MSC, they preceded given or courtesy names, as in 司馬遷 (this is why one should not call them ‘last names’). Note that titles and occu­pations also precede given or courtesy names, as in 周公旦, ‘Dan, duke of Zhou’, or 輪扁, ‘wheelwright Bian’. In some cases, it is hard to decide whether a title or an occu­pation is still meant as such or whether it has already turned into a last name.

Clan and lineage names (姓 and 氏): There were about a score of aristocratic clans in Western Zhou and Chunqiu times with names like Ji 姬, Jiang 姜, Yao 姚, Gui 媯, Ren 任, or Ying 嬴. These were normally used for women only, thus 姜氏 was a Jiang lady. This was important because of the rule of exogamy: aristocratic women were allowed to marry only outside their clan (but, of course, within the aristocracy). Compliance with this rule was then readily recognizable by their clan names.

Aristocratic men would go by their lineage name, thus 季氏 is ‘Mr. Ji’, the head of the Ji lineage. However, both clan and lineage names fell into disuse with the social trans­formations of Zhanguo times: henceforth, 姓 came to mean ‘family name’. Posthumous names (謚): The rulers of Zhou as well as the feudal lords and the Han emperors are usually referred to by their posthumous names, as in 武王, ‘king Wu’, 晉文公, ‘duke Wen of Jin’, of 漢武帝, ‘emperor Wu of Han’. Like ‘Alexander the Great’ and ‘Ivan the Terrible’, these epithets always express a characterization of the rulers mentioned: Wu means ‘Martial’, Wen ‘Civil’, etc. The order is always: polity name, post­humous name, title. References: The classic study on given and courtesy names is Wang Yinzhi 1985b, juan 22–3 (春秋名字解詁), which was followed up by Zhou Fagao 1968; for personal names in ­general, cf. Bauer 1959; Gass­mann 2006; Wilkinson 2018, 119–54.

77

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 7

Adverbial modification

7.1 7.2 7.3 7.4 7.5 7.6 7.7

Introduction78 Suppositions and questions S  其 P  S 何 P 79 Distributives S  皆 P 80 Noun phrases S  NP P 81 然 phrases S  X然 P 83 Subordination with 而 S A 而 P 84 Exercises 87

7.1 Introduction Having introduced the central elements of verbal sentences—subject, predi­ cator, and objects—this lesson continues the discussion of adverbial modifi­ cation, which was begun in Lesson 4. It introduces various ways in which adverbial modifiers may contribute to the meaning of the clause, namely: – by expressing modality, i.e. the speaker’s attitude concerning cer­ tainty, probability, possibility, or necessity of an event: especially slight doubts and suppositions will be discussed here; – by forming questions with the help of interrogative pronouns that function as adjunct; – by providing information about time, place, degree, manner, in­ strumentality, or scope of the action through distributives or noun phrases as adjuncts; – and through complex adverbial constructions that employ the verb 然 or the conjunction 而. All these adjuncts are placed in medial position, between subject and predicator. When several adverbial modifiers occur within one clause, 78

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

7.2 s u p p o s i t i o n s a n d q u e s t i o n s

their position in relation to one another follows certain rules, which will be discussed in the following sections. Preliminary questions: Before reading the following examples, think of how adverbial modification works in your mother tongue: come up with examples indicating place, time, manner, degree, means, and per­ haps other circumstances. 7. 2  Suppositions and questions Classical Chinese possesses several words that are used adverbially to  express uncertainty: suppositions, slight doubts, speculations, even questions—or modest proposals, e.g. 其, ‘probably, surely’, 殆, ‘perhaps, likely’, 蓋, ‘presumably, ought to’, and 庶幾, ‘may, hopefully’. Among these, the most important is 其. Note that this is another word than the personal pronoun written with the same character. The two are easily distinguishable, since the adverb always occurs between the subject and the predicator. ① 吳 其 亡 乎。 Wu will probably be

S  A P A

destroyed.

吳 Wú, np, a polity 其2 qí, adv, ‘probably’ 亡 wáng, vtr, ‘destroy’

其 regularly goes together with the ptint 乎 in final position, although such clauses are not interrogative but suggestive. The effect of 乎 may be compared to the high pitch in English at the end of questions and other utterances that are open or suggestive, as in ‘She was looking háppy (wasn’t she)’. In order to express questions, the interrogative pronoun 何 (which we have already discussed as an object in questions, cf. 5.3.2) and some  other interrogative pronouns may be employed as adjuncts, meaning ‘why, how’. In this function, 何 usually expresses rhetorical questions. ② 吾 何 畏 彼 . . .

S A P O

Why should I fear him?

79

畏 wèi, vn, ‘fear’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 7 : a d v e r b i a l m o d i f i c at i o n

In such sentences, the adverbial function of 何 is clearly recognizable since there is an object following the predicator: therefore, 何 cannot be the object. ③

君其勿許。



君其圖之。



殆必不可。



王庶幾改之。



仁人固如是乎。



我何愛一牛。

其誰不知。



許 xǔ, vtr, ‘agree’ 圖 tú, vn, ‘plan’ 殆 dài, vst, as adv, ‘perhaps’ 庶幾 shùjī, adv, ‘may’ 改 gǎi, vi, ‘change’ 牛 niú, nc, ‘ox’

Comments ④ 其 sometimes has a hortatory sense: ‘you had better. . .’ ⑤ A paradoxical combination of adjuncts amounting to ‘it would cer­ tainly seem to. . .’ Note that 可 is not aux but P. ⑥ 庶幾 often expresses a hope or expectation. ⑦ 固, which usually asserts something that is well known (cf. its usage in nominal clauses, 3.3.2), expresses a doubtful ‘really’, ‘seriously’ when used in questions. ⑧ ‘How could I. . .’ , meaning: ‘I would never’. The numeral 一 often does not simply mean ‘one’, but ‘a single’. ⑨ This is an exception from the usual position of modal 其: if the S is a print, it almost always precedes the latter. Can 其 be mistaken for the prper in such cases? 7.3 Distributives Distributives are adverbs that indicate the scope of the action. They typically refer to a plural subject, indicating which parts of it—‘all’, ‘some’, ‘none’, or the like—relate to the predicate. Less often, they may refer to a plural object, expressing which parts of it are affected by the predicator. In any case, they are placed in medial position, between sub­ ject and predicator: ⑩ 人 皆 有 兄弟。Men all have brothers.

S A P O

80

皆 jiē, adv, ‘all’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

7.4 n o u n p h r a s e s

As the example shows, 皆 functions much like MSC 都. Note that in sentences with a deleted subject, distributives may appear in initial position—yet they are syntactically not the subject. ⑪

或 問 子產。Someone asked about Zichan. ⟨S⟩ A P O

或 huò, adv,  ‘someone’ 子產 Zı ̌chǎn, np

Several distributives were morphologically derived by the addition of a suffix *-k: 有 *wəɁ > 或 *wə̂k, 無 *ma > 莫 *mâk, 誰 *dui > 孰 *duk, 舉 *kɨa > 各 *kâk. 獨 *dôk also ends on *-k, which seems to have been a ‘distributive suffix’ (Schuessler 2007, 70). The notable exception is 皆 *krî. Cf. Harbsmeier 1981, 78–87.



諸侯各愛其國,

不愛異國。



宋人或得玉。



百姓莫敢不順

上之法。



禮與食孰重。

一民莫非其臣也。



諸侯 zhūhóu, nc, ‘feudal lords’ 各 gè, adv, ‘each' 異 yì, vst, ‘be different’ 莫 mò, adv, ‘nobody’ 禮 lı ,̌ nc, ‘rites’

孰 shú, adv, ‘which one?’ 重 zhòng, vst, ‘be important’

Comments ⑫ Whereas 皆 expresses that the S all do the same, 各 indi­ cates what each one of the S does individually. ⑬ 或 means ‘one of ’ or ‘some of ’ the S.  ⑭ Distributives always precede auxiliaries and neg­ ations. ⑮ 孰 is etymologically related to 誰; however, it is not a pronoun but a distributive: it means ‘which (of the S)’ and may refer to both ­people or things. ⑯ This is a rare example of a distributive in a nominal clause. The S is a singular, but a collective singular: ‘the entire people’. Note also the double negation (litotes), which, like in English, expresses a strong assertion. There are several CC words—兼, 遍, 周, 舉, 盡, and others—that may be used adverbially to indicate that all objects are affected by the action expressed by the P.  Since these do not constitute a distinct class of adverbs, they are considered separately in Box 30.

7.4  Noun phrases Just as verb phrases may serve as modifiers of nouns (cf. 2.4.3), so noun phrases may function as adjuncts to modify verbal predicates. Nouns, 81

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 7 : a d v e r b i a l m o d i f i c at i o n

used as adjuncts, may serve to express an object, a place, a time, or an instrument. They are always placed before the predicator. ⑰

夜 行 晝 居。 ⟨Foxes and panthers⟩ ⟨S⟩ A P A P roam at night and rest at daytime.

夜 yè, ntemp as adv, ‘at night’ 晝 zhòu, ntemp as adv, ‘at ­daytime’ 居 jū, vi, ‘rest’

A case that deserves to be singled out is that of nominal adjuncts used to make a comparison, expressing that the subject does something like the nominal adjunct: ⑱ 豕 人 立。The pig stood up like a man.

S A P

豕 shı ̌, nc, ‘pig’ 立 lì, vi, ‘stand’

Such expressions may be loosely compared to nominal determination of adjectives in English (‘dog-­tired’, ‘ramrod straight’) or German (‘spin­ nefeind’, ‘pudelnackt’); they live on in MSC expressions like 形影不離, ‘inseparable as body and shadow’. ⑲

君子

尸居  . . . 

而龍見。



商君車裂。

主人疏食水飲。



尸shī, nc as adv, ‘like a corpse’ 龍 lóng, nc as adv, ‘like a dragon’ 見2 xiàn, vi, ‘appear’ 商 Shāng, np, a place 裂 liè, vtr, ‘rip apart’ 疏/蔬 shū, nc as adv, ‘on grain’ 飲2 yı ̌n, vn, ‘drink’

Comments ⑲ Nominal adjuncts expressing comparisons. ⑳ The ‘legal­ ist’ statesman Shang Yang (390–336 bc), a victim of his own harsh laws, was quartered: the adverbial noun indicates the instrument with which the punishment was administered. ㉑ Here, the adverbial nouns indi­ cate the object; the action thus expressed is habitual: the master ‘lives on grain and water’ as opposed to eating them on one occasion. 82

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

7.5

然 phrases

7.5  然 phrases Whereas simple verbs may be used adverbially, such adverbial expres­ sions may also be formally marked by 然. Although 然 is a word, hence a free morpheme, it may be regarded as functionally equivalent to the adverbial suffix ‘-ly’ in English. ㉒

囂囂然  曰 . . . ⟨He⟩ indifferently said: … ⟨S⟩ P A

囂囂 xiāoxiāo, vi, ‘be indifferent’ 然 rán, vst, as suffix, ‘-ly’ 曰 yuē, vtr, ‘say’

Frequently, descriptive two-­ syllable words are marked by 然. Typically describing emotions or attitudes, such constructions are often difficult to translate precisely. 然 *nan is the fusion of the verb 如 *na, ‘be like, as if ’ and a demonstrative pronoun with a phoneme *n; thus its literal meaning is ‘like that’. Note the survival of adverbial 然 in m ­ odern 忽然, 當然, and other expressions. Incidentally, 然 does not seem to occur after stative verbs.



公怫然怒。



填然鼓之。

卒然相遇。



怫然 fúrán, adv, ‘furiously’ 怒 nù, vi, ‘be angry’ 填 tián, vn as adv, onomatopoetic for the sound of drums 鼓 gǔ, nc as vn, ‘drum’ 卒 cù, adv, ‘suddenly’ 相 xiāng, adv, ‘each other’ 遇 yù, vn ‘meet’

Comments ㉓ The expression 怫然 is so conventionalized that it should be treated as a word, not a phrase. ㉔ Here, 然 serves to mark an acous­ tic impression: 填 *dîn is onomatopoetic. ㉕ 相 is a ‘pronominal adverb’: while always used adverbially, it also stands for the object of the clause (compare the usage of 何 and 誰, 5.3.2). Accordingly, the following verb is to be construed as transitive.

83

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 7 : a d v e r b i a l m o d i f i c at i o n

7.6  Subordination with 而 Just like noun phrases, verb phrases, too, may serve as adjuncts. Whereas they may be asyndetically subordinated to the predicator, they are ­frequently formally subordinated by the conjunction 而, ‘and, but’. The basic form is: ㉖

哭 而 過 市。 Wailing, ⟨she⟩ passed ⟨S⟩ A A P O the market

哭 kū, vn, ‘wail’ 過 guō , vn, ‘pass by’ 市 shì, nc, ‘market’

Sometimes such phrases can simply be translated as ‘do A and B’. But it should be borne in mind that the first verb phrase is not coordinated but subordinated to the second. Only the second is the predicator, while the first expresses a subsidiary action which precedes, accompanies, or conditions the main action; it may often be translated participially, as in ‘Wailing’. In any case, it may be removed without leaving an ungram­ matical utterance: *過市 still is a syntactically complete sentence. William Boltz has suggested that 而 *nə is related to 二 *nis, hence its etymology may be explained as ‘second, next’ > ‘then’. This would make clear that two actions connected by 而 are not simply coordinated but subordinated: the first action precedes, and then the second follows (cf. Boltz 1987). Alternatively, 而 may be explained as an unstressed form of the verb 如 *na, ‘be like, as if ’. As such, it does not precede the word it refers to but is enclitically appended to it, meaning ‘-like, -ly’. Cf. also the note at 7.6.

More complex verb phrases, consisting of a verb with an object or other elements, may also serve as adjuncts. In such cases, they must be formally subordinated by the conjunction 而: ㉗

鳥 聞之 而 飛。 S A A P

The birds heard it and flew

Again, both adjuncts may be removed, leaving the main message of the sentence intact: 鳥飛. In many cases, the expression subordinated by 而 is not concomitant with but contrary or antithetical to the ­predicator. In such cases, an adversative word should be chosen in translation:

84

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

7.6 s u b o r d i n at i o n w i t h



⟨S⟩



不教民  而 用 之。⟨If one⟩ employs the people, although A P O one has not instructed them… A

Note that in such constructions the negation 不 is part of the ­subordinated verb phrase. Hence, it does not affect the predicator but only the adjunct. As these examples indicate, 而 can have a variety of inferred meanings. These may be com­ pared to the uses of English ‘and’, as quoted in Downing and Locke 2006, 291: ‘I made the sandwiches and Jill made the salad. . .’ (simultaneity); ‘He got dressed quickly and went out. ’ (temporal sequence); ‘He was found guilty of harassment and was dismissed from his post. ’ (cause–effect); ‘You give me your telephone number and I’ll give you mine.’ (condition); ‘She came to my house and I was out.’ (inclusion: time ‘while’).



醉而臥。



保民而王。



君子和而不同,

小人同而 不和。



余聞而弗忘。

不出戶而知天下。



醉 zuì, vi, ‘be drunk’ 臥 wò, vi, ‘fall asleep’ 保 bǎo, vtr, ‘protect’ 王2 wàng, vi ‘rule (as king)’ 和 hé, vn, ‘be harmonious’ 同 tóng, vi, ‘be the same’ 忘 wàng, vn, ‘forget’ 出 chū, vn, ‘go out’ 戶 hù, nc, ‘gate’

Comments ㉙ This example well illustrates that the first VP is not coor­ dinated but subordinated: the order of the VP cannot simply be changed—*臥而醉—without distorting the meaning of the sentence. ㉚ Observe that since it is adverbially modified, 王 must be construed as a verb (pronounced in the fourth tone), ‘rule as king’. Such usage is not uncommon with nouns denoting titles, kinship terms, and the like; cf. Box 10. ㉛ In this case, 而 is best translated as ‘but’. ㉜ If both verbs have the same O, it suffices to realize the Opr 之 once, after the P. Where is it in this sentence? ㉝ Note, again, the limited scope of the negation: it does not apply to the P.

85

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 7 : a d v e r b i a l m o d i f i c at i o n

Box 10  Word class flexibility Having introduced Classical Chinese word classes and their typical syntactic behaviour in 1.2, we must now make some important additions. In principle, every verb may appear as subject or object, that is it may function the same way nouns do. Moreover, most nouns (but not pronouns or ntemp) may in principle function as predicators in verbal clauses. Consider the following examples: 今周君天下。 楚不國矣。

Now Zhou rules the realm. (ZGC 1.12) . . . Chu would no longer be a state. (Zuo 12.16)

If this usage seems perplexing, recall that there are examples of this in English as well, as in ‘He childed as I fathered’ (Shakespeare, King Lear 3.6), or ‘Don’t sir me!’ (J.  Heller, Catch-­22), or, more prosaic, ‘Bag it, bin it’. But in Classical Chinese, such flexible usage is much more common. Especially titles and nouns that express social status may freely be used as verbs, as in: 君君,臣臣,父父,子子。 . . . 不亦君子乎。

Let the ruler be ruler, let ministers be ministers, fathers be fathers, and sons be sons. (LY 12.11) . . . is that not gentlemanly? (LY 1.1)

In these examples, nouns function like stative verbs; just like the latter, they may be used in a causative sense. Such usage also applies to names, as the last examples show (compare English ‘to google sth.’ or ‘to guillotine s.o.’): 若必師之。 不法先王 . . . 爾欲吳王我乎。 居楚而楚,居越而越 . . .

You must take him as a teacher. (SJ 47) Not to take as a model the former kings. . . (Xun 6) Do you want to make me a ‘King of Wu’ ? (Zuo 11.10) Living in Chu one behaves in Chu manner; living in Yue, one behaves in Yue manner. . . (Xun 8)

In most cases, these different functions do not correspond to morphologi­ cal distinctions, thus they cannot be regarded as conversions to a new word class (as in English ‘live a good life’) but simply as different functions of the same word. 君, 臣, 君子, etc. are all nouns, but they are being used like verbs in the above sentences. Hence, they are listed as nouns in the glossary of this book; their verbal usage is noted in sub-­entries. Only cases like 王 wáng, ‘king’, vs. 王 wàng, ‘rule as king’, where there is a morphological distinction, are treated as different words in separate entries. Of course, this does not mean that Classical Chinese words could be used arbitrarily in any way an author chose; if anything was possible, Classical Chinese would not have functioned as a language. Rather, word usage was governed by rules: certain nouns, when used verbally, behave like neutral verbs, others like transitive verbs, etc.; and semantic restrictions always applied. In other words: only what was possible was possible. References: For a study of ‘Word-­Class Flexibility’, cf. Zádrapa 2011; cf. also Cikoski 1978a, 46ff.

86

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

7.7 e x e r c i s e s

7.7 Exercises Analyse the following sentences, especially the adverbial modifications they contain. ㉞

吾其何得。



民莫敢不敬。



老人兒啼。



眾人皆醉,我獨醒。



鵠不日浴而白。



顏淵喟然歎。

殺孔父而取其妻。



References  ① SY 9  ② Meng 3A1  ③ Zuo 5.7  ④ SJS 1  ⑤ Zuo 10.25  ⑥ Meng 2B12  ⑦ Meng 5A3  ⑧ Meng 1A7  ⑨ Zuo 5.32  ⑩ LY 12.5  ⑪ LY 14.9  ⑫ Mo 14  ⑬ Zuo 9.15  ⑭ Xun 12  ⑮ Meng 6B1  ⑯ Meng 2A1  ⑰ Zhuang 14  ⑱ Zuo 3.8  ⑲ Zhuang 11  ⑳ HFei 13  ㉑ Li 22  ㉒ Meng 5A7  ㉓ HFei 34  ㉔ Meng 1A3  ㉕ Wu 5  ㉖ Zuo 6.18  ㉗ Zhuang 12  ㉘ Meng 6B8  ㉙ HFei 19  ㉚ Meng 1A7  ㉛ LY 13.23  ㉜ Zuo 10.3  ㉝ HSWZ 3  ㉞ Zuo 9.28  ㉟ LY 13.4  ㊱ SJ 119  ㊲ ChuC 7  ㊳ Zhuang 14  ㊴ LY 9.11  ㊵ Zuo 2.2

87

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 6

Shujing The ‘Canon of Documents’, Shujing 書經, is perhaps the most venerated text in the Confucian canon; indeed, it was called ‘Exalted Documents’, Shangshu 尚書, as early as Han times. The awe the Documents inspired was due to their purported age: it was taken for granted that they derived from high antiquity, the oldest chapters going back to pre-­dynastic—i.e. neolithic!—times, the youngest dating from 627 bc. Legend has it that no lesser man than Confucius himself chose 100 documents from a corpus of over 3,000 and edited them. Indeed, they seem to have been an integral part of traditional lore in Zhanguo times. Thinkers of various schools would quote short snippets from ‘documents’, sometimes with a specific title, to bolster their arguments with time-­honoured wisdom. Despite such quotations, there is little evidence to suggest that the ­‘documents’ were edited in a coherent ‘Canon’ before imperial times. Indeed, the ‘documents’, having fallen victim to the burning of books by the first emperor in 213 bc, seem to have faded into obscurity by the beginning of the Han: ‘When emperor Xiaowen (r. 180–157 bc) looked for someone who could master the Shangshu, there was no such one in the empire’ (SJ 121). Finally, a hoary scholar from Shandong, Fu Sheng 伏生, over ninety years of age, presented twenty-­nine ‘documents’ that he had hidden in his wall. Among these were the ‘Yao dian’ 堯典, a record of the legendary Yao’s deeds and his succession, ‘Yu gong’ 禹貢, containing the first geographic description of the Chinese realm, ‘Lü xing’ 呂刑, an early treatise on penal law, five political ‘proclamations’ (誥) ascribed to early Zhou rulers, the ‘Hong fan’ 洪範 that lays out cosmological foundations of political philosophy, and ‘Qin shi’ 秦誓, an address by duke Mu of Qin to his officers, and others. Henceforth, these texts became the canonical foundation of Chinese political thought as well as a trusted source of ancient Chinese history. Until the twentieth century, Chinese scholars adduced the Shangshu as evidence for exemplary rulership. However, the problems that beset the Shangshu are manifold. Some years after Fu Sheng, another version of the Shangshu appeared, written 88

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 6: s h u j i n g

in ‘old script’ (古文) and containing forty-­four chapters; and in the following years, more fragmentary and complete versions followed. The earliest Chinese bibliography, compiled about 90 ad, lists ‘nine schools of the Documents, totalling 412 chapters’ (Hanshu 30). These traditions were lost by the early fourth century ad. But the Shangshu resurfaced shortly thereafter, when an alleged ‘old script’ version with fifty-­eight chapters was presented to the Jin emperor Yuan (r. 318–323). Its proven­ ance is dubious, and indeed, the ‘old script’ chapters that were not part of Fu Sheng’s text have been shown to be forgeries fabricated by ­commingling early quotations and other fragments associated with the ‘documents’. If the history of the Shangshu’s transmission is convoluted, so is that of its commentaries. Already in classical times, its arcane—or perhaps: archaizing—language eluded Chinese scholars. In early Han times, there was no one in the empire ‘who could master the Shangshu’, when new versions of the text appeared, ‘there was none among the academicians and courtiers who could understand them’ (Lunheng 20.61), and when Sima Qian used some Shangshu pieces for his history of early China in c.90 bc, he actually translated them into the language of his day. Compare the following excerpt from the ‘Yao dian’ with the Shiji rendering: ‘Yao dian’ Shiji 1 帝曰,疇咨若時,登庸。 堯曰,誰可順此事。… 放齊曰,胤子朱啟明。 放齊曰,嗣子丹朱開明。 帝曰,吁。嚚訟,可乎。 堯曰,吁。頑凶,不用。 帝曰,疇咨若予采。 堯又曰,誰可者。… 驩兜曰,都,共工方鳩僝功。 讙兜曰,. . . 共工旁聚布功,可用。 帝曰,吁,靜言庸違, 堯曰,. . . 共工善言,其用僻, 象恭滔天。 似恭漫天,不可。 To be sure, the ‘Yao dian’ is an extreme case. Already in Han times, commentaries to this short chapter ran into hundred thousands of words, and the first four characters alone—曰若稽古—supposedly elicited 300,000 words of commentaries. Nor was this the end of Shangshu philology. For centuries, commentaries were heaped upon commentaries, virtually burying the text underneath them. The Shangshu is a prime example for the symbiotic relationship of canon building and commentary—and yet, it was never properly understood. 89

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 6: s h u j i n g

Even in the twentieth century, Bernhard Karlgren considered the Shangshu ‘a collection of exceedingly difficult and, largely, very obscure texts’, doubting ‘whether it will ever be possible definitely to determine its correct form and interpretation’; and a towering scholar like Wang Guowei (1877–1927) plainly confessed that he did not understand half of the Shangshu’s texts. For more than 2,000 years, the Documents have been venerated despite—or, perhaps, precisely because of—their obscurity: ‘The reader doesn’t know exactly what they mean, but they have all the more on that account an air of sublimity; an agreeable feeling at once of awe and of exaltation comes over him as he contemplates thoughts of so immeasurable a profundity—their profundity being convincingly evidenced to him by the fact that he can see no bottom to them’ (A. Lovejoy): omne ignotum pro mirifico. Editions: Shangshu zhengyi 尚書正義, ed. by Li Xueqin 李學勤. Taipei 2001, is an updated version of Ruan Yuan’s 阮元 edition of 1816; Jinwen Shangshu kaozheng 今文尚書考証, ed. by Pi Xirui 皮錫瑞. Beijing 1998, contains only the ‘authentic’ new text chapters. Translations: There are three main translations into Western languages, none of which is satisfactory: Legge 1865; Couvreur 1897; and Karlgren 1950. Studies: Pelliot 1916; Karlgren 1948/49; Chen Mengjia 1985; Jiang 1988; Liu Qiyu 1989; Nylan 2001, 120–67; Vogelsang 2002[a]; Kern and Meyer 2017.

90

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 8

Prepositional phrases I

8.1 Introduction91 8.2 Phrases with 於, 于, or 乎92 8.2.1 General usage S P (O) 於X 於X S P (O)93 8.2.2 Comparisons S P 於X 93 8.2.3 Passive voice S P 於X 94 8.2.4 Object and prepositional phrase S P O 於X 94 8.2.5 Fusion words S P 焉 S 焉 P 95 8.3 Phrases with 自 or 至 98 8.4 Exercises 98

8.1 Introduction Besides noun and verb phrases, one may identify a third phrase category in Classical Chinese: prepositional phrases. As the name indicates, these phrases centre around a preposition, accompanied by a noun phrase (which may be regarded as its ‘object’). Compare English ‘on the table’ or Modern Standard Chinese 在北京. Since prepositions are actu­ ally ‘prepositional verbs’, originating from full verbs (cf. 1.2.4), pre­ positional phrases occur as constituents in verbal clauses only. Functionally, they serve as adjuncts, being supplementary to the ­predicator and adding to its meaning. They may occur in medial, final, or—less frequently—initial position. In contrast to verb phrases, ­prepositional phrases cannot be used in any ‘nominal’ function such as subject, object, or complement, and they cannot be modified. This lesson introduces phrases centred on some of the most important prepositions—the general ‘locative’ prepositions 於, 于, and 乎, as well as the ‘ablative’ preposition 自 and the ‘terminative’ preposition 至, 91

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 8: p r e p o s i t i o n a l p h r a s e s i

which both indicate spatial and temporal relations—and discusses their syntactic properties. For discussions of these and other prep (treated in Lesson 9), cf. Pulleyblank 1986; Unger 2019, 666–911; Wu Sue-mei 1997; Yang Bojun and He Leshi 2003, 376–452.

Preliminary questions: Do you know what the words ‘locative’, ‘ablative’, and ‘terminative’ mean? Which prepositions correspond to these notions in your mother tongue and in Modern Standard Chinese? 8.2  Phrases with 於, 于, or 乎 Of the three locative prepositions 於, 于, and 乎, which are basically synonymous, 於 is by far the most common in classical texts. 于, on the other hand, is an older word from pre-­Classical Chinese which survived as an archaizing element typically used in formal speech, especially in the texts Zuozhuan and Guoyu; and 乎 is an unstressed form of 於. All three primarily indicate local relations: ‘in, on, at’, and so on. But beyond this, they may express temporal and a variety of other relations which call for different translations. Most commonly, prepositional phrases with 於, 于, or 乎 take final position. Less frequently, they may occur in initial or medial position, but never between the predicator and an object (cf. 5.1). The main forms are: ① 天 生 德 於予。

S P O A

Heaven nurtured virtue in me.

② 於是 吳王 起 師。 Thereupon, the king

A

S

 P  O

of Wu raised in army.

於 yú, prep, ‘in, at, on’

起 qı ̌, vi, ‘rise’ 師 shī, nc, ‘army’

With an adjunct in initial position the subject moves to medial position. 于 originally meant ‘go (to a place/to do sth)’, and 於, ‘be (in a place)’. The latter still occa­ sionally functions as a full verb in classical texts (compare MSC 在 in 我在北京 and in 我 在北京工作). As prep, both are still distinct in presumably old texts such as Zuozhuan and Guoyu, but in later texts, 於 takes over the functions of 于. Today the character 于 is the simplified form of 於, but this should not obscure the fact that the two were different words in Classical Chinese, *wa and *ʔa (note that the literary pronunciation of 於 is in the first tone). References: Karlgren 1926; Kennedy 1940; Pulleyblank 1986; Guo Xiliang 1997; Petersen 2001.

92

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

8.2 p h r a s e s w i t h

於, 于,

or



8.2.1  General usage The following examples serve to give an impression of locative usages and some meanings derived therefrom. ③

鳥飛於空,魚游於淵。



王坐於堂上。



立于門外。



天子曰,天降禍于周。



於是陳亂。

居於陋巷。



空 kōng, vst, as nc, ‘air’ 游 yóu, vi, ‘swim’ 淵 yuān, nc, ‘deep water’ 坐 zuò, vi, ‘sit’ 于 yú, prep, ‘in, at, on’

天子 tiānzı ,̌ nc, ‘Son of Heaven’ 降 jiàng, vi, ‘descend’ 禍 huò, nc, ‘misfortune’ 陳 Chén, np, a polity 居 jū, vn, ‘reside’ 陋 lòu, vi, ‘be shabby’ 巷 xiàng, nc, ‘alley’

Comments ④–⑤ The locative prep simply indicate general local rela­ tions (‘in the hall’, ‘at the gate’); in order to specify them (‘up there in the hall’, ‘in front of the gate’), additional local nouns are necessary. ⑥ 于 is typically used in formal or solemn speech like in this pronouncement by the ‘Son of Heaven’, i.e. the Zhou king. ⑦ An example for initial ­position of the PrP. This often occurs when temporal relations are ­indicated, for which 於是, ‘thereupon’, is a common expression. 8.2.2 Comparisons In the absence of inflectional comparative forms of stative verbs, the prepositions 於 and 乎 also serve to express comparisons. A sentence like ‘X 高於 Y’ may be understood as ‘X is high in relation to Y’, hence ‘X is higher than Y’. Naturally, such comparisons can only be done with stative verbs.

93

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 8: p r e p o s i t i o n a l p h r a s e s i



福輕乎羽

 . . .

禍重乎地。

君富於季孫,

齊 大於魯。 . . .



福 fú, nc, ‘good fortune’ 輕 qīng, vst, ‘be light’ 乎2 hū, prep, ‘in, at, on’ 羽 yǔ, nc, ‘feather’ 富 fù, nc as vst, ‘be rich’ 季孫 Jìsūn, np, family name

8.2.3  Passive voice Following passive verbs, 於 and 乎 may serve to indicate the agent of the action, who, of course, is not expressed by the subject. ⑪ 隱公. . . 殺 於弟。 Duke Yin was . . . 

S 

P

A

killed by his brother. ⑫

食 於人。  . . .

必見欺於張儀。



食人

德蕩乎名,

知出乎爭。



隱 Yı ̌n, np, posth. name

張儀 Zhāng Yí, np 食2 sì, vtr, ‘feed’ 蕩 dàng, vtr, ‘shake’ 知2/智 zhì, nc, ‘knowledge’ 爭 zhēng, vn as nc, ‘fight’

Comments ⑬ Both 食人 and 食於人 may be interpreted in two ways— which ones? Consider the different verb classes! ⑭ 蕩 is clearly passive, but how about 出? 8.2.4  Object and prepositional phrase A special case of locative meaning occurs with certain verbs that express transfer or communication. Since only some neutral verbs may accom­ modate two objects (cf. 5.2.2), other verbs are followed by a direct object and a prepositional phrase which logically designates the indirect object, namely the person to whom something is transferred or com­ municated. Syntactically, however, these prepositional phrases are still adjuncts: 94

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

8.2 p h r a s e s w i t h

⑮ 寡人 託 國 於子。

S

P Od A

於, 于,

I entrust the country to you.

or



寡人 guǎrén, nc, ‘I, my humble self ’ 託 tuō, vtr, ‘entrust’

Likewise, the prepositional phrase may indicate the person from whom something is transferred or communicated: ⑯ 鄭人 取 貨 於印氏。  The people of

S

P Od

A

Zheng took goods from the Yin lineage. ⑰

堯欲傳天下於舜。



魏委國於王。



齊景公問政於

孔子。

仲弓問子桑伯子。



鄭 Zhèng, np, a polity 印 Yìn, np, lineage name 氏 shì, nc ‘lineage’

傳 chuán, vtr, ‘transfer’ 魏 Wèi, np, a polity 委 wěi, vn, ‘hand over’

景 Jı ň g, np, posth. name 政 zhèng, nc, ‘government’ 仲弓 Zhònggōng, np 子桑伯子 Zı̌sāng Bózı̌, np

Comments ⑰ Yao was famous for passing on the rule not to his son but to a worthy, namely Shun. ⑲–⑳ Observe the usage of 問, which resembles French ‘demander qc à qn’: the O expresses the content of the question, whereas the person asked may be added in the PrP. Can you analyse all the names in these two clauses? 8.2.5  Fusion words When locative prepositions are preceded or followed by pronouns, they are usually contracted to form a single word. Thus 之於 and 之乎 may be contracted to 諸, which stands for a pronoun and a preposition: ‘it in’, followed by the ‘object’ of the preposition. Compare the following ­sentences:

95

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 8: p r e p o s i t i o n a l p h r a s e s i



克 之 於牧野。 ⟨He⟩ defeated  ⟨S⟩ P O them at Muye. A

㉒ 公 伐 諸鄢。

克 kè, vn, ‘defeat’ 牧野 Mùyě, np, a place

The duke attacked them at Yan.

S P OA

諸 zhū, fus, ‘it in/on/at’ 鄢 Yān, np, a place

諸, as a fusion of 之 and a preposition, can be easily distinguished from 諸 as a fusion of 之 and the interrogative particle 乎 (cf. 5.3.1): the former is followed by the ‘object’ of the preposition, in this case 鄢, whereas the latter closes the sentence. Moreover, *於之 or *乎之 are not found to occur; instead, 焉 takes their place, which is a combination of 於 and a pronoun. This fusion expresses different meanings depending on its syntactic position: in final position, it stands for the preposition and a pronoun—‘in it’, ‘on it’, ‘there’, ‘from him’, or the like—and is pronounced yán. In medial ­position, it almost invari­ably stands for the preposition 於 and an ­interrogative pronoun, meaning ‘where?’, ‘who?’, or ‘how?’ In this case, it is pronounced yān. ㉓ ㉔

其君 在 焉。 S P A

Its lord is there.

焉 yán, fus, ‘there’

君 焉 在。 S A P

Where is the lord?

焉2 yān, adv, ‘where’

Note that 焉 cannot be phonologically explained as a contraction of a preposition and 之. Rather, a (demonstrative?) pronoun ending on *-n must be involved. Cf. Kennedy 1940, 1953; Goldin 2003.



子張書諸紳。



吾聞諸夫子。



邑于岐山之下,

居焉。



余焉能戰。



孔子問焉。

未知生,焉知死。



子張 Zı ž hāng, np 紳 shēn, nc, ‘sash’ 邑 yì, nc, as vi, ‘found a city’ 岐山 Qíshān, np, a mountain 戰 zhàn, vi ‘fight’ 未 wèi, adv, ‘not (yet)’

Comments ㉕ Observe that 諸 here functions as Od and part of the A.  子張 Zizhang, a disciple of Confucius, perhaps noted the words of his 96

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

8.2 p h r a s e s w i t h

於, 于,

or



master on a bamboo slip or the like which he then placed in his sash. ㉗ Mount Qi, in present-­day Shanxi, is considered the homeland of the Zhou. Indeed, numerous bronze vessels found in the area testify that it was a cultural centre in Zhou times. ㉚ On the negation 未, cf. 10.4. The second clause is a rhetorical question: ‘how could. . . ?’ Box 11  Pseudo-­transitivity Quite a few of the constructions with 於, 于, or 乎 allow an alternative para­ phrase in which the prep is deleted (compare English ‘walk on the beat’ vs. ‘walk the beat’, ‘jump over the turnstile’ vs. ‘jump the turnstile’, etc.). Especially in later Classical texts, these prep are often dropped. Consider the following pairs: 諸侯朝於天子 . . . 孟子將朝王。 行於道路。 行路 。

The feudal lords had an audience with the king . . . (Meng 1B4) Mengzi was about to have an audience with the king. (Meng 2B2) ⟨A man⟩ walks on the streets and roads. (Li 21) ⟨He⟩ walked on the road. (Lü 9.2)

Without the prep, the P—朝 and 行—appear to be followed by an O, although what follows is, in fact, still an A. This follows from the fact that 朝 and 行 are both vi: if they were followed by an O, they would have to be understood in a causative sense, which is clearly not the case in these examples. We will therefore call such construc­ tions pseudo-­transitive. Such pseudo-­transitive usage may also occur with vtr: 士勸於勇。 農夫勸其事矣。

The knights are encouraged in their audacity. (Guan 18) The peasants are encouraged in their efforts. (Guan 81)

Again, both constructions should be analysed in the same way: the P is not followed by an O but (contrary to appearances in the second clause) by an A, hence it is passive. Moreover, consider the following pairs: 齊景公問政於孔子。 Duke Jing of Qi asked Confucius about government. (LY 12.11) 景公問政孔子。 Duke Jing asked Confucius about government. (SJ 47) 晉侯復假道於虞。 The marquis of Jin repeatedly borrowed a passage from Yu. (Zuo 5.5) 不好假道人。 ⟨He⟩ does not like to borrow methods from others. (Xun 9)

The second clause of each pair appears to involve ditransitive usage of the P, in which an Oi (孔子, 人) appears to follow the Od. However, recall that the word order in ­ditransitive constructions is the opposite: the Od follows the Oi (cf. 5.2.2). In fact, the above constructions are pseudo-­ditransitive: 孔子 and 人 are not O at all but must be construed as A. Since real ditransitive usage is only possible with vn, cases involving vtr—compare, for example, 錯 諸地 (Yi 7) and 措之廟 (Li 2)—are readily identifiable as pseudo-­ditransitive.

As these examples show, some seemingly straightforward S-­P-­O(-­O) con­ structions may be deceptive. Always reckon with pseudo-­transitivity. When in doubt, check the glossary whether pseudo-­tr or pseudo-­tr2 usage is attested for the verb in question.

97

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 8: p r e p o s i t i o n a l p h r a s e s i

8.3  Phrases with 自 or 至 The conversion from verb to preposition is clearly observable in the cases of 自 and 至. As verbs, they mean ‘come from’ and ‘go to’, respectively, and as prepositions ‘from’ and ‘to’, ‘until’. Both can express local and tem­ poral relations. ㉛

其劍自舟中墜於水。



季 孟自南門入,

出自東門。



故人至暮不來。



自十月不雨至于五月。

其兄自外至。



劍 jiàn, nc, ‘sword’ 自2 zì, prep, ‘from’ 墜 zhuì, vi, ‘fall’ 季 Jì, np 孟 Mèng, np 入 rù, vi, ‘go into’ 故 gù, nc, mod, ‘familiar’ 至 zhì, vi as prep, ‘until’ 暮 mù, nc, ‘evening’ 雨 yù, vn ‘rain’

Comments ㉜ PrP with 自 may be placed in final position, especially when the P is a vi. In this sentence the option is used to create a chiasmus. ㉝ 故人 means an ‘intimate’, or ‘friend’. ㉞ ‘From . . . until’. 至 is frequently used together with one of the local prep: its verbal quality is strongly felt in such constructions, and sometimes the distinction between verb and prep is blurred. ㉟ There is only one prep in this clause. 8.4 Exercises Insert a fitting preposition or fusion word next to the dotted line. ㊱

移其民



河東。 ......

何畏 ...... 。



有朋 ...... 遠方來。

鄢。 ...... 98



決 ...... 東方。



鄭伯克段



驪姬問 ...... 。



民 ...... 老死不相

往來。

是國莫敢出言。 ......



OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

8.4 e x e r c i s e s

References ① LY 7.23 ② GY 19.9 ③ Shen 9 ④ Meng 1A7 ⑤ Zuo 7.2 ⑥ Zuo 10.32  ⑦ Zuo 2.5  ⑧ Meng 4B29  ⑨ Zhuang 4  ⑩ HFei 39  ⑪ HSWZ 7  ⑫ SJ 40  ⑬ Meng 3A4  ⑭ Zhuang 26  ⑮ HFei 34  ⑯ Zuo 9.26  ⑰ HFei 34  ⑱ ZGC 25.12  ⑲ LY 12.11  ⑳ LY 6.2  ㉑ Lü 5.5  ㉒ Zuo 1.1  ㉓ Zuo 11.9  ㉔ Zuo 9.30  ㉕ LY 15.6  ㉖ LY 19.17  ㉗ Meng 1B15  ㉘ LY 20.1  ㉙ Li 3  ㉚ LY 11.12  ㉛ Lü 15.8  ㉜ Zuo 11.6  ㉝ HFei 32  ㉞ Zuo 5.3  ㉟ Meng 3B10  ㊱ Meng 1A3  ㊲ Meng 3B5  ㊳ LY 1.1  ㊴ Meng 6A2  ㊵ Zuo 1.1  ㊶ GY 7.6  ㊷ Lao 80  ㊸ GY 1.3

99

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 7

Exemplary rulers In the preceding lessons, we have encountered quite a few references to exemplary rulers: Mengzi praises ‘the principles of Yao and Shun’ (堯舜 之道, Meng 6B2), Han Feizi reports that ‘Yao wanted to pass on the realm to Shun’ (堯欲傳天下於舜, HFei 13), and allegedly, ‘Tang discarded Jie, and king Wu attacked Zhou’ (湯放桀,武王伐紂, Meng 1B15). There are many stories about such ancient or mythical rulers in Classical Chinese literature, who were either paragons of virtue or incarnations of evil. Yao 堯, the first of three legendary sage kings that lived in times immemorial extended his wise and benevolent government over the realm, caring for the people and defeating its enemies. Allegedly, he even had an archer shoot down nine out of ten suns that scorched the earth in his times. Finally, he abdicated the throne not to his son but in favour of a filial and virtuous man from the common people, Shun 舜. Having ruled in equally exemplary manner, Shun in turn abdicated in favour of Yu 禹, who battled victoriously against the floods that inundated the North China plain in his times, thereby saving his people from turning into fish, as a classical text puts it. Yu is said to have been the first ruler who passed on the throne to his son, thus founding the first dynasty, Xia 夏. But his successors strayed from the path of virtue, until the last of them, the utterly debased Jie 桀, was violently ‘discarded’ by the worthy Tang 湯, founder of the next dynasty, Shang 商. However, history repeated itself and the Shang rulers, too, successively fell from grace until the evil last ruler, Zhou 紂, was vanquished by king Wu 武, the founder of the Zhou dynasty. 3rd millennium 2nd millennium 1st millennium Pre-­dynastic   堯 ––> 舜 Xia     ––> 禹 . . . . . . 桀 Shang     湯 . . . . . . 紂 Zhou        武 . . . …

100

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 7 : e x e m p l a ry r u l e r s

The time line indicated should be taken as merely impressionistic. The first three figures are certainly mythical, and the dates of the entire Xia—if it was historical at all—as well as the early Shang dynasty have not been ascertained. Only the very last event, the victory of Zhou over Shang, may be dated with some degree of certainty to the middle of the eleventh ­century bc.

Although the Zhou were still formally recognized as the ‘ruling’ dynasty in the classical age, they had lost power centuries before. In 771 bc, the unfortunate king You 幽 had been killed when the ‘barbarian’ Quanrong sacked his capital, and the Zhou were driven out of their homeland in Shaanxi. Another turn of the dynastic cycle was imminent. Maybe it was the waning of the old order that inspired Chinese ­thinkers to conjure up the virtuous rulers of the past. The more dis­ order­ly the present appeared, the more orderly the past was depicted; the more rational and impersonal political systems became, the more moral and personal ancient rulership was imagined; in fact: ‘the later the times, the older the ancient history that they know; and the less evidence the texts provide, the more they know about ancient history’. This observation was made by Gu Jiegang 顧頡剛 (1893–1980), the father of critical Chinese history. He noted that whereas the Western Zhou knew only of Yu, in Confucius’ times the stories of Yao and Shun gained currency, and in the Zhanguo period the lore of even more remote rulers such as Huangdi 黃帝, the ‘Yellow Thearch’, Shennong 神農, the ‘Divine Farmer’, and Fuxi 伏羲 emerged. China’s ancient history, Gu concluded, grew in successive layers, each new layer extending further back into a distant past—or rather: into the realm of myth. Historical criticism has dispelled the cherished belief in mythical ­rulers that Chinese scholars had held until the twentieth century. It has shaken our trust in the authenticity of pre-­classical and classical works, suggesting that they, too, are layered and composite in character. It has alerted us to the fact that they may not simply be taken as sources of ancient history. Many of the stories the Chinese classics tell were probably invented just like the myths of Yao and Shun: they should be treated as literary artefacts. References: Relevant studies by Gu Jiegang and his colleagues are found in Gu Jiegang 1982, esp. vol. 7, and Gu Jiegang 1996; cf. also Maspéro 1924; Karlgren 1946; Birrell 1993; Friedrich 2004.

101

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 9

Prepositional phrases II

9.1 Introduction102 9.2 Phrases with 以102 9.2.1 General usage S 以X P (O) 103 9.2.2 Disposal constructions S 以X P Oi 103 9.2.3 Anteposed prepositional ‘objects’ S X以 P (O) 104 9.2.4 The 以為 construction S 以X 為 O 105 9.3 Phrases with 為 S 為X P (O) 106 9.4 Phrases with 與 S 與X P (O) 107 9.5 Exercises108

9.1 Introduction This lesson continues the discussion of prepositional phrases. It introduces the ‘instrumental’ preposition 以 , which marks means or instruments; the ‘dative’ preposition 為 , which expresses cause, ­purpose, etc.; and the ‘comitative’ preposition 與 , which signifies accompaniment. Preliminary questions: Do you know expressions with 以 and 為 ­(pronounced wèi) from MSC? What is their function? What does 以為 mean? 9.2  Phrases with 以 以 is the most common preposition in Classical Chinese. Originally, 以 was a verb meaning ‘take’. Presumably from expressions like ‘take X to do Y’, it developed into an ‘instrumental’ preposition: ‘do Y with X’. As such, 以 marks the means or instruments by which something is done, to be translated as ‘with’, ‘by means of ’, ‘through’. Prepositional phrases with 以 usually take medial position: 102

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

9.2 p h r a s e s w i t h



①   以肱 擊 之。⟨He⟩ struck it with his arm. 以 yı ̌, prep, ‘with’

⟨S⟩ A  P  O

肱 gōng, nc, ‘(upper) arm’

擊 jī, vn, ‘strike’

By extension, 以 may also mean ‘by dint of ’, ‘by virtue of ’, ‘because of ’, ‘according to’, and so on. While 以 phrases may be followed by virtually any predicator, the frequent expression 以 X 為 Y deserves special ­mention. 9.2.1  General usage The following examples give an impression of the range of meanings expressed by 以 phrases. ②

以先王之政,治當世之民。



以道觀之,物無貴賤。



以君靈不死。



司馬以吾故亡其良子。

以十月入。



先 xiān, vi, ‘be earlier’ 治2 chí, vn, ‘order’ 當 dāng, vn, ‘face’ 世 shì, nc, ‘era, age’ 觀 guān, vn, ‘look at’ 賤 jiàn, vst, ‘be low’ 靈 líng, nc, ‘grace’ 司馬 sīmǎ, nc, ‘marshal’ 亡2 wáng, vi, ‘be lost’

Comments ② Note the punctuation in this sentence: 政治 is not a NP. ③ Here, ‘according to’ or ‘from the perspective of ’ is meant. ④ Here, ‘thanks to’ best captures the meaning. 靈 cannot be construed as S— why? ⑤ 以 X 故 is a common expression, meaning ‘for X’s sake’ or ‘because of X’. ⑥ Here, 以 is temporal: it typically expresses an appointed time, often in the future, as opposed to 於 (cf. 8.2.1–⑦), which simply indicates when something happened. 9.2.2  Disposal constructions Used with neutral verbs that take two objects, the prep 以 may function to mark the direct object. In many cases, 以 is used to move the direct object from final to medial position, thereby emphasizing it. Compare the following constructions: 103

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 9: p r e p o s i t i o n a l p h r a s e s i i

⑦   賜 母 金布。⟨He⟩ presented bronze

⟨S⟩ P  Oi   Od

and cloth to ⟨his⟩ mother.

布 bù, nc, ‘cloth’

⑧   以金 賜 其母。⟨He⟩ presented his

⟨S⟩ A   P   Oi     mother with bronze.

Observe that 金 in ⑧ is part of the prepositional phrase, hence it is not to be construed as Od of the clause but as part of the A. Incidentally, the construction in ⑧ is very similar to MSC disposal constructions with 把 or 將.



教人以善…



授之以政。



堯以天下與舜。



以此教人。

子貢以告孔子。



授 shòu, vn, ‘hand over’

子貢 Zı ̌gòng, np

Comments ⑨–⑩ In these constructions, the order of the O remains unchanged. ⑬ The Opr 之 is almost never realized after 以 (cf. also 5.2.3), so 以 followed directly by a verb is to be understood as *以之. 子貢 was the courtesy name of Duanmu Ci 端木賜, a disciple of Confucius. Do you see a relation between his given name and his courtesy name (cf. Focus 5)? 9.2.3  Anteposed prepositional ‘objects’ In quite a few cases, the ‘object’ of 以 is anteposed. This is the rule with the print 何 (cf. 5.3.2), and it is frequent with the prdem 是 (compare English ‘what for’ or ‘thereby’). Occasionally, such anteposition also occurs with regular nouns. In analogy to 5.3.2, anteposition is to be expected with 誰, as well. However, the only example seems to be 誰以易之 (LY 18.6), whereas 以誰 occurs several times. Anteposition of nouns also—rarely—occurs with 於, as in 君於何有 (Zuo 9.23), ‘What does it have to do with the ▲ ruler?’

104

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

9.2 p h r a s e s w i t h

⑭ 子 何以 知 之。

S   A  P  O

How do you know that? ⑮

以德報怨

報 bào, vtr, ‘requite’ 怨 yuàn, nc, ‘hatred’ 邪2 xié, vi as nc, ‘evil’ 貫 guàn, vtr, ‘string together’

 . . .

何以報德。



是以亂作。



政以治民,

刑以正邪。

予一以貫之。





Comments ⑯ Whereas 以是 can have a variety of meanings, 是以 almost always expresses causality, ‘therefore’. The other prdem are usually placed after 以; for an example, cf. ⑫. 是以 frequently takes initial ­position while the S moves into medial position. ⑰ Examples of nc in anteposition, placing emphasis on them. Note the opposition ‘by ­government . . . by punishments’. 9.2.4 The 以為 construction The construction 以 X 為 Y, which occurs very frequently in classical texts (and later, surviving into Modern Chinese), may in principle express any of the meanings attributed to 以 above. Among these, three usages are so common that they may be singled out for special discussion. These are:

1. ‘use X to make Y / make Y out of X’

⑲   以木 為 車。 ⟨He⟩ made carts out of wood.



⟨S⟩ A  P  O

2. ‘take X as Y / make X function as Y’

⑳   以堯 為 君。 ⟨They⟩ made Yao ruler.



⟨S⟩ A  P  O

3. ‘take X to be Y / regard X as Y’

㉑ 吾 以子 為 鬼。 I took you for a ghost.

S 

A  P  O

105

鬼 guı ̌, nc, ‘ghost’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 9: p r e p o s i t i o n a l p h r a s e s i i

The following are examples of all three usages. ㉒

天將以夫子為木鐸。



吾以夫子為天地。



吾以女為死…



君…以臣為有罪。



君子義以為上。

玉人以為寶也。



將 jiāng, adv, ‘about to’ 鐸 duó, nc, ‘bell’ 義 yì, vst as nc, ‘righteousness’ 寶 bǎo, nc, ‘treasure’

Comments ㉒ 木鐸 is a bell with a wooden clapper: metaphorically, it means that the master proclaims Heaven’s will. ㉓ This may be understood as usage 2 or 3: the border is sometimes not clearly defined. ㉔–㉕ Note that with usage 3, the object of 為 may be a VP. ㉖ Here, too, the ‘object’ of 以 is anteposed. ㉗ Note the deletion of 之 after 以. This construction seems to have led to the MSC usage of 以為. The VP is marked by 也: it expresses a categorical assurance (cf. 4.4.2). 9.3  Phrases with 為 The preposition 為 wèi is derived from the verb 為 wéi, perhaps being the perfective form of the latter, marked by a *-h. As such, it became grammaticalized as a preposition meaning ‘for’, or ‘because of ’ (which has survived in the modern 為 什 麼 ). In many cases, its ‘object’ ­corresponds to a dative object in English. Prepositional phrases with 為 , as a rule, take medial position before the predicator: ㉘ 吾 將 為女 問 之。I will inquire about it for you.

為2 wèi, prep, ‘for’

S  A  A  P  O

106



臣…為王言樂。



獻公為之不寐。



最為天下貴也。

子盍為我言之。



獻 Xiàn, np, posth. name 寐 mèi, vn, ‘sleep’ 最 zuì, adv, ‘most’ 盍 hé, adv, ‘why not?’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

9.4 p h r a s e s w i t h



Comments ㉙ Due to its meaning, 為 is usually followed by pronouns or other forms of address. ㉚ 之 is usually realized after 為 : ‘for that (reason) . . . ’ ㉛ The adv 最 expresses a superlative. Its position in this clause is remarkable: logically it modifies 貴, but placed before the PrP it acquires more emphasis; compare clauses like 三 以 天 下 讓 (LY 8.1) ▲ or 吾 不 與 子 為 昆 弟 (HFei 23). ㉜ 盍 is a fusion of the adv 胡 , ‘why’, ▲ and 不 (cf. Box 16), meaning ‘why not’. Like in English, this is meant as an encouragement to do something. Again, the adverbial mod (for which compare adv 何 , 7.2) is placed before the PrP, although it refers to the P. 9.4  Phrases with 與 The conjunction 與 , ‘and/or’ (cf. 2.3), may also function as a preposition meaning ‘together with’. Although it is written with the same character as the verb ‘give’, and an interrogative particle (cf. 3.4.1), it is easily distinguishable from the two in syntactic context. The verb, as predicator, is usually followed by two objects (cf. 5.2.2); the interrogative particle takes final position; the preposition 與 , however, is followed by an ‘object’ (completing the PrP) and the predicator. The basic structure is: ㉝ 鄒人  與楚人    戰。If the people of Zou waged

  S

A

P

war with the people of Chu … (Meng 1A7)

鄒 Zōu, np, a place

As the example shows, the preposition is not as easily differentiated from the conjunction written with the same character. Thus sentence ㉝ could also be translated as ‘If the people of Zou and the people of Chu waged war’. The following examples will show some usages of the preposition 與 and how it may be distinguished from the conjunction.

107



吾與回言…



與朋友交而不信乎。



夫子與之遊。

齊王誰與為其國。



交 jiāo, vn, ‘mingle’ 信 xìn, nc as vst, ‘be trustworthy’ 遊 yóu, vn, ‘go for a walk’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 9: p r e p o s i t i o n a l p h r a s e s i i

Comments ㉞ Is 與 to be construed as conj or prep? ㉟ In cases where the S is deleted, 與 can only be a prep. ㊱ The pr 之 may be deleted after 與 , but it need not; in any case, it is clear that 與 represents the prep, not the conj, because it is followed by 之 . ㊲ Again, the print precedes the prep, just like it precedes verbs. Box 12  Verb classes Having discussed various uses of verbs, it would seem useful to review the respective properties of transitive, intransitive, stative, and neutral verbs. These may be summarized as follows: vtr

vi

vst

vn

Without

passive

active

active

active

object

天下定

鳥飛

任重

不易民而教

With one

transitive

factitive

factitive/putative

transitive

object

定天下

飛之

重罰 / 重死

不教民

With two

–––

–––

–––

ditransitive

objects As mod

父教之忠 partp

parta

descriptive

定術

飛鳥

重罰

With 於/

agent/local

local

comparison/local

local

于/乎

定于一/定於內

鳥飛於空

禍重乎地/重於喪祭

教於國

–––

Of course, there are exceptions to these general rules. For example, a few neutral verbs (e.g. 往, 來, or 死) may be used as modifiers; and pseudo-transitive or pseudo-ditransitive usage may occur with different verbs (cf. Box 11).

9.5 Exercises Translate the following sentences containing prepositional phrases. In ㊴ and ㊵, is 與 a preposition or a conjunction—and why?

108

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

9.5 e x e r c i s e s



君子不以言舉人,

不以人廢言。



吾不與子為昆弟矣。



孟子嘗與我言於宋。



禮以行義,義以生利,

利以平民。

是以民為皆如老聃也。



References  ① Zuo 8.2  ② HFei 49  ③ Zhuang 17  ④ Zuo 10.14  ⑤ Zuo 10.21  ⑥ Zuo 3.16  ⑦ LNZ 6.2  ⑧ HSWZ 9  ⑨ Meng 3A4  ⑩ LY 13.5  ⑪ Meng 5A5  ⑫ Zhuang 33  ⑬ HSWZ 9 ⑭ Zuo 9.31  ⑮ LY 14.34  ⑯ Zuo 10.2  ⑰ Zuo 1.11  ⑱ LY 15.3  ⑲ SHJ 18  ⑳ Meng 6A6  ㉑ Zhuang 19   ㉒ LY 3.24  ㉓ Zhuang 5  ㉔ LY 11.21  ㉕ Zuo 10.31  ㉖ LY 17.20  ㉗ Zuo 9.15  ㉘ Xun 29  ㉙ Meng 1B1  ㉚ CQFL 14.1  ㉛ Xun 9  ㉜ Meng 2B10  ㉝ Meng 1A7  ㉞ LY 2.9  ㉟ LY 1.4  ㊱ Meng 4B30  ㊲ ZGC 1.16  ㊳ LY 15.23  ㊴ HFei 23  ㊵ Meng 3A2  ㊶ Zuo 8.2  ㊷ HFei 46

109

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Review 2

Analysing clauses Since Classical Chinese is a dead language, it is difficult—perhaps impossible—to reach the kind of intuitive understanding that enables us to comprehend utterances of a living language without conscious reflection. Instead, even experienced readers of Clas­sical Chinese need to analyse complicated sentences systematically in order to fully under­ stand them. In the following paragraphs, we will review some main grammatical rules encoun­tered so far and discuss some practical conse­ quences for the analysis of Classical Chinese clauses. 1  Ground rules 1.1  Modifiers always precede that which is modified. It follows that the core of phrases as well as clauses is likely to be found at the end. Thus, as a rule, always start analysing phrases and clauses at the back. For example, a noun phrase like 先王之道 (Xun 19) may be analysed as follows:   What is it? 道, a ‘principle’.     > Which principle? 王之道, ‘the kings’ principle’.        > Which kings’ principle? 先王之道, ‘the former kings’ principle’. At first sight, we may identify the head, being the central part of the phrase: 道. Everything else is ‘just’ additional information. Simple nominal clauses (3.2) and verbal clauses with intransitive or passive predicators may be ana­lysed in the same way; take 朋友死 (LY 10.22):   What is being said? 死 , someone ‘died’.    > Who died? 朋友死 , ‘comrades and friends died.’ Again, the predicator—someone ‘died’—may be readily found in final position. Of course, most clauses are more elaborate. For their analysis, we need further strategies derived from other grammatical rules. 110

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 2: a n a lys i n g c l au s e s

1.2  The subject is usually in initial position. So after having identified the predicator at the end of the clause, it is useful to next consider the beginning in order to find the subject. Take 鵠不日浴而白 (Zhuang 14):   What is being said? 白, someone or something ‘is white’.    > Who is white? 鵠 . . .白, ‘The swan . . . is white.’ Note that this already captures the core of the utterance—everything else is just supplementary information—adjunct—which may be then analysed from back to front:   > How is it being white? 鵠 . . . 浴而白, ‘The swan . . . is white by ­bathing.’   > How exactly? 鵠 . . . 日浴而白, ‘The swan . . . is white by bathing daily.’      > More exactly? 鵠不日浴而白, ‘The swan is white although it does not bathe daily’. Thus, after the basic elements—predicator and subject—have been identified, the full meaning of the clause may be filled in step by step. 2  Using signal words Many clauses are not quite as straightforward as the ones discussed above. But fortunately, the more intricate clauses are, the more ‘signal words’—particles, preposi­tions, con­junctions, adverbs, or modal auxil­ iaries—they tend to contain. Note that ‘signal words’ do not constitute a word class; the term does not denote a gram­ matical category at all, but refers to a practical tool kit of sorts: it simply means all those words that may serve to analyse a sentence.

These are invaluable signposts for syntactical analysis. In fact, the first thing students of Classical Chinese should do when confronted with a complicated clause is to mark the signal words it contains: in many cases, they will provide all the information needed to analyse the clause. The following is a checklist with points to be considered about certain characters and the signal words they may stand for as well as the con­ clusions to be drawn for the analysis of the clause. 之 → in final position? yes → Opr → verbal clause → preceded by a transitive P. 111

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 2 : a n a lys i n g c l au s e s

no → pt → part of a NP including the words before and after 之. or → Opr → preceded by a P, followed by another O or PrP. Exception: in some cases, 之 may be a verb.

與 → in final position? yes → ptint → marks a question. no → conj → NP, including the words before and after 與. or → prep → followed by NP and P. or → a verb which may be followed by two O. 也 → in final position? yes → pt → very likely marks a nominal P. no → marks a topic (cf. 13.2) Exception: in some cases, 也 may follow a verbal P.

矣 → in final position? yes → pt → verbal clause, perfective aspect (cf. 10.4). no → marks an inverted P (cf. 14.3) 非 → pt → negates nominal P → between S and P. Exception: in some cases, 非 may negate a verbal P. Moreover, it may be used as a nc or a conj.

不 → adv → followed by a verb, likely the P.

Note: after 不, a Opr may be deleted, in which case it functions like 弗 (see below).

而 → conj → preceded and followed by a VP, the latter likely being the P, the former, A. or → conj → connects two clauses (cf. Lesson 16). or → prper → followed by a NP.

Exception: 而 may represent a conditional conj, in which case it is preceded by a NP (cf. 15.4.1–㉒).

可, 見, 被, 足 → modal auxiliaries → likely preceding the P, which is a passive verb. 能, 敢, 肯, 知, 得, 務, 欲, 願, 好, 宜 → typical aux → likely preceding the verbal P. Exception: These verbs may also—less often—be used as regular verbs, followed by an O.

必, 甚, 益, 忽, 果, 固 → adv → verbal clause → between S and P. Exception: 必 and 固 may also occur in nominal clauses.

皆, 俱, 各, 或, 莫, 孰, 獨 → adv → verbal clause → between S and P → S plural. Exception: 皆 may also occur in nominal clauses; 獨 follows a S in the singular.

112

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 2: a n a lys i n g c l au s e s

其 → adv → verbal clause → between S and P. or → prper → followed by a NP. or → prper → followed by a nominalized clause (cf. 12.3). 於,于,乎 → prep → verbal clause → followed by NP, very likely preceded by the P. Exception: 乎 in final position → ptint.

以,與,為 → prep → followed by NP and – very likely – the P. 弗,勿 → fus → imply an Opr → followed by a transitive P. 諸 → fus → implies an Opr → verbal clause with a transitive P. → in final position? yes → implies ptint → marks a question. no → implies prep → preceded by verbal P, followed by NP. 焉 → fus → implies a pr → verbal clause. → in final position? yes → implies prdem → preceded by the P. no → implies print → a question.

3  Using parallelism The parallelismus membrorum—syntactical concordance of successive clauses—is probably the most important stylistic device in Classical Chinese texts (cf. Focus 21). For our understanding of these texts, it is an invaluable tool, since it allows us to infer—with due caution—the mean­ ing of an opaque clause from the analysis of a parallel clause. Take, for example: 冬與夏不能兩刑, Winter and summer cannot take shape together, 草與稼不能兩成。 grass and crops cannot ripen together. (Lü 24.5) If the meaning of 刑—the character usually stands for ‘punishment’—is not clear in the first clause, it may be concluded from the parallel 成, ‘ripen’, that it here stands for the semantically parallel 形, ‘take shape’. In many cases, it is useful to first analyse the last member of a series of parallel clauses, since due to the stylistic law of increasing constituents it is often the most explicit. For example, in

113

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 2 : a n a lys i n g c l au s e s

吾始聞之懼, 復聞之怠, 卒聞之而惑

When I started hearing it, I was scared; When I continued hearing it, I was sluggish; When I concluded hearing it, I was confused. (Zhuang 14)

only the last line has the conj 而, explicitly marking the adverbial ­modification: in the first two lines, the syntactical relations are not as clearly stated. 4 Exercises Analyse the following sentences by first marking all the signal words and then identifying the constituents as far as possible. Try to do this without looking up any words in the glossary: the focus here is not on exact translations but on grasping syntactic structures.



江河

見國之利而忘君

是誰之過與。







禮 國 之 幹 也, 敬 禮 之 輿 也。

是非君子之道。



之水非一源之水也。

安。



移其

汝身非





禮與食孰重。

誰非君之子。







吾與汝共之。



民於河東。

汝有也。 楚不敢伐。 惠而不知為政。



不出於戶而知天下,

大國



人皆有兄弟,我





牛鳴於門外。

不窺於牖而知天道。



四海之內皆兄弟也。





獨無。



不敢以陳於王前。



之樹必巨。

君子 可 贊 衣 補 衣 而 見 荊 王。  . . . 欺 也, 黃帝遊乎赤水之北,登乎

不可罔也。

㉒ ㉑

萬 民 弗 忍, 居 王 于 彘。

虛其心,實其腹,

親親,仁也,敬長,



崑 崙 之 丘。

百姓皆以我為愛。

弱其志,強其骨。

義也。





References  ① Zuo 5.11 ② Mo1.1 ③ XX 7.10 ④ Meng 3B8 ⑤ LY 16.1 ⑥ Zhuang 6  ⑦ Meng 6B1  ⑧ Meng 1A3  ⑨ Zuo 12.6  ⑩ Zhuang 22  ⑪ Zuo 2.6  ⑫ Meng 4B4  ⑬ HFei 20  ⑭ Lü 17.2 ⑮ LY 12.5 ⑯ LY 12.5 ⑰ SY 12 ⑱ Meng 2B2 ⑲ Lü 15.5 ⑳ LY 6.26  ㉑ Zhuang 12  ㉒ Zuo 10.26  ㉓ Meng 1A7  ㉔ Lao 3  ㉕ Meng 7A15

114

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 10

Reference to time

10.1 10.2 10.3 10.4 10.5 10.6

Introduction115 Past and present          A S  P (O) 116 Duration             S  A  P (O)    S P (O) A 117 Imperfective and perfective aspect S (A) P (O) 矣119 Other aspectual distinctions    S  A  P (O) 121 Exercises122

10.1 Introduction Verbal sentences, by their very nature, refer to time. In particular, they may express tense, relating the verbal action to the time of utterance or another point of reference; or aspect, reflecting the way the action is regarded with respect to inherent properties such as duration, completion, repetition, and the like. In Classical Chinese, neither tense—present, past, or future—nor aspect is expressed morphologically; rather, they are indicated with the help of adjuncts. Thus prepositional phrases or temporal nouns, occurring in initial position, may indicate the tem­ poral setting of the sentence; and adverbial elements may express the aspect of the predicator. This lesson discusses the expression of present and past tense with temporal nouns, the expression of dur­ation, imper­ fect­ive and perfective aspect, as well as further as­pect­ual distinc­ tions. Whereas in other works the treatment of aspect is usually limited to the distinction between imperfective and perfective aspect, the con­ cept is used here in a broader sense, including ‘prospective’, ‘progressive’, ‘occasional’, ‘continuative’, and ‘temporary’ aspects. For an up-­to-­date overview including many references, cf. Meisterernst 2015.

Preliminary questions: How are tenses and aspects expressed in the English language? How in Modern Standard Chinese? 115

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 10 : r e f e r e n c e to t i m e

10.2  Past and present The location of events in time is not expressed by inflectional verb forms in Classical Chinese. In fact, the temporal setting of the verbal action is usually not indicated at all. A sentence like 王坐於堂上 (Meng 1A7) states no specific time in relation to the utterance; only the con­ text—Mengzi reporting what he has heard—suggests that it should be translated in the past tense. Instead, the temporal setting may be encoded by temporal nouns (cf. 1.2.1) functioning as adjuncts. They are usually placed in initial position:

① 昔 文公 與秦 伐 鄭。 Formerly, duke Wen 昔 xī, n , ‘ancient times‘ temp

A  S  A P O

together with Qin attacked Zheng.

文 Wén, nc as np, posth. name

秦 Qín, np, a polity

Such noun phrases point to a time before, simultaneous with, or after a certain point of reference. This point of reference may be the utterance itself (as in ‘Tomorrow, I will go shopping.’) or the context described by the utterance (as in ‘The next day, Peter went back home.’). Only ntemp function in this way; nloc or other nouns may not be placed in initial position when used as A (cf. Cikoski 1971, 124–5). For PrP expressing the temporal setting, cf. 8.2.1–⑦ and 9.2.1–⑥.



九月衛穆公卒。



秋八月丁卯,大事于大廟。



昔者疾,今日愈。



今周德既衰。



當堯之時,天下猶未平。



初,晉武公伐夷。

明日,王出。



衛 Wèi, np, a polity 穆 Mù, np, posth. name 卒2 zú, vi,  ‘die’ 丁 dīng, ntemp, 4th ‘stem’ 卯 mǎo, ntemp, 4th ‘branch’ 廟 miào, nc, ‘temple’ 疾 jí, vst, ‘be ill’ 愈 yù, vst, ‘be better,’ 今 jīn, ntemp, ‘now’ 既 jì, vi as adv,  ‘already’ 衰 shuāi, vn,  ‘diminish’ 猶yóu, vn as adv,  ‘still’ 平 píng, vi,  ‘be peaceful’ 初 chū, ntemp,  ‘formerly’ 晉 Jìn, np, a state 夷 Yí, np, a people

116

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

10.3 d u r at i o n

Comments ② A concrete date, ‘the ninth month’ (not ‘September’, be it noted), serving as temporal adjunct. ③ Dates can include any of the fol­ lowing elements (in this order): year, season, month, day, ganzhi-date (cf. Focus 8), in asyndetic coordination. Adverbial 大 may often be translated more elegantly as modifier: ‘made a great sacrifice’ rather than ‘sacrificed greatly’. ④ 昔 is a temporal noun pointing to an un­defined point in the past (in this case, the day before); it may be followed by 者 without apparent change in meaning. Note the contrast to 今日. ⑥ Specific times may be expressed by modified 時; such constructions are often preceded by 當, used as a prep. ⑦ Initial 初 always points to an undated event in the past; it functions almost like ‘once upon a time’ in fairy tales. ⑧ This sentence is part of a historical narrative, so 明日 points to a ‘next day’ in the past. The sentence must accordingly be translated in the past tense. 10.3 Duration The predicator may also be modified with regard to its duration. This is expressed by nominal adjuncts either in medial or—if they are to be stressed—in final position. The forms are: ⑨   終日 飲 酒。⟨They⟩ drink ale all day

終 zhōng, vtr,  ‘complete’

⑩ 武 留 十餘歲。   Wu stayed for more than

餘 yú, nc,  ‘surplus’

⟨S⟩ A  P  O

ten years.

A

S P

留 liú, vi,  ‘stay’

In clauses like ⑩, the A appears in the position of an O, which may create the impression that the P is used transitively. This is not the case. In ⑩, it is clearly the S that stays, whereas with an O, 留would be causative, that is the S would make the O stay. Nor can this construc­ tion be called pseudo-­transitive, since duration is never expressed in PrP; hence this is not a case of an omitted prep (cf. Box 11).



吾嘗終日不食,

終夜不寢 。



弟子事夫子有

年矣。



臣不睹臣之母

三年矣。



吾與回言終日。

人生幾何。



嘗 cháng, adv,  ‘once’ 寢 qı̌n, vi,  ‘sleep’ 年 nián, ntemp,  ‘year’ 矣 yı̌, pt, marks perfective aspect 睹 dǔ, vn,  ‘see’ 幾何 jı̌hé, print,  ‘how long?’

117

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 10 : r e f e r e n c e to t i m e

Comments ⑫–⑭ The temporal expression in final position is stressed; indeed, it may be taken as the core of the utterance. ⑮ The print 幾何, ‘how much/many/ big/long?’, seems to derive from the idiomatic co­ord­ in­ation of two print. Questions concerning duration may also be posed with 幾日/幾年, or the like. These are always placed in final position.

Box 13  Numerals and measures In quite a few cases, adjuncts containing numerical expressions are placed in final pos­ition, that is behind the phrase they relate to. This occurs regularly with temporal expressions, as in the above examples; it may also occur when things or actions are counted, as in: 季桓子微服往觀再三。 Ji Huanzi went in plain clothes to look two or three times. (SJ 47) 燭鄒有罪三。

Zhu Zou is guilty on three accounts. (Yan 7.13)

In quite a few cases, the count is specified by measure words, which spe­ cify a unit in which things are counted. In the case of ‘container’ measures, these are placed as modifiers before the head: 一杯水 . . . 一車薪

one cup of water . . . one cartload of firewood (Meng 6A18)

束錦

a bundle of brocade (Zuo 12.12)

In other cases, however, especially when numerals are accompanied by measure words, they are placed behind the phrases they refer to: 文馬三十駟

thirty quadrigas of embellished horses (SJ 47)

舜有臣五人。

Shun had five ministers. (LY 8.20)

孔子長九尺有六寸。

Confucius was nine feet and six inches tall. (SJ 47)

田常乃選齊國中女子

Tian Chang thus selected women from within Qi that were

長七尺以上為後宮。

more than seven feet tall to make up his harem. (SJ 46)

Measure words, although already found in the earliest stages of OC, remained rare throughout the Classical period. Moreover, they are only used with numerals, not with demonstratives, as for example in MSC 這個. Note also that in CC measures are only used with nouns, not with verbs, as in MSC 去一趟.

Since, logically, these adjuncts modify the preceding phrase, these are remarkable exceptions to the ground rule that modifiers always precede that which is modified. Such post-­modification is called for when a numerical expression provides new ­information, which is placed in a prominently

118

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

10.4 i m p e r f e c t i v e a n d p e r f e c t i v e a s p e c t

stressed position. In many cases, it ­contains the central message of the clause and, indeed, functions not as adjunct but as ­predicator: 凡天下戰國七。

Altogether, the warring states in the realm were seven. (SJ 69)

牛飲者三千人。

Those drinking like oxen were 3,000 men. (HSWZ 4)

今臣之刀十九年矣。

Now my knife is already nineteen years old. (Zhuang 3)

死已三千歲矣。

⟨It⟩ has been dead already for 3,000 years. (Zhuang 17)

In the last example, 三千 must be construed as P, since it is placed between an A, 已, and a final particle. Note that syntactically 歲 functions as the O of 三千; however, this is almost impossible to reflect in translation.

Recall that numerals are intransitive verbs: as such, they may well function as predicator, and they may even take an object. Thus certain cases of appar­ ent post-­modification may actually be analysed—if not translated—quite differently.

10.4  Imperfective and perfective aspect The main aspectual distinction of Classical Chinese is that between imperfective aspect, which views the action as not (yet) completed, i.e. in progress, and perfective aspect, which regards the action as completed. These aspects are not to be mistaken for tenses, nor are they intrinsically bound to certain tenses. The imperfective aspect simply presents the action as ongoing—be it in the past, present, or future. Similarly, the perfective aspect presents the action as complete in the sense of having reached a certain point—regardless of when this ­happens in relation to the time of the utterance. In many cases, the perfective aspect seems to have been mor­pho­logic­al­ly distinguished in OC by a suffix *-s/-h. There are scores of OC verbs that possess two forms, one of which has a suffix *-s/-h (which evolved into the fourth tone in MSC). This distinction indicates imper­ fect­ive and perfective forms, respectively. Consider the following pairs: 治 chí < *drə, ‘order’, vs. 治 zhì < *d-­ləh, ‘be ordered’; 去 qǔ < *khaʔ, ‘remove’, vs. 去 qù < *khah, ‘leave, be removed’.

While the imperfective aspect is usually unmarked, the perfective aspect is indicated by the temporal adverbs 既 or 已 and/or by the particle 矣 in final position. The basic forms are: ⑯ 宋人 既 成 列。 The people of Song had

  S  A  P O

already closed ranks.

S  P  O A

I have already heard of that.

⑰ 吾 聞 之 矣。

119

列 liè, nc, ‘rank’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 10 : r e f e r e n c e to t i m e

The perfective aspect may also be explicitly negated by the adverb 未. It emphatically expresses that a point has ‘not yet’ or ‘never’ been reached, in other words that the predicator is in the imperfective aspect. Instead of 矣, which never occurs after 未, such clauses are usually closed by 也. Contrast the following example with ⑰, which is from the same passage: ⑱ 吾 未 之 聞 也。

I have never heard of that.

S  A O  P A

The particle 矣 *ləɁ is functionally equivalent to MSC 了 le. It is almost completely absent from pre-­classical language. For a pioneering study, cf. Simon 1934.

其馬力竭矣。

門已閉矣。

吾既許之矣。

子未學禮乎。

曰未可︙曰可矣。

未之有也。

吾不反矣。

自今子其無事矣。

吾罪大矣。

㉗ ㉖ ㉕ ㉔ ㉓ ㉒ ㉑ ⑳ ⑲

馬 mǎ, nc,  ‘horse’ 力 lì, nc,  ‘strength’ 竭 jié, vtr,  ‘exhaust’ 已 yı̌, vi as adv,  ‘already’ 閉 bì, vtr,  ‘close’ 反 fǎn, vi,  ‘return’

Comments ⑳ The adv 已, ‘already’ (cf. MSC 已經), marks a past action the effects of which reach into the present: the door has been closed and still is. 已 is a frequent counterpart to 矣, perhaps even a different form of the same word. ㉑ The adv 既, which expresses the completion of an action, may be used independently, as in ⑤ and ⑯; but it mostly occurs with 矣 reinforcing its meaning. ㉓ These sentences nicely capture the meaning of the perfective aspect: ‘not yet . . . now we can’, the crucial point in time has been reached. ㉔ A frequently used expression that became a stock phrase: it means that something has never occurred, implying that it is impossible. As usual in negated sentences, the Opr 之 is placed before the P (cf. 5.2.3). ㉕ Negations with 不 or 無 in com­bin­ ation with final 矣 are quite different from negations with 未. Whereas the latter negates the perfective aspect, meaning that an action has not yet reached a certain point, the former express a negated P in the per­ fect­ive aspect: the action has reached the point where it will no longer occur. ㉖ The PrP makes the expression even clearer: ‘from now on . . . no longer’. Here, the perfective aspect goes together with future tense. ㉗ 矣 may also occur after vst, indicating that the quality has now been 120

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

10.5 ot h e r a s p e c t ua l d i s t i n c t i o n s

achieved; or that the quality has a certain degree: ‘quite big’, ‘so big’, ‘big enough’, or ‘too big’. In this sense, 矣 is not so much aspectual but rather modal, compare 了 in MSC 太好了. 10.5  Other aspectual distinctions Besides perfective and imperfective, Classical Chinese makes use of other aspectual distinctions. These include prospective (‘going to do’), progressive (‘be doing’), occasional (‘do once’), continuative (‘still do’), and temporary (‘do for a while’) aspects, which characterize the inher­ ent state of the action regardless of tense. These aspects, which may be combined with the perfective or imperfective, are marked by adverbs. The form is: ㉘ 孟子 將 朝 王。 Master Meng was about to see

    S  A  P  O

the king in audience.

朝 cháo, nc, as vi ‘attend court’

To be sure, a clear distinction between time and aspect in CC cannot always be maintained. In fact, most grammarians take the constructions discussed in this section as expressions of tense, not aspect. While Norman 1988, 94, writes that ‘time adverbs indicate various tem­ poral and aspectual relationships’, Unger speaks of ‘temporal particles’, and Pulleyblank 1996, 119–22, calls them ‘time words’. The view adopted here that all the adv in this section are indicative of aspect is supported by Dobson 1959, 48–59.



吾將問之。



齊王方大飲左右。



回嘗聞諸夫子。



告子未嘗知義。



吾猶有望。

子姑待之。



方 fāng, adv, ‘just then’ 左右 zuǒyòu, nc, ‘entourage’ 回 Huí, np 望 wàng, vn, as nc, ‘hope’ 姑 gū, adv, ‘for a moment’ 待 dài, vn, ‘wait (for)’

Comments ㉘–㉙ 將 is the most common adv marking the (imminent) prospective aspect: it indicates that an action is about to be taken, will be taken, or that there is an intention to take it. This does not say anything about the action’s temporal relation to the utterance: ㉘ must be trans­ lated in past, ㉙ in future tense. ㉚ The adv 方 indicates the progressive 121

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 10 : r e f e r e n c e to t i m e

aspect: the action is ‘just in the process of ’ happening. It often occurs in temporal clauses, to be translated as ‘When . . .’ 方 may be loosely com­ pared to MSC 在, as in 她在睡覺. ㉛ The adv 嘗 expresses actions taken on one occasion: ‘once’ or ‘at some time in the past’. In most cases, it refers to actions completed in the past: ‘I have once heard . . .’ ㉜ 嘗 is mostly negated with 未 in the sense of ‘never ever’, ‘not once’. ㉝ The adv 猶 expresses the continuative aspect, indicating that an action or a state is still ongoing, its effects are continuous; accompanied by 未, it expresses that the process or action has ‘not yet’ or ‘still not’ reached a certain point (cf. ⑥). ㉞ The adv 姑 expresses the temporary aspect, meaning that something is or should be done ‘for a while’. For similar aspectual distinctions in MSC, cf. Li Chor-­Shing 1991 and Xiao and McEnery 2004.

10.6 Exercises Translate the following sentences; analyse all expressions referring to time or aspect. ㉟

諸侯方睦於晉。



子路終身誦之。



魯平公將出。



夫子既聖矣乎。



今我已亡之矣。



吾未能也。

猶未足也。



References  ① Zuo 9.14  ② Zuo 8.2  ③ Zuo 6.2  ④ Meng 2B2  ⑤ Zuo 5.24  ⑥ Meng 3A4  ⑦ Zuo 3.16  ⑧ HFei 31  ⑨ Li 19  ⑩ XX 7.29  ⑪ LY 15.31  ⑫ HSWZ 8  ⑬ Lü 9.5  ⑭ LY 2.9  ⑮ Zuo 9.31  ⑯ Zuo 5.22  ⑰ Meng 4A19  ⑱ Meng 4A19  ⑲ Zhuang 19  ⑳ Zuo 12.15  ㉑ Zuo 9.9  ㉒ Meng 3B2  ㉓ Zuo 3.10  ㉔ Meng 4A1  ㉕ Zuo 8.15  ㉖ Zuo 10.3  ㉗ Zuo 8.15  ㉘ Meng 2B2  ㉙ LY 7.15  ㉚ Lü 20.6  ㉛ Zhuang 22  ㉜ Meng 2A2  ㉝ Zuo 5.24  ㉞ Zuo 1.1  ㉟ Zuo 9.18  ㊱ LY 9.27  ㊲ Meng 1B16  ㊳ Meng 2A2  ㊴ HFei 22  ㊵ Zuo 10.1  ㊶ Xun 19

122

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 8

Telling time The task of dating particular events and establishing a chronology for ancient Chinese history has occupied generations of scholars. It is not that calendrical information is lacking in the sources, quite the op­pos­ ite. Ancient Chinese texts, both transmitted and excavated, contain an abundance of dates. Already the Shang inscriptions are dated, so are many Zhou bronze inscriptions, and classical literature is full of refer­ ences to the time of events. What makes dating such a daunting task for historians of ancient China is the difficulty of converting Shang and Zhou dates into the Gregorian calendar. The earliest dating system, used in Shang oracular inscription specifies days only. It employs two series of ten and twelve symbols, respectively, which were later called ‘Heavenly stems’, 天干, and ‘Earthly branches’, 地支: 1 天干: 甲 地支: 子

2 乙 丑

3 丙 寅

4 丁 卯

5 戊 辰

6 己 巳

7 庚 午

8 辛 未

9 壬 申

10 癸 酉

11

12





These symbols are combined to yield a cycle of sixty in the following way: 甲子 = 1, 乙丑 = 2, 丙寅 = 3, and so on, every series recommencing after it reached its end, thus 甲戌 = 11, 乙亥 = 12, 丙子 = 13, and so on, until they both end in 癸亥 = 60. Since odds and evens (e.g. *甲丑) are not combined, only half of the possible total of 120 combinations is used. The complete cycle is:

1 甲子 2 乙丑 3 丙寅 4 丁卯 5 戊辰  6 己巳 7 庚午 8 辛未 9 壬申  10 癸酉 11 甲戌 12 乙亥 13 丙子  14 丁丑 15 戊寅  16 己卯 17 庚辰  18 辛巳 19 壬午 20 癸未  21 甲申 22 乙酉  23 丙戌 24 丁亥  25 戊子  26 己丑  27 庚寅 28 辛卯 29 壬辰 30 癸巳 31 甲午  32 乙未 33 丙申  34 丁酉  35 戊戌 36 己亥  37 庚子  38 辛丑  39 壬寅 40 癸卯  41 甲辰 42 乙巳 43 丙午 44 丁未 45 戊申 46 己酉  47 庚戌 48 辛亥 49 壬子 50 癸丑 51 甲寅  52 乙卯  53 丙辰  54 丁巳  55 戊午 56 己未 57 庚申  58 辛酉  59 壬戌 60 癸亥 123

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 8: t e l l i n g t i m e

After a complete sexagenary cycle, the next cycle would start anew with 甲子, and so on: in this way, days were counted continuously without reference to months or years. Significantly, the oracular inscriptions never specified which cycle their dates referred to: for the Shang, time was cyclical, repeating itself after sixty days. They noted days for their qualities rather than for their position on a time line: it was the kairos that mattered, not the chronology. At some time, perhaps as early as the Western Zhou, the twelve ‘Earthly branches’ were associated with the twelve animals of the Chinese zodiac, 鼠, 牛, 虎, 兔, 龍, 蛇, 馬, 羊, 猴, 雞, 狗, and 豬. These, of course, were also associated with certain qualities. This tradition of Chinese astrology is still very much alive, as attested in the numerous almanacs (萬年曆) published in China, Hongkong, and Taiwan.

Years and months—indeed, even lunar phases—became a regular feature of dating only with the bronze inscriptions of the Western Zhou. The Zhou divided the year into twelve months of twenty-­nine or thirty days length, adding an intercalary month every few years in order to adjust to the solar year. Tradition has it that the Zhou diverged from previous practice by their definition of new year’s day. Allegedly, the legendary Xia rulers had begun the first month of the year on the day of second new moon after the winter solstice (which, corresponding to a date between 21 January and 20 February in the Gregorian calendar, has since 1914 been celebrated as ‘Chinese New Year’); their successors, the Shang, supposedly moved this to the first new moon after the winter solstice; and the Zhou finally moved it back another month to the new moon before the winter solstice. These divergent traditions were continued (or perhaps: invented) in Chunqiu and Zhanguo times, when different regional states main­ tained their own calendars. For example, the states of Lu and Qi ­followed the Zhou calendar, while Jin followed the Xia calendar, and Song apparently adhered to the Shang calendar; Qin had a calendar of its own. The situation is further complicated by the fact that there was no common starting point for the year count as in the Christian tradition. Instead, every regional polity counted the reign years of their of rulers, starting over with every new ruler. Thus the first year recorded in the Chunqiu, the annals of Lu, which is there called 隱公元年, ‘first year of duke Yin’, corresponded in Zhou to the forty-­ninth year of king Ping, in Qi to the ninth year of duke Xi, in Jin to the second year of marquis E, in Qin to the forty-­fourth year of duke Wen, and so on. In the Gregorian 124

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 8: t e l l i n g t i m e

calendar, it corresponds to the year 722 bc—at least roughly, since the Lu year began not on 1 January, but sometime in December. Finally, an example for a date containing all of the above elements: 三年,春,王三月,壬戌

Third year ⟨of duke Yin of Lu⟩, spring, the king’s third month, 59th day of the cycle. (Zuo 1.3)

References: Synoptical tables of the reign years of regional rulers are provided in SJ 14, and Loewe and Shaughnessy 1999, 26–9. The Academia Sinica computing center provides a use­ ful calculator for the conversion of dates ad at . Gassmann 2002 is a detailed study of the Zhou/Lu calendar. Two comprehensive concordances are Tchang 1905 and Hoang 1910. For a detailed overview and further references, cf. Wilkinson 2018, 545–99.

125

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 11

Complements

11.1 11.2 11.3 11.4 11.5

Introduction126 Subject complements  S 為 C   S  為 O C 127 Object complements  S 謂 O  C  S  使 O C 128 Comparisons     S 如 C   S  P  O 如 C 130 Exercises132

11.1 Introduction Having discussed subjects, predicators, objects, and adjuncts, this ­lesson will introduce complements, thus completing our inventory of constituents (cf. 1.3). Some verbal predicators in Classical Chinese may be followed by an element that cannot be classified as their object. This constituent is not affected by the action expressed by the predicator; instead, it complements the subject or the object of the sentence by applying some attribute or definition to it. The syntactic behaviour of complements differs in certain ways from that of objects. Subject complements also occur in English sentences with copular verbs like ‘Peter is a student’. Here, ‘a student’ specifies the identity of the subject; passivation of the sentence is impossible. Similarly, in English sentences like ‘They made him president’, or ‘Call me Peter’, the constitu­ ents in italics are not objects but complements (in contrast to ‘They made him breakfast’, or ‘Call me a taxi’). This can be seen by the fact that they cannot be made the subject of a passive sentence (in contrast to ‘Breakfast was made’, or ‘A taxi was called’). This lesson will introduce some important constructions involving subject complements and object complements.

126

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

11.2 s u b j e c t co m p l e m e n t s

Preliminary questions: What kind of verbs do you expect to be followed by a complement? Do you already know such words in Classical Chinese? 11.2  Subject complements We have already encountered the verb 為 as a neutral verb in the broad sense of ‘make’ (cf. Box 14). Perhaps derived from this meaning, 為 also functions as a copular verb linking the subject and a complement in the sense of ‘to be’ (compare English sentences like ‘They make a charm­ ing couple’, meaning ‘They are a charming couple’, and the Modern Standard Chinese usage of 做 in 他做老師). Since such complements refer to the subject, we shall call them subject complements. The basic form is: ① 孔子 為  魯司寇 。 Master Kong was  

S

P

C

Lu’s minister of justice

司寇 sīkòu, nc, ‘minister of justice’

Observe that 魯司寇 cannot be made the subject of a passive clause: it is not an object (that is a ‘patient’, affected by the action of the predica­ tor), but a complement. Besides a complement, the copula 為 may also be followed by an indirect object, yielding the sixth type of canonical clause (cf. 1.4): ② 箕子為 之 奴。 Jizi was a slave to him.

S  P  Oi C

Apart from 為, which is the prototypical copular verb, a few other verbs, like 如, and 若, both meaning ‘resemble’, may function in the same way. One may note that nominal clauses make similar statements, the P expressing the identity, or category of the S (cf. 3.2). However, nominal clauses are strictly verbless, so what is logic­ al­ly a C must syntactically be construed as P (or else there would be no P). Moreover, nom­ inal clauses are timeless, hence unconditional and categoric, whereas clauses with cop are temporally conditioned: Confucius was minister of justice in Lu for a certain time—but before and after that, he was not.

127

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 11 : co m p l e m e n t s

孟子為卿於齊。

子為誰。

子之子為何。

是宜為君。

化而為鳥,其名為鵬。

嘗為孔子門人。

身為宋國笑。

為天下戮。

吾誰與為鄰。

⑪ ⑩ ⑨ ⑧ ⑦ ⑥ ⑤ ④ ③

卿 qīng, nc, ‘minister’ 宜 yí, vst, ‘be appropriate’ 化 huà, vi, ‘change’ 鵬 Péng, np, mythical bird 門人 ménrén, nc, ‘disciple’ 身 shēn, nc, ‘himself ’ 笑 xiào, vn, as nc, ‘laughing stock’ 戮 lù, vtr, as nc, ‘disgrace’ 鄰 lín, nc, ‘neighbour’

Comments ③ The subject C may be further specified by a PrP. ④ Observe that whereas print are always placed before the P when they serve as O (cf. 5.3.2), they remain in situ when they serve as C, following a cop: this distinguishes C from O. This question, which asks about a name, is not couched in a nominal clause (*子誰也): apparently, names were considered an impermanent trait (cf. Focus 5). ⑤ The answer is: 為孫. ⑥ The cop may be modified by aux or by adverbial phrases, as in the following. ⑦ 為 often conveys an inchoative aspect, meaning that the S ‘becomes’ or ‘turns into’ the C (compare the proximity of ‘be’ and ‘become’ in English, as in ‘He will be a senior next year’). This is another characteristic that sets such statements apart from nominal clauses. ⑧ 為 with Oi and C. This should be distinguished from constructions like 為之椁 (LY 11.8), ‘make him an outer coffin’, in which 為 accom­ modates two O (cf. 5.2.2). ⑨–⑩ Such constructions with verbs like 笑, 戮, 用, or the like (mostly conveying an undesirable action) may be con­ veniently translated in the passive voice: ‘was the laughing stock of. . .’ > ‘was laughed at by…’ They should not, however, be analysed as passive. ⑪ Compare this sentence to 9.4–㊲: what is the difference? 11.3  Object complements If a complement specifies not the subject but the direct object of a sen­ tence, we shall call it an object complement. Object complements appear after certain verbs like 謂, 使, 令, and others. The basic forms are:

128

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

11.3 o b j e c t co m p l e m e n t s

⑫   謂 之 吳孟子。⟨He⟩ called her Wu Mengzi 謂 wèi, vtr ‘call (sb. sth.)’

⟨S⟩ P O   C

吳孟子 Wú Mèngzı ̌, np

⑬ 王 使 人 問疾。

S  P  O   C

The king sent someone to ask about the disease.

Note that object complements differ markedly from indirect objects (cf. 5.2.2). Firstly, only neutral verbs can take two objects, whereas 謂 in ⑫ is a transitive verb. Secondly, an indirect object precedes the direct object, whereas the complement follows it. Object complement also dif­ fer from direct objects insofar as they cannot be replaced by 之 and cannot be made into a subject to form a passive sentence. 王使人來。

齊侯使請戰。

予助苗長矣。

吾必謂之學矣。

是謂聖人。

吾誰使正之。

子路使門人為臣。

楚人謂此鳥何。

女子許嫁。

天帝令我長百獸。

㉓ ㉒ ㉑ ⑳ ⑲ ⑱ ⑰ ⑯ ⑮ ⑭

侯 hóu, nc, ‘marquis’ 請 qı̌ng, vn, ‘ask’ 助 zhù, vn, ‘help’ 苗 miáo, nc, ‘shoots’ 長3 zhǎng, ‘vi, ‘grow’ 子路 Zı ľ ù, np

嫁 jià, vtr, ‘marry off ’ 帝 dì, nc, ‘god’ 令 líng, vtr, ‘cause’ 獸 shòu, nc, ‘animal’

Comments ⑭ Usually, the C of 使 is a verb phrase (also in ⑬); the O of 使, then, is logically the S to this verb phrase: *人來. ⑮ 使 may also be used intransitively, only with a C; as a vn, it remains in the active voice. ⑯ 助 functions the same way as 使; the sentence is from the story about a man from Song who wanted to make his crops grow by pulling them out. ⑰ 謂, ‘to call’ someone something, also requires an O and a C. ⑱ Observe that this is a passive clause: it has no O, only a C. ⑲ As usual, Print precede the P when they function as O. ⑳ Zilu (c.542–480 bc) was a prominent disciple of Confucius, known for his bravery. ㉑ Note the position of 何: Print do not precede the P when they function as C.

129

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 11 : co m p l e m e n t s

㉒ 許 functions just like 使, albeit with a slightly different meaning, ‘allow’. Is the clause active or passive? ㉓ This ­sentence, spoken by a fox

to a tiger, has an interesting parallel in ZGC 14.3: 天帝使我長百獸. 11.4 Comparisons

In verbal sentences, comparisons are usually expressed by means of the verbs 如 or 若, which both mean ‘be like, equal; as if ’. For compari­ sons, they either take a subject complement—in which case 如 and 若 function as copular verbs—or an object complement. The basic forms are: ㉔ 大巧 若 拙。The greatest skill

  S  P  C

巧 qiǎo, vn as nc, ‘skill’

resembles clumsiness.

若 ruò, vn, ‘resemble’ 拙 zhuō, vi as nc, ‘clumsiness’

㉕ 文王 視 民 如傷。 King Wen regarded the

  S 

­people as if they were injured.

P   O C

視 shì, vn, ‘regard as’

Both words are clearly related, 若 *nak perhaps being derived from 如 *na(h) by addition of the ‘distributive’ suffix *-k (for which cf. 7.3). Note also the relation to 然 mentioned in 7.6.



十仞之木若箸。



視生如死,

視富如貧。



人君之道如何。

其小大何如。



仞 rèn, nc, unit of length, approx. 8 feet 箸 zhù, nc, ‘chopsticks’ 貧 pín, vst as nc, ‘poverty’

Comments ㉘–㉙ 何, functioning as C, may follow or precede 如; the latter position may have been influenced by the behaviour of 何 as O.

130

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

11.4 co m pa r i s o n s

Box 14  The verb 為 為 is probably the most versatile verb in Classical Chinese. It has an amazingly broad range of meanings. Depending on the object it takes, it may be trans­ lated variously as ‘do’, ‘make’, ‘deal with’, ‘produce’, ‘create’, ‘establish’, ‘carry out’, ‘contrive’, ‘effect’, ‘perform’, etc. Consider the following examples: 為國

rule the state (LY 4.13)

為黍

prepare millet (LY 18.7)

為箕

weave a winnowing fan (Li 18)

為政

carry out government (Meng 4B4)

為詩

compose odes (Zuo 9.14)

In general, 為 means doing what is usually done with the given object. Although it is the most general of verbs, it is well to choose a more specific verb for translation. In connection with the prep 以 (cf. 9.2.4), 為 may also take on a putative sense, ‘consider, deem’, etc.: 吾以子為鬼。

I took you for a ghost. (Zhuang 19)

吾以女為死 . . .

I thought you were dead . . . (LY 11.21)

Moreover, 為 may take a complement (cf. 11.2), in which case it means ‘act as’, ‘be’, or ‘become’: 箕子為之奴。

Jizi was a slave to him. (LY 18.1)

化而為鳥。

It transforms and becomes a bird. (Zhuang 1)

Finally, 為 may serve as a modal auxiliary: 苟為後義而先利。 If one does place righteousness last and profit first . . . (Meng 1A1)

Used with stative verbs, it may express either the inchoative aspect or the superlative: 唯天為大。

Heaven, indeed, is the greatest. (LY 8.19)

江 河合水而為大。 The Jiang and the He become big by uniting small rivers. (Zuang 25) References: Most uses of 為 are succinctly summarized in Kroll 2015, 472. For prepositional use of 為, cf. 9.3.

131

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 11 : co m p l e m e n t s

11.5 Exercises Translate the following sentences; what kind of C do they include? ㉚

孟嘗君為相數十年。



民可令農戰。



或謂寡人勿取,

或謂寡人取之。



視人如豕,視吾如人。

愛之如父母,仰之如日

月,敬之如神明。



References  ① Meng 6B6 ② LY 18.1 ③ Meng 2B6 ④ LY 18.6 ⑤ SJ 75 ⑥ Zuo 3.11  ⑦ Zhuang 1  ⑧ Yan 7.11  ⑨ HFei 49  ⑩ Lü 9.2  ⑪ Zhuang 20  ⑫ LY 7.31  ⑬ Meng 2B2  ⑭ Meng 2A2  ⑮ Zuo 8.2  ⑯ Meng 2A2  ⑰ LY 1.7  ⑱ HFei 8  ⑲ Zhuang 2  ⑳ LY 9.12  ㉑ ZGC 27.21  ㉒ Li 1  ㉓ XX 2  ㉔ Zhuang 10  ㉕ Meng 4B20  ㉖ Xun 21  ㉗ Lie 4  ㉘ SY 1  ㉙ GY 18.2  ㉚ ZGC 4.4  ㉛ SJS 23  ㉜ Meng 1B10  ㉝ Lie 4  ㉞ Zuo 9.14

132

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 9

Two-­syllable words In Classical Chinese, the great majority of words are monosyllabic and written with one single character. However, this does not make Classical Chinese a ‘monosyllabic language’, as is sometimes asserted. In fact, there are quite a few disyllabic words in the Classical Chinese lexicon. In previous lessons, we have already encountered some of them, e.g. 聖人, 君子, 天下, and 諸侯. These are examples of compound words, which consist of two free words. Such compound words originated as phrases and, through frequent usage, took on the quality of words: thus a dictionary of Classical Chinese should record them as such. There are, moreover, also ‘true’ two-­syllable words in Classical Chinese: binomes (lianmian zi 聯緜字/連綿字) that cannot be broken down into independent words. Many of these binomes consist of two syllables that share at least one phonological feature: they are either rhyming (dieyun 疊韻), alliterative (shuangsheng 雙聲), or reduplicative (chongyan 重言). Examples for such binomes are: 逍遙 踸踔 囂囂 濯濯

xiāoyáo, *siau-­jau: ‘be carefree, unfettered’ chěnchuō, *thrəm-­thrâuk: ‘skip’ xiāoxiāo, *hâu-­hâu: ‘be indifferent’ zhuózhuó, *driauk-­driauk: ‘be bare, denuded’

Such binomes are typically descriptive verbs, expressing attributes, emotions, or manners in which something is done or happens; often, they are followed by the suffix 然 (cf. 7.5). They have a poetic, impressionistic quality, and their precise meaning is often difficult to grasp. Even though their constituent morphemes may occur independently, the meaning of the binomes is not necessarily derived from these independent meanings. Thus reduplicative binomes are not simply intensifications of a monosyllabic word; for example, 濯 means ‘to wash’, and 囂 means ‘clamour’: clearly, the meaning of binomes 濯濯 and 囂囂 cannot be logically inferred from these. This becomes even more obvious if one considers the fact that the orthography of binomes is subject to considerable variation. William Boltz (2015b) cites the case of weiyi, *ʔoiʔ-­lai,

133

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 9 : t w o - s yl l a b l e w o r d s

‘describing a kind of “serpentine movement”, written variously as 委施, 委蛇, 委移, 委佗, 逶 迤, 逶移, 逶迆, 逶蛇 and 蜲蛇, yet the actual binome as a lexical and linguistic unit read wěi yī ~ wēi yī is the same in all cases. . .’ Obviously, the various characters used to write the binome are used for their phonetic value only.

Moreover, many of these binomes are so rare that their interpretation often rests on one or two occurrences. Consider, for example, the following passage, which contains an early assessment of student behaviour: 其容簡連 填填然,狄狄然 莫莫然,瞡瞡然 瞿瞿然,盡盡然 盱盱然。

Their manner is insolent and rude. They seem smug and pretentious as they amble about, but their eyes dart nervously around. They may seem complacent, comfortable, and settled, but their gaze is confused and frightened.

(Xunz 6, tr.: Knoblock 1988–94, vol. 1, 221)

In order to understand the flurry of binomes in this description, students of Classical Chinese should make use of dictionaries such as the four-­volume Lianmian zidian 聯緜字典, that specializes in binomes. However, they may be in for a surprise. The entry for 簡連 is: 簡連  連力延切 ㊀ 傲慢不前之貌。【荀子非十二子】其容簡連。注。 簡連。 傲慢不前之貌。 The only citation given for the binome is none other than the above Xunzi passage: it is a hapax legomenon, a word attested only once. This is hardly a firm basis for determining its meaning (especially since the commentary that explains it dates from the ninth century, that is more than a millennium later than the text). For 狄狄, 瞡瞡, and many other binomes, one gets the same circular results: the dictionary entry refers back to the passage which was the starting point of the enquiry. Looking up binomes can be frustrating business. Another type of binome has been no less troublesome for lexico­ graphers. These binomes, which do not necessarily rhyme or alliterate, are nominal and they typically represent names of animals (especially insects) or plants. Examples include: 蝴蝶 麒麟 芙蓉

húdié, *gâ-­lêp: ‘butterfly’ qílín, *gə-­rin: ‘unicorn’ fúróng, *ba-­loŋ: ‘lotus’

134

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 9 : t w o - s yl l a b l e w o r d s

These binomes do not consist of two independent morphemes. They are, strictly speaking, monomorphemic. But the fact that they are written with two characters has caused considerable confusion. If one looks up the ‘butterfly’ in a dictionary, one is likely to find something like the following entries: 蝴 蝶

The butterfly. 蝴蝶 the butterfly. A butterfly.               (Mathews 1931, Nos. 2174 and 6321)

So what was the Classical Chinese word for ‘butterfly’: 蝴, 蝶, or 蝴蝶? The answer: 蝴蝶, and nothing else. The fact that it is a binome written with two characters has misled lexicographers to treat both characters as independent, which they are not. In some cases, philologists have come up with ingenious explanations. For example, 麒 is usually glossed as the male unicorn, and 麟 as the female. While this may hold some allure, it is nonetheless false: the unicorn was simply called 麒麟: the two syllables have no ­independent meaning.

Dictionaries of (Classical) Chinese are usually arranged by characters, not by words. It is well to remember the difference between the two. While characters always represent one syllable, Classical Chinese words may consist of two. References: Kennedy 1951, 1955; DeFrancis 1984, 177–88; Boltz 2015b. The Lianmian zidian 聯緜字典. 4 vols. Beijing 1954, is a specialized dictionary for binomes.

135

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 12

Nominalization

12.1 12.2 12.3 12.4

Introduction136 Object nominalization 137 12.2.1  所 with verbs 所 VP 137 12.2.2  所 with prepositions 所 prep  VP139 Clause nominalization     S 之 P  (O) (也)140 Exercises142

12.1 Introduction Whereas the preceding lessons have dealt with increasingly complex clauses, this lesson will introduce complex noun phrases which result from the nominalization of verb phrases. In 6.3, we have already dis­ cussed nominalization with the particle 者, which, logi­cally, stands for the subject of a (hypothetical) corre­sponding verbal clause, as in: *宋人  耕  田。 A man from Song S    P   O ploughed the field. 耕田者 ⟨the man⟩ who ploughed the field

耕 gēng, vn, ‘plough’ 田 tián, nc, ‘field’

NP

These rather frequent constructions are functionally equivalent to rela­ tive clauses, which do not exist in Classical Chinese. The same is true for another kind of nomi­nalization by which a verb phrase is converted into a noun phrase that refers to the object of the clause. In our example: *宋人 耕 田。   S  P O  

*所耕

   NP

A man from Song ploughed the field. what ⟨the man from Song⟩ ploughed 136

所 suǒ, nc as prrel, ‘what’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

12.2 o b j e c t n o m i n a l i z at i o n

As can be seen, object nominalization is brought about by the relative pronoun 所. Finally, entire clauses may be nominalized, as in: *宋人  耕 田。 S

A man from Song ploughed the field.

PO

*宋人之耕田也 NP

⟨the fact⟩ that the man from Song ploughed the field

the man from Song’s plough­ing the field /

Clause nominalization makes use of the particle 之, and, in many cases, the particle 也. Such nominalized clauses, being noun phrases, may themselves function as con­stituents in other clauses, as in ‘The one who ploughed the field is my father’, ‘I don’t know what the man from Song ploughed ’, or ‘There is nothing wrong with the man from Song’s plough­ ing the field ’. Having already intro­duced subject nominalization (cf. 6.3), we will now turn to the latter two types of nominaliza­tion, discussing their internal structure as well as their syntactical functions. Preliminary questions: Reconsider 者 phrases: which syn­ tactical ­functions can they fulfil, and which not? Is the same to be expected of nom­in­al­ized clauses? 12.2  Object nominalization The term ‘object nominalization’ refers to the nominalization of verbal expressions with the help of the relative pro­noun 所. Just like nom­ in­al­iza­tion with 者, this is a very fre­quent phenomenon in Classical Chinese (as well as in later stages of the language). But whereas 者 stands for the subject of verbal expressions, 所 always substitutes for the object of a verbal or prepositional expression. The vestiges of these constructions are to be found in MSC 所有, 所謂, 所在, 所以, 無所謂, and others. Look up 所 in a MSC dictionary: you will likely be surprised at the number of phrases beginning with 所.

12.2.1  所 with verbs Substituting for the object of a verb, 所 always precedes the verb, as in: ① 所為

NP

what ⟨the subject⟩ does

137

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 12 : n o m i n a l i z at i o n

Semantically, 所 phrases are equivalent to English relative clauses (‘the things that ⟨the sub­ject⟩ does’), which do not exist in Classical Chinese. It should be borne in mind, however, that 所 phrases are not clauses but noun phrases. As such, they may be further modified. The modifier— mostly subordinated with 之—precedes the 所 phrase, yielding a ­subjective genitive, as in: ② 百工 之 所為

what all the artisans produce

NP® NP    NP

工 gōng, nc, ‘artisan’

Instead of a noun phrase and 之, 所 phrases may also be modified by the prper 其—which, of course, is the semantic equivalent of a noun phrase and 之 (cf. Box 5). ③ 其 所不能   

what he cannot do

NP ® NP   NP

In turn, 所 phrases may also serve to modify other noun phrases, gen­ erating a descriptive genitive (cf. Box 9): ④

所不召  之 臣 NP ® NP    NP

ministers, whom ⟨he⟩ does not call

召 zhào, vn, ‘call’

As a noun, 所 means ‘place’ (cf. Zuo 1.7: 鄭公子忽在王所). Thus 所居 (HSWZ 8) means ‘the place he lives’ > ‘where he lives’; 所之 (Meng 1B23) means ‘the place he goes’ > ‘where he goes’. By extension, 所 may also appear in constructions which, in English, require other whpronouns, as in: 其所為, ‘what he does’ (Meng 1A7), and the other ex­amples given in this lesson. Compare dialectal German usage, as in ‘das schlechteste Messer, wo er hat’ (Hebel).



上之所為,民之

138

歸也。

富,人之所欲。

⑦ ⑥

無為其所不為,

無欲其所不欲。

臣以所學者觀之。

其所善者  . . .  其所惡者 此天之所與我者。

仲子所居之室,

伯夷之所築與。

⑪ ⑩ ⑨ ⑧

歸 guī, vi as nc, ‘refuge’ 居2 jū, vn, ‘dwell (in)’ 室 shì, nc, ‘house’ 築 zhú, vn, ‘build’

仲子 Zhòngzı,̌ np, a name 伯夷 Bóyí, np, a name

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

12.2 o b j e c t n o m i n a l i z at i o n

Comments ⑤–⑥ Bear in mind that the punctuation of CC texts does not necessarily correspond to English usage (cf. Box 7); in these cases, the comma separates S and P. Can you see why this may be helpful? ⑦ In some cases, 所 phrases express what the S should do; compare 惡其所好 (GY 10.2), ‘to hate what one should like’. Recall that 其 substitutes for a NP and 之. ⑧ 所 phrases may be completed by 者. This does not change the fact that they are cases of object nominalization; in fact, it does not seem to alter the meaning of the NP at all. For a translation, one may choose the formula ‘that which’ rather than simple ‘what’ in order to reflect the addition of 者. ⑨ Since 所 replaces an O, the follow­ing verb must be transi­tive: what does that imply for the verb 善 in this phrase? ⑩ If a verb accommodates two O, 所 may replace either the Od or the Oi. The other O then appears after the verb, being part of the 所 phrase. Just like in ⑥, 也 is omitted in this nom­ in­al clause. ⑪ If a 所 phrase modifies another NP, it functions like a re­strict­ ive relative clause: ‘the house which Zhongzi inhabited’; cf. ④. Boyi 伯夷 was a subject of the Shang dynasty, famous for his virtue and loyalty. 12.2.2  所 with prepositions Prepositions, too, may appear in 所 phrases in various ways. We have already seen, in ⑩, that 所 may nominalize verb phrases that include an object. By the same token, it may nominalize a verb phrase including a preposi­tional phrase: ⑫ 吾 

所問於女  

what I have asked you

NP ® NP  NP Moreover, 所 can indeed substitute for a prepositional phrase, as in the following construction:

⑬ 聖人  有  所遊。

  S 

 P 

 O

The sage has ⟨some place⟩ where he wanders.

Since 遊 never takes an object but is usually followed by a prepositional phrase, indi­cating a location, 所 here substitutes for this prepositional phrase. Finally, 所 may substitute for the object of a preposition. Most pre­ posi­tions occur with 所, the only notable exception being 於/于/乎, 139

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 12 : n o m i n a l i z at i o n

which are very rarely used this way. In such phrases, 所 is followed directly by the preposition: ⑭

⟨those⟩ with whom ⟨he⟩ stands

所與立    NP

Note that not only the preposition but also the following verb is nom­in­ al­ized by 所: the entire verb phrase serves as part of a noun phrase. 吾國…虎狼之所棲也。

所惡於上

不知其所自來。

不知過之所由生。

博學而無所成名。

有所得犬。

此所為有名也。

㉒ ㉑ ⑳ ⑲ ⑱ ⑰ ⑯ ⑮

上君所與居,

皆其所畏也。

舜之所以事堯



虎 hǔ, nc, ‘tiger’ 狼 láng, nc, ‘wolf ’ 棲 qī, vi, ‘dwell’ 過2 guò, nc, ‘mistake’ 由 yóu, vn, as prep, ‘from’ 博 bó, vi, ‘be broad’

Comments ⑱ In this clause, 由 must be construed as prep, since it is fol­ lowed by 生; but what if the clause ended on 由? ⑲–⑳ In the con­ structions 無所VP and 有所VP, which are very frequent, 所 mostly substitutes for a 於 phrase in the sense of ‘somewhere’ or ‘nowhere’ (cf. ⑬); some­times more generally, ‘some’, ‘somehow’, or the like. ㉓ 所以 is a faux ami: it does not mean ‘therefore’, as in MSC, but ‘that with which’, or ‘the way in which’. 12.3  Clause nominalization Entire verbal clauses may be formally nominalized by inserting the par­ ticle 之 be­tween subject and predicator. The clause thus acquires the form of a noun phrase—X 之 Y (cf. 6.2)—and is used in the same way: such constructions typically function as subject or object in the sense of that-clauses (cf. English sentences like ‘That he failed the test came as no surprise’, or ‘I know that my redeemer liveth.’). We will accordingly call them subject clause or object clause. In most cases, they are add­ ition­al­ly marked by 也, resulting in the construction NP 之 VP 也. 140

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

12.3 c l au s e n o m i n a l i z at i o n

天下之無道也 ㉔   久 矣。

That the realm lacks the right



⟨He⟩ knew that the state was about to be destroyed.

S

知 ⟨S⟩ P

P  A

國之將亡也   。 O

principle, has already lasted long.

久 jiǔ, vi, ‘last long’

In these constructions, the logical subject of the nominalized clause (here, 天下 and 國, respectively) is syntactically the modifier of a verb phrase (cf. 6.2.2). If this modifier is omitted, this leaves a verb phrase as the subject or object of a sentence. Such con­struc­tions may be trans­ lated as infinitive subjects/objects or as -ing participle clauses: ㉖ 為君 難。 ㉗

S

⟨S⟩

To be a ruler is difficult.

P

為人師 好   。 O P

⟨They⟩ like to be the teachers of others.

It bears emphasis that the term ‘infinitive’ refers only to the translation, not to the CC phrase. Since there are no morphologically marked finite verbs in CC, it makes no sense to speak of infinite verbs, either.

治國者欲民之農也。

夫子何以知其將見殺。

何由知吾可也。

言之易,行之難。

㉜ ㉛ ㉚ ㉙ ㉘

不患人之不己知,

患不知人也。

無望民之多於鄰國也。

吾之不遇魯侯,天也。

㉞ ㉝

農 nóng, nc as vn, ‘farm’ 己 jı̌, prrefl, ‘my-/your-/ him-/her-/itself ’

Comments ㉘ 欲 is often followed by an object clause. Compare English ‘wish that…’ ㉙ 知, too, is very frequently followed by an object clause: ‘know, that. . .’ Again, 其 substitutes for a noun and 之. ㉚ Compare 何由 with 何以 in ㉙: the meaning is almost the same. ㉜ In negated con­ structions, the prrefl 己 as O is placed between the negation and the P (cf. 5.2.3); used in an object clause, it always refers to the S of the entire clause (which in this case is deleted). Contrast the usage of 自, which refers to the S of the object clause: 故明主不窮烏獲,以其不能自舉 141

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 12 : n o m i n a l i z at i o n

(HFei 24). ㉞ Review the different kinds of statements made by nominal sentences (cf. 3.2)—which kind is this? Box 15  Titles Since Classical Chinese literature is mainly concerned with the activities of the elite, it abounds with titles, aristocratic and professional. The highest of these, used in addressing the Zhou rulers, is 王, ‘king’. Below the king, Classical texts name five ranks of nobility, in descending order: 公, usually translated as ‘duke’, 侯, ‘marquis’, 伯, ‘earl’, 子, ‘viscount’, and 男, ‘baron’. We do not know when this hierarchy was established—Western Zhou inscriptions do not bear it out—nor how strictly it applied. These translations, taken from European history, are not meant to suggest structural similarities between medieval European and ancient Chinese societies; they are simply used faute de mieux. Nor must they be taken seriously in every instance: for example, 公 may be used simply to mean ‘noble’ or as a polite form of address; and it is used as a posthumous title for any feudal lord, no matter what his rank was. For 伯 and 子 in courtesy names, which is a different matter, cf. Focus 5.

Starting in the fourth century bc, regional rulers began appropriating the title 王, which was once the prerogative of Zhou rulers. Perhaps as a consequence of this inflationary use, the ruler of Qin, when he created the Chinese empire in 221 bc, chose for himself the new title: 皇帝, ‘emperor’. All aristocratic titles may be preceded by the name of the fief and/or the posthumous name of its bearer, as in 齊侯, ‘marquis of Qi’, 武王, ‘king Wu’, or 齊桓公, ‘duke Huan of Qi’. Syntactically, these are NP in which two or three nouns are subordinated, the title being the head of the phrase. Care should be taken in translation: 齊侯 is not ‘marquis Qi’, and 武王 is not the ‘king of Wu’. References: Hucker 1985 is the foremost resource for titles and government structure in pre-­ imperial and imperial China.

12.4 Exercises Translate the following sentences into Classical Chinese.

㉟ I do not see what the duke does, but I hear what he says. ㊱ I am afraid that people do not know me. ㊲ To learn the Odes is what I desire. ㊳ The wife saw that her husband was going out of the door. ㊴ The rituals that people like are not those that they perform.

References  ① SY 1.15  ② Meng 3A4  ③ Meng 6B15  ④ Meng 2B2  ⑤ Zuo 10.21  ⑥ Meng 5A1  ⑦ Meng 7A17 ⑧ ZGC 30.9 ⑨ XX 4 ⑩ Meng 6A15 ⑪ Meng 3B10 ⑫ HFei 30  ⑬ Zhuang 5  ⑭ LY 10.3  ⑮ Guan 80  ⑯ Li 42  ⑰ Zhuang 12  ⑱ Zhuang 12  ⑲ LY 9.2  ⑳ GY 15.13  ㉑ Xun 22  ㉒ HFei 33  ㉓ Meng 4A2  ㉔ LY 3.24  ㉕ SY 13  ㉖ LY 13.15  ㉗ Meng 4A23  ㉘ SJS 3 ㉙ Meng 7B29  ㉚ Meng 1A7  ㉛ Lü 24.1  ㉜ LY 1.16  ㉝ Meng 1A3  ㉞ Meng 1B16

142

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Review 3

Analysing complex noun phrases We have now encountered several types of complex noun phrases which may be challeng­ing to analyse. Since they are very frequent in classical texts, it is useful to re­view their composition and to deduce some practical rules of thumb for their analysis. 1  者 phrases 者 nominalizes verb phrases: the phrase as a whole has nominal value, while its content (except 者 itself) is verbal. VP  知我者  NP Analysing the phrase from back to front, as usual, we must therefore look for a verb phrase (possibly including an object and/or adverbial modifiers), being the content of the 者 phrase. Thus the above example may be analysed as follows: 知我者 > 我 is a prper, not a verb, so there must be more to the 者 phrase.   知我者 > 知 is a verb: we have now identified the 者 phrase. When dealing with entire clauses, we should also look for an element that may syntactically precede a noun phrase, thus marking it off. Take, for example, the following clause containing a 者 phrase: 吾未見剛者 (LY 5.12). Starting with the 者, we can analyse as follows:  未見剛者。> 剛 is a vst, so it may well be the content of the 者 phrase. 吾    吾未見剛者。> 見 is a vn, which may be followed by a NP, its O: we have fully identified the 者 phrase—剛者—and can move on to the rest of the sentence.   Another example: 為此詩者,其知道乎。(Meng 6A6)   為此詩者 > Since 詩 is a nc, not a verb, we must look further. 143

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 3 : a n a lys i n g co m p l e x n o u n p h r a s e s

  為此詩者 > Since 此 is a prdem, not a verb, we must look further.     為此詩者 > 為 is the first verb we encounter: together with the following NP, it may constitute the content of the 者 phrase.      > There is nothing to the left, so we have fully identified the 者 phrase: it is the S of the sentence. Another example: 富而不驕者,鮮。 (Zuo 11.13) 富而不驕者 > 驕 is a v, so this may well be the content of the 者 phrase.    富而不驕者 > But this is not all: since 不 negates verbs, it cannot precede a 者 phrase, which is nominal. It follows that 不驕 both belong to the content of the 者 phrase. Have we fully identified it now?      富而不驕者 > No, we have not: since 而 subordinates verb phrases, it cannot precede a 者 phrase but must be part of its content. Moreover, since 而 subordinates two verb phrases, the next verb to the left must also be part of the 者 phrase.        富而不驕者 > This is it: 富 is a vst, completing the content of the 者 phrase. Since there is nothing to the left, we have fully identified the 者 phrase. A very obvious case is the construction 有/無 . . . 者, ‘there was somebody/nobody who . . .’ where 有/無 precedes the 者 phrase. In such cases, the latter may be identified at first sight, as in: 宋人有耕田者 (HFei 49). Thus before embarking on the sort of analysis described above, one should first check for a 有/無 to the left of 者. 2  所 phrases 所 also nominalizes verb phrases. This means that 所 is followed by a verb phrase (for the object of which it substitutes), and that the entire phrase—所 and the verb phrase—has nominal value. VP  所不知    NP Whereas in the case of 者, the problem was to find the beginning of the phrase, the challenge in analysing 所 phrases lies in identifying the end. Again, we must look for a verb phrase (possibly including another object and/or adverbial modifiers); in the above example: 144

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 3 : a n a lys i n g co m p l e x n o u n p h r a s e s

  所  不知 > 不 is an adv: it cannot, by itself, be a complete VP, so we must look further.    所不知 > This is it: 不知 is a VP, the O of which is represented by 所. When dealing with entire clauses, we should also look for an element that syntacti­cally follows the 所 phrase. Take the following example: 天之所興誰能廢之 (GY 10.1). This time, we have to proceed from left to right, analysing as follows:   所  興誰能廢之 > Since 興 is a vi, we may already have fully identified the 所 phrase.    所興誰能廢之 > 誰 cannot be part of the verb phrase: as an O, it would have to occur before the verb, besides 興 can only take one object, which is 所. Thus we have identified the 所 phrase: 所興.     We may now turn to the words surrounding the 所 phrase, analysing them accord­ing to the methods we have learned in Review 2: To the left, 之 marks a subordinate NP; and to the right other constituents follow. Another example: 是仁人之所羞而不為也。(Xun 16) 所羞而不為也 > 羞 is a vn, so this may already be the entire 所 phrase.     所羞而不為也 > But it is not, since 而 subordinates VP and does not follow NP: we have to look further for at least one more VP.      所羞而不為也 > This may well be it: the 所 phrase looks complete.        所羞而不為也 > This is the second element we were looking for: 也 marks nominal P, thus it naturally follows NP: the 所 phrase is thus clearly marked off: 所羞 而不為.         We may then analyse the words left of the 所 phrase as above. Finally, the most reliable signal word—analogous to 有/無 in the case of 者 phrases—is 者 to the right of 所. In such cases, 者 functions as the last element of the phrase, marking its end, as in: 人之所不學而能者,  其良能也 (Meng 7A15). Thus, always look for a 者 before further analys­is. As we have seen in 12.2, despite the addition of 者 such constructions are nevertheless 所 phrases. However, syn­tactical analysis (not translation!) works just as well if we treat them like 者 phrases and proceed from right to left:

145

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 3 : a n a lys i n g co m p l e x n o u n p h r a s e s

  人之所不學而能者 > By the rationale outlined above, all this must belong to the (pseudo-) 者 phrase.    人之所不學而能者 > Since 所 substitutes for an object of a v, which it precedes, it cannot precede a 者 phrase. It follows that 所 also belongs to the content of the (pseudo-) 者 phrase.     人之所不學而能者 > 之 subordinates NP, so it may perfectly well precede a (pseudo-) 者 phrase. We have thus identified it: 所不學而能者.

3  Nominalized clauses Nominalized clauses, functioning as subject or object, are frequent ­phenomena, and it is crucial to recognize them. Watch for the telltale construction NP 之 VP 也 or其VP 也, which are reliable indicators of a nominalized clause. Especially subject clauses are easily identifi­able by a 也 which is not in final position: 西河之 為秦也不久矣。

That Western He belongs to Qin will not last long. (Lü 20.8)

Once the signal words 之 and 也 have been identified, so has the subject clause: it includes everything from 西河 to 也. Bear in mind that the construction NP 之 may be replaced by 其: 其   誣也   久矣。

Their fraudulence has been going on for a long time. (Zuo 10.29)

Observe the expression 久矣, which follows the subject clause in both examples. Such favourite expressions may also serve as indicators for nominalized clauses. This is especially useful for object clauses, which usually appear after 知, 見, or verbs expres­sing hope and desire, or their opposites, fear and loathing, such as 欲, 望, 患, 恐, 惡, or 怨. Even in the absence of 之 or 也, such predicators should alert you to the possibility of an object clause: 不患   人之   不己知。 夫子何以知  其   將見殺。 安知 我不知魚之樂。

Do not fear that others do not know you. (LY 14.30) How did you know that he was going to be killed? (Meng 7B29) How do you know that I don’t know the fish’s joy? (Zhuang 17) 146

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 3 : a n a lys i n g co m p l e x n o u n p h r a s e s

English usage is very similar. Compare expression like ‘know that. . .’, ‘see that. . .’, ‘hope that. . .’, or ‘fear that. . .’ which are also typically followed by object clauses.

4 Exercises The following exercises may be more tricky than the examples given above: sometimes there are several possibilities for defining the extent of 者 or 所 phrases. Try to analyse the entire sentences, making use of the signal words discussed in Review 2. 王之所大欲可得聞與。

吾未見好德如好色者也。

③ ② ①

宋人有為其君以象為楮

葉者。



明主救天下之禍,安

天下之危者也。



賜也,女以予為多學

而識之者與。

民惟恐王之不好勇也。

所愛美女欲以其子為後。

是知其不可而為之者與。

吾怨其棄吾言。

此非臣之所敢任也。

⑩ ⑨ ⑧ ⑦ ⑥

References  ① Meng 1A7  ② LY 9.18  ③ HFei 21  ④ Guan 64  ⑤ LY 15.3  ⑥ Meng 1B3  ⑦ HFei 31  ⑧ LY 14.38  ⑨ GY 18.9  ⑩ SJ 76

147

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 13

Topics

13.1 Introduction148 13.2 Subject topicalization      S 者  P  (O) 149 13.3 Object exposure 149 13.3.1  Regular objects      O  S  P  之 149 13.3.2  Prepositional objects    NP S  P  (O)  焉 150 13.4 Absolute topics         NP S  P (O) 151 13.5 Exercises154

13.1 Introduction This lesson is concerned with clauses that topicalize a certain element in order to highlight it as the point of departure of the message. The topic typically occupies initial position in a sentence. Syntactically, it usually coincides with the subject, in which case the topic is generally unmarked. In some cases, however, a subject may be especially marked in order to give it special prominence as topic (compare sentences like ‘That car, is it yours?’). Moreover, the object of a sentence may be topicalized by exposure, that is by placing it at the beginning of a sentence (compare English ‘Scotch I like, but Bourbon I can do without.’). Such clauses deviate from the canonical word order of S P O, since the object is placed in front of the subject: they are non-­canonical clauses. Finally, a noun phrase may precede the subject without being grammatically connected to it (compare English ‘As for Germany, the beer is the best.’). Such ‘detached’ or absolute topics provide a frame for the following sentence, defining the extent of the subject matter to which it is relevant. In this lesson, we will discuss marked top­ic­ al­iza­tion of the subject, exposure of the object, and absolute topics. Preliminary questions: In your own everyday language, you probably use topics quite often. Can you think of examples? Are topics used in MSC? 148

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

13.3 o b j e c t e x p o s u r e

13.2  Subject topicalization A subject may be marked as topic—albeit in its normal initial pos­ition— by the particles 也, 者, or 則. Following the subject, these par­ticles cause a pause in the flow of the sentence, thereby stressing the subject and topicalizing it. The basic structure is: ① 鳥 則 擇 木。 The bird chooses the

S  A P  O

則 zé, conj as ptemp

tree.

擇 zé, vn, ‘choose’

則 here serves a contrastive function. It topicalizes the subject (‘the bird’) in order to distinguish it from that of the following sentence, which is: ‘How could the tree choose the bird?’ Observe the special usage of the particles: 也, exceptionally, does not mark a nominal clause; 者, unlike its standard usage, follows a noun phrase, not a verb phrase. ②

天地者,萬物之

父母也。



道者,令民與上

同意也。



凡官者,以治為

任,以亂為罪。



孝弟也者,其為

仁之本與。

此則寡人之罪也。



意 yì, nc, ‘will’ 凡 fán, pt, ‘generally’ 任 rèn, nc, ‘duty’ 孝 xiào, nc, ‘filial piety’ 弟/悌 tì, nc, ‘fraternal love’ 本 běn, nc, ‘root’

Comments ②–③ Clauses with topicalized S frequently have the character of definitions: therefore, they are often closed with 也. ④ Topicalized S may be preceded by the pt 凡, ‘in general’, which functions as an A indicating the scope of the utterance—it is not a mod to the S. ⑤ 也 and 者 may be combined without change in meaning. The expression 其 . . . 與 is equivalent to 其 . . . 乎, on which cf. 7.2. 13.3  Object exposure 13.3.1  Regular objects An object may be exposed in order to stress that the predicator applies to it and not to something else. In such cases, the object proper is placed 149

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 13 : to p i c s

in initial position, before the subject, while a coreferential ‘dummy’ 之 takes its place after the predicator. ⑦ 夏禮,吾 能 言 之。 The rites of the Xia

夏2 Xià, np, a dynasty

  O   S  A P O= I can talk about.

This sentence makes the same claim as the regular *吾能言夏禮. But by organizing the information in a different manner, it ‘says the same thing’ with another emphasis. ⑧

愛人者人恆愛之,敬人者

人恆敬之。

人之所教,我亦教之。

治國去之,亂國就之。

⑪ ⑩ ⑨

飢者食之,寒者衣之,

不資者振之。



不好犯上,而好作亂者,

未之有也。

子女玉帛,則君有之。

忠信者士死之。

⑭ ⑬

恆 héng, vi, ‘be constant’ 敬 jìng, nc as vn, ‘respect’ 亦 yì, conj, ‘also’ 就 jiù, vn, ‘approach’ 飢 jī, vi, ‘starve’ 寒 hán, vst, ‘be cold’ 衣2 yì, vn, ‘clothe’ 資 zī, vn, ‘provide’ 振 zhèn, vtr, ‘support’ 帛 bó, nc, ‘silk’

Comments ⑧ Cf. the alternative wording in Mo 16: 愛人者必見愛也. ⑨ As this example shows, O are often exposed in order to sidestep ‘bottom-­ heavy’ constructions like *我亦教人之所教. ⑩–⑪ If the S is missing, an exposed O may appear to be the S. Would this make sense in these examples? ⑫ Again, a very long O is exposed rather than placed in its usual position. ⑬ An exposed O may be followed by the ‘topicalizing’ particle 則. Note how the comma marks a rhythmic—not syntactic— caesura, dividing the clause into two halves of four-­syllables. ⑭ 死 is a vn, thus 死之 is not causative (‘make him die’), but pseudo-transitive: ‘die for it/him’ (compare 民死上, ‘the people die for their ­superiors’). 13.3.2  Prepositional objects The objects of prepositional phrases (cf. 8.1) may be exposed in the same way as regular objects. This occurs occasionally with the prepositions 為 and 以, but it is most common with 於. In such cases, the exposed 150

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

13.4 a b s o lu t e to p i c s

object is also resumed by a ‘dummy’ pronoun after 於, fusing with the latter into 焉 (on which cf. 8.2.5): ⑮ 萬   取 千 焉。Ten thousand—⟨if one⟩ takes a thousand from it…

NP ⟨S⟩ P  O  A

Note that 萬 in ⑮ is not a constituent in itself but only an exposed part of a constituent, namely the A; therefore, it is simply labelled as NP.



兩者大王何居焉。



雖小道,必有可

觀者焉。



諸侯之門,義士

存焉。

所惡於上,毋以

使下,所惡於下,

毋以事上。



雖 suī, conj as ptemp, ‘even’ 存 cún, vi, ‘exist’

Comments ⑯ 者, while serving to nominalize 兩, also marks the exposed O.  居, ‘dwell, settle’, here means ‘settle on/for’, hence ‘choose, accept’. ⑰ Before a NP, 雖 means ‘even’; it also serves to mark topical elements. ⑲ Here, the O of 以 is exposed; can you tell where it is resumed? On the force of 毋, cf. 4.3. 13.4  Absolute topics While topicalized subjects and exposed objects are syntactically integrated into the sentence, sometimes noun phrases can function as ­topics without any syntactic connection to the sentence. Such absolute ­topics (‘absolved’ from the syntactical structure, as it were) are still prevalent in Modern Standard Chinese (as in 花, 我最喜歡玫瑰花, ‘As for flowers, I like roses best’). They always precede the subject, without functioning as a modifier to the latter. Rather, they provide a frame of reference for the entire following sentence: ⑳ 里  仁 為 美。 As for the neighbourhood,

NP   S  P  C

kindness is the most excellent.

里 lı̌, nc, ‘neighbourhood’ 美 měi, vst, ‘be excellent’

As the example shows, absolute topics define the subject matter to  which the statement of the following clause—or, indeed, several clauses—applies. Such topics may also consist of nominalized clauses: 151

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 13 : to p i c s

禹之治天下 ㉑    ,   使 民心 變。As for the way 禹 Yǔ, np, mythical ruler ⟨S⟩ P 

NP

Yu governed 變 biàn, vtr, ‘change’ the realm, ⟨he⟩ made the people’s hearts change.

O  C

As the arrow indicates, the modifier of the absolute topic is the implicit subject of the clause. Thus the clause may also be translated ‘When Yu governed. . .’ , but it is important to bear in mind that this is not simply a temporal adjunct but that the entire nominalized clause is an absolute topic. Finally, the special case of nominalized clauses with 於 as predicator (!) deserves to be singled out: ㉒ 君子之於禮,  

敬 而 安 之。 As for the gentleman’s relation ⟨S⟩ A  A  P  O to the rites: ⟨he⟩ respects them and finds peace in them.

 NP

In this case, 於 retains its original status as a verb (cf. 8.2). Note the elegant pattern in this construction: the modifier of the absolute topic is the implicit subject, and its head is the object of the clause. ㉓

漢東之國,隨為大。



馬,蹄可以踐霜雪,

毛可以禦風寒。



喪之法,不過三年。



君子之學也,以美其身。



人之少也愚。



下之事上也,不從其所令,

從其所行。

古之聖王,其行若何。



漢 Hàn, np, a river 隨 Suí, np, a polity 蹄 tí, nc, ‘hoof ’ 踐 jiàn, vn, ‘tread on’ 霜 shuāng, nc, ‘frost’ 雪 xuě, nc, ‘snow’ 毛 máo, nc, ‘fur’ 禦 yù, vn, ‘repel’ 風 fēng, nc, ‘wind’ 喪 sāng, nc, ‘mourning’ 少 shào, vst, ‘be young’ 從 cóng, vn, ‘follow’ 令 lìng, vn, ‘command’ 古 gǔ, ntemp, ‘olden times’

152

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

13.4 a b s o lu t e to p i c s

Comments ㉓ 大 is a superlative. Just like the comparative (cf. 8.2.2), the superlative is not expressed morphologically but by relating the S to a reference set. To be ‘big’ with reference to a group means to be the biggest among it. ㉔ While at first sight, 馬蹄 could be construed as S, the second sentence makes clear that 馬 functions as a topic for both ­sentences. ㉕–㉖ Again, the subordinated element (here, 喪 and 君子, respectively) within the absolute topic is logically the S of the following sentence: it is the mourning that should not exceed three years, and the gentleman who adorns himself. Note also that in ㉖ the head (學) of the topical NP is resumed as O of the prep 以. ㉗–㉘ How are the elem­ents of the absolute topic logically connected to these sentences? ㉙ In some cases, absolute topics serve to avoid cumbersome double sub­or­din­ ations like *古之聖王之行.

Box 16  Fusion words Classical Chinese possesses a few words which result from the fusion of two morphemes. Like English ‘it’s’ < ‘it is’ or Modern Standard Chinese 這個 < 這 一個, or 別 < 不要, this is a result of fast speech. Recall that written Classical Chinese still ­cor­res­pond­ed rather closely to the spoken language! Fusion words are morphologically one word, but semantically they represent two words, and syntactically often two c­ onstituents. In such cases, they may not be assigned to a single word class; we can only call them ‘fusion words’: 弗 < 不 + 之 (*pət < *pəʔ + *tə), 之 being object pronoun 勿 < 毋 + 之 (*mət < *mə + *tə), 之 being object pronoun 焉 < 於 + prdem (*ʔan < *ʔa + *-n) 諸 < 之 + 於/乎 (*ta < *tə + *ɦâ), 乎 being prep or ptint

Other fusion words which only represent one constituent (namely, an adjunct) may be assigned to a word class. Just like 而已 may be treated as one pt, and 然則 as one conj, so the following fusion words may be conveniently classified as particles: 非 < 不 + 唯 (*pəi < *pə + *wi) 與/歟 < 也 + 乎 (*la < *laɁ + *ɦâ), 乎 being interrogative particle 耳 < 而 + 已 (*nəʔ < *nə + *ləʔ) 耶/邪 < 也 + 乎 (*la/ja < *laiʔ + *ɦâ) 夫 < 不 + 乎 (*ba < *pə + *ɦâ)

153

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 13 : to p i c s

By the same token, the following two, which are always used adverbially, should simply be called adverbs: 焉 < 於 + print (*ʔan < *ʔa + *-n) 盍 < 胡 + 不 (*gâp < *gâ + *pə)

Finally, there is the case of 然: 然 < 如 + prdem (*nan < *na + *-n)

Since it is regularly used like a stative verb, it should be classified thus. This reflects the fact that even in Classical times awareness of the origins of some words as fusion forms was fading and they were used like simple words (cf. note at 5.2.3). References: Kennedy 1940, 196–207; Yang Bojun and He Leshi 2003, 13–35.

13.5 Exercises Some more sentences to hone your analytical skills: ㉚

飢者弗食,勞者弗息。



君子之於物也,愛之

而弗仁;於民也,

仁之而弗親。



明君之道,使智者

盡其慮。



萬乘之國弒其君者,

必千乘之家。

北海則有走馬吠犬焉。



References  ① Zuo 12.11  ② Zhuang 19  ③ Sun 1  ④ Lü 17.3  ⑤ LY 1.2  ⑥ Meng 2B4  ⑦ LY 3.9  ⑧ Meng 4B28  ⑨ Lao 42  ⑩ Zhuang 4  ⑪ Guan 80  ⑫ LY 1.2  ⑬ Zuo 5.23  ⑭ SY 16  ⑮ Meng 1A1  ⑯ ZGC 14.17  ⑰ LY 19.4  ⑱ Zhuang 29  ⑲ Li 42  ⑳ LY 4.1  ㉑ Zhuang 14  ㉒ Xun 12  ㉓ Zuo 2.6  ㉔ Zhuang 9  ㉕ CQFL 2  ㉖ Xun 1  ㉗ Lü 1.4  ㉘ Li 33  ㉙ Yan 3.5  ㉚ Meng 1B11  ㉛ Meng 7A45  ㉜ HFei 5  ㉝ Meng 1A1  ㉞ Xun 9

154

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 10

Chunqiu The Chunqiu 春秋, ‘Springs and Autumns’, is probably the most per­ plexing of the Five Canonical Texts (cf. Focus 1). On the face of it, the Chunqiu is nothing more than anonymous, terse annals of the small polity of Lu, covering the years 722–481 bc. But since—centuries after its presumptive completion—it became associated with the name of Confucius, it has not ceased to fascinate the minds of Chinese scholars. Mengzi states it dramatically: ‘When Confucius completed the Springs and Autumns, unruly ministers and villainous sons were terrified’ (Meng 3B9). Indeed, the impact of these annals was enormous. Depite its brevity—less than 17,000 characters—the Chunqiu became eponymous for an entire historical period, and despite its laconic style it was a model for generations of later historians. In classical times, chunqiu became a byword for historical writing. The text itself, however, contrasts strikingly with this high esteem. Take, for example, the entries for the first year of duke Yin of Lu, 722 bc: First year, spring, the king’s first month. 元年,春,王正月。 三月,公及邾儀父盟于蔑。 Third month, the duke made a covenant with Yifu of Zhu at Mie.

夏,五月,鄭伯克段于鄢。 Summer, fifth month, the earl of Zheng defeated Duan in Yan.

秋,七月,天王使宰咺來歸 Autumn, seventh month, the Heavenly 惠公 仲子之賵。 九月,及宋人盟于宿。 冬,十有二月,祭伯來。 公子益師卒。

king sent the superintendent Xuan to come and present funeral presents for duke Hui and Zhongzi. Ninth month, ⟨the duke⟩ made a covenant with the people of Song at Su. Winter, twelfth month, the earl of Zhai came. Prince Yishi died.

And so it continues. Year after year is filled with such scanty entries concerning affairs of Lu and other polities, sometimes leaving the reader at a loss: ‘Fifth year, spring. The duke arrayed the fishermen in Tang’, ‘The 155

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 10 : c h u n q i u

Heavenly King hunted in Heyang’, ‘Caterpillars’, or ‘Autumn, seventh month’. Even the great translator James Legge (1815–97) could not sub­ due his ‘intense feeling of disappointment’ at this famed text: Instead of a history of events woven artistically together, we find a congeries of the briefest possible intimations of matters . . . without the slightest tinscture of literary abil­ ity in the composition, or the slightest indication of judicial opinion on the part of the writer. . . . The notices, for we cannot call them narratives, are absolutely unimpassioned. A base murder and a shining act of heroism are chronicled just as the eclipse of the sun is chronicled. So and so took place;—that’s all. (Legge 1872, 3)

This is all the more puzzling if one considers the stories concerning these ‘notices’ that are narrated in the Zuozhuan 左傳, an extensive work of history arranged as a commentary to the Chunqiu (cf. Lesson 20). For example, Zuozhuan gives us the background story to the note ‘the earl of Zheng defeated Duan in Yan’: a veritable drama including a breech birth, a court intrigue, an attempt at rebellion, and a civil war in Zheng. And the truth behind the Chunqiu entry ‘The Heavenly King hunted in Heyang’, according to Zuozhuan, is that the king was actually compelled by the marquis of Jin to visit him, which was nothing less than an ignominy. None of this is narrated in the Chunqiu. In fact, the Chunqiu is not a narrative text at all—it is not a history. It is an annalistic text that entirely lacks the kind of ‘emplotment’ through which ‘the chronicle is organized into a story’ (H. White). What is more, the Chunqiu is not even particu­ larly concerned with events. Dozens of its entries contain no event at all but simply note the season: ‘Autumn, seventh month.’ Whereas major events of Chunqiu history go unnoted in the Chunqiu, the text hardly ever fails to mention dates: all 242 years from 722 to 481 and almost all  seasons (except for two autums and four winters) are duly noted, no  matter whether anything happened or not. Almost a  quarter of the text—4,069 characters—consists entirely of dates. Arguably, these dates—not the events—are at the core of the text: the Chunqiu is not historiography, but chronography, it recounts the orderly passage of time. Not just any time, be it noted, but the dukes’ years and the kings’ months: chronography is not an end in itself, it seems to have had a legitimizing function. In this scheme, events are of secondary im­port­ ance; they are noted to mark the continuous passage of time, as it were.

156

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 10 : c h u n q i u

To be sure, this interpretation runs counter to two millennia of Chunqiu exegesis. Starting in Han times, scholars set themselves the task of explaining the ‘great meaning of the insignificant words’ (微言大義). Why exactly does the Chunqiu write ‘the earl of Zheng defeated Duan in Yan’? Two great commentaries, Gongyang zhuan 公羊傳 and Guliang zhuan 榖梁傳, are almost exclusively devoted to such questions. And, they come up with remarkable answers: it says ‘defeated’ because he killed him; he is called ‘Duan’ and not ‘younger brother’ or ‘prince’ in order to blame him; and it says ‘in Yan’, to express the distance: ‘this is like saying he took him from his mother’s bosom and killed him—it considers it terrible’. Arguably, it was through such commentaries that the Chunqiu rose to canonical status. By admitting ‘an excess of the signified over the signi­ fier’, as Foucault put it, the Chunqiu commentaries ‘give voice to a con­ tent that was not explicitly signified’, dooming us ‘to an endless task that nothing can limit’ (Foucault 2003, xvi). Until today, scholars debate the question of whether—and how—the Chunqiu conveys ‘praise and blame’. After more than 2,000 years, the Chunqiu has lost none of its fascination. Editions: The Chunqiu always comes with one of its three major commentaries. All three are contained in Shisan jing zhushu 十三經注疏, ed. by Ruan Yuan 阮元. repr. Shanghai 2009. Translations: The same applies. The classic translations by Legge 1872 and Couvreur 1914 have now been superceded by Durrant, Li, and Schaberg 2016. Studies: Kennedy 1964; Gassmann 1988; Nylan 2001, 253–306; Van Auken 2007, 2016.

157

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 14

Anteposition and inversion

14.1 14.2 14.3 14.4

Introduction158 Anteposition of objects      S O 之 P 159 Inversion of subject and predicate  P  A  S 161 Exercises163

14.1 Introduction Continuing the discussion of non-­canonical emphatic structures, we now turn to the anteposition of objects and the inversion of subject and predicate. Both constructions serve to stress a constituent by moving it further to the front of the sentence. Thus an object may be anteposed between the subject and the predicator in order to emphasize it, often in contrast to another conceivable object. (Compare, roughly, English sentences like ‘He has a job to do.’) Inversion we shall call the reversal of subject and predicate (not just the predicator, but the entire predicate, cf. 1.3). This typically occurs in spontaneous and/or evaluative comments, approximately like in the English ‘A total disaster it was!’ Note that both constructions differ from the topics discussed in the previous lesson: anteposed objects are not topics at all, nor do they take initial position; and inverted predicates, while topical by position, are not only the point of departure of the message but also its main point, the following subject appearing like a mere afterthought. This lesson will first discuss anteposed objects and then inverted predicates. The classic article on these constructions is Yang Shuda 1942; cf. also Yang Bojun and He Leshi 2003, 784–813.

Preliminary questions: Review the constructions discussed in Lesson 13. Are you sure you have understood those, so you can distinguish them 158

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

14.2 a n t e p o s i t i o n o f o b j e c t s

from what follows? Do you remember other cases of objects being placed between the subject and the predicator in Classical Chinese? 14.2  Anteposition of objects We have already seen how objects of prepositional phrases, in order to emphasize them, may be placed before instead of after the preposition (cf. 9.2.3). In a similar way, the object of a clause may be anteposed for reasons of emphasis by placing it before the predicator, in medial position. The anteposed object is subsequently resumed by a ‘dummy’ pronoun 之, 是, or 實 with purely syntactic function. The construction is:

① 子 是 之 學。

S  O O= P

You study that.

Note the syntactic difference between exposed (cf. 13.3.1) and anteposed objects: object placed before the subject; ‘dummy’ placed after the predicator. Anteposed: object placed after the subject; ‘dummy’ placed before the predicator. Exposed:

   O S P 之   S O 之  P

Anteposed objects seem to be a phenomenon of early CC; they are largely limited to a few texts, notably Zuozhuan, Guoyu, and Lunyu. By Han times, they seem to disappear from the language, surviving only in some stereotypical expressions like X 之謂, or 何 X 之 有 (cf. ⑬ and ⑭). They are enjoying a renaissance in MSC, however, where expressions like 唯利 是圖, 唯命是從, 唯德是舉, and others have become quite common.

Moreover, there is a decisive difference in meaning between the two. Exposed objects function as topics which are then followed by the main message; anteposed objects, quite to the contrary, are themselves the main part of the message. ‘You study that’—and nothing else. This message is often reinforced by addition of the particle 唯/惟, ‘only’:

②   唯 敵 是 求。Only the enemy ⟨we⟩ sought.

⟨S⟩ A O O= P

唯/惟 wéi, ptemp, ‘only’

In such sentences, the predicator is not negated with simple 不, but with 不唯 or 非 (which is a fusion of the former; cf. 3.3.1): 159

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 14 : a n t e p o s i t i o n a n d i n v e r s i o n

③ 是夫 也 將 不 唯 衛國 之 敗。 This man will

  S  A A A  A   O O= P

敗 bài, vn, ‘destroy’

not only destroy the polity of Wei.

Note the co-­occurrence of subject topicalization (是夫也) and an anteposed object (衛國): the former initiates the message, the latter is its nucleus. 吾不免是懼。

我且賢之用,能之使。

小國將君是望,敢不唯命是聽。

君人者將禍是務去。

何必公山氏之之也。

舍其舊而新是謀。

今王非越是圖。

鬼神非人實親,惟德是依。

非夫人之為慟而誰為。

夫子之謂也。

何懼之有。

⑭ ⑬ ⑫ ⑪ ⑩ ⑨ ⑧ ⑦ ⑥ ⑤ ④

且 qiě, conj as adv,‘be about to’ 務 wù, vn, ‘strive to’ 去2 qǔ, vn, ‘eliminate’ 公山 Gōngshān, np 之3 zhī, vn, ‘go to’ 舍 shě, vn, ‘discard’ 舊 jiù, vi as nc, ‘sth. ancient’ 新 xīn, vst as nc, ‘sth. new’ 謀 móu, vn, ‘plan’ 越 Yuè, np, a polity 神 shén, nc, ‘spirit’ 實3 shí, prdem, ‘it’ 親 qīn, nc as vst, ‘be close to’ 依 yī, vn, ‘rely on’ 夫 fú, prdem, ‘this’ 慟 tòng, vi, ‘grieve’

Comments ④ In this clause, 不 does not negate the P but is part of the O. What kind of an O is it? For the answer, reconsider 12.3. ⑤ If an adv is present, the anteposed O is placed between it and the P. 且 is a synonym of 將, expressing an immediate imminence of the action. ⑥ The second sentence is a rhetorical question: ‘How could we dare. . .’ 唯命是聽 is a stock phrase used to acknowledge an order. ⑦ In predicates with an adv and an aux, the anteposed O precedes the latter. ⑧ Note the force of this anteposition: ‘why, of all people, Mr. Gongshan?’ Can you make sense of the two 之? ⑨ Here, the anteposition is used to create a chiasmus. ⑩ The ntemp 今, ‘now’, in initial position is almost always an A, not 160

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

14.3 i n v e r s i o n o f s u b j e c t a n d p r e d i c at e

a mod of the following S (cf. 10.2–⑤). It often takes on a modal meaning, drawing attention to the point that follows; cf. English ‘Now, if I were you. . .’ Here, a verbal P is negated by 非: this implies a categoric denial, ‘by no means’, virtually dismissing the case (compare, loosely, the use of 不是 in MSC clauses like 我不是罵你). ⑫ This is a special case in which not the O of the clause is anteposed but the O of the prep 為, that is part of an A (compare 13.3.2). ⑬ The expression X 之 謂, ‘⟨this⟩ means/applies to X’, is one of a few stock phrases in which anteposed O survived in late CC and beyond. ⑭ Another stock phrase that lived on is 何 X 之 有, ‘what X could there be?’, or rather ‘how could there be X?’, construing 何 as an adv to 有 (for this usage, cf. 7.2). 14.3  Inversion of subject and predicate Inversion of subject and predicate is another type of non-­canonical structure in which the two superordinate constituents (cf. 1.3) simply change places. This construction typically appears in exclamations, evaluative comments, or spontaneous utterances. In verbal sentences, such inversions occur almost exclusively when the predicator is a stative verb. They are typically indicated by the fact that an emphatic particle appears in medial position. The structure is: ⑮ 大 哉,  堯之為君也。   Great, indeed, was the way

P A

S

哉 zāi, ptemp

Yao ruled!

Often, as in this case, the inverted subject is a subject clause (cf. 12.3). Thus the fronting of the predicate not only serves to emphasize it, but also to avoid unwieldy ‘top-­heavy’ clauses like *堯之為君也大哉 (compare constructions with an anticipatory ‘it’ in English: ‘It is delightful that you came to visit me.’). In nominal clauses, inversion is rare and largely restricted to clauses with interrogative pronouns as predicator: ⑯ 何 哉, 爾所謂達者。  What is it that you call

P A

S

achievement?!

達 dá, vi as nc, ‘achievement’

Again, a lengthy subject is being inverted in order to avoid ‘top-­heavy’ constructions. 161

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 14 : a n t e p o s i t i o n a n d i n v e r s i o n

甚矣汝之不惠。

⑱ ⑰

久矣哉,由之

行詐也。



異哉君之名

子也。

善哉問也。

易哉為君。

㉒ ㉑ ⑳

巍巍乎,其有

成功也。

君子哉若人。



惠 huì, nc as vi, ‘be benevolent’ 巍巍 wéiwéi, vst, ‘be magnificent’ 功 gōng, nc, ‘merit’

Comments ⑰ The perfective pt 矣, too, often marks an inverted predicate. ⑱ Here, two coordinated pt—矣哉—mark the predicate: they should both be reflected in translation. What is the function of 也 in this construction? 由 was the given name of Confucius’ disciple 子路 (cf. 11.3–⑳). ⑲ The pt 哉 may express surprise, disapproval (as in this case) or approval. ⑳ The inverted S may also be marked by a ptemp. ㉑ What kind of a S is this? ㉒ 乎 marks the inverted predicate; note that this is not a question. ㉓ How is 君子 in this clause to be analysed? Reconsider Box 10! Note that 若 here represents a prdem, not the prper ‘you’. Box 17  Mandatory anteposition Quite a few CC words, used as objects, regularly precede the predicator. In contrast to the constructions discussed in this section, such antepositions are not rhetorical devices but grammatically mandatory. Interrogative pronouns are a case in point: 吾誰欺。

Whom could I cheat? (LY 9.12)

子將奚先。

What will you do first? (LY 13.3)

The fusion word 焉, which is a PrP containing a print, also precedes the P, as in 君焉在, ‘Where is the lord?’ (Zuo 9.30).

The same rule applies when interrogative pronouns function as the object of ­prepositions: 天下惡乎定。

Whereby can the realm be settled? (Meng 1A6)

子何以知之。

How do you know that? (Zuo 9.31)

Personal pronouns are placed before the predicator only if the latter is negated: 君不吾聽。

The lord did not listen to me. (SY 1)

未之有也。

This has never occurred. (Meng 1A3)

162

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

14.4 e x e r c i s e s

The indefinite pronouns 有, ‘something, somebody’, and 無, ‘nothing, nobody’ always precede the predicator: 諸侯有討於鄭。

The feudal lords demanded something from Zheng. (Zuo 5.7)

耳無聞,目無見也。 His ears heard nothing, and his eyes saw nothing. (Meng 3B10)

Finally, one may include pronominal adverbs in this discussion, which may ­logically represent the object but always function adverbially: 乃歸自殺。

Then he withdrew and killed himself. (SY 4)

有無相生。

Being and non-­being produce one another. (Lao 2)

所 phrases are a somewhat different case. Although 所 substitutes for the object of a verb, which it precedes, a 所 phrase is one single NP. Hence it cannot be analysed as two distinct ­constituents, as in the cases discussed above.

14.4 Exercises Try to transform the following sentences into their canonical form. The last one is tricky; for a hint, cf. 11.2–⑧. ㉔

父母唯其疾之憂。



吾斯之未能信。



惟令是視。



何事之不勝。



野哉由也 。



久矣,吾不復夢見周公。

我文王之為子,武王之為弟。



References  ① Meng 3A4  ② Zuo 7.12  ③ Zuo 8.14  ④ Zuo 9.22  ⑤ HFei 33  ⑥ Zuo 9.28  ⑦ Zuo 1.3  ⑧ LY 17.4  ⑨ Zuo 5.28  ⑩ GY 19.3  ⑪ Zuo 5.5  ⑫ LY 11.10  ⑬ Meng 1A7  ⑭ Zhuang 12  ⑮ LY 8.19  ⑯ LY 12.20  ⑰ Lie 5  ⑱ LY 9.12  ⑲ Zuo 2.2  ⑳ Meng 1B4  ㉑ XX 4  ㉒ LY 8.19  ㉓ LY 5.3  ㉔ LY 2.6  ㉕ LY 5.6  ㉖ Guan 15  ㉗ Lü 3.4  ㉘ LY 13.3  ㉙ LY 7.5  ㉚ Xun 32

163

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Review 4

Non-­canonical clauses Non-­canonical clauses, such as discussed in Lessons 13 and 14, are among the most analytically challenging structures in Classical Chinese. All the more important it is to recognize and correctly construe them wherever they occur. This review will discuss some analytical strategies that may be helpful in dealing with absolute topics, exposed objects, anteposed objects, and inverted predicates. Before embarking on these, recall that non-­ canonical clauses are exceptional. Most clauses are canonical, and one should begin analysis on this assumption. Only if a clause cannot be construed in terms of canonical syntax, should a non-­ canonical construction be considered. 1  Adjacent noun phrases Watch for two adjacent noun phrases in initial position which are neither co- nor subordinated. The obvious case of a nominal clause (S P 也) aside, this is a reliable indicator of non-­canonical clauses. Since two such noun phrases must represent two distinct constituents, only one of them can function as subject. The other one can be either exposed object or absolute topic. Take the following sentence: 萬乘之國弒其君者,必千乘之家。 (Meng 1A1) 萬乘之國 and 弒其君者 are clearly two NP: they are neither ­coordinated (which would require a 與), nor are they subordinated (by  之). There is no explanation for this with reference to canonical clause ­structure.  > It follows that the second NP is the S, while the first NP is either exposed O or absolute topic.   > Since an exposed O would require a ‘dummy’ Opr in final ­position, which is not there, we have an absolute topic. Another example: 164

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 4 : n o n - c a n o n i c a l c l au s e s

不忠信之士,我將罪賤之。 (Mo 13) The NP 不忠信之士 cannot be construed as a mod to 我; neither are the two NP coordinated. They must represent distinct constituents.  > 我 must be the S, while 不忠信之士 is either exposed O or absolute topic.   > The Opr 之 in final position is likely to be a ‘dummy’. Since it makes good sense to construe it as referring to 不忠信之士, the latter is an exposed O. The construal is less obvious in clauses where the subject is not expressed after an exposed object or absolute topic. In such cases, there is no clear syntactic indicator of a non-­canonical structure: on the face  of it, 治國去之 (Zhuang 4) looks just like a canon­ical verbal clause. Only from the lexical meaning it becomes clear that 治國 is an exposed O. 2  Signal words Whereas exposed objects and absolute topics serve to topicalize certain elements, anteposed objects and inversions of subject and predicate are emphatic structures that lay stress on a constituent. For this purpose, the constituent is often not only transposed further to the front of the clause but also marked by emphatic particles. These may serve as signal words for identifying non-­canonical clauses. 唯/惟, in CC, is almost exclusively used to mark anteposed O.  Whenever you encounter 唯/惟 in a clause, look for a ‘dummy’ 之/是/實 to the right. If there is one, you may pinpoint an anteposed O between the two signal words. For example, in 父母唯其疾之憂 (LY 2.6), 其疾 must be an anteposed O. 之/是/實 may serve as signal words even in the absence of 唯/惟. When you encounter any of these in medial position, you should reckon with an anteposed O. Check whether they are preceded by a NP and followed by a VP, as in 君人者將禍是務去 (Zuo 1.3). In such cases, the NP preceding 之/是/實 can only be construed as an anteposed O. 哉 always marks the predicate, hence it usually belongs in final position. If it occurs anywhere else, the only explanation is that the predicate and the subject have been inversed, as in 野哉 ,由也 (LY 13.3). 165

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

r e v i e w 4 : n o n - c a n o n i c a l c l au s e s

矣 also always marks the predicate. Hence, the same logic applies: in 甚矣汝之不惠 (Lie 5), 甚矣 may be unambiguously identified as the inverted predicate. 3 Exercises Try to apply the criteria discussed above in order to analyse the following clauses. 彼唯人言之惡聞。

民不見德而唯戮是聞。

鳥之將死也其嗚也哀。

④ ③ ② ①

其子之肉尚食之,其誰

不食。

諸侯之禮吾未之學也。

晉大夫其誰先亡。

遠矣全德之君子。

何敵之不服。

不仁哉梁惠王也。

余唯利是視。

聖人,吾不得而見之矣。

久矣吾不復夢見周公。

⑬ ⑫ ⑪ ⑩ ⑨ ⑧ ⑦ ⑥ ⑤

昨日之事,子為制,今日

之事,我為制。

其三人則予忘之矣。



References  ① Zhuang 18  ② Zuo 5.23  ③ LY 8.4  ④ ZGC 22.3  ⑤ Meng 3A2  ⑥ Zuo 10.14  ⑦ Zhuang 54  ⑧ Lü 15.6  ⑨ Meng 7B1  ⑩ Zuo 8.13  ⑪ LY 7.26  ⑫ LY 7.5  ⑬ Lü 16.6  ⑭ Meng B3

166

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 15

Complex sentences I

15.1 15.2 15.3 15.4

Introduction167 Coordination          S P S 亦 P  S P 且 S  P 168 Subordination with 而       S P 而  S  P 170 Conditional clauses 172 15.4.1  Conditional conjunctions   若 S  P S P  S 如 P S P 172 15.4.2  Consecutive conjunctions   S  P  則 S  P 174 15.5 Exercises175

15.1 Introduction So far, we have focused on clauses, that is simple sentences consisting of one subject and one predicate. We now turn to units that consist of two or more clauses, which we shall call complex sentences. Complex sentences are the highest-­ranking unit of grammatical analysis. Just like the elements of noun or verb phrases, the clauses of a complex sentence may be coordinated or subordinated. In the first case, clauses of equal rank are simply juxtaposed, as in English ‘He tried hard, but he failed’. In the case of subordination, the two clauses enter into a hierarchical relationship, as in ‘Although he tried hard, he failed.’ In this case, the sentence consists of an independent main clause (‘he failed’) and a subordinate clause (‘Although he tried hard’). Just like adjuncts, such subordinate clauses provide information about circumstances—conditions, time, and the like—which may or may not be indicated by conjunctions. This lesson will first discuss coordinated clauses, before turning to simple subordination with 而 and conditional/consecutive clauses. Preliminary questions: What types of subordinate clauses (conditional, relative, etc.) do you know from your mother tongue? Which ones would you expect in Classical Chinese and which ones not? 167

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 15 : co m p l e x s e n t e n c e s i

15.2 Coordination In principle, adjacent sentences which are not subordinated are co­ord­ in­ated. Such coordination is not necessarily indicated by conjunctions or other explicit features. In fact, the preceding lessons have already stealthily included such sentences among the examples and exercises; reconsider the following case: ① 君子 和 而 不 同,

小人 同 而 不 和。   S  A A  A P

Gentlemen harmonize without being alike; petty men are alike without harmonizing.

These are clearly two clauses which are coordinated to form a complex sentence. Observe how the parallelismus membrorum serves to link the clauses together: parallelism is an important device for constructing complex sentences (cf. Focus 21). Moreover, absolute topics, which ­pertain to successive clauses indicate that these should be construed as  one complex sentence; typically these clauses also exhibit parallel structures: ② 馬,蹄 可以 踐 霜雪,

  毛 可以 禦 風寒。 NP S  A P  O

As for horses: their hoofs can tread on frost and snow; their fur can resist wind and chill.

Having reviewed these features, the rest of this section will focus on the use of coordinative conjunctions that link successive clauses. Such conjunctions include 且, 亦, and 又, meaning ‘and’, ‘also’, and ‘moreover’: ③ 舜 人 也,我 亦 人 也。

S  P A  S A  P A

Shun was a man; I, too, am a man.

The coordinative conjunction occurs in the second (or following) clause(s) of a complex sentence. In the case of 亦 and 又, it follows the subject.

168

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

15.2 co o r d i n at i o n



知者不失人,亦不失言。



彼其道遠而險,又有江山。



事不可必也,且政在焉。



我不欲人之加諸我也,

吾亦欲無加諸人。

學,行之,上也。言之,

次也。教人,又其次也。



險 xiǎn, vst, ‘be dangerous’ 又 yòu, conj, ‘also’ 江 jiāng, np, as nc, ‘river’ 加 jiā, vn, ‘impose’ 次 cì, nc, ‘the next’

Comments ④ In addition to the common topic—知者—these clauses are additionally linked by 亦 to form a complex sentence. Whereas in ③ it referred to the S (‘I, too’), it here refers to the O of the second clause (‘nor words’). ⑤ Here, 又, ‘and also’, coordinates two clauses with a common S.  In this NP, we encounter another function of 其: it serves to subordinate prdem (which is not done with 之), expressing a possessive genitive: 彼其道, ‘its way > the way thither’; cf. Box 18. ⑥ The conj 且 introduces an additional aspect to the preceding clause in the sense of ‘moreover, besides, in addition’. It is distinguished from the adv 且 (cf. 14.2–⑤) by its position before the S.  ⑦ In addition to the conj, the two clauses are connected by the Opr, which in both clauses (where is it?) refers to the same thing. Confucius’ Golden Rule may be translated as subordinated clauses (‘If I do not want . . ., then . . .’), but syntactically, they are coordinated. ⑧ As in ③, these are nominal clauses linked to form a complex sentence. The last clause is the longest and the only one, in which a conj appears, capping off the complex sentence: this stylistic device follows the law of increasing constituents.

169

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 15 : co m p l e x s e n t e n c e s i

Box 18  Functions of 其 The character 其 may represent two different words, both of which have important syntactic functions. Firstly, and most importantly, it represents a personal pronoun which only functions as a modifier: 勿奪其時。

Do not steal their time. (Meng 1A3)

In this function, 其 expresses a ‘possessive genitive’: the head ‘belongs’ to an aforementioned noun, in other words, it is ‘proper’ to it. Indeed, it can also express a sense of appropriateness, as in 非其人, ‘. . . is not the right man [for the job]’ (Meng 7B11), or 非其道, ‘. . . is not the right way [to do it]’ (Xun 1). Being the equivalent of NP 之, the prper 其 may also serve to nominalize verbal clauses: 夫子何以知其將見殺。

How did you know, master, that he would be killed? (Meng 7B29)

The same word may also function as a particle which subordinates demonstrative pronouns, expressing a ‘possessive genitive’: 是其故何也。

What is its reason? (Mo 8)

此其道出乎一。

His way proceeds from oneness. (Xun 8)

若是其大乎。

Does it reach its size? > Is it that big? (Meng 1B2)

In this function, 其 is easily recognizable as a particle by its position after a demonstrative pronoun. Secondly, 其 may represent a modal adverb which is regularly placed between the subject and the predicator: 吳其亡乎。

Wu will probably be destroyed. (SY 9)

In this function, 其 regularly goes together with the interrogative particles 乎 or 與, expressing suggestive statements. References: Cikoski 1978b; Malmqvist 1981; Unger 1995–97, vol. 7: 60–70.

15.3  Subordination with 而 We have already seen that within a clause the conjunction 而 sub­or­din­ ates verb phrases (cf. 7.6). By the same token, it may also subordinate entire clauses, including a subject. The decisive difference between both constructions is that in the case of subordinated clauses, the subject changes. The basic form is: 170

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

15.3 s u b o r d i n at i o n w i t h

⑨ 民 安 而 國 治。

 S  P  A  S  P



⟨If ⟩ the people are at peace, the state is orderly.

If the main clause has no overt subject, the construction may resemble a simple sentence with an adjunct subordinated by 而. But if the context suggests that the subject changed, we have to reckon with a complex sentence. Sentences in which the object of the subordinate clause is identical to the subject of the main clause are a case in point; in such constructions, the latter is frequently deleted:

⑩   買 肉 而  臭。 ⟨One⟩ buys meat,

⟨S⟩ P  O  A ⟨S⟩ P

and ⟨it⟩ stinks.

臭 chòu, nc as vi, ‘stink’

Since Classical Chinese had no 3rd person pr to be used as S (cf. Box 5), the alternative was to repeat or—more elegantly—to omit the S. While 而, as a rule, subordinates clauses, it does not specify their ­logical relation: it might be conditional (as in ⑨) or temporal, and it often has an adversative notion, meaning ‘but, however’ (cf. 7.6–㉘). The specific meaning can only be determined by the context. ⑪

本立而道生。



人其盡死,而我獨存乎。



蒼梧之弟娶妻而美好。

七十者衣帛食肉,黎民

不飢不寒,然而不王者,

未之有也。



盡 jìn, vi as adv, ‘all’ 獨 dú, vi as adv, ‘alone’ 蒼梧 Cāng Wú, np 娶 qǔ, vn, ‘marry’ 好2 hǎo, vst, ‘be good’ 衣3 yì, vn, ‘wear’ 黎民 límín, nc, ‘the (blackhaired) people’ 然而 rán’ér, conj, ‘this being so . . . but’

Comments ⑫ Observe how 其 . . . 乎 spans the entire complex sentence, linking subordinate and main clause. ⑬ The woman was beautiful, not the man: the S of the main clause is omitted. (Compare ⑩.) ⑭ Frequently, 然而 is used as a reinforcement of simple 而, always in an adversative sense: ‘(this being so, but >) however, nevertheless’. Note that everything from 七十 to 不王 is nominalized by 者: what follows from that for the analysis of the complex sentence? 171

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 15 : co m p l e x s e n t e n c e s i

15.4  Conditional clauses Conditional clauses express a state of affairs that is necessary before the statement of the main clause is possible. They may be subordinated to the main clause asyndetically (compare English ‘No pain, no gain.’) or, more typically, by the use of conjunctions. In the first case, sub­or­din­ ate clause and main clause are simply juxtaposed: ⑮ 脣 亡 齒 寒。

S  P  S  P

⟨If ⟩ the lips are gone, the teeth freeze.

脣 chún, nc, ‘lips’

齒 chı,̌ nc, ‘teeth’

While in such cases the logical relationship between the two clauses must be inferred, their syntactical relation is determined by the rule that modifiers precede that which is modified, hence the subordinate clause precedes the main clause. Mostly, however, the hierarchy of clauses is explicitly marked by conditional and/or consecutive conjunction. In what follows, these cases will be discussed separately. For studies of conditional clauses, cf. Harbsmeier 1981, 229–87, and Herforth 1994.

15.4.1  Conditional conjunctions The most important conditional conjunctions are 如, 若, and 而, which are etymologically related, as well as 苟 and 使 (compare the MSC equivalents 如果, 假若, and 假使). Expressing different nuances of ‘if, provided (that), supposing (that)’ and the like, they may occur before or after the subject of a conditional clause: ⑯ 若 我 往,晉 必 患 我。If we go, Jin will surely

A  S  P  S  A  P O

cause us trouble.

⑰ 彼 如 出 之,  可以 得 荊。 If they hand

S  A  P  O ⟨S⟩  A  P  O

him over, ⟨we⟩ can obtain Jing.

晉 Jìn, np, a polity

出2 chū, vtr, ‘hand over’ 荊 Jīng, np, a polity

若 *nak is derived from 如 *na(h) by addition of a suffix *-k. and The latter seems to have been a Lu dialect word, since it is frequently used as a conjunction mainly in Lunyu (which, in turn, never uses 若) and Mengzi. In all other classical texts, 若 is the conjunction of choice. Cf. Karlgren 1926.

172

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

15.4 co n d i t i o n a l c l au s e s

Negative conditions are expressed by the conjunctions 非 or 微, which both mean ‘if not, unless’ or ‘if it were not for, without’. Whereas 微 is always placed in initial position, 非 may either precede or follow the subject. ⑱ 非 吳 喪 越, 越 必 喪 吳。 If Wu does not

destroy Yue, Yue will certainly destroy Wu.

A  S   P O  S  A  P O



若子死,將誰使代子。



子產而死,誰其嗣之。



人而無信,不知其可也。



君苟有信,諸侯不貳,

何患焉。



使死者無知,則已矣。



微君言,臣故將謁之。

微禹,吾其魚乎。



非2 fēi, vst as conj, ‘if not’ 喪2 sàng, vtr, ‘destroy’

代 dài, vn, ‘replace’ 嗣 sì, vn, ‘succeed’

苟 goˇu, conj, ‘if really’ 貳 èr, vn, ‘split loyalty’ 微 wēi, conj, ‘if (it were) not (for)’ 謁 yè, vn, ‘announce’

Comments ⑲ For the structure of the main clause, cf. 11.3–⑲. ⑳ Being an un-­emphatic form of 如 (cf. note at 7.6), 而 is always used enclit­ ically, following the S. Note that in this function 而 follows a NP, not a  VP. There is an interesting parallel to this sentence in Lü 16.5: 子產若死,其使誰嗣之。㉑ In some cases, conditional clauses with 而 may simply be translated as NP modified by a relative clause: ‘if a man lacks sincerity’ > ‘a man who lacks sincerity’. (Compare ⑬, which may well be rendered ‘. . . a wife that was beautiful’.) ㉒ The conj 苟 may precede or—more frequently—follow the S. It marks a minimal condition, ‘once’, ‘if only’, often expressing a wish. In this case, two clauses f­ollow the conditional clause, which may either be construed as sub­or­din­ated (‘and if . . . then’) or as coordinated (‘then . . . and then’). ㉓ As vn, 使 means ‘cause, make s.o. do’; from this, it became used as a conj meaning ‘provided that, assuming that’, or simply ‘if ’ and usually marking counterfactual suppositions. Rare in early classical texts, this usage is frequent in late Zhanguo and Han texts. ㉔ 微 expresses negative counterfactual suppositions: ‘if not’. The corresponding main clause often has adv like 173

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 15 : co m p l e x s e n t e n c e s i

將, 且, 故, stating that something ‘would (surely)’ happen or have happened. ㉕ 微 may simply be followed by a NP, in which case it may be translated as ‘if it were not for, without’. The ­legendary ruler Yu 禹, successor of Yao and Shun and the alleged founder of the Xia dynasty, was credited with having drained a disastrous deluge that flooded North China. Note that 魚 should probably be ­construed as a verb in this ­sentence. 15.4.2  Consecutive conjunctions Relations between clauses may also be expressed by consecutive conjunctions like 則, 乃, and 斯, which syntactically belong to the main clause. They may express consecutive relations in the double sense of ‘following from’ or ‘following after’, just like the English ‘then’ may ­indicate logical or temporal relations.

㉖ 名不正,則言不順。 ⟨If ⟩ names are not correct, then words are

S  A  P 

not appropriate.

 A S A  P

Similar to the English adv ‘thus’ (< OE pr ‘thæt’), OC prdem seem to have acquired consecutive sense: 斯 is still used as a pr in classical texts, 則 may derive from the pre-­classical prdem 茲, and 此 is also sometimes used as a consecutive conj. 必 could be construed as a conj in ⑯ and ⑱; but considering its purely adverbial usage in clauses like 大國之樹必巨 (SY 12), we shall treat it as an adv (cf. 4.3).



 . . .

上好禮,則民易使也。



王如善之,則何為不行。



何為則民服。



禹往見之,則耕在野。



我欲仁,斯仁至矣。



得其民,斯得天下矣

得其心,斯得民矣。

數奪民時,大饑乃來。



服 fú, vi,  ‘submit’ 耕 gēng, vn,  ‘plough’ 野 yě, vi, as nc,  ‘uncultivated land’ 斯 sī, prdem as conj,  ‘then’ 數 shuò, adv,  ‘repeatedly’ 饑 jī, nc,  ‘famine’ 乃 nǎi, conj,  ‘thereupon’

Comments ㉗ The conj 則 always precedes the S, while the ptemp 則 follows it (cf. 13.2). As a noun, 則 means ‘rule’, hence as a consecutive conj it often expresses consequences that follow ‘as a rule’; it typically follows 174

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

15.5 e x e r c i s e s

unspecific, generic conditional clauses (cf. also ㉖), quite unlike the ones with 如 or 若, which usually mark specific conditions. ㉘ Here, both a conditional and a consecutive conj are used. If the S is omitted, 則 must be construed as a conj (since a ptemp could hardly topicalize a constituent that is omitted). ㉙ Just like the consecutive clause (in ㉘), the conditional clause may be a question; in such cases, 則 should be translated as ‘so that’ or the like. What is the difference between the two 為 in ㉘ and ㉙? ㉚ Here, 則 takes on a temporal meaning which translates as: ‘When Yu went, . . .’ This seems to be a rare example of 在 used as a prep. ㉛ Mainly in Lunyu and Mengzi—texts which were probably influenced by Lu dialect—斯 functions as a consecutive conj. Like 則, it precedes the S, but its force seems somewhat stronger than 則; in this example, one could even translate ‘et voilà’. ㉜ Again, 斯 has a forceful ring: ‘Once you gain its people, then. . .’ Note how the clauses are linked by the pr 其 in the subordinate clause, which refers to the O of the main clause. ㉝ 乃 *nə̂ʔ is perhaps an emphatic form of 而 *nə, hence its meaning may have developed from ‘thus (in this manner)’ to ‘thus (as a result)’. It is used as a conj especially in later classical texts; in c­ ontrast to 則 and 斯, it follows the S. Note that the adv 數 is pronounced shuò. 15.5 Exercises Insert fitting conjunctions or adverbs next to the dotted lines. ㉞

君好之,民



不辜。 ......

欲之。 ......

周宣王殺其臣杜伯

不吾告。 ......

用天下而有餘; ...... 175



吾問焉



無為也,

小子何述焉。 ......

為天下用而不足。 ......

不言, ......

有為也,



趙孟死,為政者其韓子乎。 ......





OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 15 : co m p l e x s e n t e n c e s i

References  ① LY 13.23  ② Zhuang 9  ③ Meng 4B28  ④ LY 15.8  ⑤ Zhuang 20  ⑥ Zuo 10.25  ⑦ LY 5.13  ⑧ FY 1  ⑨ HFei 6  ⑩ HNan 17  ⑪ LY 1.2  ⑫ Zhuang 11  ⑬ SY 3.21  ⑭ Meng 1A3  ⑮ Zuo 5.5  ⑯ GY 5.8  ⑰ HFei 23  ⑱ Lü 14.5  ⑲ HFei 22  ⑳ Zuo 9.30  ㉑ LY 2.22  ㉒ Zuo 10.13  ㉓ GY 19.9  ㉔ HFei 36  ㉕ Zuo 10.1  ㉖ LY 13.3  ㉗ LY 14.41  ㉘ Meng 1B5  ㉙ LY 2.19  ㉚ Zhuang 12  ㉛ LY 7.30  ㉜ Meng 4A10  ㉝ Lü 26.3  ㉞ Li 33  ㉟ Mo 8  ㊱ Zhuang 19  ㊲ Zhuang 13  ㊳ Zuo 9.31  ㊴ LY 17.16

176

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 11

Shijing The Canon of Odes, Shijing 詩經, is perhaps the most important, the most widely quoted and certainly the most delightful of the Five Canonical Texts. It is a collection of 305 songs traditionally dated to the eleventh to seventh centuries bc. However, recent scholarship suggests that it more likely dates to the Chunqiu period. The received text is divided into four parts, namely: Nos. 1–160 Nos. 161–234 Nos. 235–265 Nos. 266–305

Guofeng 國風, ‘Airs of the States’ Xiaoya 小雅, ‘Minor Odes’ Daya 大雅, ‘Major Odes’ Song 頌 ‘Hymns’

The Guofeng are the most ‘popular’ among the odes. They resemble folk songs about love, grief, everyday life, and festivals from fifteen different regional polities. The following sample is No. 6, ‘Tao yao’ 桃夭: 桃之夭夭,  灼灼其華。 The peach tree is young and elegant; Brilliant are its flowers.

之子于歸,  宜其室家。 This young lady is going to her future home, And will order well her chamber and house.

桃之夭夭,  有蕡其實。 The peach tree is young and elegant; Abundant will be its fruits.

之子于歸,  宜其家室。 This young lady is going to her future home, And will order well her house and chamber.

桃之夭夭,  其葉蓁蓁。 The peach tree is young and elegant; Luxuriant are its leaves.

之子于歸,  宜其家人。 This young lady is going to her future home, And will order well her family.

(tr. Legge 1871, 12f )

177

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 11 : s h i j i n g

This little ditty about a girl which is going to be married displays several main features that are characteristic of the Odes in general. (1) The lines normally consist of four syllables (a preference that is reflected even in classical prose, cf. 6.2.2); (2) stanzas often consist of quatrains, in this case with a rhyme in every other line; (3) rhymes usually change with every new stanza; in this case the rhyme scheme is: 1. 華 *wrâ 家 *krâ

2. 實 *m-­lit 室 *lhit

3. 蓁 *tsrin 人 *nin

For the importance of these rhymes in reconstructing OC, cf. Focus 3. While the rhymes may still apply to CC, the grammar of the Shijing certainly does not. Note, for example the use of 之 as a prdem in 之子于歸, as well as that of 于, which apparently functions as P. This is pre-­Classical Chinese, or rather: the language of the Shijing, which is sui generis.

The same features also apply to the other parts of the Shijing. The two Ya sections contain courtly odes, often in praise of the early Zhou kings. The Song are hymns of praise from Zhou, Lu, and Shang, which were perhaps sung during ceremonies and ancestor worship. We can only speculate about their original sitz im leben, since the uses of the Odes had changed entirely by classical times. The anthology was allegedly compiled by Confucius from a corpus of 3,000 songs (note the similarity to Shujing lore, Focus 6). Although this is rather unlikely, the songs were indeed very popular in the Confucian school. ‘My children,’ Confucius reportedly admonished his disciples, why do none of you study the Odes? The Odes may serve to stimulate, to contemplate, to come together, and to express resentment. Keeping them close, you may serve your father, taking them far, you may serve your ruler. You may get to know the names of many birds, beasts, and plants. (LY 17.9)

Indeed, the Shijing contains a plethora of zoological and botanical names (like the ‘peach tree’) from different regions of China which may have been useful for aspiring officers. But there was much more to the songs. On another occasion, the master allegedly scolded his son: ‘If you do not learn the Odes, you will have no means of discourse’ (LY 16.13). Apparently, the master did not intend the odes to be sung but to be quoted in (political) discourse. Indeed, the Odes came to be an important rhetorical device in Zhanguo politics. Especially texts of the Confucian tradition like Mengzi and Xunzi frequently adduce lines from the Odes in order to bolster their arguments, and historical texts provide ­countless examples of such quotations in court debates. For example, during a meeting in Zheng in 546 bc, seven nobles recited (parts of) odes, which 178

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 11 : s h i j i n g

were then commented upon. One of these ­comments was: “Boyou will be executed: the odes express one’s intention, and his intention was to slander the ruler” (Zuo 9.27). This episode may illustrate the changed function of such odes in the classical age. Rather than being sung ­during festivities or ceremonies, they were recited or merely cited in political discourse. It came to be understood that ‘the odes express one’s intention’. With usage in political discourse came an entirely new interpretation of the odes. They were often linked to historical personalities or events and understood in terms of praise and blame not unlike the Chunqiu (cf. Focus 10). In this way, even the simple ‘Tao yao’ acquired deeper sense: it was thought to praise the halcyon days of king Wen in which all people married at the right age and in the proper season. It was certainly due to their political role that the Odes fell victim to the Qin burning of books. After the fall of the Qin, they resurfaced in four different versions: those from the states of Lu (魯詩) and Qi (齊詩) as well as those of a certain Han Ying (韓詩) and of Mao Heng (毛詩). Of these, only the last one has come down to us in its entirety. It has become the standard version of the Odes, which every Chinese knows. More than two millennia after Confucius, Chinese children still study the Odes, and they learn quite a few of them by heart. The Shijing is still the most important, the most widely quoted and the most delightful canonical text. Editions: Maoshi zhengyi 毛詩正義. In: Shisan jing zhushu 十三經注疏, ed. by Ruan Yuan 阮元. repr. Shanghai 2009. Translations that have stood the test of time include: Legge 1871; Waley 1937; von Strauß 1880; and Couvreur 1892. Studies: Granet 1919 is a classic; Karlgren 1942–46 assembles scholarly notes from millennia. Recent studies include Yu 1983; Saussy 1997; Schaberg 1999; Kern 2000; Hong Zhanhou 2002; Chen Zhi 2007; and Rusk 2012.

179

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 16

Complex sentences II

16.1 16.2 16.3 16.4 16.5 16.6

Introduction180 Concessive clauses      S 雖 P  S  P 180 Temporal clauses      S 既  P  S  P   S  P  S  遂 P 182 Causal clauses       S P S 故 P  S P 是以 S P 184 Final and consecutive clauses  S  P  使 S P  S P  以  P 185 Exercises186

16.1 Introduction Continuing the discussion of complex sentences, this lesson will intro­ duce other types of subordinate clauses that enter into a hierarchical relation with main clauses, namely: concessive clauses, which denote a situation or action that is contrary to expectation in the light of the main clause; temporal clauses, which refer to a time previous or simultaneous to that of the main clause (often also implying a causal relation); causal clauses, which express causes, reasons, or mo­tiv­ ations for the action or situation denoted in the main clause; as well as final and conclusive clauses, two closely related clause types which express the (putative or factual) result of an action. Preliminary question: While these lessons treat different types of sub­or­ din­ate clauses separately, in connected discourse they may occur in complex and variable combinations. Can you think of ways to combine different subordinate clauses in your mother tongue? 16.2  Concessive clauses Concessive clauses express a point that seems to run counter to the proposition of the main clause—it concedes that point, as it were—without 180

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

16.2 co n c e s s i v e c l au s e s

affecting the validity of that proposition (compare English ‘Although he failed, he never gave up trying.’). In Classical Chinese, such clauses are introduced chiefly by the conjunction 雖, usually placed after the sub­ ject (just like MSC 雖然). 雖 *swi is probably related to the ptemp 唯 *wi (Pulleyblank 1996, 156). Indeed, it typically functions as a ptemp when placed before the S or an exposed O, as in 雖我亦將非之 (Mo 16) ‘Even I would reject that’ (cf. also 13.3.2–⑰). Compare the relation between English ‘even though’ and ‘even’.

① 事 雖 大,   必 濟。Although the task is big, ⟨we⟩

S  A  P ⟨S ⟩ A P

濟 jì, vi, ‘succeed’

will certainly succeed.

Assertive adverbs like 必 often occur in the main clause, affirming the proposition in spite of the concessive clause. Other, less frequently used concessive conjunctions include 縱 (compare MSC 縱使), which is ­typically placed before the subject, and 猶, which usually occurs after the subject. In any case, the concessive clause always precedes the main clause. ②

齊 楚雖大,何畏焉。



寡人雖死,亦無悔焉。



予之,雖亡地亦得信。



今君雖終,言猶在耳。



雖楚有材,晉實用之。



雖與之縣,不亦可乎。



雖然,未聞道也。



縱子忘之,山川鬼神其忘諸乎。

臣猶知之,而況君乎。



悔 huı̌, vn, ‘regret’ 材 cái, nc, ‘(natural) resources’ 縣 xiàn, nc, ‘county’ 縱 zòng, conj, ‘even if ’ 川 chuān, nc, ‘river’ 況 kuàng, vn, ‘compare with’

Comments ② There are two possibilities to construe 何: which ones? ③ After concessive clauses, the main clause often contains an adversative conjunction: in this usage, 亦 means ‘nevertheless’. ④ Note that this sentence consists of three clauses: the first—and most sub­or­din­ate—is conditional, the second is concessive, the third is the main clause. ⑤ 猶, 181

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 16 : co m p l e x s e n t e n c e s i i

too, often occurs after concessive clauses. It may be construed either as an adv indicating the continuative aspect (cf. 10.5–㉝) or as an adversa­ tive conj, meaning ‘but, yet’ (cf. English ‘yet’, which can mean ‘still’ or ‘nevertheless’). Which interpretation seems preferable in this case? ⑥ An example for the conj 雖 in initial position. This construction may be chosen to emphasize the S; note that in the main clause, the S is equally emphasized: 實 here functions as a ‘dummy’ that reiterates the S (not the O, as in 14.2). ⑦ Observe the rather frequent expression 不亦…乎, which expresses a rhetorical question: ‘wouldn’t it be . . . though?’ In  this usage, 亦 functions as an adv expressing modality, similar to German ‘denn’. ⑧ 雖然, ‘although this is so’, is actually a concessive clause; however, it is so frequent in most classical texts (except Lunyu and Laozi) that it may have already been understood as a conj, ‘nev­ ertheless’, that introduces the main clause. ⑨ What do you notice about 諸乎? ⑩ In the subordinate clause, 猶 has a concessive mean­ ing, ‘if even’; in this usage, it is often followed by 況, in the main clause, expressing a  rhetorical question: ‘and does [the minister] compare to the ruler?’, meaning ‘how much more [does this apply to] the ruler?’ 16.3  Temporal clauses Temporal clauses relate the time of the situation denoted in the main clause to another time, expressed in the subordinate clause. The latter may be simultaneous with or previous to the time of the main clause. In both cases, a causal relation may also be implied (compare English ‘since’, French ‘comme’, and German ‘weil’); but this is secondary, whereas the primary meaning is temporal. The most important con­ junctions used in temporal clauses are 既, ‘after’, 未, ‘before’, and 及, ‘when, as soon as’. ⑪ 君 既 定 矣, 又 何 求。After you have been established, my lord,

S  A  P  A ⟨S⟩  A O  P

what more could ⟨I⟩ ask for?

Note that 既 appears in the same position within the clause as the adverb (cf. 10.4). Its func­ tion as a conj can only be inferred from its syntactic role within the complex sentence, namely its position within the subordinate clause.

Alternatively, temporal succession may be expressed by consecutive conjunctions in the main clause indicating that the situation or action is 182

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

16.3 t e m p o r a l c l au s e s

subsequent to that denoted in the subordinate clause. Such consecutive conjunctions include 遂, 乃, 既而, 然後, and 而後. ⑫ 王 怒, 遂 圍 蕭。The king was furious

S  P ⟨S⟩ A P O

and thereupon laid siege to Xiao.

遂 suì, vi as conj, ‘thereupon’ 圍 wéi, vtr, ‘lay siege to’ 蕭 Xiāo, np, a polity

It is no coincidence that most examples for temporal clauses are taken from historical ­literature, since temporal succession is a defining characteristic of historical narratives.



既欲其生,又欲其死。



且既為人君,而又為人臣,不如死。



楚子賜之金,既而悔之。



公未言,榮成子曰︙



及生,有文在掌曰友,遂以名之。



武王又再拜稽首,乃出。

事必成然後舉,身必安而後行。



既而 jì'ér, conj, ‘afterwards’ 榮成子 Róngchéng Zı,̌ np 及 jí, vn as conj, ‘when’ 文 wén, nc, ‘writing’ 掌 zhǎng, nc, ‘palm of the hand’ 拜 bài, vn, ‘bow’ 稽 qı ,̌ vn, ‘bow’

首 shǒu, nc, ‘head’ 然後 ránhòu, conj, ‘(only) then’ 舉 jǔ, vtr, ‘take up’ 而後 érhòu, conj, ‘(only) then’

Comments ⑬ As in ⑪, the conj 又 is opposed to the temporal conj 既. One may translate ‘first . . . and then’ or even ‘on the one hand . . . on the other hand’, construing the two clauses not as subordinated but as co­ord­ in­ated (which is the usual function of 又). ⑭ Note that the first two clauses are both subordinated to the third: what is the logical relation? For 不如, recall that S 不如 C means ‘S does not equal C’ (cf. 11.4), often in the sense of ‘S is not as good as C’. If the S is omitted, this leaves 不如 C, meaning ‘C is better’ or ‘it is better to C’. ⑮ As seen in ⑭, the clause with 既 may be formally subordinated by 而; perhaps from such con­ structions, the conj 既而, ‘afterwards, later on’, developed, which is placed in the main clause. ⑯ The conj 未, ‘before’, is the negative coun­ terpart of 既. ⑰ A double subordination with 及, ‘when’, in the first sub­ or­din­ate clause and 遂 in the main clause. Note the disposal construction 183

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 16 : co m p l e x s e n t e n c e s i i

with a C instead of an O (cf. 9.2.2). ⑱ 拜稽 means to ‘kowtow’. 乃 has already been discussed as a consecutive conj in a logical sense (cf. 15.4.2– ㉝); in historical narratives, it is mainly used in a temporal sense, ‘and then’. It marks an action that not simply follows but is consciously taken following the situation or action denoted in the subordinate clause. ⑲ 然後 and 而後 may be explained in analogy to 雖然 and 既而 (cf. ⑧ and ⑮); they often mark subsequent actions in the emphatic sense of ‘only then’. In connection with 必, one may translate ‘the project must be mature before it is undertaken’. 16.4  Causal clauses Causal clauses express relationships of cause and effect, reasons/cir­ cumstances and consequences, or motivations and results. Whereas in English, causal conjunctions like ‘because’ or ‘since’ are placed in the causal clause proper, Classical Chinese makes use of resultive con­ junctions in the main clause. The main resultive conjunctions are 故 and 因, which are both placed in initial position: ⑳

不 勝 其哀,故 哭 也。 ⟨I⟩ could not overcome the ⟨S⟩ A P O A ⟨S⟩ P A sorrow, therefore ⟨I⟩ wailed.

勝 shèng, vn, ‘overcome’ 哀 āi, nc, ‘sorrow’

Note that the ‘subordinate’ clause may actually function as an independ­ ent clause; therefore, one may equally consider the clauses as co­ord­in­ated.

184



吾少也賤,故多能鄙事。



王唯信子,故處子於蔡。



吾 唯 不 知 務 而 輕 用 吾 身, 吾是以亡足。



今者吾忘吾答,因失吾問。

三日不去朝,紂因而殺之。



鄙bı ̌, vst, ‘be lowly’ 處 chǔ, vi, ‘reside’ 足2 zú, nc, ‘foot’ 答 dá, nc, ‘answer’ 因 yīn, vn as conj, ­‘accordingly’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

16.5 f i n a l a n d co n s e c u t i v e c l au s e s

Comments ㉑ Note that 多, which logically must be taken as mod of the O, is syntactically used as A. This may be a case of hypallage, the deliberate misplacement of a modifier for stylistic reasons, as in ‘I lighted a thoughtful cigarette’ (P.G.  Wodehouse). This stylistic device is occasionally used to achieve a more elegantly balanced sentence; compare Li 4: 大為我棺,   ‘make me a big coffin’, or SJ 46: 秦又多予金, ‘Qin also gave ⟨him⟩ a lot of gold’. ㉒ The predicate of the causal clause may be emphasized by the ptemp 唯, ‘only (because)’, thus marking it as the cause for what follows. ㉓ Recall that the PrP 是以, too, means ‘therefore’ (cf. 9.2.3–⑯); therefore, it can replace a resultive conjunction. ㉔ The verb 因 means ‘accord with’, whence the usage as a conj, ‘accordingly, thereupon’; as such, it may have logical or temporal meaning. ㉕ 因 may be followed by 而 without appar­ ent change in meaning (in fact, the function as conj may have developed from the verbal usage in this way: 因 X 而P > 因 而 P > 因 P). 16.5  Final and consecutive clauses Final clauses express an end, that is a purpose or a result that one seeks, whereas consecutive clauses express the actual result of an action. Both types of clauses overlap in meaning and also in the conjunctions used (compare English ‘so that’, which may mark an intended as well as an achieved result). In Classical Chinese, these are 以, which seems to be used equally in a final or consecutive sense, and 使, which mostly denotes a purpose. Since final and consecutive clauses express circum­ stances or actions that result from those denoted in the main clause, they are placed behind the latter: ㉖

屏 王之耳目   , 使 不 聰明。 ⟨He⟩ shielded the king’s ears ⟨S⟩ P      A ⟨S⟩ A  P O and eyes, so that ⟨he⟩ could not see or hear clearly.

屏 bı ̌ng, vn, ‘shield’ 目 mù, nc, ‘eye’ 聰 cōng, vst, ‘have acute hearing’

The function of 以 as a conj may be explained from its prep meaning ‘by, with’ (cf. 9.2), in which function it may imply an Opr. This Opr refers to the predicate of the preceding sen­ tence, hence ‘thereby’, ‘by that means’ or ‘as a result of that’. The function of 使 as conj clearly derives from its verbal meaning ‘cause to’.

185

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 16 : co m p l e x s e n t e n c e s i i

後之人法,使無陷于惡。



臣聞聖王公之先封者,遺



鄭人游于鄉校,以論執政。



回也聞一以知十。

彼奪其民時,使不得

耕耨,以養其父母。



封 fēng, vtr, ‘enfeoff ’ 遺2 wèi, vn, ‘bequeath’ 後 hòu, nloc, ‘later’ 陷 xiàn, vi, ‘fall into’ 鄉 xiāng, nc, ‘district’ 校 xiào, nc, ‘barracks’ 論 lún, vn, ‘discuss’ 執政 zhízhèng, nc, ‘adminis­ trators’ 耨 nòu, nc as vn, ‘weed’ 養 yǎng, vtr, ‘feed’

Comments ㉗ The use of 使 implies a change of S. In general, its function as a conj overlaps with its function as a verb followed by O and C (cf. 11.3). ㉘ With 以 as conj, the S stays the same (and it is not repeated): S does something and thereby achieves something, or S does something in order to achieve something. ㉙ This is clearly a consecutive clause. ㉚ This may be construed as a double subordination of a consecutive and a final clause. But note that 以 occasionally also subordinates VP, as in 古之民樸以厚, 今之民巧以偽 (SJS 7), or 敬以知微 (GY 8.8). Hence, 使 . . . 父母 could also be construed as a single clause. Which makes more sense? 16.6 Exercises Translate the following complex sentences, starting with a structural analysis. ㉛ 雖有詩書,  . . .  猶無益於治也 。



愛人,故惡人之害之也。



若民,則無恆產,因無恆心。

186



是故三代雖亡,治法猶存。



不奪不饜。



苟不志於仁,終身辱,以陷

於死亡。

棄田以為園囿,使民不得衣食。



OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

16.6 e x e r c i s e s

References  ① Zuo 8.9  ② Meng 3B5  ③ Zuo 1.3  ④ Lü 19.7  ⑤ Zuo 6.7  ⑥ Zuo 9.26  ⑦ Zuo 10.28  ⑧ Meng 3A2  ⑨ Zuo 11.1  ⑩ GY 10.1  ⑪ Zuo 11.4  ⑫ Zuo 7.12  ⑬ LY 12.10  ⑭ Zuo 6.16  ⑮ Zuo 5.18  ⑯ GY 5.5  ⑰ SJ 33  ⑱ SJ 4  ⑲ YDan 1  ⑳ SY 6  ㉑ LY 9.6  ㉒ Zuo 10.15 ㉓ Zhuang 5  ㉔ Zhuang 23  ㉕ XX 7  ㉖ Zuo 10.27  ㉗ GY 4.3  ㉘ Zuo 9.31  ㉙ LY 5.9  ㉚ Meng 1A5  ㉛ SJS 3  ㉜ Xun 15  ㉝ Meng 1A7  ㉞ Xun  ㉟ Meng 1A1  ㊱ Meng 4A9  ㊲ Meng 3B9

187

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 12

The unification of writing Standard editions of Classical Chinese texts are usually set in modern type and in standardized orthography. However, texts of Zhanguo and Han times looked nothing like that. Not only were they written by hand on narrow slips of bamboo, displaying numerous calligraphic variations; but their orthography, too, differed signifi­cantly from modern standards. In fact, the very idea of a standard orthography was absent from pre-­Qin texts. With the spread of writing from the Zhou court to the regional polities of Northern and Southern China, orthography became increasingly diverse. By Zhanguo times, characters from Qi, Chu, Qin, and other states displayed marked structural differ­ences, and even within these states, scribal traditions would vary. Consider the specimens of characters from seven different states shown in Figure 5. As the examples show, these differences are not simply such of ­calligraphy, associat­ed with individual scribal hands. Rather, the internal structure of the characters—the number of strokes, the components, and their arrangement—differs in many cases: these are ortho­graphic variations. Such varia­tions even occur within single manu­scripts or text corpo­ra. For example, within a related group of cov­enant texts from Jin, the pronoun 其 appears in five variations; the verb 事 in seven ­variations; and the word ‘abdomen’ 腹 in at least twenty-­three variations (Galambos 2006, 46–54; cf. also Focus 19). In short, the associa­tion of graph to word was still very flexible. For later scholars this was a cause of some chagrin. Editing the Zhanguo ce, the Han bibliographer Liu Xiang (79–78 bc) complained that ‘the original charac­ters are often wrong, leaving out half of the character: thus 趙 is rendered 肖, 齊 is rendered 立, and many other characters are like this’ (ZGC, ‘Liu Xiang shulu’). In fact, ‘leaving out half of the character’ was rather common in pre-­Qin writing. As the example of 趙 and, in Figure 5, 陽 and 城 illustrate, characters were often written with their phonetic element only, leaving out the semantic classifier; or rather: these classifiers had not yet become an essential com­ponent of the characters in question. Recall that the so-­called ‘radicals’ 188

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 12 : t h e u n i f i c at i o n o f w r i t i n g

Qi character Chu character Yan character Han character Zhao character Wei character Qin character

Figure 5  Regional characters Source: Yu Weichao (ed.). A Journey into China’s Antiquity. Beijing: Morning Glory Publishers, 1997, vol. 2, 74.

were not at the root of character forma­tion (cf. Focus 4). Rather, the phonetic element was primary, and in Zhan­guo times it may have seemed conceivable that Chinese writing would develop into a purely phonetic script. But it never came that far. After the Qin had conquered all other regional states in 221 bc, one of the first measures they imposed was the unification of writing. They introduced the so-­called ‘small seal script’ (小篆), a tidy, well-­balanced, slightly curved script with evenly thin strokes. But the reform was not merely one of calligraphy but also of orthography: the diverse characters of the seven states were unified, ‘discard­ing those that differed from the writing of Qin’ (Shuowen, ‘Xu’). Now the components of every char­acter—semantic and phonetic— were supposedly standardized. While this did not affect the structure of simple characters very much, more complex characters were ‘sometimes greatly abbreviated or changed’ (ibid.). This reform caused quite a stir in Han times. Since the Qin not only abandoned old orthography but in 213 bc supposedly burned most ancient writings, virtually all surviving texts were rewritten in ‘new script’ (今文). When decades after the fall of the Qin some texts— Lunyu, Shangshu, Xiaojing, Zuozhuan, and others—written in ‘old script’ (古文) reappeared, they seem to have differed significantly from their ‘new script’ counterparts. Already in Han times, the question of which versions were more author­itative led to major scholarly controversies (with decidedly political dimensions), which would have repercus­sions even after

189

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 12 : t h e u n i f i c at i o n o f w r i t i n g

2,000 years. Critical scholars of Qing times revealed numerous forgeries within ‘old script’ editions, shaking the foundations of imperial orthodoxy; and in the nineteenth century, radical political reforms were advocated with explicit recourse to ‘new script’ texts.

To be sure, the Qin reform did not unify writing in one fell swoop; rather, the process of standardization continued well into Han times. But its cumulative effects were dramatic. As many as a quarter of the pre-­Qin characters were completely discarded, leaving no successors at all in the new script; others were significantly changed in their structure, obliterating differentiations that the old scripts had possessed. Since we do not know on which principles this stand­ardization was done, it is in many cases impos­sible to extrapolate ‘original’ character forms from texts that have been edited after the Qin unification of writing. By the same token, excavated texts from Zhanguo times are so hard to decipher: many characters defy transcription into modern regular script be­cause one or more of their components no longer exist or because the combination there­of has no equivalent in modern writing. Any attempt to modernize such old character forms carries the risk of seriously distorting them. Yet standard editions of Classical Chinese texts are usually set in modern type and in standardized orthography. While there seems to be no alternative to this editorial prac­tice, students of Classical Chinese should be aware that it is by no means unproblematic. Beneath the standardized facade of transmitted texts lies an unfathomable diversity of char­acter forms which, in pre-­Qin manuscripts, may have represented quite dif­ferent words. References: Barnard 1978; Galambos 2004, 2006.

190

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Part II

Having discussed the most important features of Classical Chinese grammar in Part I of this book, we will now move from sentences to texts and from the language to the literature written in Classical Chinese. Part II introduces seminal texts of Zhanguo and Han times, providing brief accounts of their history, structure, and contents, followed by selected passages representa­tive of the texts in question. The texts are presented as they come in modern editions (with occasional omissions, indi­cated by suspension dots), including full punctuation as well as trad­ition­al or modern commentaries. The focus thus shifts from intensive grammatical ana­lysis to more extensive reading. No longer centred around one grammatical feature, the following lessons will explain difficult constructions as they appear, relating them to the basic rules discussed in Part I. There will be few genuinely new constructions in these texts; rather, we will observe well-known phenomena, but also exceptional, anomalous or even corrupt passages. None of the texts presented in the following lessons are specially tailored ‘textbook examples’ that faithfully adhere to linguistic standards. While Part I introduced basic rules of grammar, Part II covers the whole range of linguistic expres­sion in Classical Chinese, including violation of the rules and—more im­port­ ant­ly—their evolution over time. The dating of almost all of the classical works is highly contested. Since they most likely were not written by single authors at one particular point in time but rather grew in an extended process of ‘productive transmission’, it often makes little sense to assign a date to entire works. In many cases, chapters or even paragraphs would seem to have been written at very different times (cf. Focus 17). Critical scholarship on these issues is assembled in Zhang Xincheng 1939.

All of this is meant to prepare students for dealing with such texts on their own. For this purpose, important tools for the study of Classical Chinese texts are introduced in this part. Commentaries—albeit significantly abridged for the present purposes—accompany every text in Part II; they are introduced as an all-important aid for dealing with classical texts. Commentaries tell us how eminent Chinese scholars

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

pa r t II

under­stood these texts (indeed, sometimes, how little they understood them): students of Classical Chinese cannot afford to ignore them. While the main phrases of commentary language will be explained in order to facilitate their usage (cf. Focus 14), the vocabulary of the commentaries is not included in the glossary or vocabulary lists. Since commentaries to Classical Chinese texts are written in Literary Chinese or MSC, the details of their grammar and lexicon are not within the purview of this book.

Moreover, several Foci will introduce dictionaries, the historical background, and methods of textual scholarship: these are the ­prerequisites for dealing in a scholarly way with the mani­fold problems that Classical Chinese texts pose at any level of analysis. Loewe 1993 gives concise introduc­tions to the contents, structure, history, and editions of sixty-four classi­cal works. Knechtges and Chang 2010–14 has entries for books, authors/ compilers, and genres and is espe­cially valuable for its bibliographies of scholarly studies. Translated selections from classical literature are pro­vided in De Bary and Bloom 1999.

192

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 13

Modern editions of classical texts Not a single classical text has come down to us the way it was first written down. Rather, all transmitted texts have undergone a long process of copying, commenting, and editing before they were first printed. But even those first prints, which mostly date to Song times (960–1279), are rarely available to students of Classical Chinese. Instead, we usually rely on readily accessible modern editions. Many standard editions of classical texts used by scholars are published by Zhonghua shuju 中華書局 in Beijing, especially those in the series Xinbian zhuzi jicheng 新編諸子 集成. These are mostly based on excellent editions by Qing scholars, critically revised, punctuated, and arranged in a way that makes them more easily readable than traditional editions. Figure 6 gives an e­ xample of the Xunzi jijie 荀子集解, Col­lected Explanations to the Xunzi, an ­edition originally compiled by Wang Xianqian 王先謙 (1842–1917) in 1891 and re-edited in 1988. Several points are to be noted when look­ing at this page: ① The book title appears on the right outer margin; the number of

the juan (卷) as well as the chapter (pian 篇) number and title appear on the left outer margin (not pictured). ② Page numbers are on the lower outer margins. ③ At the beginning of a juan, its number appears in boldface. ④ The smaller title gives the name and number of the chapter: these are either numbered continually from the beginning of a book (as in this case) or starting anew with every juan. A note on bibliographical units: Before the age of paper, juan probably referred to a silk roll, and pian meant a bundle of bamboo slips. In print editions, juan is the main division of a work, a ‘book’, as in ‘Book of Genesis’ (for reasons of clarity, however, we shall simply call it juan), and pian is the next subdivision, equivalent to a ‘chapter’. While juan are only numbered, pian usually have numbers and titles; they may be further subdivided into sections called zhang 章. When quoting classical texts, never simply refer to a page number, since these differ widely between editions. Rather, references should contain both the juan and the chapter (or, if applicable, chapter and section) numbers. In this example, a full reference, includ­ing page num­ber, would read Xunzi 17.23, 434.

193

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 13 : m o d e r n e d i t i o n s o f c l a s s i c a l t e x t s

Figure 6  A page from the Xunzi jijie Source: Wang Xianqian 王先謙 (ed.), Xunzi jijie 荀子集解. 2 vols. Beijing: Zhonghua shuju, 1996, 434.

194

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 13 : m o d e r n e d i t i o n s o f c l a s s i c a l t e x t s

References: The classic introduction to ancient Chinese books is Tsien Tsuen-hsuin 2004; Wilkinson 2018, 1033–7, gives a convenient glossary of bibliographical terms.

Note that this chapter title is immediately followed by a paragraph of commentary which, in this case, serves as an introduction to the chapter. ⑤ In modern editions, proper names are marked by a single solid

line.

⑥ In editions that contain more than one commentary (note the

title: Col­lected Expla­nations), the principal commentary may be separated from subsequent commen­taries by a circle. In this case, the principal com­mentary is by Yang Liang 楊倞 (ninth cent.), ­fol­lowed by that of Lu Wenchao 盧文弨 (1717–96), whose name is marked here.

The editorial principles as well as the works cited are always explained in the introductions to modern edi­tions: these should be read carefully in order to make the best use of these editions. For more detailed information about the structure of commentaries, cf. Focus 14.

⑦ This is a standard gloss, explaining the character 偽 in the sen-

tence above. 偽,為也: ‘ “artificial” means “made” ’. This kind of paronomastic explanation is often seen in traditional commentaries. ⑧ Here, the editor himself, Wang Xianqian, adds his own opinion. Such commen­taries by the editor are typically introduced by the expression 案, ‘I argue’. ⑨ Book titles and chapter titles are marked by an undulating line.

195

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 17

Lunyu

17.1 Introduction196 17.2 Learning (LY 1.1, 2.4, 16.13) 198 17.3 Morals (LY 4.15, 12.2, 17.5, 13.19) 202

17.4 Ritual (LY 12.1, 6.27, 3.15) 17.5 Junzi vs. xiaoren (LY 4.11, 13.23, 12.5)

204 206

17.1 Introduction Lunyu 論語, the Judgments and Sayings attributed to Confucius (551– 479 bc), is perhaps the most important text in Chinese or even East Asian tradition. The relatively short book—it contains just under 16,000 characters—is divided into twenty chapters and about 500 sections. These consist of sayings by or about Confucius as well as dialogues between him and his disciples or other contemporaries. Most of these pieces are presented out of context, typically beginning with a simple 子曰, ‘The master has said.’ However, there is no information about the Lunyu from Zhanguo sources. The text gained currency only in Han times, when it was attributed to Confucius’ immediate disciples who supposedly compiled it from their notes or memories after the  master’s death. Henceforth, the Lunyu was revered as a fundamental text of ‘Confucian’ thought: it was widely used as a primer, and by the second century ad it was officially ranked among the ‘seven canonical texts’ of the orthodox tradition. In the following centuries, dozens of commentaries to the Lunyu were produced, of which those by Zheng Xuan (鄭玄, 127–200), He Yan (何晏, 190–249), and Huang Kan (皇侃, 488–545) were the most influential. The veneration of the Lunyu reached its apogee when it was selected by Zhu Xi (朱熹, 1130–1200) as one of the ‘Four Books’ (四書) that were to 196

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

17 . 1 i n t r o d u c t i o n

become the canon of Neo-­C onfucian thought. From the Yuan period (1271–1368) onward, memorizing the Lunyu became a fundamental precondition for all scholarship. Modern Western scholarship takes a more critical stance on the Lunyu. Building on the achievements of critical Qing scholars like Cui Shu 崔述 (1740–1816) and others, it has made clear that the text was not compiled by a single author at one point in time. Rather, numerous inconsistencies and repetitions suggest that it is a collection of snippets from different sources that perhaps grew in successive layers over several centuries. As a book, the Lunyu is likely not older than 150 bc (the oldest manuscripts, found in Dingzhou, Hebei, and Lelang, North Korea, are dated roughly a century later). Moreover, at this time there were three different versions of the text—one of Lu, one of Qi, and one written in ancient script—which were later conflated into the transmitted edition. These findings have diminished the Lunyu’s value as a source for the life and thought of Confucius: not only does it contain pieces that probably originated centuries after Confucius’ death, but sayings of the ‘master’ in other works might be equally or even more authentic witnesses to his teachings. While modern scholarship thus endeavours to deconstruct the Lunyu, the book has made an overwhelming comeback in the Chinese public: popular interpretations like Yu Dan’s 于丹 Lunyu xinde 論語心得 (2006) have brought it to a mass audience, adding yet another chapter to the multi-­faceted history of this classic. The following selections, taken from the Lunyu zhengyi 論語正義, ed. by Liu Baonan 劉寶楠 (1791–1855), are representative of the book’s structure, and they provide samples of some of its main themes: learning, morals, ritual, and the junzi. It includes two levels of commentary: the early glosses assembled by He Yan, here called zhu 注, and Liu Baonan’s own commentary, called zhengyi 正義, which draws on a host of others from the Six Dynasties until the Qing period. Editions: Two editions stand out: Lunyu zhengyi 論語正義, ed. by Liu Baonan 劉寶楠. 2  vols. Beijing: Zhonghua 1998, and Lunyu jishi 論語集釋, ed. by Cheng Shude 程樹德. 4 vols. Beijing: Zhonghua 1997. A very readable edition with modern Chinese translation is Yang Bojun 楊伯峻, Lunyu yizhu 論語譯注. Beijing: Zhonghua 1980. The Western Han bamboo manuscript from Dingzhou has been edited as Dingzhou Hanmu zhujian: Lunyu 定州漢墓竹簡:論語. Beijing: Wenwu 1997. Translations of the Lunyu are legion. In English, Legge 1893 and Lau 1992a, have been highly influential; for a translation based on the Dingzhou manuscript and other archaeological

197

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 17 : lu ny u

witnesses, cf. Ames and Rosemont 1998. In German, Wilhelm 1921a, peerless for a long time, has been supplemented by Schwarz 1985 and Moritz 1998. In French, cf. Cheng 1981 and Levi 2018. Studies: Makeham 1996 is essential reading; Brooks and Brooks 1998 is a pioneer attempt to reconstruct the different layers of the Lunyu. Simson 2006 is a study in textual transmission and textual criticism; Vogelsang 2010 proposes a sociologically based interpretation of the Lunyu; Leys 2014 contains an extensive introduction as well as studies on various aspects of the book. Hunter 2017 and Hunter and Kern 2018 offer studies of the Lunyu’s origins.

17.2  Learning (LY 1.1, 2.4, 16.13) Confucius lived in a time of transition, when the aristocratic order of late Western Zhou gave way to a society in which, increasingly, commoners could rise to higher positions. The members of a newly emerging intermediary class, called shi 士, were not qualified by birth but by learning. The spread of writing now made it possible to acquire the knowledge necessary for success in a trans-­local elite society. It is no coincidence that Confucius was called China’s ‘first teacher’ (先師) in imperial times: he is said to have devoted his later years to the instruction of his disciples. This preoccupation remains evident throughout the Lunyu. In fact, its very first words address learning, a topic which reappears many times, as the following selections exemplify. The Lunyu thus set the tone for much of Chinese cultural history, throughout which learning was almost always regarded as particularly im­port­ ant. The following pieces come without any context: we mostly do not know on what occasions, and to whom, the ‘master’ imparted his ­wisdom.



198

論語正義卷一 學而第一

凡十六章

子 曰 :「 學 而 時 習 之, 不 亦 說 乎? 【 注 】 馬 曰 :「 子 者, 男

子之通稱,謂孔子也。」王曰:「時者,

學 者 以 時 誦 習 之。」 正 義 曰 : 爾 雅

釋 詁 :「 說, 樂 也。」 皇 本 凡「 說 」

皆作「悅」。說文有「說」無「悅」,

「悅」是俗體。 有朋自遠方來, 不 亦 樂 乎? 【注】包曰:「同門

曰 朋。」 正 義 曰 : 宋 氏 翔 鳳 樸 學 齊



說/悅 yuè, vi, ‘be joyful’ 方2 fāng, nc, ‘region’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

札記:「史記 世家:『定公五年,魯自大夫以下,

皆僭離於正道,故孔子不仕,退而修詩書禮樂,

弟子彌眾,至自遠方,莫不受業焉。』弟子至自

遠方,即『有朋自遠方來』也。『朋』即指弟子。」

按 : 宋 說 是 也。 釋 文 云 :「 有 或 作 友, 非。」



人不知而不慍,不亦君子乎?」【注】

慍,怒也。凡人有所不知,君子不怒。正義曰:

「人不知」者,謂當時君卿大夫不知己學有成舉

用之也。



17 . 2 l e a r n i n g

慍 yùn, vi, ‘be angry’

Comments ① The first lines give the title of the book, juan number, chapter title and number, as well as the number of sections the chapter contains. Like those of all Lunyu chapters, the title 學而 is an incipit, resuming the first characters of the text. It should not be taken to summarize the contents—although learning, of course, is a major topic in the chapter. ② The commentary is marked by the word zhu 注, the standard expression for intralinear commentaries. It begins by quoting the early commentaries by Ma Rong 馬融 (79–166) and Wang Su 王肅 (195–256); note that only their family names are given—the reader is expected to know who is meant (which he does, if he has read the introduction). Then follows the main commentary, 正義, which points out that 說 here represents the word yuè, ‘be joyful’. It quotes three works to support this: the Erya 爾雅, an old collection of glosses, the Lunyu edition of Huang Kan, and the Shuowen jiezi 說文解字, a character dictionary from Han times. Traditional commentaries, while discussing factual, orthographic, and semantic issues at length, usually provide little help with syntactic analysis. In this case, for example, it does not mention that 學而時習之 is an infinitive subject (cf. 12.3). Also note that the conj 亦 often does not simply mean ‘also’, but rather expresses an assertive ‘indeed, after all’. ③ The prind 有 may be used as mod when new persons or facts are ­introduced; it may simply be translated as an indefinite article, ‘a’. The 199

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 17 : lu ny u

commentary adduces a work by the Qing scholar Song Xiangfeng (1777–1860), which in turn quotes the Shiji in order to explain the meaning of 朋. The commentary agrees and then goes on to quote the Jingdian shiwen (cf. Focus 2), which dismisses the reading 友朋 instead of 有朋 that is found in some editions: on what grounds could this reading be rejected? ④ Being unknown to others is a recurrent theme in the Lunyu; cf. 不患人之不己知,患不知人也 (LY 1.16), 不患莫己知,求為可知也 (LY 4.14), and 不患人之不己知,患其不能也 (LY 14.30): the Warring States period was a time of increasing social mobility in which one had to reckon with strangers—and with being a stranger oneself. For the usage of the word 君子 in this clause, cf. Box 10. 子 曰 :「 吾 十 有 五 而 志 于 學, 三十而立,四十而不惑,五十而 知天命,六十而耳順,七十而從 心 所 欲, 不 踰 矩。」 正 義 曰 :「 有 」

之 言「 又 」 也。 白 虎 通 辟 雍 篇 :「 古 者

所 以 年 十 五 入 大 學 何? 以 為 八 歲 毀 齒,

始有識知,入學學書計。七八十五陰陽備,

故 十 五 成 童 志 明, 入 大 學, 學 經 術。」 漢

書 藝 文 志 :「 古 之 學 者, 且 耕 且 養, 三

年而通一經,用日少而畜德多,三十而五

經立。」



惑 huò, vi, ‘be in doubt’ 踰 yú, vn, ‘transgress’ 矩 jǔ, nc, ‘right measure’

Comments ⑤ For numerals, review Box 3. This passage is so well-­known that people later referred to their age by alluding to it: 而立 came to mean ‘to be thirty years old’. The commentary quotes two texts from the Later Han, Baihu tong 白虎通 and Hanshu 漢書: note how they construct a highly systematic ritual order around this passage, which originally might have just been a casual remark.

200

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

子 不 可 以 不 學, 見 人 不 可 以 不 飾。 不 飾 則 無 根, 無 根 則 失 理, 失

理 則 不 忠, 不 忠 則 失 禮, 失 禮 則 不 立。』」 說 苑 所 述, 疑 即 過 庭 學

禮之訓,而文較詳。



陳亢問於伯魚曰:「子亦有異聞乎?」【注】馬曰:「以 為伯魚,孔子之子,所聞當有異。」對曰:「未也。嘗獨立, 【注】孔曰:「獨立謂孔子。」 鯉趨而過庭。曰:『學詩乎?』 對曰:『未也。』 『不學詩,無以言。』鯉退而學詩。 他日,又獨立,鯉趨而過庭。曰:『學禮乎?』對曰: 『未也。』『不學禮,無以立。』鯉退而學禮。聞斯二者。」 陳亢退而喜曰:「問一得三,聞詩,聞禮,又聞君 子 之 遠 其 子 也。」 正義曰:說苑 建本篇:「孔子曰:『鯉君

17 . 2 l e a r n i n g



陳亢 Chén Gāng, np 伯魚 Bóyú, np 對 duì, vn, ‘answer’ 鯉 Lı ,̌ np

趨 qū, vn, ‘hasten’ 庭 tíng, nc, ‘courtyard’ 退 tuì, vn, ‘withdraw’ 他 tā, prdem, ‘another’ 喜 xı̌, vi, ‘be happy’

Comments ⑥ This is one of the longer passages in the Lunyu, involving a certain Chen Gang and Boyu, Confucius’ son (cf. Focus 5). Again 亦 is used in an assertive sense, ‘certainly’. 對 *tûps, which is related to 答 *tûp, ‘answer’, is usually subordinated to 曰. Like English ‘not yet’, the negative response 未也 omits the P.  The commentary quotes Kong Anguo (孔安國, second cent. bc), a distant descendant of the master who ­supposedly wrote a commentary to the Lunyu. ⑦ Songs and ritual— namely, the way to behave in an elite society—were two core elements of Confucius’ teaching. The passage suggests that both already existed as written works, since Boyu retreated to study them. Recall that 他 is not a prper but a prdem meaning ‘other, another’ (compare MSC 其他). 斯 is a Lu dialect word (cf. 15.4.2–㉛) used instead of 此. Again, a parallel transmission—from the Shuoyuan—is quoted in the commentary: note how much more elaborate it is.

201

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 17 : lu ny u

17.3  Morals (LY 4.15, 12.2, 17.5, 13.19)

子曰:「參乎!吾道一以貫之。」 曾子曰:「唯。」 正義曰:「參」者,

曾子名。說文「森」字,「讀若曾參之參」,



則「參」、「森」音同。其字子輿,則取

三 人 同 輿 義 也。 子 出, 門 人 問 曰:「何謂也?」曾子曰:「夫 子之道,忠恕而已矣。」 正義曰:

「門人」者,謂受學於夫子之門之人也。

賈子 道術:「以己量人謂之恕。」



「忠恕」者,周語云:「中能應外,忠也。」

The learning Confucius had in mind was not simply the acquisition of knowledge but foremost moral learning: the transition to a trans-­local elite society required the formalization of morals. The Lunyu introduces a number of formalized moral concepts that are encapsulated in the term 仁, ‘humaneness, benevolence, kindness, compassion’. These moral precepts culminate in the Golden Rule, which is stated thrice in the Lunyu: ‘Do not do unto others what you do not wish for yourself.’ 參 Cān, np 曾子 Zēngzı̌, np 唯2 wěi, interj, ‘yes’ 恕 shù, vn, ‘feel empathy’ 而已 éryı̌, ptemp, ‘only’

Comments ⑧ Zeng Can 曾參 was one of Confucius’ most prominent disciples and an eminent personality of the early Confucian tradition. He is credited with two other Confucian classics, the Daxue and the Xiaojing. Since he is one of a few disciples who are also being called ‘master’ in the Lunyu, he is supposed to have been among the compilers of the text. The pronunciation of his given name is uncertain (cf. the commentary). The reading ‘Can’ is adopted here following U. Unger, who pointed out that 參 might be an orthographic variant of 驂, a ‘team of three horses’ used to draw a carriage; and ‘carriage’, 輿, happens to be part of Zeng Can’s courtesy name. For 一以貫之, cf. 9.2.3–⑱. The interj 唯, ‘yes’, acknowledges that one has understood, similar to Japanese はい. ⑨ Zeng Can, obviously a privileged disciple, here transmits Confucius’ teachings to other disciples. 而已, originally a verbal expression, ‘. . . and that’s the end of it / and that’s all’, has been grammaticalized to become a pt. Typically used after after P, it functions as A, meaning ‘only, simply’; the combination with 矣 is rather common. The commentary quotes two titles which require explanation: ‘Zhouyu’ is not a book, but the title of a chapter in the Guoyu 國語—the 202

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

17 . 3 m o r a l s

仲弓問仁。子曰:「出門 如見大賓,使民如承大祭。



【 注 】 孔 曰 :「 為 仁 之 道, 莫 尚

正 義 曰 : 史 記 弟 子 傳 作「 仲 弓

問 政 」。 然 前 後 章 皆 是 問 仁, 不

應此為問政,史記誤也。



乎 敬。」 己 所 不 欲, 勿 施 於人。在邦無怨,在家無 怨。」【注】包曰:「在邦為諸侯, 在家為卿大夫。」 仲弓曰:「雍 雖 不 敏, 請 事 斯 語 矣。」

reader is expected to know his classics. Jiazi is an alternative title of the Xinshu 新書 by Jia Yi 賈誼. Again, the reader is expected to know this. 賓 bīn, nc, ‘guest’ 承 chéng, vtr, ‘offer’ 祭 jì, nc, ‘sacrifice’ 施 shī, vn, ‘spread out’ 邦 bāng, nc, ‘country’ 雍 Yōng, np 敏 mı̌n, vn, ‘be clever’ 語 yǔ, vn as nc, ‘saying’

Comments ⑩ 仲弓 is the courtesy name of Confucius’ disciple Ran Yong 冉雍. For the usage of 如, review 11.4. ⑪ The prrefl 己 here functions as mod to the 所 phrase. Recall that 勿 implies an Opr 之; what does that Opr refer to? Interestingly, the Shiji, quoted in the commentary, has Zhonggong ask about government, not benevolence, and it records the same answer, but without the Golden Rule. We will never know what the master really said. 趙氏佑溫故錄:「惠,順也。仁者待人,務順乎人情。」此義

子張問仁於孔子。孔子曰:「能行五者於天下, 為仁矣。」「請問之。」曰:「恭,寬,信,敏,惠。 恭則不侮,【注】孔曰:「不見侮慢。」寬則得眾,信 則人任焉,敏則有功,惠則足以使人。」 正義曰:

亦通。

樊遲問仁。子曰:「居處恭,執事敬,與人忠。 雖之夷狄,不可棄也。」 正義曰:「居處」謂所居之處。

「 執 」 猶 行 也。 此 章 所 言, 亦「 克 己 復 禮 為 仁 」 之 意。 爾 雅

釋詁:「恭,敬也。」二字訓同,此對文稱異。漢書 五行志:

「內曰恭,外曰敬。」





203

恭 gōng, vn, ‘be respectful’ 寬 kuān, vn, ‘be indulgent’ 侮 wǔ, vtr, ‘insult’ 眾 zhòng, vst as nc, ‘mass’ 樊遲 Fán Chí, np 執 zhí, vtr, ‘take charge of ’ 狄 Dí, np, a people 棄 qì, vtr, ‘abandon’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 17 : lu ny u

Comments ⑫ Significantly, the five virtues mentioned here are to be practised in the realm ‘under heaven’: morals are not so important in one’s family or hometown but in public, ‘political’ life. Is 不侮 active or passive voice? Consult the commentary! 足以 is a near synonym to 可以 (cf. 4.4). ⑬ There is an interesting parallel to this passage in LY 12.22: 樊遲問仁,子曰:愛人. Confucius is famed for having ‘conveyed his teachings according to the talent’ of the student (因材施教): to the same question, he gave different answers, adjusting his explanations to changing needs and situations. This is typical for oral, context-­ bound instruction. The commentary associates this passage with LY 12.1, quoted below: note how this kind of association creates coherence in a corpus that may not have been so coherent at all. 17.4  Ritual (LY 12.1, 6.27, 3.15) Ritual, 禮, another major pillar of Confucius’ teachings, is intimately linked to morals. Rituals—this includes solemn rites as well as the ­rituals of everyday life—are morals enacted, as it were; by performing ­rituals, the performers accept and reaffirm moral conventions. ‘A man without benevolence, what has he to do with rituals?’ (LY 3.3), the master is quoted, and the connection between morals and rituals is evident throughout the Lunyu.



204

顏 淵 問 仁。 子 曰 :「 克 己 復 禮 為 仁。



【注】馬曰:「克己,約身。」孔曰:「復,反也。

身 能 反 禮, 則 為 仁 矣。」 一 日 克 己 復 禮, 天 下 歸 仁 焉。 為 仁 由 己, 而 由 人 乎 哉?」【注】孔曰:「行善在己,不在人也。」 顏淵曰:「請問其目。」子曰:「非 禮勿視,非禮勿聽,非禮勿言,非禮 勿動。」【注】鄭曰:「此四者,克己復禮之目。」 顏 淵 曰 :「 回 雖 不 敏, 請 事 斯 語 矣。」 正義曰:「勿」者,禁止之辭。視、聽、

言、動,皆在己不在人,故為仁由己,不由人也。

「動」猶行也,謂所行事也。



顏淵 Yán Yuān, np 復 fù, vi, ‘return’ 歸2 guī, vn, ‘assign (to)’ 動 dòng, vi, ‘act’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

17 . 4 r i t ua l

君 子 博 學 於 文, 約 之 以 禮, 亦 可 以 弗 畔 矣 夫。【 注 】

鄭曰:「弗畔,不違道。」

正義曰:臧氏琳經義

雜記:「君子乃成德之

稱, 不 嫌 其 違 畔 於 道。

顏淵篇此章再見,無『君



子 』 字。」 知 此 亦 無 有

者為得也。

Comments ⑭ Yet another definition of 仁. How is the S of Confucius’ answer to be analysed? ⑮ The last clause is a rhetorical question; the particle combination 乎哉 corresponds exactly to a combination of question mark and exclamation mark. ⑯ For the use of 非, cf. 15.4.1. 約 yuē, vst, ‘be restrained’ 畔 pàn, vn, ‘transgress’ 夫2 fú, fus, ‘isn't it?’

Comments ⑰ 矣夫 is a combination of final pt. 夫 is a fusion of 不 and 乎; occurring exclusively in final position to form tag questions: ‘…, isn’t it?’, ‘…, n’est-­ce pas?’. The commentary argues that the word 君子 should not appear in this passage: do you understand the line of reasoning? 子入太廟,【注】包曰:「太廟,周

公廟。孔子仕魯,魯祭周公而助祭也。」

紇所治邑。時人多言孔子知禮,或人以

每事問。 或曰:「孰謂鄹人 之 子 知 禮 乎? 入 太 廟, 每 事 問。」【注】孔曰:「鄹,孔子父叔梁

為知禮者不當復問。」子聞之,曰: 「 是 禮 也。」【 注 】 孔 曰 :「 雖 知 之,

當復問,慎之至也。」正義曰:三蒼云:

「事,謂犧牲復器及禮儀諸事也。」





太 tài, vst, ‘be great’ 每 měi, prind, ‘each’ 鄹 Zōu, np, a place

Comments ⑱ 每事問 seems to be a relic of pre-­Classical Chinese: an anteposed O without a resuming ‘dummy’. ⑲ 鄹, in the polity of Lu, was Confucius’ home town. 三蒼, quoted in the commentary, refers to three early character dictionaries which were later collectively called Cangjie pian 蒼頡篇.

205

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 17 : lu ny u

17.5  Junzi vs. xiaoren (LY 4.11, 13.23, 12.5) Confucius’ educational ideal was the junzi, a man who possessed the knowledge, morals, and ritual refinement that made him fit for public life. The word ‘gentleman’ would seem apt as a translation for junzi: not simply a morally ‘superior man’, but a man who had acquired a certain habitus due to his role in elite society. It was this habitus that distinguished the junzi from the xiaoren, the ‘petty man’ who never transcended the confines of his village or kin group. 子 曰 :「 君 子 懷 德, 小 人 懷 土 ; 君 子 懷 刑, 小 人 懷 惠。」 正 義 曰 : 爾 雅 釋 詁 :

「 懷, 思 也。」 說 文 :「 懷, 思

念也。」

子 曰 :「 君 子 和 而 不 同, 小 人 同 而 不 和。」【 注 】 君 子

心 和,然 其所 見 各異 , 故曰「 不

同」。小人所嗜好者則同,然各爭

利,故曰「不和」 。





懷 huái, vn, ‘cherish’

Comments ⑳ The junzi cherishes the abstract and general, while the xiaoren cherishes the concrete and personal. For the texts cited in the commentary, cf. ②. ㉑ The xiaoren, living among equals, ‘are alike without harmonizing’, while for men of public life the opposite is true. What does the commentary say? 司馬牛憂曰:「人 皆 有 兄 弟, 我 獨 無。」子夏曰:「商 聞之矣:死生有 命,富貴在天。 君子敬而無失,與 人恭而有禮,四海 之內皆兄弟也,君 子何患乎無兄弟 也?」【注】包曰:「君

子 疏 惡 而 友 賢, 九 州 之

人,皆可以禮親。」



司馬牛 Sīmǎ Niú, np 子夏 Z ı x̌ ià, np



Comments ㉒ A conversation between two of Confucius’ disciples, 司馬牛 (whose courtesy name was 耕), and 卜商, whose courtesy name was 子夏. What do you notice about their last names, and what about 206

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

17 . 5 j u n z i v s . x i ao r e n

the relation between their given and courtesy names? ㉓ The junzi, quite unlike the xiaoren, transcends kin ties: for him everybody (within the elite, one should probably add) is a brother. This is a description of the emerging trans-­local elite society in a nutshell. Just like ‘within the four seas’ designates the area of Great Britain, 四海之內 means the entire Chinese realm (cf. 6.2.2–⑨).

207

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 14

Commentaries It may come as relief to students of Classical Chinese that the Chinese themselves, at least since Han times, apparently had difficulties understanding classical texts. This is evidenced by the fact that Han scholars began to produce great amounts of commentaries on these texts. The earliest commentaries were probably the ‘Ten Wings’ (十翼) appended to the hexagrams of the Yijing as well as three works—Guliang zhuan, Gongyang zhuan, and Zuozhuan—that elucidated the ‘great meaning behind the subtle words’ of the Chunqiu attributed to Confucius. While these may be special cases, scholars soon began to explicate other classical works not in a philosophical but in a philo­logic­al manner. Instead of elucidating the ‘great meaning’, a Later Han text complained, ‘unrefined scholars expand their learning by applying themselves to the names of things, considering the details of tools and implements, stressing glosses and explanations, extracting chapters and cola’ (Zhonglun 1). Indeed, ‘extracting chapters and cola’ seems to have been among the first concerns of Han philologists: works like Zhang Yu’s 張禹 Lunyu zhangju 論語章句 (c.40 bc) and others suggest that there was an entire genre of commentaries devoted to demarcating ‘chapters’ (章) and ‘cola’ (句). Considering that classical texts had been transmitted in scriptio continua (cf. Box 7), determining the basic structure of texts was an essential task for early commentators. Many other ‘glosses and explanations’ followed: in one of the greatest endeavours in the history of scholarship, enormous amounts of commentaries and sub-­commentaries were produced for virtually all clas­ sic­al and canonical texts, exceeding the volume of the texts themselves by far. In one case, the commentary to a four-­character phrase is said to have amounted to 300,000 words! Arguably, commentaries were the most important form of scholarly publication in imperial China; and some of the ‘unrefined scholars’ that wrote these commentaries rank among the greatest names in Chinese intellectual history:

208

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 14 : co m m e n ta r i e s

Zheng Xuan 鄭玄 (127–200), com. to Shijing, Shujing, Zhouli, Liji, Lunyu etc.; He Yan 何晏 (190–249), com. to Lunyu; Wang Bi 王弼 (226–49), com. to Yijing and Laozi; Guo Xiang 郭象 (d. 312 ad), com. to Zhuangzi; Guo Pu 郭璞 (276–324), com. to Erya, Shanhai jing, Mu Tianzi zhuan, Chuci; Lu Deming 陸德明 (c.550–630), author of the Jingdian shiwen. The influence of these scholars can hardly be overstated. For centuries they have determined how classical and canonical texts—that is the core of Chinese tradition—were to be understood. No classical or canonical text was studied without a commentary, in fact they never came without a commentary. Even today, it is well-­nigh impossible to find editions of these texts without a commentary. The very titles of such editions—Lunyu jijie, Mengzi zhengyi, and the like—are telling: they are NP in which the name of the commentary is the head, while the title of the classical text is mod: hence, the commentary, not the classical text is the main element. Indeed, trad­ition­al scholars consulted such editions mainly for the commentary—the classical text was well known, anyway.

Even today, no Sinologist can afford to ignore these commentaries; and to students of Classical Chinese they are invaluable tools for reading texts: they give us explanations of difficult characters and words, occasional syntactical analyses, text-­critical notes, historical facts, parallel passages, philosophical interpretations, and other useful information at our fingertips. However, it takes some skill to unlock this wealth of knowledge. Commentaries were written by erudite scholars for erudite scholars. Their language is not Classical but a genre of Literary Chinese: certainly a specialist language with a rather technical vocabulary. While a detailed discussion of commentary language is not within the purview of this book, basic knowledge of their phraseology is indispensable in order make proper use of commentaries. Simple explanations of words and phrases are mostly couched in straightforward nominal sentences: X Y 也, ‘X means Y’. Other common formulas are: X 謂 Y means ‘X means Y’, X being the term that is explained, as for example 門人 in the following passage: 209

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 14 : co m m e n ta r i e s

① 門人問曰:「何謂也。」正義曰:「門人」者,謂受學於夫子之門 之人也。

X 曰 Y / X 為 Y / X 謂之 Y also mean ‘X means Y’, but the term that is explained is Y, as for example 恕 in the same passage: ② 夫子之道,忠恕而已矣。 賈子 道術:「以己量人謂之恕。」

X 之言 Y / X 之為言 Y, ‘what X expresses is Y’, is used when a word is explained by a (nearly) homophonous word: ③ 吾十有五而志于學… 正義曰:「有」之言「又」也。

X 讀為 Y / X 讀曰 Y or X 讀如 Y / X 讀若 Y, ‘X should be read as/like Y’, or X 音 Y, ‘X is pronounced Y’ are used for explaining characters that are used to represent a word other than the one they usually represent, as in: ④ 帝王不相復 . . .

禮鴻案:復讀如复。說文:「复,行故道也。」

⑤ 生莊公及共叔段。 共音恭。

For 案, cf. Focus 13–⑧. Another way of indicating the pronunciation of a character and, hence, the word it represents, is of course the fanqie method introduced in Focus 2.

X Y 通用 is a general statement that ‘X and Y may be used interchangeably’; it supports the interpretation that X, in a particular instance, is used in the sense of Y: ⑥ 情不知其不義也 . . .

王云:「情、誠通用。」

⑦ 守國之度,在飾四維 . . . 俞樾云:禮義廉恥非由修飾,「飾」當讀為 飭。「飭」與 「飾」古通用。

Besides such simple glosses, commentaries may offer text critical notes, that is observations on variant readings in different editions or speculations about the original form of a passage. Typical formulas used for such notes are: X A 本作 Y means that in a given passage ‘X is written Y in the e­ dition A’, as in: ⑧ 雖有槁曓 . . . 盧文弨曰:「曓」,舊本作「暴」,非。

210

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 14 : co m m e n ta r i e s

Unfortunately, instead of a precise reference to a specific edition, such notes often point to a vague ‘old edition’, as above, 一本,‘one ­edition’, or to 或, ‘some (edition)’, as in: ⑨ 有朋自遠方來 . . .

釋文云:「有或作友,非。」

Instead of X A 本作 Y, the formula may also be changed to A 本 X 作 Y, as in 舊 本「知」作「之」—a syntactical construction that would not be possible in CC.

衍, literally ‘superfluous’, refers to (presumptively) redundant characters, as in: ⑩ 今至大為不義攻國,則弗知非… 舊本「知」作「之」,下又 衍「而」字。 ⑪ 至殺不辜人也,扡其衣裘 . . . 王云:「『也』即『扡』字之 誤而衍者。

Sometimes, as in the last example, possible reasons for textual errors are given: The character 也 may have slipped into the text because the copyist mistook it (誤) for the graphically similar 扡 which follows.

脫, in turn, means that a character has (probably) been ‘left out’, as in: ⑫ 此可謂知義與不義之辯乎。 舊本「可」上脫「此」字, 又「謂」誤「為」。

倒 refers to another type of textual error, namely the (accidental) transposition of word order or entire sentences. ⑬ 從而譽之,謂之義。 畢云:「舊『之謂』二字倒,一本如此。」

Besides simply explaining words or characters and noting textual variants, commentaries may also pass judgment on certain readings or on other commentaries, employing words like 是, ‘right’, 非, ‘wrong’, 當, ‘should be’, 誤, ‘mistaken’, or 通, ‘passable, acceptable’. ⑭ 敏則有功,

惠則足以使人。

正義曰:趙氏佑温故錄:「惠,順也。」 此義 亦通。

⑮ 野蕪曠則民乃菅 . . . 菅,當為姦。安井衡云:「菅」當為「營」字 之誤耳。. . . 翔鳳案:安井說是也。

Note that, in the above example, although the commentator is convinced that 菅 should actually be 營, he leaves the wrong character in possession

211

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 14 : co m m e n ta r i e s

of the text. In fact, this is the rule in Chinese textual criticism. Only in exceptional cases, editors actually emend a text in order to restore it to its original form, correcting (正) or deleting (刪) wrong characters or supplying (增/補) putatively correct ones: ⑯ 大為非攻國,則不知非… 舊本「不知」下衍「而」字,今 據王、蘇校刪。 ⑰ 此可謂知義與不義之別乎。可,舊本作「何」。畢云:「一本 作『可』,是。」今據正。 ⑱ 當此,天下之君子… 畢云:「舊脫此字,據後文增。」

So do commentaries provide us with all we need? By no means. While a commentary may be highly useful when first reading a text, it should never simply be taken as the ‘right’ interpretation of that text. Nor can we find out the ‘true meaning’ of a classical text by consulting as many commentaries as possible. Quite the opposite: the more commentaries one reads, the more obvious it will become how ambiguous and contested many passages in classical literature are—and how uncertain Chinese scholars themselves were about them. After all, commentaries were written precisely because classical texts were so difficult to understand.

References:  ① LY 4.15  ② LY 4.15  ③ LY 2.4  ④ SJS 1  ⑤ Zuo 1.1  ⑥ Mo 17  ⑦ Guan 1  ⑧ Xun 1  ⑨ LY 1.1  ⑩–⑬ Mo 17  ⑭ LY 17.5  ⑮ Guan 1  ⑯–⑱ Mo 17

For scholarly discussions of commentaries, cf. Henderson 1991; Wagner 2000; Gentz 2001; Wang Li 1981, vol. 2, 604–19.

212

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 18

Mengzi

18.1 Introduction 18.2 Profit, benevolence, and righteousness (Meng 1A1)

18.3 Benevolent government (Meng 1A5) 18.4 Human nature (Meng 6A2)

213 214

217 219

18.1 Introduction Apart from the Lunyu, Mengzi is probably the most widely read Classical Chinese text. It is the main source for the thought of Meng Ke 孟 軻 (trad. 372–289 bc), the ‘second sage’ of the Confucian school. Meng Ke, who supposedly stood in the line of Confucius’ disciple Zeng Can 曾 參 (505–436) and his grandson Kong Ji 孔 伋 , also known as Zisi 子 思 (483–402), represented the ‘idealistic’ wing of the Confucian school. Convinced that human nature was fundamentally good, Meng Ke argued that a ruler could govern a state successfully by dint of ‘benevo­ lence and righteousness’ (仁 義 ): caring for the people was his le­git­im­ ation. By the same token, Meng Ke argued that the people had the right to resist and, indeed, overthrow an immoral ruler. This line of reasoning has made the Mengzi a canonical text of new Confucianism since Song times—one of the Four Books (四 書 )—and a text dreaded by despotic rulers of late imperial China. The received text of Mengzi, approx. 35,000 characters long, is gener­ ally believed to have been written by Meng Ke’s disciples after the mas­ ter’s death. It has been arranged by its first commentator, Zhao Qi 趙 歧 (d. ad 201), in seven chapters, each divided into two parts, which are further divided into sections (thus the standard way to quote it is by chapter, part A or B, and section, e.g. Mengzi 1A7). The earliest Chinese book catalogue, the Hanshu ‘Yiwenzhi’, records the Mengzi with eleven pian. In fact, a collection called Mengzi waishu 孟 子 外 書 , Outer Writings of Mengzi,

213

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 18 : m e n g z i

­contains four pian, which would add up with the seven of the Mengzi proper to exactly eleven. However, the Mengzi waishu are considered to be a forgery dating from postclassical times.

It contains dialogues of Meng Ke with the rulers of his time or with his disciples and other contemporaries. These dialogues show him con­ tending with rival schools—especially those of Mo Di and the ‘hedonist’ Yang Zhu—in the discussion of good government, which, for Meng Ke, puts the people first and the ruler last. Not only its edifying contents but also its clear and polished language have made the Mengzi a ­favourite among textbook editors as well as students of Classical Chinese lan­ guage and culture. Its style has been widely imitated in the literature of imperial times. The following excerpts are taken from the edition Mengzi zhengyi 孟 子 正 義 by Sun Shi 孫 奭 (962–1033), re-edited by Jiao Xun 焦 循 (1763–1820). Editions: The standard scholarly edition is Mengzi zhengyi 孟 子 正 義 , ed. by Jiao Xun 焦 循 . 2 vols. Beijing 1998; a very readable edition with a modern translation and commen­ tary has been edited by Yang Bojun 楊 伯 浚 , Mengzi yizhu 孟 子 譯 注 . 2 vols. Beijing 1960. Translations: Lau 1970; Bloom 2009; Wilhelm 1921b; Lévy 2003; Van Norden 2008. Studies: Ivanhoe 1990; Huang Junjie 1991–97; Shun Kwong-loi 1997; Chan 2002.

18.2  Profit, benevolence, and righteousness (Meng 1A1) The very beginning of Mengzi records the famous dialogue between Meng Ke and king Hui of Wei about ‘profit’ versus ‘benevolence and righteousness’. A good example of Mengzi’s style of argumentation, it sets the tone for the entire work. 梁惠王章句上 ︻注︼梁惠王者,魏惠

214

王也。魏,國名。惠,謚也。王,號也。



時天下有七王,皆僭號者也。魏惠王居

於大梁,故號曰梁王。

孟 子 見 梁 惠 王,︻ 疏 ︼ 正 義 曰 : 魏

世 家 云 :「 惠 王 數 被 軍 旅, 卑 禮 厚 幣,

以 招 賢 者, 鄒 衍、 淳 于 髡、 孟 軻 皆 至

③ ②

梁。」 六 國 表 云 :「 魏 惠 王 三 十 五 年,

孟 子 來, 王 問 利 國。 王 曰 :「 叟 不遠千里而來, 亦將有以利 吾 國 乎?」 ︻ 注 ︼ 叟, 長 老 之 稱



梁 Liáng, np, a city 叟 sǒu, nc, ‘old man’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

也。猶父也。孟子去齊,老而之魏,故王尊禮之曰父。

︻疏︼正義曰:史記 孟子列傳云:「孟子,騶人也。

受業子思之門人。道既通,游事齊宣王,宣王不能用,



適粱。」時秦用商君,富國強兵,惠王所以遷粱,故



曰亦將有以利 吾國,謂亦如商 君之於秦,俾 富國強



兵也。論衡 刺孟篇述此文,作「將何以利吾國乎」。

論語曰:「放於利而行,多怨。」故不欲使王以利為

名也。︻疏︼高誘注云:「危,亡也。」

孟子對曰:「王何必曰利,亦有仁義 而已矣。 王曰『何以利吾國』,大夫曰 『何以利吾家』,士庶人曰『何以利吾身』, 上下交征利,而國危矣! ︻注︼征,取也。

18.2 p r o f i t , b e n e v o l e n c e , a n d r i g h t e o u s n e s s

大夫 dàfū, nc, ‘dignitary’ 庶人 shùrén, nc, ‘common people’ 征 zhēng, vn, ‘take’

Comments ① Zhao Qi’s Mengzi zhangju 孟 子 章 句 , Chapters and cola of the Mengzi, is an example of the genre of zhangju commentaries (cf. Focus 14). The commentary first gives a detailed explanation of the title 梁 惠 王 —how much of it do you understand? ② The zhengyi com­ mentary refers to a Shiji chapter, which provides the historical back­ ground to Meng Ke’s visit in Liang (located in the area of present Kaifeng). ③ 叟 , as a form of address, may be rendered as ‘venerable sir’; this address has given rise to learned debates about Meng Ke’s age and the year of his visit to Wei. Most scholars agree that it must have been toward the very end of king Hui’s reign, around 320 bc, when both he and Meng Ke were old. But what date does the Shiji, quoted in the pre­ ceding commentary, suggest? ④ For the use of 亦 , cf. 17.2–②. In 將 有 以 利 吾 國 ,有 clearly cannot be a vn, since it is followed by a prep, not a NP. Rather, 有 represents a prind, ‘something’. Just like it pre­ cedes the P if used as an O (cf. 5.2.1 and Box 17), it also precedes a prep if used as O to the latter. ⑤ The commentary here gives us a biography of Meng Ke in a nutshell; can you understand its gist? ⑥ 對 indicates that an inferior person responds to a superior. It is no mere whim that Meng Ke fulminates against the king’s talk of ‘profit’ (利 ): this was a key term in the teachings of the Mohists, the main rivals of the Confucian school at the time. ‘Benevolence and righteousness’, on the other hand, were at the heart of Meng Ke’s teaching. For 而 已 矣 , cf. 17.3–⑨. ⑦ Observe that this is a complex sentence including several conditional clauses. The 215

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 18 : m e n g z i

adv 交 , ‘mutually’ is sometimes used instead of 相 (cf. 7.5–㉕); just like the latter, it may stand for the O of the following verb, as in 交 惡 , ‘hated one another’. Is this the case here? 萬乘之國,弒其君者,必千乘之家。 ︻注︼萬乘,兵車萬乘,謂 天 子 也。 千 乘, 兵 車 千 乘, 謂 諸 侯 也。千 乘 之 國, 弒 其 君 者, 必 百 乘之家。︻注︼天子建國, 諸侯立家。百乘之家,謂大國之卿,食釆邑有

兵車百乘之賦者也,若齊崔、衛甯、晉六卿等是。以其終亦皆弒其君,此以



︻疏︼經文



百乘取千乘也。萬取千焉,千取百焉,不為不多矣。



承上萬乘千乘百乘,則萬千百仍指乘言。是諸侯於天子,萬乘中取其千。大

夫於諸侯,千乘中取其百。苟為後義而先利, 不奪不饜。 ︻注︼苟, 誠也。誠令大臣皆後仁義而先自利,則不篡奪君位,不足自饜飽其欲矣。未 有仁而遺其親者也,未有義而後其君者也。 ︻注︼仁者親親, 義者尊尊。人無行仁而遺棄其親、行義而忽後其君者。王亦曰仁義而已 矣,何必曰利。」 章指言:治國之道明,當以仁義為名,然後上下和親,

君臣集穆。天經地義,不易之道,故以建篇立始也。



乘 shèng, nc, ‘chariot’ 饜 yàn, vn, ‘be satisfied’ 遺 yí, vtr, ‘leave behind’



Comments ⑧ Note the absolute topic at the beginning. For an explana­ tion of what is meant, see the commentary and Focus 15. ⑨ For the analysis of 萬 取 千 焉 and 千 取 百 焉 , cf. 13.3.2. The double negation, or litotes, in the last clause gives emphasis to the assertion; for further analysis of this construction, cf. 11.2. ⑩ 為 here serves as an aux (cf. Box 14). Note the succinct conditional construction in 不 奪 不 饜 . ⑪ Two long 者 phrases; for their analysis, cf. Review 3. ⑫ In an effective rhe­ torical finish, Mengzi’s last sentence reiterates the first. The final com­ mentary, also by Zhao Qi, summarizes the essence (指 ) of this paragraph and acknowledges its prominent position at the beginning of the book. 216

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

18.3 b e n e v o l e n t g o v e r n m e n t

18.3  Benevolent government (Meng 1A5) The beginning of Mengzi displays an interesting pattern: the first four sections contain more or less overt polemics against the rival teachings of Mohists (1A1, cf. ⑥), hedonists (1A2), natural philosophers (1A3), and political realists (1A4), before section 1A5 turns to Mengzi’s own teach­ ings. The line of argument presented in this section is typical for Mengzi—and, indeed, for ‘Confucian’ thinkers in general—insofar as it ascribes political success to personal qualities, not to institutions: if only the ruler is benevolent, all will be well. This principle has later been called 人 治 , ‘government by people’, as opposed to 法 治 , ‘government by laws’.

注 】

217

梁 惠 王 曰 :「 晉 國, 天 下 莫 強 焉, 叟 之 所 知 也。【



韓 魏 趙 本 晉 六 卿, 當 此 時 號 三 晉, 故 惠 王 言 晉 國 天 下 強 也。



【 疏 】 注「 韓 魏 」 至「 強 也 」 ○ 正 義 曰 : 史 記 六 國 表 云 :「 六 卿 擅 晉 權, 征 伐 會 盟, 威 重 於 諸 侯, 終 之 卒 分 晉。 量 秦 之 兵, 不 如 三 晉 之 強 也。」

及 寡 人 之 身, 東 敗 於 齊, 長 子 死 焉 ; 西 喪 地 於 秦 七 百 里, 南 辱 於 楚, 寡 人 恥 之, 願 比 死 者 壹 洒 之, 如 之 何 則 可?」 ︻

疏︼正義曰:史記 魏世家:「惠王︙三十年,魏伐趙,趙告急齊,齊宣王用孫子計,救趙

擊魏。魏遂大興師,使龐涓將,而令太子申為上將軍,與齊人戰,敗於馬陵。齊虜魏太子申,

殺將軍涓,軍遂大破。」「三十一年,秦 趙 齊共伐我。秦將商君,詐我將軍公子卬而襲奪

其軍,破之。秦用商君,東地至河;而齊 趙數破我。安邑近秦,於是徙治大梁。」音義云:

「洒之,丁音洗,謂洗雪其恥也。」孟子對曰:「地方百里而可以王。【注】言古 王如施仁政於民,省刑罰,薄稅斂, 聖人以百里之地,以致王天下,謂文王也。





辱 rù, vtr, ‘humiliate’ 恥 chı̌, nc as vn, ‘be ashamed of ’ 願 yuàn, vn, ‘wish to’ 比 bì, conj as prep, ‘for’ 壹 yī, vi as adv, ‘once and for all’ 洒 xı̌, vn, ‘cleanse’ 省 shěng, vtr, ‘reduce’ 罰 fá, vn as nc, ‘penalty’ 薄 bó, vst, ‘be sparse’ 稅 shuì, nc, ‘taxes’ 斂 liǎn, vn as nc, ‘levy’ 壯 zhuàng, vst, ‘be ablebodied’ 暇 xiá, nc, ‘leisure’ 修 xiū, vst, ‘be refined’ 梃 tı ň g, nc, ‘club’

撻 tà, vn, ‘beat’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

深耕易耨,壯者以暇日,修其 孝弟忠信,入以事其父兄,出 以事其長上,可使制梃以撻秦、 楚 之 堅 甲 利 兵 矣。 彼 奪 其 民時,使不得耕耨,以養其父 母 ; 父 母 凍 餓, 兄 弟 妻 子 離 散。彼陷溺其民,王往而征之, 夫誰與王敵? 【注】彼,謂齊、秦、 楚也。彼困其民,願王往征之也。故 曰 : 仁 者 無 敵, 王 請 勿 疑!」

【注】鄰國暴虐,己脩仁政,則無敵矣。

王請行之,勿有疑也。



l e s s o n 18 : m e n g z i



堅 jiān, vst, ‘be firm’ 甲 jiǎ, nc, ‘armour’ 利2 lì, vi, ‘be sharp’ 凍 dòng, vi, ‘freeze’ 餓 è, vi, ‘starve’ 離 lí, vn, ‘separate’ 散 sàn, vtr, ‘scatter’ 溺 nì, vi, ‘drown’ 夫3 fú, prdem as ptemp, ‘well’ 疑 yí, vn, ‘doubt’

Comments ⑬ As the commentary remarks, the polity of Jin no longer existed at the time of king Hui; what he refers to are the combined forces of Jin’s successor states Han, Wei, and Zhao. ⑭ Try to figure out the his­ torical facts the king alludes to: the commentary is of some help, but you may want to consult scholarly literature such as mentioned in Focus 15. ⑮ The temporal conj 比 is here used as a prep. 壹 (a variant of 一 , nowadays used on banknotes) is often used in an emphatic way: ‘once and for all’, or ‘wholly’. 如 之 何 is an idiomatic expression in which 如 , ‘to approach’, takes an O and a C.  It expresses a rhetorical question: ‘when it comes to this, what’ > ‘what (to do) about it?’ ⑯ Note that 地 方 , unlike in MSC, is not a NP; rather, the expression may be construed as S and A, followed by the P (cf. Box 13): ‘a hundred miles square’, in other words, 10,000 Chinese square miles, which roughly equals 1,700 square kilometers. The commentary surmises that this refers to the territory king Wen once controlled: its borders could be reached within a day. ⑰ This is Mengzi’s political ideal: a benevolent government that induces its populace to practice Confucian virtues and thus strengthen them to overcome the great powers of Qin and Chu. 入 and 出 , used adverbi­ ally, may be rendered as ‘on the one hand . . . on the other hand’ (but note how much more specific the CC expression is). ⑱ The ptemp 夫 , ‘Well, . . .’ , which usually occurs as A in initial position, serves to introduce and emphasize a new point, typically a general statement, an abstract prin­ ciple, or a judgment. The expression 故 曰 , ‘therefore it may be said’, introduces quotations of common sayings, which serve as a kind of quod erat demonstrandum. 敵 means ‘be a match for’, or ‘rival’ in the sense of (being) an equal opponent. 218

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

18.4 h u ma n n at u r e

18.4  Human nature (Meng 6A2) Chapter six of the Mengzi contains conversations with a certain Gaozi, whose full name supposedly was Gao Buhai 告 不 害 . Gaozi was prob­ ably not a disciple of Mengzi but an independent philosopher, allegedly also connected to the Mohist school, who disagreed with Mengzi on quite a few points. Their most famous debate is that about human nature. 告 子 曰 :「 性 猶 湍 水 也, 決 諸 東 方 則 東 流, 決 諸 西 方 則 西 流, 人 性 之 無 分 於 善 不 善 也, 猶 水 之無分於東西也。」【注】湍者,圜也。謂湍湍瀠水也。 告子以喻人性若是水也。善惡隨物而化,無本善不善之性也。 孟子曰:「水信無分於東西, 無分於上下乎? 人 性 之 善 也, 猶 水 之 就 下 也。 人 無 有 不 善, 水無有不下。 今夫水,搏而躍之,可使過顙; 激 而 行 之, 可 使 在 山 : 是 豈 水 之 性 哉? 其 勢 則 然也。 人之可使為不善,其性亦猶是也。」

於上下乎,水性但欲下耳。人性生而有善,猶水欲下也。所

【 注 】 孟 子 曰, 水 誠 無 分 於 東 西, 故 決 之 而 往 也。 水 豈 無 分

以知人皆有善性,似水無有不下者也。躍,跳。顙,額也。

⑳ ⑲





性 xìng, nc, ‘(human) nature’ 湍 tuān, vi, ‘flow rapidly’ 決 jué, vtr, ‘breach’ 流 liú, vi, ‘flow’ 分 fēn, vtr, ‘divide’ 下2 xià, vi, ‘descend’ 搏 bó, vn, ‘beat’ 躍 yuè, vi, ‘leap’ 顙 sǎng, nc, ‘forehead’ 激 jī, vtr, ‘pump’

豈 qı ̌, adv, ‘surely not’



Comments ⑲ Here, the vn 猶 functions as cop. The word 性 , ‘(human) nature’ is clearly related to 生 (cf. statements like Meng 6A3: 生 之 謂 性 ), it designates inborn qualities. Gaozi argues that man, by nature, is dis­ posed neither to good nor to evil. ⑳ 信 , used adverbially, may be trans­ lated as ‘truly, indeed’ (interestingly, the commentary uses 誠 instead). Recall that 無 may represent a prind, which as O always precedes the P (cf. ④); thus, rather than ‘has no division’, the expression 無 分 should be translated as ‘makes no difference’. ㉑ Mengzi imitates the sentence structure from Gaozi’s initial statement. In the following complex sen­ tence, 人 and 水 should be construed as locative S (cf. 6.2.2–⑱). Note that 有 , unlike in MSC, is negated by 不 or, as in this case, 無 : the expression 無 有 is more forceful than simple 無 . What do you note 219

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 18 : m e n g z i

about the O of the two 有 ? ㉒ In this case, 夫 may be construed either as ptemp (coordinated with 今 ) or as prdem. Note the absolute topic, 水 , that connects the following four clauses. ㉓ The entire construction from 人 之 to 是 也 may be construed as one nominal clause: can you analyse its constituents? Box 19  Combinations of particles As stated in 1.2.7, particles add modal, aspectual, accentuating, interrogative, or other meanings to sentences. Since many sentences may express more than one of these meanings, it is not surprising that such particles may be com­ bined. This is especially true of particles in final position. While the two most common final particles, 也 and 矣, are mutually exclusive, they can be indi­ vidually coordinated with others: 莫我知也夫。

Nobody knows me, do they? (LY 14.35)

吾已矣夫。

Now it is over with me, isn’t it? (LY 9.9)

位其不可不慎也乎。

The position should certainly be treated with care! (Zuo 8.2)

可矣乎。

Is it possible now? (Zuo 5.27)

此豈山之性也哉。

Now should that be the nature of the mountain? (Meng 6A8)

久矣哉,由之行詐也。

It has been long, indeed, that You has practised deceit. (LY 9.12)

可與事君也與哉。

Could one possibly serve the lord together with them?! (LY 17.15)

夫子之道,忠恕而已矣。 The master’s way consists in integrity and empathy, that’s it. (LY 4.15) Whenever possible, every particle should be reflected in translation. However, 也 is an exception: since it is mandatory with a nominal P, it does not add a translatable meaning to the sentence.

The two particles 乎 and 哉 may also be combined: 何為不易乎哉。

Why should it not be easy?! (HFei 37)

吾罪也乎哉。

Should that be my fault?! (Zuo 9.25)

In all these cases, the order in which these particles are combined is fixed: they, too, seem to follow euphonic rules.

220

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 15

The Warring States At the beginning of the third century bc, the political landscape of what was to be­come ‘China’ no longer bore any similarity to that of early Zhou times. The quasi-­feudal order established by the early Zhou rulers (cf. Focus 7), which centred around the Zhou king, had long since disintegrated. Although the Zhou kings were still formally acknowledged as rulers of  ‘All under Heaven’, by the seventh century bc their erstwhile allies had become independent polities within a system led by changing ‘hege­mons’. But this order, too, collapsed when in many places the old aristocracy was toppled by new elites: in Lu, the ‘three Huan’ assumed control of the government in the sixth century; Jin was divided into three states by the families Zhao, Wei, and Han between 453 and 403 bc, and in Qi the Tian family usurped the throne in 386 bc. Other states were entirely annihilated: Zheng was destroyed by Han, one of the successor states of Jin, in 375 bc; Chen and Cai were both extinguished by Chu in the fifth century bc, Cao was conquered by Song in 487 bc, and Song itself was wiped out by Qi in 286 bc. At this time, only seven big states com­manding ‘10,000 chariots’—Han, Wei, Zhao, Yan, Qi, Chu, and Qin—and several smaller polities com­ manding ‘1,000 chariots’ remained (see Map 1). But these, too, would not survive much longer. When Meng Ke visited Wei around 320 bc, it had already suffered crushing defeats against Qi and Qin which forced Wei into quasi-­submission to the latter. Thus began the rise of Qin and a series of conquests which by 221 bc were to extinguish all of the ‘Middle States’. It is no coincidence that this age saw the beginnings of Chinese political thought. The heavenly legitimized rule of the Zhou kings was clearly at an end; instead, rulers of other states, like Hui of Wei, began calling themselves ‘king’ (王). But they could neither claim the same legitimacy as the Zhou kings nor could they rule the radically altered society by the same means as the latter had done. The art of government had to be re-­ formu­lated, or rather: it had to be formulated for the very first time. Whereas before, the order of the world was taken for granted, it now appeared contingent and questionable. Times had changed and the ­traditional rules inherited from the ancestors no longer worked; they themselves had to be changed to adapt to the times (SJS 1, cf. 27.2). This 221

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 15 : t h e wa r r i n g s tat e s

search for new forms of legitimation and principles of rulership drove the political discussion of the Warring States, in which thinkers like Mo Di, Meng Ke, Xun Kuang, Zhuang Zhou, Shang Yang, Han Fei, and the other ‘Hundred Schools’ (百家) contended to offer the best solution for the government of a new age. Significantly, they appeared at the courts of Qi, Wei, Han, or Qin—but not at the court of Zhou, where the last rulers of the age-­old dynasty still believed in the incontestable ‘Mandate of Heaven’. The Warring States period were the best of times and the worst of times. While the land was ravaged by wars, cities were laid waste and people uprooted on an unprece­dented scale, political philosophy flowered like never before. The Warring States were China’s classical age: the ‘Hundred Schools’ laid the intellectual foundations of Chinese trad­ ition, they developed systems of thought that Chinese thinkers would turn to for more than two millennia, until the present day. References: Classic accounts of Zhanguo thought are Hsiao Kung-­chuan 1979, Graham 1989, and Schwartz 1985; more recent works that take into account latest archaeological sources are Pines 2009 and Major and Cook 2017. Zhanguo history, archaeology, and thought are treated extensively in Loewe and Shaughnessy 1999, chs. 9–12.

YAN

Hua nghe

ZHAO

Handan

QI Linzi

WEI

Yellow Sea

LU SONG

QIN Xianyang

SHU

HAN

Xinzheng

Yangzi

ai

Hu

CHU Ying

Map 1  The Warring States Adapted from Caroline Blunden and Mark Elvin (eds.), Cultural Atlas of China. Oxford: Phaidon, 1983.

222

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 19

Xunzi

19.1 Introduction223 19.2 Encouraging learning (Xun 1) 224

19.3 On human nature (Xun 23)228 19.4 Social order (Xun 9)230

19.1 Introduction Xunzi, attributed to a certain Xun Kuang 荀況 (c.313–238 bc) of Zhao 趙, is widely recognized as the third major Confucian work of classical times, besides Lunyu and Mengzi. For a long time, however, Xunzi was hardly known to Chinese scholars. The great literatus Han Yu (768–824) described his surprise after having obtained a copy of the book, and its earliest surviving commentary was written by Yang Liang 楊倞 in the ninth century. This neglect may be due to the fact that among Xun Kuang’s students were two major exponents of the ill-­reputed realist tradition, Han Fei 韓非 and Li Si 李斯, and that many of the Xunzi’s tenets are indeed akin to realist thought. Indeed, the claim that ‘human nature is evil’ (人之性惡, Xun 23) challenges Mengzi’s cherished conviction ‘that human nature is good’ (人性之善也, Meng 6A2); and the advice to ‘take as models the later kings’ (法後王, Xun 8) does not tally well with the Confucians’ reverence for antiquity. However, in the chapter ‘Against the Twelve Master’ (非十二子), which critically summarizes the thought of other philosophers, the advice is given to ‘take as a model the purport of Zhongni [i.e. Confucius] and Zigong [one of his disciples]’ (法仲尼子弓之義). In fact, main ­topics in Xunzi are clearly inspired by Confucian thought: chapters like ‘Encouraging Learning’ (勸學), ‘Rectification of Names’ (正名), ‘Not Careless’ (不苟), or ‘On Rites’ (禮論) expand on themes expressed in the Lunyu. These chapters amount to veritable treatises on subjects that had 223

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 19 : x u n z i

only been touched upon in succinct remarks by the ‘master’. Xunzi, much more than the other works so far discussed, presents sophisticated, carefully structured essays rather than dialogues, aperçus, or disparate anecdotes. These are perhaps the first deliberately composed works in Chinese literature. Like so many other works, the received edition of Xunzi goes back to Liu Xiang who edited the text in thirty-­two chapters (cf. Focus 18). However, it is by no means clear how many of these sections are reliable representations of Xun Kuang’s thought. Differences in style and content suggest that quite a few chapters are later fabrications that were attached to the text in Han times. Especially the last seven chapters, containing a collection of poetic riddles and anecdotes, would seem to be late additions. The following passages are taken from Wang Xianqian’s Xunzi jijie 荀子集解. Editions: Xunzi jijie 荀子集解, ed. by Wang Xianqian 王先謙. 2 vols. Beijing 1996. Xunzi jianshi 荀子簡釋, ed. by Liang Qixiong 梁啟雄. Beijing 1983. Translations: Watson 1963a; Köster 1967; Knoblock 1988–94; Hutton 2014. Studies: Knoblock 1982–83; Goldin 1999, 2000; Kline and Ivanhoe 2000; Satō Masayuki 2003; Hutton 2016.

19.2  Encouraging learning (Xun 1) The following excerpt is taken from the very beginning of Xunzi; it expresses one of the book’s main points, the importance of education for the proper development of man’s faculties—and for the correction of his amoral nature. ①





224

荀子卷第一 勸學篇第一

君子曰:學不可以已。 青, 取 之 於 藍 而 青 於 藍 ; 冰, 水 為 之 而 寒於水。 以喻學則才過 其 本 性 也。 木 直 中 繩, 輮 以 為 輪, 其 曲 中 規, 雖 有 槁 曓, 不 復 挺 者,輮使之然也。 輮,

屈。 槁, 枯。 曓, 乾。 挺,



勸 quàn, vtr, ‘encourage’ 青 qīng, vst, ‘be blue’ 藍 lán, nc, ‘indigo plant’ 冰 bīng, nc, ‘ice’ 直 zhí, vi as nc, ‘straightness’ 中2 zhòng, vn, ‘match’ 繩 shéng, nc, ‘rope’ 輮 róu, vn, ‘bend’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

19.2 e n co u r ag i n g l e a r n i n g

直 也。 晏 子 春 秋 作「 不 復 贏 矣 」。 ○ 盧 文 弨 曰 :「 曓 」, 舊 本 作「 暴 」, 非。故 木 受 繩則直,金就礪則利,君子博學而日參省乎己,則知明而行無過矣。

參,三也。曾子曰:「日三省吾身。」知,讀為智。行,下孟反。 ○ 俞樾曰:「省乎」 二字,後人所加也。荀子原文蓋作「君子博學而日參己」。參者,驗也。君子博學而

日參驗之於己,故知明而行無過也。後人不得「參」字之義,妄據論語「三省吾身」

之文,增「省乎」二字,陋矣。大戴記 勸學篇作「君子博學如日參己焉」,「如」「而」

古 通 用, 無「 省 乎 」 二 字, 可 據 以 訂 正。故 不 登 高 山, 不 知 天 之 高 也 ; 不 臨深谿,不知地之厚也;不聞先王之遺言,不知學問之大也。干、越、 夷、貉之子,生而同聲,長而異俗,教使之然也。 干、越,猶言吳、越。

呂氏春秋「荊有次非得寶劍於干、越」,高誘曰:「吳邑也。」貉,東北夷。同聲,謂

啼 聲 同。 貉, 莫 革 反。 先 謙 案 : 吳、 干 先 為 敵 國, 後 干 併 於 吳, 管 子「 吳、 干 戰 」

及左傳「吳城邗」,即其明證。干為吳滅,而吳一稱干,猶鄭為韓滅而韓亦稱鄭。





曲 qū, vi as nc, ‘curve’ 規 guī, nc, ‘pair of compasses’ 槁 gǎo, vi, ‘dry up’ 曓 pù, vtr, ‘dry’ 挺 tı̌ng, vi, ‘be straight’ 礪 lì, nc, ‘whetstone’ 受 shòu, vn, ‘accept’ 參 cān, vi, ‘be three’ 省2 xı̌ng, vn, ‘examine’ 登 dēng, vn, ‘climb’ 臨 lín, vn, ‘overlook’ 谿 xī, nc, ‘gorge’ 厚 hòu, vst, ‘be thick’ 干 Gān, np, a city 貉 Mò, np, a tribe 聲 shēng, nc, ‘sound’ 俗 sú, nc, ‘custom’

Comments ① Not Xunzi speaks, but a junzi: like in many other passages of classical literature, he is quoted as a moral authority whose words are to be heeded. Observe that 學 is not the S of this sentence—what is it? ② Two parallel sentences with exposed O; for the function of 於 in both cases, review 8.2.2. In 水為, 水 may be construed as S or adverbial

225

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 19 : x u n z i

­ odification: in the latter case, what does it express? ③ 中 here stands m for a verb, ‘to hit (a target) > match’. In fact, the ancient form of the character represents an arrow piercing the centre of a target. ④ 有 may seem ­irritating in this sentence: which word does it represent? Note that 者 nominalizes the entire phrase starting from 雖. What is the syntactic function of this 者 phrase? ⑤ Three parallel sentences, the last of which is—as usual—the most elaborate: its clauses are much longer and it is concluded by 矣. The expression 日參省乎己 has been much commented upon. It is crucial to identify 省 as the P, then all the rest falls into place (but Yu Yue disagrees—what is his opinion?). The commentary compares this to 日三省吾身 (LY 1.4); what difference do you note in the construction? ⑥ This is a reminder of the cultural diversity of ancient China: not only the above four peoples had ‘different customs’ but more or less all regions of what was to become China; as a Later Han texts put it, ‘a hundred miles ⟨away⟩, habits are not the same, and a thousand miles ⟨away⟩, customs are not the same’ (百里不同 風,千里不同俗, Fengsu tongyi). Arguably, these differences were not the result of education but they made education necessary in the first place. 詩 曰 :「 嗟 爾 君 子, 無 恆 安 息。 靖 共 爾 位, 好 是 正 直。 神 之 聽 之, 介 爾景福。」 詩,小雅小明之篇。靖,謀。介,

為學則自

助。景,大也。引此詩以喻勤學也。神莫大

於化道,福莫長於無禍。

化道,故神莫大焉。修身則自無禍,故福莫長

焉。 ○ 俞 樾 曰 : 上 引 詩 云「 神 之 聽 之, 介 爾 景 福 」, 此 文「 神 」 字「 福 」 字 即 本 詩 文 也。

今本此二句提行,屬下節,非是。吾嘗終日 先謙案:大戴記「吾」上有「孔 而思矣 , ○ 子曰」三字。不如須臾之所學也, 吾 嘗 跂 而 望 矣, 不 如 登 高 之 博 見 也。







226

思 sī, vn, ‘ponder’ 須臾 xūyú, nc, ‘moment’ 跂 qı̌, vn, ‘stand on tiptoe’ 高2 gāo, nc, ‘high place’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

19.2 e n co u r ag i n g l e a r n i n g

念孫曰:「江河」本作「江海」,「海」與「里」

登 高 而 招, 臂 非 加 長 也, 而 見 者 遠 ; 順 風 而 呼, 聲 非 加 疾 也, 而聞者彰。 假輿馬者,非利足也, 而致千里;假舟檝者,非能水也, 能, 善。 絕, 過。 ○ 王 而 絕 江 河。

為韻,下文「不積小流,無以成江海」,亦

與「里」為韻,今本「海」作「河」,則失

皆 以 喻 修 身 在 假 於 學。 生 非 異,

其韻矣。君 子 生 非 異 也, 善 假 於 物

也。

言與眾人同也。 ○ 王念孫曰:生,讀為性,

大戴記作「性」。









招 zhāo, vn, ‘wave’ 臂 bì, nc, ‘arm’ 呼 hū, vn, ‘call (out)’ 疾2 jí, vst, ‘be intense’ 彰 zhāng, vst, ‘be clear’ 假 jiǎ, vn, ‘make use of ’ 輿 yú, nc, ‘cart’ 檝 jí, nc, ‘oar’ 絕 jué, vtr, ‘cross’

⑦ This is a stanza from the Shijing, ode Nr. 207; Legge translates it as:

‘Ah! ye gentlemen! / Do not reckon on your repose being permanent. / Quietly fulfil the duties of your offices, / Loving the correct and upright. / So shall the Spirits hearken to you, / And give you large measures of bright happiness’ (remember that this is pre-­Classical language, thus we need not concern ourselves with the grammar or vocabulary). The commentary tries to explain how this fits with the chapter’s line of argument. ⑧ These parallel sentences expound the two key words from the last two verses of the quoted stanza. Knoblock translates 化道 as ‘the transformation of the self with the Way’: do you agree? ⑨ This echoes LY 15.31, and the commentary points to yet another version of the saying; for 終日, cf. 10.3–⑭; since it is subordinated with 而, it should be construed as a VP. ⑩ 非 here negates a verbal P, implying an emphatic denial: ‘by no means’. ⑪ Whereas 見者遠 made good sense, 聞者彰 is somewhat odd; 所聞彰 would seem more logical—but not as nicely parallel. ⑫ In 能水, ‘master the water’, hence ‘swim’, 能 retains its original status as a verb. ⑬ 生 here functions as an adverbial modifier—or perhaps, as the commentary suggests, it stands for 性. Consider the difference between 假於物 and *假物; does 假 mean ‘borrow’ or ‘lend’ in this construction?

227

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 19 : x u n z i

19.3  On human nature (Xun 23) Like Mengzi, Xunzi also discussed human nature. But his stance on the question is the very opposite of Mengzi’s: ‘human nature is evil’, Xunzi maintains, it must be actively shaped to become good. This brings Xunzi closer to the realist tradition of Chinese thought than to the humanist Confucian tradition.

 性惡篇第二十三

228

荀子卷第十七

當戰國時,競為貪亂,不脩仁義,而荀卿明於治道,知其可化,無勢位以臨之,故激憤而著此論。

書曰「惟天生民,有欲無主,乃亂,惟聰明時乂」,亦與此義同也。舊弟二十六,今以是荀卿論



議之語,故亦升在上。

人 之 性 惡, 其 善 者 偽 也 。偽,為也,矯也,矯其本性也。 ○ 郝 懿 行 曰 :性,自 然 也。 偽, 作為也。「偽」與「為」,古字通。楊氏不了,而訓為矯,全書皆然,是其蔽也。 今人之性, 生 而 有 好 利 焉, 順 是, 故 爭 奪 生 而 辭 讓 亡 焉 ; 生 而 有 疾 惡 焉, 順 是, 故 殘 先 賊 生 而 忠 信 亡 焉 ; 疾 與 嫉 同。 惡, 烏 路 反。生 而 有 耳 目 之 欲, 有 好 聲 色 焉 , ○ 然則從人之性 , ○ 謙 案 : 下「 有 」 字 疑 衍。順 是, 故 淫 亂 生 而 禮 義 文 理 亡 焉。

俞樾曰:「犯分」,當作「犯文」。此本以「文」「理」相對。 ○

先 謙 案 : 論 語 八 佾 篇 集 解 :「 從, 讀 曰 縱。」 下 同。順 人 之 情, 必 出 於 爭 奪, 合 於 犯

分亂理而歸於暴 。 ⑯



偽 wěi, vtr, ‘contrive’ 辭 cí, vn as nc, ‘courtesy’ 讓 ràng, vn as nc, ‘deference’ 殘 cán, vtr as nc, ‘destruction’ 賊 zéi, vn as nc, ‘murder’ 色 sè, nc, ‘beauty’ 淫 yín, nc, ‘depravity’ 理 lı̌, nc, ‘order’ 然則 ránzé, conj, ‘consequently’ 情 qíng, nc, ‘emotion’ 合 hé, vi, ‘combine’ 暴 bào, vi as nc, ‘violence’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

19.3 o n h u ma n n at u r e

故必將有師法之化,禮義之道 ,道與導同。 然 後 出 於 辭 讓, 合 於 文 理, 而 歸 於 治。 用此觀之,然則人之性惡明矣,其善者  . . . 孟 子 曰 :「 人 之 學 者, 其 偽 也。 性善。」曰:是不然。是不及知人之性, 而不察乎人之性、偽之分者也。 凡性 者, 天 之 就 也, 不 可 學, 不 可 事 ; 禮 義 者, 聖 人 之 所 生 也, 人 之 所 學 而 能, 所事而成者也 。聖人之所生,明非天性也。事, 為 也, 任 也。不 可 學、 不 可 事 而 在 人 者 謂 之性,可學而能、可事而成之在人者謂 之偽。是性、偽之分也。

道2 dǎo, vn as nc, ‘guidance’ 察 chá, vn, ‘be clear about’







Comments ⑭ The ambivalent word 偽, which is hotly debated in the commentaries, is perhaps best rendered as ‘forged’, which also means ‘made, shaped’ as well as ‘false’. ⑮ A complex sentence consisting of three parallel sentences (the last one being the longest) linked together by the same topic. Note the use of 故 as a consecutive conj, just like 則; the parallelism of the two is nicely illustrated in SJS 6: 小人不避其 禁,故刑煩,君子不設其令,則罰行. ⑯ Human nature, according to Xunzi, is quarrelsome, rapacious, and violent. This certainly reflects the experience of a time that was not without reason called ‘Warring States’. Note the sequence 出 . . . 合 . . . 歸, which is repeated in the next sentence; 爭奪 and 辭讓, too, are recurring phrases that structure this carefully composed piece. ⑰ 用 was a common prep in pre-­Classical texts; in CC, it occurs as a more dignified synonym of 以 (cf. 9.2.1–③). 師法, the ‘models of a teacher’ which appears several times in Xunzi, apparently means transmitted knowledge; in other words, it is through education that man is refined. What kind of genitive is expressed in 師法之化 and 禮義之道 (cf. Box 9)? ⑱ Xunzi now turns to a direct criticism of Mengzi. However, the quotation he gives is not to be found

229

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 19 : x u n z i

in the Mengzi (and it is syntactically puzzling, as well); the following analysis of the difference between ‘nature’ and ‘nurture’, penetrating though it is, seems to miss Mengzi’s argument—or perhaps this argument simply has not been transmitted. ⑲ For the use of 凡, cf. 13.2–④. Is there anything amiss about the construction 可事而成之在人? 19.4  Social order (Xun 9) For all their talk about morals, learning, and ritual, Confucian thinkers had a keen sense of social order—an order based not on equality and universal rights, but on strict hierarchy. Xunzi is the most outspoken about this social order, which assigns every member of society a specific lot (分) and function that he has to fulfil. The chapter on the ‘Kingly system’ (王制) not only stresses that ‘the lord should act as lord, a minister as minister, a father as father, a son as son, an older brother as older brother, a younger brother as younger brother . . . a peasant as peasant, an officer as officer, an artisan as artisan, and a trader as trader’ (expanding a saying of Confucius, LY 12.11); it actually identifies the hierarchical order as the trait that sets humans apart from all other beings.

230

水火有氣而無生,草木有生而無知 ,知,謂性識。

禽獸有知而無義 , ○ 郝懿行曰:釋詰:「知者,匹也。」 詩曰:「樂子之無知。」此草木有生無知之說也。曲禮曰:「禽

獸無禮,故父子聚麀。」此禽獸有知無義之說也。楊注「知,





謂性識」,是已。蓋因有性識然後有匹偶,故此二義兼之乃備。

人 有 氣、 有 生、 有 知, 亦 且 有 義, 故 最 為 天 下 貴 也。 力 不 若 牛, 走 不 若 馬, 而 牛 馬 為 用, 何也?曰:人能羣,彼不能羣也。人何以能羣? 曰:分 。無分則爭,爭則不能羣也。分何以能行?曰: 義。 故義以分則和 ,言分義相須也。義,謂裁斷也。 和 則 一, 一 則 多 力, 多 力 則 彊, 彊 則 勝 物, 故 宮室可得而居也。 ㉑

氣 qì, nc, ‘energy’ 草 cǎo, nc, ‘grass’ 禽獸 qínshòu, nc, ‘birds and beasts’ 走 zoˇu, vi as nc, ‘running’ 羣 qún, nc as vi, ‘form a community’ 分2 fèn, vn as nc, ‘division’ 宮 gōng, nc, ‘residence’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

19.4 s o c i a l o r d e r

Comments ⑳ This is Xunzi’s version of the ‘great chain of being’ that places all things in the world in a hierarchical order, the elements at the bottom, then plants, animals, and humans at the top. 氣, immaterial pneuma or energy, permeates all beings, and at every level another quality is added. Note that the distinctive quality of humans is not cognitive but moral. For 且, cf. 15.2–⑥. For the last clause, cf. 9.3–㉛. ㉑ Humans can form differentiated communities (or ‘social systems’—but certainly not ‘herds’, despite the character form). The word 分 means ‘division’ as well as the particular ‘lot’ or ‘situation’ determined by this division. Differentiation of social systems requires that every member accepts his lot, in other words: that they act with ‘propriety’ (義). ㉒ This is the clincher: propriety leads to harmony, harmony leads to unity, and unity leads to strength. Although Mengzi also argues that the realm is ‘stabilized through unity’ (Meng 1A6), this line of reasoning seems much closer to political realism; compare SJS 5: 作一則力摶,力摶則強. Although Xunzi only talks about ‘overwhelming creatures’, it is but a small step to overwhelming enemies. In the construction 可得而居, 可 passivizes the P 居, despite the subordination by 而, which seems to be purely stylistic: compare 可得聞 (Meng 1A7) and 可得而聞 (LY 5.13). 故 序 四 時, 裁 萬 物 , ○ 先 謙 案 : 裁, 亦 成 也, 說 見非十二子篇。兼利天下,無它故焉,得之分 義也。 故人生不能無羣,羣而無分則爭, 爭則亂,亂則離,離則弱,弱則不能勝物, 故 宮 室 不 可 得 而 居 也, 不 可 少 頃 舍 禮 義 之 謂也。能以事親謂之孝,能以事兄謂之弟, 能 以 事 上 謂 之 順, 能 以 使 下 謂 之 君 。 能 以, 皆謂能以禮義也。君者,善羣也 。善能使人為羣也。 羣道當則萬物皆得其宜,六畜皆得其長, 羣 生 皆 得 其 命。 故 養 長 時 則 六 畜 育, 殺 生 時 則 草 木 殖 , 殺 生, 斬 伐。政 令 時 則 百 姓 一, 賢良服。 ㉔





231

序 xù, nc as vtr, ‘order’ 裁 cái, vn, ‘control’ 兼 jiān, vtr, as adv, ‘all’ 頃 qı̌ng, nc, ‘moment’ 畜 chù, nc, ‘domestic animal’ 育 yù, vtr, ‘breed’ 殖 zhí, vtr, ‘cultivate’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 19 : x u n z i

Comments ㉓ 兼 here functions as an adverb, referring to the O in the sense of ‘all’ (cf. Box 30). Note that Xunzi speaks of 利—Mengzi would not have approved. ㉔ There seems to be a caesura in the text: 故宮室不 可得而居也, which echoes the almost identical clause before, concludes the argument. What follows, appears as a coda with a different tone: 禮 義 is a new expression, so are the moral concepts that follow as well as 謂, which occurs five times. All these sentences provide definitions: what do they remind you of? Paronomastic associations like 君 and 羣 are not uncommon in old commentaries (compare 政者正也, LY 12.17, 仁者人也 or 義者宜也, both Li 31): they should be taken as puns rather than serious etymologies. ㉕ Recall that 其, as a mod, may mean ‘their proper’ (cf. Box 18). 殺生, ‘kill living beings’, here specifically refers to ‘felling (trees)’. Note the realist turn in the final sentence: not only the concern about ‘governmental orders’, but the unification of the people and the subjugation of worthies are characteristic for the discourse of political realism (cf. Lesson 27).

Box 20  Taboo characters The Hanshu yiwenzhi, China’s first bibliography, does not list a work entitled Xunzi. Instead, there is an entry that runs 孫卿子三十三篇: the name 荀 is rendered 孫, because 荀 was a taboo character in Han times. Being part of the Han emperor Xuan’s (r. 74–48 bc) personal name Liu Xun 劉詢, it was forbidden to be spoken or written. The custom of tabooing the given names of emperors (and one’s ancestors) probably originated in Qin times. While the exact rules and practices of dealing with taboos—when to start, how long to abide by it, and what names to avoid—are not entirely clear, the ways in which taboo characters were treated should be noted by students of Classical Chinese. The character in question could simply be omitted, it could be altered (by omitting or adding strokes), or—most commonly—it could be replaced by a homophonous or synonymous character. For example, Qin texts avoided the use of 楚, which was the given name of king Zhuangxiang of Qin (秦莊襄王, r. 250–247 bc), and wrote 荊, instead; they also avoided the given name of the First Emperor, 政, even replacing the character 正 by the synonym 端. Another famous example is the name of the Han dynasty’s founder, Liu Bang 劉邦 (r. 202–195 bc), which necessitated the replacement of the character 邦 by the synonymous 國 in Han texts. And 荀

232

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

19.4 s o c i a l o r d e r

(in late Han: *suin), as in the example above, was replaced by a close homophone, 孫 (*suən), in Han texts. Reckoning with taboo characters thus becomes part of textual criticism: one must always consider the possibility of taboo characters having been avoided or—even more tricky—mistakenly reinserted by Han or later editors. References: The use of almost a thousand taboo characters throughout Chinese history is discussed in Wang Yankun 1997; for a general overview, cf. Wilkinson 2018, 290–3; for the intricate case of 荀, cf. Knoblock 1988–94, vol. 1, 233–9.

233

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 16

Dictionaries Dictionaries are indispensable tools for students of Classical Chinese. They can be of great help, but they can also be very misleading. Therefore, the choice of the right dictionary is crucial. A good dictionary for our purposes should derive its definitions from Classical Chinese texts, it should give citations, and it should indicate the sources quoted. It goes without saying that dictionaries of Modern Standard Chinese are useless for dealing with classical texts (recall the problem of faux amis, Box 2). Moreover, diachronic dictionaries that draw on sources from different periods but do not indicate their sources are equally inadequate. These include R.  H.  Mathews’ immensely popular Chinese-­English Dictionary (Mathews 1931), Herbert A. Giles’ massive Chinese-­English Dictionary (Giles 1892), and even the otherwise impressive Grand dictionnaire Ricci de la langue chinoise (Institut Ricci 2001). All these works offer translations of a great many words and phrases, but they do not indicate the texts or periods for which these translations apply.

What one needs is a dictionary of Classical Chinese. There are only three Western works that qualify for this. The first is Séraphin Couvreur’s Dictionnaire classique de la langue chinoise (Couvreur 1904), which, albeit not limited to classical texts, supports most of its translations with example sentences and their sources. The second is Ulrich Unger’s Glossar des Klassischen Chinesisch (Unger 1989) which is strictly limited to the vocabulary of Zhanguo texts; however, it does not provide ­citations or sources and it is not comprehensive. The most recent work is A Student’s Dictionary of Classical and Medieval Chinese (Kroll 2015), which provides careful and precise translations and is especially strong in its discussion of function words; however, since it combines Classical and medieval Chinese without, in most cases, providing citations or sources, it is often not easy to keep the two apart. Two rather specialized dictionaries deserve to be mentioned: Bernhard Karlgren’s Grammata Serica Recensa (Karlgren 1957), a pioneering work that provides phonological reconstructions as well as translations (with sources, but without citations) for more than 8,000 characters from pre-­Classical and Classical texts. Although this is a classic, Karlgren’s reconstructions are largely outdated (for a complete update, cf. Schuessler 2009), and the arrangement of the dictionary by phonetic series makes it somewhat cumbersome to use. Axel Schuessler’s ABC Etymological Dictionary of Old Chinese (Schuessler 2007), arranged by pinyin, is more up to date, but less comprehensive; its emphasis is on phonology and

234

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 16 : d i c t i o n a r i e s

etymology, while translations are brief and citations are not given. Finally, John Cikoski’s Lexicon of Classical Chinese (Cikoski 1994–2011) should be noted, an overwhelming work in three volumes; however, the Lexicon’s structure is highly idiosyncratic and, despite its title, it does not differentiate between Classical and Literary Chinese.

There are several useful modern Chinese dictionaries of Classical Chinese. Two affordable and convenient works for quick reference may be singled out: the Gu Hanyu zidian 古汉语字典 (Wang Li 2000) and the Gudai Hanyu cidian 古代汉语词典 (Chen Fuhua 2002), both of which provide pronunciations, translations, and carefully chosen c­ itations of ten thousands of Classical Chinese words in a single volume. At some point, however, every serious student of Classical Chinese will turn to the comprehensive scholarly dictionaries of the Chinese language, the Zhongwen da cidian 中文大辭典 (Zhang Qiyun 1990) and the Hanyu da cidian 漢語大詞典 (Luo Zhufeng 1986–94). Both are multi-­ volume works of great erudition, compiled by editorial teams over many years, comparable in scope to the Oxford English Dictionary or Grimm’s Deutsches Wörterbuch. But they go further than these latter works insofar as they are encyclopaedic dictionaries, dealing not only with words, but also with Chinese places, people, books, flora and fauna, etc. These are truly comprehensive resources for the study of Chinese culture. Due credit needs to be given to the mother of all encyclopaedic Chinese dictionaries, Morohashi Tetsuji’s 諸橋轍次 magnificent Dai kanwa jiten 大漢和辭典 (Morohashi 1955–60), which contains more than half a million entries culled from classical and imperial texts along with Japanese readings and translations. The Zhongwen da cidian is, for the most part, a translation of this groundbreaking work.

This being said, significant differences between the two need to be pointed out. While the Hanyu da cidian is especially strong for Early Modern and Modern Chinese—more than half of the 1430 texts it quotes date from Qing and Republican times—the Zhongwen da cidian lays more emphasis on Old Chinese and Middle Chinese. Moreover, it is vastly superior in its encyclopaedic material. The comparison of the respective entry 莊子, given in Figure 7, may serve to illustrate this. The Hanyu da cidian explains this as (1) ‘village’, (2) ‘manor’, or (3) ‘Zhuang Zhou. Philosopher from Zhanguo times’. It does not give any information whatsoever about the book Zhuangzi, nor does it provide a separate entry on Zhuang Zhou. The Zhongwen da cidian, on the other hand, has a long article about the book, providing the supposed author’s name and the alternative title, a summary of its contents and philosophical thought, information about editions and commentaries, its structure, and the names of all thirty-­three chapters as well as a bibliography containing dozens of 235

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 16 : d i c t i o n a r i e s

r­ eference works. Moreover, Zhuang Zhou is treated in a separate article which provides a short biography (including references) and even a ­picture of the man. This example should suffice to indicate the wealth of information offered in the Zhongwen da cidian. Although its sheer size and the arrangement of characters by traditional classifiers may pose a challenge to students, it is well worth the effort to get acquainted with this encyclopaedic dictionary: it is the premier tool for dealing with Classical Chinese texts. Both Zhongwen da cidian and Hanyu da cidian as well as an array of other dictionaries— including the above-­mentioned Wang Li 2000, traditional works like Shuowen jiezi 說文解 字 and Kangxi zidian 康熙字典, and several specialized dictionaries—can be electronically searched at .

Figure 7  Three dictionary entries for 莊子 and 莊周 Sources: a) Zhang, Qiyun 張其昀et al. (eds.), Zhongwen da cidian 中文大辭典. 10 vols. Taipei: Zhongguo wenhua daxue, 1990, vol. 7, 1554;

236

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 16 : d i c t i o n a r i e s

b) Zhang, Qiyun 張其昀et al. (eds.), Zhongwen da cidian 中文大辭典. 10 vols. Taipei: Zhongguo wenhua daxue, 1990, vol. 7, 1556; c) Luo Zhufeng 羅竹風 et al. (eds.). Hanyu da cidian 漢語大詞典. 13 vols. Shanghai: Hanyu da cidian chubanshe, 1986–94, vol. 9, 425.

237

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 20

Zuozhuan

20.1 Introduction238 20.2 A family feud in Zheng (Zuo 1.1) 239

20.3 The story of Zhao Dun (Zuo 7.2)

244

20.1 Introduction The Zuozhuan 左傳, or Zuo Tradition, is China’s oldest narrative history, covering the period from 722 to 468 bc, and the longest pre-­imperial Chinese text we have. Yet, surprisingly little is known about its origins. Tradition has attributed the text to a certain Zuoqiu Ming 左丘明, al­leged­ly a contemporary of Confucius. Moreover, it is thought to be a com­mentary to the Chunqiu, the annals of Lu attributed to Confucius, and as such has been regarded as a canonical work. However, the Zuozhuan is very much a work in its own right. At about 180,000 characters, it is far bigger, richer, and more complex than the Chunqiu, and most of its narratives are concerned not with Lu but with the state of Jin. The Zuozhuan was firmly connected to the Chunqiu only with the ­editorial work of Liu Xin 劉歆 (46 bc–ad 23) and Du Yu 杜預 (222–284), who wrote an influential Zuozhuan commentary. Ever since, the two works have been interwoven: the short Chunqiu comments to an entire year are followed by the long, sometimes epic narra­tions of the Zuozhuan. Therefore, the Zuozhuan is usually quoted not by juan and pian but by ruler and year; the passage below would be quoted as Zuozhuan Xuan 2 or—as in the present book—Zuo 7.2 (Duke Xuan being the seventh of twelve dukes treated in the Chunqiu).

These narrations include strategic and moralistic discussions, pol­it­ ical schemes, stories of love and hate, succession struggles, as well as conspiracies, personal vengeance, warfare, and other atrocities. All this is clearly more than just a commentary to the Chunqiu, and modern 238

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

20.2 a fa m i ly f e u d i n z h e n g

scholars have come to treat the Zuozhuan as an independent his­tor­ ic­al source. However, this view seems equally problematic, since the Zuozhuan certainly is not a source contemporary to the events it narrates; in fact, its main parts are now believed to have been composed around the fourth century bc. Moreover, numerous passages in the Zuozhuan, including many of its elaborate speeches, seem to be not so much historical records as literary artefacts: the Zuozhuan may well be studied as a work of literature. Such literary analysis serves to show how heterogeneous a work the Zuozhuan is: it consists of short annalistic entries, didactic speeches, battle scenes, songs, divinations, juicy anecdotes, and moral­izing commentaries ascribed to a ‘gentleman’ or even Confucius himself. In all this, the Zuozhuan is testimony to the vibrant and multifaceted intellectual climate of the classical age. The following passages are taken from Yang Bojun’s Chunqiu Zuozhuan zhu. Editions: The best edition, including an excellent modern commentary, is Chunqiu Zuozhuan zhu 春秋左傳注, ed. by Yang Bojun 楊伯峻. 4 vols. Beijing: Zhonghua, 1981. Two fine Japanese editions, based on the oldest complete manu­script version, were done by Takezoe Shin’ichirō 竹添進一郎, Saden kaisen 左傳會箋. 2 vols. Tōkyō: Fuzanbō, 1911; and Kamata Tadashi 鎌田正, Shunjū sashiden 春秋左氏伝. 4 vols. Tōkyō: Meiji Shoin, 1971. Translations: For a long time, the only Western translations of the Zuozhuan were those by Legge 1872 and Couvreur 1914. They have now been superseded by Durrant, Li, and Schaberg 2016. Studies: Karlgren 1926, Maspéro 1931–32, Kamata Tadashi 1963, and Tong Shuye 1980 are essential; Xu Renfu 1981 is a useful collection of tex­tual parallels; Shen Yucheng and Liu Ning 2000 give an overview of Chinese scholar­ship; for Zuozhuan syntax, cf. Guan Xiechu 1994; among recent Western monographic studies, Schaberg 2001, Pines 2002, and Li Wai-­ yee 2007 stand out; for philo­logical studies of textual transmission and integrity, cf. Egan 1977; Vogelsang 2007, 2011; Petersen 2015.

20.2  A family feud in Zheng (Zuo 1.1) The very first story in the Zuozhuan refers to the Chunqiu entry 夏,五月,鄭伯克段于鄢, ‘In summer, in the fifth month, the Earl of  Zheng defeated Duan in Yan’. The Zuo­zhuan provides an exciting background narrative to this dry entry: this includes the breech birth of the Zheng heir apparent, his mother’s animosity and jealousy, growing power of his younger brother, speeches by wise counsellors, and a ­veritable civil war in Zheng; even the romantic reconcili­ation of mother and son is recounted. It is a good example of how the Zuozhuan turns annalistic records into history. 239

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 20 : z u o z h ua n





初,鄭武公娶于申,曰武姜, 初, 乃表追敘前事副詞。鄭武公



名掘突。 經稱鄭伯,傳公者,公是諸侯之通稱,無分於公、侯、伯、子、



男。申,國名,伯夷之後,姜姓。後為楚所滅。故城在今河南省 南陽市。

武是武公之諡,姜為姓。 生 莊 公 及 共 叔 段。 共音恭。杜注以為 段出奔共,故曰共叔。叔是排行,段是名。 莊公寤生,驚姜氏, 故名曰寤生,遂惡之。 寤字當屬莊公言,乃「牾」之借字,寤生 猶言逆生,現代謂之足先出。愛共叔段,欲立之。 立之為太子。 亟請於武公,公弗許。 亟音器,屢也。弗,義同于「不 ⋯⋯ 之」, 動 詞 下 賓 語 包 含 在 內。 及 莊 公 即 位, 為 之 請 制。 制, 地 名, 即今河南省 滎陽縣 汜水公社。公曰:「制,巖邑也,巖邑即險邑。 虢 叔 死 焉。 佗 邑 唯 命。」 當 言 佗 邑 唯 命 是 聽, 此 是 語 言 省 略, 請京,使居之,謂之京城大叔。 乃當日常語。



申 Shēn, np, a polity 姜 Jiāng, np, a clan 莊 Zhuāng, np, posth. name 共叔段 Gōng Shūduàn, np 寤 wù, vn, ‘go against’ 驚 jīng, vi, ‘be shocked’ 亟 qì, adv, ‘repeatedly’ 即 jí, vn, ‘arrive’ 巖 yán, vi, ‘be dangerous’ 虢叔 Guó Shū, np 京 Jīng, np, a place

Comments ① The ntemp 初, placed at the beginning of a clause, refers to an undated time in the past; much like ‘once upon a time’ in fairy tales, it typically introduces historical anecdotes that were probably transmitted orally. Such anecdotes are quite distinct from annalistic entries and moralistic commentaries in the Zuozhuan—unlike those latter, they often contain juicy gossip about love and intrigue: they are the stuff that makes the Zuozhuan such an entertaining book. For the names in this sentence, all necessary information is given in the commentary. ② 及 here represents a coordinating conj, ‘as well as’, which is somewhat more formal than 與; moreover, it seems to imply that the second noun is less important than the first. Do you agree with the analysis of the name 共叔段 given in the commentary? ③ Here we have the typical ingredients of a family feud: two sons, the elder being the rightful heir to the throne, but the younger one being favoured by his mother. The meaning of 寤生 has been hotly debated; suffice it to say that it was an unpleasant birth. For the relationship between events 240

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

20.2 a fa m i ly f e u d i n z h e n g

at birth and given names, cf. Focus 5. ④ Note the syntactical information given in the commen­tary: on this account, modern commentaries are much better than traditional glosses. ⑤ 即位 means ‘accede the throne’. 虢叔 was the ruler of a small state about whom not much is known. For 佗邑唯命, cf. the ­commentary.



祭仲曰: 祭仲,鄭大夫。 「都城過百雉,國之害也。



都、城為兩詞,都為都邑,城謂城垣。雉高一丈長三丈。管

子云「國小而都大者弒」,可為「國之害也」注腳。 先王 之制:大都,不過參國之一;中,五之一;小, 九 之 一。 參 同 三。 國, 國 都 ; 與 國 家 之 國 意 義 不 同。 今 京 不 度, 非 制 也, 不 度, 猶 言 不 合 法 度。君 將 不堪。」公曰:「姜氏欲之,焉辟害?」 焉,何處。

辟,同避,逃避。對曰:「姜氏何厭之有? 猶言有何



厭;厭,足也。此是賓語(何厭)置於動詞(有)前之結構。



不如早為之所,無使滋蔓!蔓,難圖也。 滋蔓, 同 義 連 綿 詞。蔓 草 猶 不 可 除, 況 君 之 寵 弟 乎!」 公曰:「多行不義,必自斃,子姑待之。」



祭仲 Zhài Zhòng, np 都 dū, nc, ‘city’ 雉 zhì, nc, a unit of meas­urement 度 dù, nc as vi, ‘comply with (a measure)’ 堪 kān, vn, ‘endure’ 辟/避 bì, vn, ‘avoid’ 厭 yàn, vn as nc, ‘satisfac­tion’ 早 zǎo, vi, ‘be early’ 滋 zī, vi, ‘increase’ 蔓 màn, vi, ‘spread’ 除 chú, vtr, ‘eliminate’

寵 choˇng, nc, ‘favour’

自 zì, adv, ‘him-/herself ’ 斃 bì, vi, ‘topple’

Comments ⑥ An otherwise unknown person objects; note that the first colon of his statement is a subject clause without 之 or 也. ⑦ Fractions may be expressed by simple subordination of numerals, as in 三一, or with the addition of an nominal head to the first numeral, as in 三分之一 or 參國之一 (cf. Box 3). The following parallel clauses are highly elliptical: they supply only the new information, leav­ing out aforementioned heads of NP as well as the P. ⑧ The construction 何厭之有 should look familiar; if it does not, consider the commentary for an explanation. ⑨ The clause 不如早為之所 merits detailed syntactical analysis: try to determine the function of every single word. For the construction 猶 . . . 況, cf. 16.2–⑩. ⑩ For the meaning of 姑, cf. 10.5–㉞. 241

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 20 : z u o z h ua n

既而大叔命西鄙、北鄙貳於己。 既而,猶言不久。西鄙,北鄙,鄭國西部與北部邊境一帶地。 公子呂 曰:「國不堪貳,君將若之何?欲與大叔, 謂以國君之位讓與大叔。臣請事之;若弗與,則請 除之,無生民心。」 公曰:「無庸,將自及。」 將自及,謂禍將自及。大叔又收貳以為己邑, 至於廩延。子封曰:「可矣,厚將得眾。」 厚,謂勢力雄厚。此為語言緊縮後之分句,以意義言,宜作 一逗。公曰:「不義,不暱。厚將崩。 」 不義不暱。當解為不義則不暱。 大叔完、聚,完,謂完城郭, 孟子 裡樓上「城郭不完」可證。繕甲、兵,具卒、乘,將襲鄭,夫人將啟之。 啟,開也。此謂開城門。 公聞其期,曰:「可矣。」 命子封帥車二百乘以伐京。 春秋時多以車戰,車一輛謂之一乘。一車 用甲士十人。京叛大叔段。段入於鄢,公伐諸鄢。五月辛丑, 古以干支紀日,自甲骨鐘鼎皆如此。辛 丑為二十三日。大叔出奔共。 共即今河南省 輝縣。 書曰:「鄭伯克段于鄢。 」段不弟,故不言弟; 弟, 或 破 讀 為 悌, 亦 通。如 二 君, 故 曰 克 ; 稱 鄭 伯, 譏 失 教 也 : 此 言 兄 本 有 教 弟 之 責, 莊 公 於 弟 不 加 教 誨, 養 成 其 惡, 故 不 言 兄, 而 書 其 爵。謂 之 鄭 志。 不 言 出 奔, 難 之 也。 遂 寘 姜 氏 于 城 潁, 而誓之曰:「不及黃泉,無相見也!」 黃泉,地下之泉。此二句猶言不死不相見。 ⑫













242

大叔 Dàshū, np 西鄙 Xībı̌, np, a city

北鄙 Běibı ̌, np, a city 呂 Lüˇ , np

庸 yōng, vtr, ‘need’ 收 shōu, vtr, ‘receive’

廩延 Lı ̌nyán, np, a city 子封 Zıf̌ ēng, np

暱 nì, vst, ‘be close’ 崩 bēng, vi, ‘collapse’ 完 wán, vi, ‘be ready’ 聚 jù, vtr, as nc, ‘troops’ 繕 shàn, vn, ‘put in order’ 具 jù, vi, ‘be complete’ 襲 xí, vn, ‘attack’ 夫人 fūrén, nc, ‘lady’ 啟 qı̌, vtr, ‘open (up)’ 期 qī, vn, as nc, ‘appointed time’ 帥 shuài, vn, ‘lead’ 叛 pàn, vn, ‘rebel’ 辛 xīn, ntemp, 8th ‘stem’ 丑 chǒu, ntemp, 2nd ‘branch’ 奔 bēn, vi, ‘flee’ 稱 chēng, vtr, ‘call’ 譏 jī, vn, ‘criticize’ 寘 zhì, vn, ‘place’

城潁 Chéngyı ̌ng, np, a city 誓 shì, vn, ‘vow’ 黃泉 huángquán, nc, ‘the netherworld’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

20.2 a fa m i ly f e u d i n z h e n g

Comments ⑪ For 既而, which means ‘that being finished’ > ‘afterwards, later’, cf. 16.3. 貳 is a technical term referring to a second allegiance: Dashu is build­ing up a local power base. For the reference of 己, cf. 12.3–㉜. ⑫ The vn 與 may be used with either Oi or Od, or with two O; which is

the case here? Note the rare use of the conj 若 and 則: this contrasts with most of the other sentences that use no conj at all. ⑬ Three advisors say more or less the same thing, and the duke thrice gives more or less the same answer: how could this apparent redundancy be explained? ⑭ Only after his son actually prepares an attack, the duke says 可矣: ‘now we can’. ⑮ Now the story has arrived at the point of the Chunqiu entry: 鄭伯克段于鄢. Note how briefly the battle is described. The Zuozhuan places far more emphasis on the debates surrounding warfare than on the actual engagements. ⑯ This is an entirely different sort of commentary: not an historical narrative but a philological commentary that meticulously explains every word of the Chunqiu (referred to as 書) entry, trying to discern the judgments they supposedly ex­press. The Gongyang zhuan and Guliang zhuan consist almost exclusively of this sort of commentar­ies (cf. Focus 10). ⑰ The conj 遂 leads back to the narration; note that it connects to the last ­sentence before the commentary. The implicit S is the duke. In the absence of a legal system, oaths, 誓, were quite common in Chunqiu times; hundreds of them have been found on jade and stone tablets in  the area of modern Houma and Wenxian (Shanxi), and many appear in the Zuozhuan. 既而悔之。潁考叔為潁谷封人,聞之, 有 獻 於 公。 公 賜 之 食。 食 舍 肉。 舍 音捨,置也。公問之, 對曰:「小人有母, 皆 嘗 小 人 之 食 矣 ; 未 嘗 君 之 羹, 請 以遺之。」 遺,去聲,饋也,與也。公曰: 「 爾 有 母 遺, 繄 我 獨 無!」 潁 考 叔 曰 :「 敢 問 何 謂 也?」 公 語 之 故, 且 告 之 悔。 對 曰 :「 君 何 患 焉? 若 闕 地 及 泉, 隧 而 相 見, 闕, 掘 也。 隧, 動 詞, 掘 作 隧 道。其 誰 曰 不 然?」 其, 語 氣 副 詞, 此表疑問語氣。公從之。公入而賦:「大 隧 之 中, 其 樂 也 融 融。」 中、融為韻,古 ⑱



243

潁考叔 Yı ̌ng Kǎoshū, np 潁谷 Yı ̌nggǔ, np, a place

獻 xiàn, vn, ‘present (to)’ 羹 gēng, nc, ‘broth’ 繄 yī, interj, ‘alas’ 語 yù, vn, ‘tell (s.o. sth.)’ 闕/掘 jué, vn, ‘dig’ 隧 suì, nc, as vn, ‘dig a tunnel’ 賦 fù, vn, ‘re­cite’ 融融 róngróng, vi, ‘be united’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

音同在冬部。 姜出而賦:「大 隧 之 外, 其 樂 也 洩 洩。」

作「 泄 」。 今「 洩 」 已 通 行, 故



不 改。 外、 洩 為 韻, 古 音 同 在

曷 部。遂 為 母 子 如 初。 君 子 曰 :「 潁 考 叔 純 孝 也,愛其母,施及莊公。施,

去 聲, 讀 難 易 之 易, 延 及 也。

詩 曰 :『 孝 子 不 匱, 永 錫 爾 類。』 詩 見 今 詩 經 大 雅

既醉。其是之謂乎!」



「 洩 」 本 作「 泄 」。 金 澤 文 庫 本

l e s s o n 20 : z u o z h ua n

洩洩 yìyì, vi, ‘spread’ 純 chún, vi, ‘be upright’ 施2 yì, vi, ‘spread’

Comments ⑱ 既而 introduces another break in the narrative progression; moreover, yet another new person is introduced: these are indications of an originally independent textual unit. ⑲ This is one of the rare cases in which a distributive, 皆, refers not to the S but to the O (cf. 7.3). The clause 爾有母遺 is remarkable: it has 有 followed by an O and a C (cf. Box 25). Compare English constructions like ‘I have a job to do’. Note the rather informal usage of the prper 爾 and 我: this corresponds to the message of the utterance, in which the duke ignores all etiquette and vents his personal emotions. ⑳ Here, the story takes quite a surprising turn. The duke and his mother both recite a poem, as is often done to express one’s feelings. The reduplicative binomes (cf. Focus 9) further emphasize the emotive quality of the couplets. ㉑ A final commentary concludes this episode. Such commentaries by a 君子 or by 孔子 are rather frequent in the Zuozhuan. They are not philological but express moral judgments. The quote from Shijing 247, in James Legge’s translation, reads: ‘For such filial piety, without ceasing, / There will ever be conferred blessing on you.’ Remarkably, the junzi comment refers only to the very last part of this story, ignoring the entire drama that preceded Ying Kaoshu’s appearance. 20.3  The story of Zhao Dun (Zuo 7.2) The following passage is taken from the second year of Duke Xuan of Lu (r. 608–591 bc). It corresponds to the Chunqiu entry of the same year which runs: 秋九月乙丑,晉趙盾弒其君夷皋, ‘In autumn, in the ninth month, on the day yichou [i.e. the 26th day of the 9th month], 244

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

20.3 t h e s to ry o f z h ao d u n

Zhao Dun of Jin assassinated his ruler Yigao’. The narrative is a good example of how the Zuozhuan expands on such scanty Chunqiu entries.







晉靈公不君: 論語 顏淵「父不父,子不子」,此則君不君。猶言在 君 位 而 言 行 不 合 為 君 之 道。厚 斂 以 彫 牆 ; 從 臺 上 彈 人, 而 觀 其 辟丸也;宰夫胹熊蹯不熟, 熊蹯即孟子 告子上之熊掌,其味甚美, 然 難 熟。殺 之, 寘 諸 畚, 畚 音 本, 說 文 云 :「 蒲 器 也。」 使 婦 人 載 以 過 朝。 趙 盾、 士 季 見 其 手, 死屍之手露于外。問 其 故, 宣 子驟諫, 公患之, 晉語五云:「靈公虐,趙宣子 而患之。 . . .



驟諫,公患之。」韋注:「患,疾也。」呂氏春秋 過理篇云:「趙盾驟諫而

不 聽, 公 惡 之。」 患、 疾、 惡 三 義 相 近。 使 鉏 麑 賊 之。 晨 往, 寢 門闢矣,盛服將朝。 尚早,坐而假寐。 假寐,不解衣冠而 睡。麑 退, 歎 而 言 曰 :「 不 忘 恭 敬, 恭 敬 指 早 起 盛 服 將 朝 之 事。 民之主也。賊民之主,不忠;棄君之命,不信。有一於此, 不如死也。」 觸槐而死。 公羊傳云「遂刎頸而死」,與諸書異。













彫 diāo, vtr, ‘carve’ 牆 qiáng, nc, ‘wall’ 臺 tái, nc, ‘terrace’ 彈 tán, vn, ‘shoot’ 丸 wán, nc, ‘pellet ’ 宰夫 zǎifū, nc, ‘cook’ 胹 ér, vtr, ‘boil in water’ 熊 xióng, nc, ‘bear’ 蹯 fán, nc, ‘paw’ 熟 shú, vi, ‘be well cooked’ 畚 běn, nc, ‘basket’ 婦人 fùrén, nc, ‘woman’ 載 zài, vn, ‘carry’ 趙盾 Zhào Dùn, np 士季 Shì Jì, np

宣子 Xuānzı ̌, np 驟 zòu, adv, ‘repeatedly’ 諫 jiàn, vn, ‘protest’ 鉏麑 Chú Ní, np 晨 chén, ntemp, ‘morning’ 辟/闢 pì, vtr, ‘open’ 盛 shèng, vst, ‘be full’ 服2 fú, nc, ‘clothes’ 尚 shàng, adv, ‘still’ 寐 mèi, vn, ‘sleep’ 歎 tàn, vn, ‘sigh’ 觸 chù, vn, ‘butt’ 槐 huái, nc, ‘Pagoda Tree’

Comments ㉒ Most of the Zuozhuan narratives are concerned with the state of Jin. Duke Ling of Jin reigned 620–607 bc, the story is set in 607 bc (being the second year of Duke Xuan of Lu). The posthumous name Ling, ‘Numinous’, implies a negative judg­ment, as this case exemplifies. In this sentence, 君 must be a verb, since it is negated by 不; on this usage, cf. Box 10 and the parallel given in the commentary. ㉓ Bear’s paws were one of 245

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 20 : z u o z h ua n

the greatest delicacies in ancient China. Mengzi, to whom the commentary alludes, says: 魚, 我所欲也, 熊掌, 亦我所欲也。 二者不可 得兼, 舍魚而取熊掌者也 (Meng 6A10). The construction of this sentence is unusual insofar as 而 is to be expected before 不. ㉔ The rare character 畚 is explained by a standard commentary phrase: A音B, meaning ‘A is pronounced like B’; the commentary then goes on to quote the Han dictionary Shuowen jiezi 說文解字, which defines the word 畚 as ‘a reed container’. ㉕ Being implicit in the following 以, the O is omitted after 載 for stylistic reasons. For the analysis of以, cf. 16.5–㉚. ㉖ Xuanzi, so the commentators tell us, is the same person as the afore­mentioned Zhao Dun. Can you explain the difference in names? ㉗ The commentary quotes parallels to this story from the ‘Jin yu’, which is a chapter of the Guoyu (note that only the chapter title is quoted: the reader is expected to know which book it belongs to), including the third century commentary by Wei Zhao 韋昭, and from the Lüshi chunqiu, section ‘Guo li’. Many historical anecdotes are transmitted not only in one but in several texts, and there are often significant differences between them. ㉘ 盛服 is a NP which serves as an adverbial modifier; compare 子路盛服而見孔子 (Xun 29); for 將朝 cf. 10.5–㉘. ㉙ The adv 尚 marks the continuative aspect, indicat­ing that a process or state is ‘still’ ongoing. 假寐 is an idiomatic expression: ‘catch a nap’. ㉚ The last sentence has the structure S 不 如 C 也, both S and C being ‘infinitive’ con­structions (cf. 12.3). ㉛ Again, a parallel transmission, from Gongyang zhuan, is quoted.



246

秋 九 月, 晉 侯 飲 趙 盾 酒, 伏 甲, 將攻之。 公羊傳接「遂刎頸而死」下云:「靈



公聞之,怒,滋欲殺之甚。衆莫可使往者,於



是伏甲于宮中,召趙盾而食之。」 其右提 彌明知之, 趙盾之車右也。趨登,曰: 「 臣 侍 君 宴, 過 三 爵, 非 禮 也。」



禮 記 玉 藻 所 謂 … 君 子 之 飲 酒 也, 受 一 爵 而 色





洒如也,二爵而言言斯,禮已三爵而油油,以



退」是也。 遂扶以下。 公嗾夫獒焉, 獒,杜注:「猛犬也。」 明搏而殺之。 搏, 鬪也。盾曰: 「棄人用犬,雖猛何為!」 鬪且出。 提彌明死之。





伏 fú, vi, ‘lie in ambush’ 攻 gōng, vn, ‘attack’ 右/佑 yòu, vn as nc, ‘assistant’ 提彌明 Tí Mímíng, np 侍 shì, vn, ‘be present’ 宴 yàn, nc, ‘feast’ 爵 jué, nc, ‘goblet’ 扶 fú, vtr, ‘support’

嗾 soˇu, vn, ‘set (a dog on s.o.)

獒 áo, nc, ‘hound’ 猛 měng, vst, ‘be wild’ 鬪/鬬 dòu, vi, ‘fight’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

20.3 t h e s to ry o f z h ao d u n

Comments ㉜ The story continues in autumn; what do you notice about the names? ㉝ The story line of the Gongyang zhuan is continued in the commentary: what do you notice about it? ㉞ The NP 其右提彌明 ­consists of two NP in apposition. They are juxta­posed and they have the same referent: ‘His assistant, Ti Miming’; cf. Box 21. For 登, con­sider the setting: the banquet hall was raised, perhaps on a mound (cf. 2.4.1–㉗). ㉟ The commentary explains Ti Miming’s pretext by reference to the Han ritual compendium Liji, which actually details how a gentleman is to behave after one, two, and three goblets of liquor. ㊱ For this construction, cf. ㉕. ㊲ The prdem 夫 (*ba or *pa, perhaps etymologically related to 彼 *paiɁ), used as mod, may simply be translated as a definite article. ㊳ A complex sentence consisting of a conditional clause, a ­concessive clause (introduced by 雖) and a main clause expressing a rhetorical ques­ tion. ㊴ The conj 且 here coordinates not two sentences (cf. 15.2–⑥) but two verbs. ㊵ For 死之, cf. 13.3.1–⑭.

Figure 8  A zun tripod Source:  William Watson, Ancient Chinese Bronzes. London: Faber and Faber, 1962 (2nd ed. 1977): fig. 10a.

初,宣子田于首山, ㊶

首山即首

247

陽山,亦即雷首山,在今山西省永濟縣

東南。舍于翳桑,杜預此注云「翳桑,

桑 之 多 蔭 翳 者 」。 然 江 永 考 實 則 以 翳 桑



當是首山間地名,王引之述聞亦云:「下



文曰『翳桑之餓人也』,翳桑當是地名。

見 靈 輒 餓, 問 其 病。 曰 :「 不 食 三 日 矣。」 食 之, 趙 盾 與 之 食。 舍 其 半。 靈 輒 餘 其 半 而 另 置 之。問 之。 曰:「宦三年矣,杜注云: 「 宦, 學 也。」未 知 母 之 存 否, 今近焉, 請以遺之。」 使盡之, ㊷

Ceremonial bronze wine vessels, like that shown in Figure 8—tripods with elongated spouts and a pointed ‘tail’ opposing it, two capped uprights, and a handle on the side—have been called jue 爵 by Song scholars. Elegantly decorated jue have come down to us from Shang and early Zhou times. However, it is doubtful whether the duke of Jin enter­tained his guests with such vessels; in classical times, jue was probably a generic word for wine ‘goblets’.

田2 /佃 tián, vi, ‘hunt’ 首山 Shoˇ ushān, np

舍2 shè, vn, ‘spend the night’ 翳桑 Yìsāng, np 靈輒 Líng Zhé, np 半 bàn, nc, ‘half ’ 宦 huàn, vi, ‘serve’

否 foˇu, adv, ‘(or) not’ 近 jìn, vst, ‘be close’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 20 : z u o z h ua n





而為之簞食與肉, 簞音單,古

代盛飯食之圓形筐,故論語 雍也

孟 子 離 婁 下 俱 言「 一 簞 食 」。寘 諸 橐 以 與 之。 既 而 與 為 公介, 杜注:「靈輒為公甲士。」 與讀去聲,參與也。倒戟以禦公 徒 而 免 之。 免 之, 免 趙 盾 於 禍。 問何故。對曰:「翳桑之餓 人也。」問其名居,不告而退, 遂自亡也。 王引之述聞云:「此

謂盾亡,非輒亡也。呂氏春秋 報更

篇謂靈輒「還鬬而死」,與傳異。



簞 dān, nc, ‘basket’ 橐 tuó, nc, ‘bag’ 與4 yù, vn, ‘join’ 介 jiè, nc, ‘armoured soldier’ 倒 dǎo, vtr ‘turn’ 戟 jı̌, nc, ‘halberd’

Comments ㊶ In the Zuozhuan, 初 often introduces flashbacks, interrupting the nar­rative in order to recount the back­ground of its events; cf. ①. The location of Shoushan is discussed in the commentary. ㊷ Note how the S changes with every sentence with­out being expres­sed. ㊸ The negation 否 probably represents the same word as 不, which is only written this way in tag questions, in which the P is deleted; 存否 thus means *存不存. For a similar utterance, cf. 10.3–⑬. ㊹ For this disposal construction, cf. 9.2.2. ㊺ 簞食 is a NP with a partitive measure as mod (cf. Box 13). ㊻ Here, 既而 ends the flashback and brings the narrative back to its original time. 乙丑,趙穿殺靈公於桃園。 各本「殺」作 「攻」,惟金澤文庫本作「煞」,煞即殺。宣子未出 山而復。 晉世家云:「盾遂奔,未出晉境。乙丑,

盾昆弟將軍趙穿襲殺靈公於桃園而迎趙盾,趙盾素貴,

得民和;靈公少,侈,民不附,故為弒易。盾復位。」

大 史 書 曰「 趙 盾 弒 其 君 」, 以 示 於 朝。 宣 子 曰 :「 不 然。」 對 曰 :「 子 為 正 卿, 亡不越竟,反不討賊,非子而誰?」宣子 曰:「嗚呼!詩曰『我之懷矣,自詒伊慼。』

各 本 無「 詩曰」 二 字, 杜 注云「逸詩也 」,則杜所 據

本有「詩曰」二字,今從金澤文庫本增。杜注以此二

句 為 逸 詩, 今 詩 邶 風 雄 雉 有 句 云,「 我 之 懷 矣, 自





248

趙穿 Zhào Chuān, np 桃 táo, nc, ‘peach’ 園 yuán, nc, ‘garden’ 示 shì, vn, ‘show’ 越 yuè, vn, ‘transgress’ 竟/境 jìng, nc, ‘border’ 討 tǎo, vn, ‘hold accountable‘ 嗚呼 wūhū, interj, ‘alas!’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi





詒伊阻」,與引詩

僅 一 字 之 異。其 我 之 謂 矣。」 孔子曰: 「 董 狐, 董 狐, 即 太 史。 古 之 良史也,書法 趙 不 隱。 宣子,古之良 大夫也,為法 受 惡。 惜 也, 越竟乃免。」

20.3 t h e s to ry o f z h ao d u n

董狐 Doˇ ng Hú, np 隱 yı̌n, vn, ‘conceal’

趙宣子 Zhào Xuānzı ̌, np 惜 xī, vi, ‘be a pity’

㊼ This is practically all the information the Chunqiu gives us about this

entire story! However, this entry seems to derive from different annals than the received Chunqiu. For the dating, cf. Focus 8. ㊽ This story has become famous for exempli­fying the virtue of Chinese historians who persist in writing the truth even in the face of danger. However, it is highly questionable whether a 大史 is a ‘grand historian’, as the title is often translated; and did he actually write the truth? ㊾ For 非子而誰, cf. 15.4.1. What follows is a typical example of the way the Odes are used in the Zuozhuan: two verses from Ode 33 are quoted (with one variant), which James Legge translates as: ‘The man of my heart!—/ He has brought on us this separation.’ Then this quotation is related to the present circumstances: ‘That applies to me!’ ㊿ This comment attributed to Confucius, just like the junzi commentary in Zuo 1.1 (cf. ㉑) only refers to the last episode, totally ignoring the background story. Its second part has been translated as ‘he did not conceal anything in his rules of writing’, and such ‘rules of writing’ have ever since been treated as a cornerstone of Chinese historio­graphy. Do you agree with the translation, however? Box 21  Appositions Besides subordination (n1→n2) and coordination (n1 + n2), noun phrases may also be constructed by apposition. In appositions, two nouns or noun phrases are juxta­posed that have the same referent: n1 = n2. This construction is typically used with titles and names, for example: 周公旦

Dan, the duke of Zhou (Li 25)

王子比干

Prince Bigan (Meng 2A1)

帝堯

Thearch Yao (Xun 5)

秦王政

Zheng, the king of Qin (SJ 5)

249

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 20 : z u o z h ua n

These phrases cannot be construed as coordinations or subordinations because their respective parts refer to the same person. The same construction is used with occupational or other designations that are not strictly titles: 墨者夷之

the Mohist Yi Zhi (Meng 3A5)

其右提彌明

His assistant, Ti Miming (Zuo 7.2)

盜跖

Zhi, the robber (Zhuang 29)

Examples for appositions that do not refer to social position or status include: 周宅酆鎬

the Zhou residences Feng and Hao (Lü 22.3)

春秋冬夏四時

the four seasons spring, summer, fall, and winter (Zhuang 18)

In rare cases, multiple apposition may occur: 魯大夫亂政者少正卯 the Lu dignitary, and political disruptor, Shaozheng Mao (SJ 47)

250

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 17

Composite texts Chinese literature of pre-­imperial times is quite different from what modern readers might expect. None of the books that have come down to us from this time are homogeneous works composed and written by a single author as a coherent whole. Rather, Classical Chinese books are composite texts. Not unlike Biblical literature, they are text collections. Some may have been put together by one editor from previously independent material. Others are thought to have grown, layer by layer, over decades or even centuries in which anonymous contributors added to or took from them, rearranged or pieced them together before they reached a stable form. Guanzi, for example, which has been called ‘an amorphous and vast repository of ancient literature’ may have been composed as shown in Figure  9. Although most of these books bear the name of illustrious authors—Meng Ke, Guan Zhong, Shang Yang, Mo Di, etc.—their real origins are obscure. Independent essays, 4th and 3d cent. BC

Historical romances 5th and 4th cent. BC Essays of Jixia scholars Proto-Guanzi c.250 BC

Independent essays, late 3d cent. BC

Materials from the court of Liu An; independent essays, 2d cent. BC

Explanatory essays; other essays 1st cent. BC Liu Xiang Guanzi c.26 BC

Figure 9  Composite texts: Guanzi Adapted from Rickett 1985–98, vol. 1, 37.

251

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 17 : co m p o s i t e t e x t s

The traces of this anonymous process of re-­working and accretion that lies behind ancient Chinese books are still recognizable in many places. Classical texts exhibit repetitions, gaps, sharp breaks, changes in vocabulary, style, or perspective, inconsistencies, and outright contra­ dic­tions that testify to their heterogeneous nature. The Lunyu repeats the master’s words verbatim in different places; the Mozi consists of radically different parts; and in the Zuozhuan it is recounted that during the battle between Chin and Ch’u at Yen-­ling, Lü Hsiang, a Chin dignitary, shoots the Ch’u king, hitting him in the eye. In the very next incident, we are told that the Ch’u king noticed the polite and chivalrous behavior of another Chin dignitary (who descended from his chariot and took off his helmet whenever he crossed the Ch’u King’s path) and sent a present to him. (Egan 1977, 349)

Similar inconsistencies and redundancies may be found in the Zuozhuan excerpt discussed in 20.3. The protagonist is sometimes called Zhao Dun, sometimes Xuanzi; duke Ling of Jin is called marquis of Jin at one point; the note that he was ‘angry with’ (患) Zhao Dun is repeated, and he tries to murder him twice; moreover, most strikingly, Zhao Dun or Xuanzi is saved twice by two different people, Ti Miming and Ling Zhe! One notices, also, that some paragraphs are dated, others are not, that the flashback introduced by 初 takes the narrative back to a point long before the primary sequence of events, that different people (Chu Ni, Ti Miming, Ling Zhe, the scribe) are introduced in different parts without being mentioned at all in others, and that the judgment by Confucius relates only to the final part of the story, taking no account of the earlier parts. All these observations suggest that the story of Zhao Dun is not a homogeneous narrative but a pastiche of several independent units: at least two—perhaps three—independent anecdotes about the attempts to murder Zhao Dun, followed by an annalistic entry and yet another short narrative at the end of which Confucius somewhat surprisingly passes judgment on Zhao Dun who had before been portrayed as a paragon of virtue. Besides the aforementioned criteria—differences in names and wording, changing time frames, redundancies, and contradictions—the evidence of parallel transmissions may help to identify independent textual units. As the commentary indicates, the text has parallels in Guoyu, Lüshi chunqiu, and Gongyang zhuan. Interestingly, the Guoyu 252

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 17 : co m p o s i t e t e x t s

only records the first episode, including Chu Ni, but none of the others. The Lüshi chunqiu, in turn, contains the Chu Ni episode and the last part, including the starving man—but it records them independently (Lü 23.4 and 15.4, respectively). Another Han text, the Shuoyuan, also records the anecdote about the starving man but no other part. None of these texts contain the Ti Miming episode or the judgment by Confucius. These parallels, then, lend further support to the assumption that all of these parts are independent units pieced together in the Zuozhuan. Similar observations could be made in most other works of early Chinese literature. Given their composite nature, it becomes clear that close reading and syntactical analysis are not enough to fully understand these texts. Rather, it is essential to recognize and analyse their distinct textual units or layers. Only the critical examination of their  intricate structure will enable us to appreciate these texts in all their historical depth and wealth of meaning. References: Egan 1977; Rickett 1985–98; Brooks and Brooks 1998; Stumpfeldt 2002; Boltz 2005; Vogelsang 2011.

253

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 21

Guoyu

21.1 Introduction254 21.2 The voice of the people (GY 1.3) 255

21.3 A femme fatale (GY 7.3, 7.6)

258

21.1 Introduction Guoyu 國 語 , Discourses of the States, is an historical work much like the Zuozhuan and yet quite unlike it. Like the latter, it has been attributed to Zuoqiu Ming; both texts are linguistically similar, both place heavy emphasis on the state of Jin, and they share more than seventy anecdotes (cf. Focus 24). But unlike Zuozhuan, the Guoyu consists of 240 anecdotal pieces with no explicit connection between them. Its twentyone juan are arranged not in chronological order but according to eight states. Unsurprisingly, stories from the northern Chinese states Zhou (three juan), Lu (two juan), Qi (one juan), Jin (nine juan), and Zheng (one juan) make up the bulk of the work, but the southern states Chu (two juan), Wu (one juan), and Yue (two juan) also receive considerable attention. While most of the Guoyu anecdotes are set in Chunqiu times, the first pieces from the Zhou chapters are set in the tenth and ninth centuries bc. The texts themselves, however, are certainly not as old. Guoyu was likely produced in a long process involving several different compilers, perhaps starting as early as the late fifth century bc and ending in the late fourth or third century bc. Just like the Zuozhuan, it cannot be treated as a source of Chunqiu (much less Western Zhou) history. Rather, it should be regarded as a piece of historical literature that ‘embraces heaven and earth, profoundly gauges fortune and disaster, reveals the obscure and subtle, and clarifies the good and the evil’, as the 254

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

21.2 t h e v o i c e o f t h e p e o p l e

Guoyu’s first commentator, Wei Zhao 韋 昭 (204–73 ad), put it. In other words, it is didactic history, concerned less with historical facts ‘as they actually were’ than with their ethical significance. The stories it tells about the last Zhou rulers, about the femme fatale Li Ji, about the ­odyssey of Chong Er, who was to become Duke Wen of Jin, and the epic  battles between Wu and Yue have become an integral part of Chinese historical lore. Yet, they should not simply be taken for ­historical truth. The following selections are taken from the Guoyu jijie 國 語 集 解 . Editions: Guoyu jijie 國 語 集 解 , ed. by Xu Yuangao 徐 元 誥 . Beijing 2002; Guoyu 國 語 . 2 vols. Shanghai 1988. Translations: There are only partial translations into Western languages. de Harlez 1893/94 is a translation of the ‘Zhou yu’, as is d’Hormon/Mathieu 1985; Imber 1975 contains translations of the ‘Jin yu’; Meisterernst 2002 is a translation of the ‘Yue yu’; Wang Hong et al. 2012 has translations of selected pieces; a full translation of the Guoyu is currently being ­prepared by Eric Henry (Chapel Hill, NC). Studies have mainly focussed on Guoyu’s relation to the Zuozhuan: Zhang Yiren 1962 is very thorough; Boltz 1990 is an ingenious ‘isocolometrical’ analysis of parallels in the two texts. See also the dissertations by Hart 1973 and Sargent 1975.

21.2  The voice of the people (GY 1.3) The following excerpt, taken from the ‘Zhou yu’ chapters, deals with the decline of the Western Zhou. It recounts how King Li (r. 857–842 bc) ruined his authority by ignoring the remonstrances of the people: in a typically ‘Confucian’ style, it implies that heeding the voice of the ­people is essential to good and successful government.

255

厲 王 虐, 國 人 謗 王 。厲王,恭



王之曾孫,夷王之子厲王 胡也。謗,

誹 也。 邵 公 告 王 曰 :「 民 不 堪 命 矣!」 邵 公, 為 王 卿 士, 言 民 不 堪 暴 虐 之 政 令。 王 怒, 得 衛 巫, 使 監 謗 者 ,衛巫,衛



國之巫也。監,察也。以巫有神靈,

有謗必知之。 以告,則殺之。 國 人 莫 敢 言, 道 路 以 目 。 不 敢發言,以目相眄而已。 王喜, 告 邵 公 曰 :「 吾 能 弭 謗 矣, 乃 不 敢 言。」 邵 公 曰 :「 是





厲 Lì, np, posth. name 虐 nüè, vi, ‘be cruel’ 謗 bàng, vn, ‘complain about’ 邵 Shào, np, a place 巫 wū, nc, ‘shaman’ 監 jiān, vn, ‘oversee’ 路 lù, nc, ‘road’

弭 mı,̌ vi, ‘stop’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 21 : g u oy u

障之也。防民之口,甚於防川。川壅而潰,傷人必多。民亦如之。 是 故 為 川 者 决 之 使 導, 為 民 者 宣 之 使 言。 故 天 子 聽 政, 使 公卿至於列士獻詩, 瞽獻曲 ,無目曰瞽。瞽,樂師。曲,樂曲也。 元 誥按:今本「曲」作「典」,蓋因二字形似而誤。史獻書 而 後王斟酌  . . . 焉, 是 以 事 行 而 不 悖。 悖, 逆 也。 民 之 有 口 也, 猶 土 之 有 山 川 也,財用於是乎出 。猶,若也。 元誥按:上「也」字依宋庠本。 猶其 有 原 隰 衍 沃 也, 衣 食 於 是 乎 生 。廣平曰原,下溼曰隰。下平曰衍,有溉 曰沃。 口之宣言也,善敗於是乎興。行善而備敗,其所以阜財 用衣食者也 。 夫民慮之於心而宣之於口,成而行之,胡可壅也! 若壅其口,其與能幾何?」「邵氏云:『此倒語也,若曰其有幾人能靖 者與?』」此句法相同。「與」乃詠歎之辭也。 王不聽。於是國莫敢出言, 三年乃流王於彘 。流,放也。彘,晉地,漢為縣,屬河東,今曰永安。 ⑬

























障 zhàng, vn, ‘obstruct’ 防 fáng, vtr, ‘block’ 口 kǒu, nc, ‘mouth’ 壅 yōng, vtr, ‘block’ 潰 kuì, vi, ‘overflow’ 導 dǎo, vn, ‘steer’ 宣 xuān, vi, ‘proclaim’ 曲2 qǔ, nc, ‘tune’ 斟酌 zhēnzhuó, vn, ‘ponder’ 悖 bèi, vi, ‘be adverse’ 財 cái, nc, ‘resources’ 原 yuán, nc, ‘plain’ 隰 xí, nc, ‘marshland’ 衍 yǎn, nc, ‘lowlands’ 沃 wò, nc, ‘grasslands’ 備 bèi, nc as vn, ‘take precautions (against)’ 阜 fù, vi, ‘be massive’ 慮 lǜ, vn, ‘think about’ 胡 hú, adv, ‘how, why’ 彘 Zhì, np, a place

Comments ① Note that the king’s posthumous name, Li, means ‘oppressive’. 國 人 does not mean the people in general, but especially the inhabitants of the capital, as opposed to the 野 人 , the rustics that lived in the backcountry: the latter were much less respected, their opinions obviously did not matter. ② The duke of Shao, head of a prominent lineage that goes back to king Wen, hence a relative of the Zhou king, was a high minister (卿 ) at court. ③ King Li commits a grave transgression: he hires a shaman (supposedly 巫 meant a female shaman, which may have made things worse) in order to circumvent his regular advisers. This was sure to incur the wrath of his minister and later historians. ④ 道 路 以 目 is either an incomplete clause (as the commentary suggests), or 以 must be construed as P.  ⑤ The aquatic metaphor is ­reminiscent of Mengzi (cf. 18.4); note that 障 and 防 both originally

256

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

21.2 t h e v o i c e o f t h e p e o p l e

meant ‘to dike’, relating to water. ⑥ Two further examples for the diverse uses of 為 (cf. Box 14); perhaps ‘control’ best catches the meaning in these cases (cf. the parallel in Lü 20.5: 治 川 者 . . . 治 民 者 ). Note also the somewhat odd usage of 宣 . Incidentally, the posthumous name of King Li’s successor, who reigned long and successfully, was none other than Xuan 宣 . ⑦ 聽 政 means ‘attend to government’. The phrase 公 卿 至 於 列 士 is remarkable: it must be construed as a NP that functions as O to 使 . It follows that 至 於 (cf. 8.3–㉞), like a conj, serves to connect two NP: ‘[everyone from] the dukes and ministers down to the ranked servicemen’. The ‘presenting of songs’, which allegedly reflected the feelings of the people, became a commonplace in Chinese political discourse, symbolizing ‘participative’ government. ⑧ In the idealized Zhou order, the blind were supposedly in charge of music, and the 史 in charge of documents; this retrospective description, however, is hardly borne out by contemporary sources. The following lengthy list of diverse groups of people—officers, artisans, commoners, etc.—that voice opinions before the ruler is omitted here. ⑨ 斟 酌 is an example of an alliterating two-syllable word (cf. Focus 9). ⑩ Again, the rhetoric of the first clause is similar to Meng 6A2. PrP with 於 may be concluded by 乎 , as in 於 是 乎 (cf. Box 23). Such emphatic PrP are commonly placed in medial position, before the P, or in initial position. In the latter case, they always express temporal relations. ⑪ The complex sentence runs all the way from 民 之 有 口 也 to 生 . Can you determine the reference of 其 ? ⑫ Here is a beautiful 所 phrase; for 所 以 , cf. 12.2.2–㉓. Again, 其 is tricky: what does it represent here? ⑬ 胡 is cognate to 何 , but restricted to adverbial usage. Should it be mere coincidence that King Li’s given name was also胡 (cf. the commentary in the first line as well as ⑥)? ⑭ Despite the suggestion in the commentary, 與 does not seem to make sense here; for 幾 何 , cf. 10.3–⑮. ⑮ The exile of king Li in 842 bc was the beginning of the end of Zhou rule. For the next fourteen years, tradition has it, there was an interregnum led by feudal lords, before the Zhou experienced a restoration under King Xuan. 842 bc is the earliest reliable date in Chinese history, since the chronology of the feudal lords in the Shiji begins with this year; in fact, it may well be considered the beginning of Chinese history: the assumption of power by the feudal lords was the first step in a process that led to the unification of China in 221 bc.

257

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 21 : g u oy u

21.3 A femme fatale (GY 7.3, 7.6) The following episodes are taken from the ‘Jin yu’ chapters. They recount the infamous story of Li Ji 驪 姬 who, through her intrigues, wreaked havoc in the state of Jin. Traditional Chinese historians have taken this story as an example of the malign influence of women in politics, but the background reveals a sordid reality that often lay behind such ­disruptions. 獻公伐驪戎,克之,滅驪子, 驪子,驪戎之君也。 本爵男,此云子者,猶言男子也。獲驪姬以歸,立以 為夫人,生奚齊。其娣生卓子 。吳曾祺曰:「古者, 人君無再娶之義,嫁女多以女弟為徒,故謂之娣。」 驪 姬請使申生主曲沃以速縣,申生,獻公太子恭君也。

獻公娶於賈, 無子。蒸於齊姜,生申生。淮南 主術訓:「其

重耳處蒲城,

於御兵刃縣矣。」高注云:「縣,遠也。」疑速縣當順「速

遠」,謂使申生主曲沃,以促之遠也。

元 夷 吾 處 屈, 蒲, 今 蒲 坂 ; 屈, 北 屈, 皆 在 河 東。 ○ 誥 按 : 今 山 西 隰 縣 北 四 十 五 里 有 蒲 子 故 城, 又 吉 縣 東 北

二 十 一 里 有 北 屈 廢 縣。奚 齊 處 絳, ○ 元誥按:絳當今 山西 絳縣北。以儆無辱之故。公許之。







驪戎 Lí Róng, np, a people 滅 miè, vtr, ‘destroy’ 獲 huò, vn, ‘capture’ 驪姬 Lí Jī, np 娣 dì, nc, ‘younger sister’

卓子 Zhuōzı,̌ np

申生 Shēnshēng, np 主2 zhǔ, vn, ‘control’ 曲沃 Qūwò, np, a place 速 sù, vst, ‘be fast’ 重耳 Chóngěr, np 蒲城 Púchéng, np, a place 夷吾 Yíwú, np 屈 Qū, np, a place 絳 Jiàng, np, a place 儆/警 jı̌ng, vi, ‘be alert’

Comments ⑯ Duke Xian of Jin (r. 677–651 bc) is said to have attacked the neighbouring Rong people (cf. Map 2) in 672 bc. This was no chivalrous matter: he ‘extinguished’ the men—and that surely means: the entire clan—of Li and abducted at least two women, Li Ji and her younger sister. In other words, he committed mass murder and rape. Just like in ancient European history (consider the story of Helena, the rape of the Sabine women, or the myth of Europa), abduction of women was a recurring element in ancient Chinese warfare. ⑰ Li Ji now begins to scheme to remove Duke Xian’s sons and make her own son his successor. Note that Duke Xian’s other sons do not all have an impeccable pedigree, either: Shensheng, so the commentary informs us, was born 258

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

21.3 a f e m m e fata l e

through illicit intercourse (蒸 ) with Lady Jiang of Qi. 速 縣 is problematic: the commentary suggests to read 遠 for 縣 —is that convincing? ⑱ 夷 吾 was to rule Jin as Duke Hui from 650–637 bc, and 重 耳 —whose name probably means ‘swollen (腫 ) ear’—was to become Duke Wen of Jin (r. 636–628 bc) and the hegemon of his time. For the location of 曲 沃 , 蒲 城 , 屈 , and 絳 , cf. Map 2. 公之優曰施,通於驪姬。 優, 俳也。 施, 其名也。 旁淫曰通。 驪姬問焉,曰:「吾欲作大 王引之曰:「 『 之徒』 二字衍文 事,而難三公子之徒,如何?」 三公子, 申生、 重耳、 夷吾也。 ○ 也。 」 對曰:「早處之,使知其極 。當早定申生,分之都城而位以卿,使自知其位所極至也。夫人 知極,鮮有慢心,俞樾曰:「韋注非也。鮮當讀為斯,此言人知其位已極,斯有怠慢之心也。鮮與斯古音相近, 說文:『𩆵,徒雨鮮聲,讀若斯。』雖其慢,乃易殘也。 」 ○ 元誥按:雖當讀為唯,雖、唯古字通用,因同 徒隹聲也。 驪姬曰:「吾欲為難,安始而可?」 ○ 元誥按:難,去聲。 優施曰:「必於申 生。其為人也,小心精潔,而大志重,又不忍人。 不忍施惡於人。 精潔易辱,重僨可疾, 是故先施讒於申生。 驪姬賂二五, 不忍人,必自忍也。 自忍, 忍能自殺也。辱之近行。」  . . .  元誥按:內傳作「 賂外嬖梁五與東關嬖五」 。曰:「夫 使言於公, 二五, 獻公嬖大夫梁五與東關五也。 ○ 曲沃,君之宗也;蒲與二屈,君之疆也, 疆, 境也。 二屈, 屈有南北也。 今河東有北屈, 則是時

復有南屈也。 蒲,今平陽 蒲子縣。不可以無主。宗邑無主,則民不威;疆埸無主,則啟戎心 。啟,

開也,開戎侵盜之心也。晉北有山戎,二屈接之。戎之生心,民慢其政,國之患也。若使大子主曲













259

優 yōu, nc, ‘entertainer’ 通2 tōng, vn, ‘have an affair (with)’ 極 jí, nc, ‘limit’ 鮮 xiǎn, vi, ‘be rare’ 慢 màn, vn, ‘disrespect’ 難2 nàn, nc, ‘trouble’ 安2 ān, adv, ‘how?’

始 shı ,̌ ntemp, as vi, ‘begin’ 精 jīng, vst, ‘be pure’

潔 jié, vi, ‘be clean’ 僨 fèn, vi as nc, ‘downfall’ 讒 chán, nc, ‘slander’ 賂 lù, vn, ‘bribe’ 宗 zōng, nc, ‘ancestral home’ 疆 jiāng, nc, ‘border’ 大子/太子 tàizı̌, nc, ‘crown prince’ 威 wēi, nc as vst, ‘be awesome’ 埸 yì, nc, ‘border’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 21 : g u oy u

沃, 而 二 公 子 主 蒲 與 屈, 乃可以威民而懼戎,且旌 君伐。」 旌,章也。伐,功也。 使俱曰:「狄之廣莫, 於 晉 為 都, 廣 莫, 北 狄 沙 漠 也。 下邑曰都, 使如為晉下邑。 晉 之 啟 土, 不 亦 宜 乎?」 公說,乃城曲沃,大子處 焉;又城蒲,重耳處焉; 又城二屈,公子夷吾處焉。 驪姬既遠大子,乃生之言, 大子由是得罪。 ㉕



懼2 jù, vtr, ‘intimidate’ 旌 jīng, vn, ‘honour’ 俱 jù, adv, ‘both’ 莫/漠 mò, nc, ‘expanse’ 城 chéng, nc as vn, ‘fortify’

Comments ⑲ 通 , ‘to have an affair’, is another frequently encountered term in Chunqiu history: in Zuozhuan and Guoyu anecdotes, such illicit affairs often provide the background for intrigues. ⑳ The commentary suggests that the characters 之 徒 are redundant, but this is not necessarily so. X 之徒 can mean ‘people like X’, ‘the likes of X’; compare enumerations like . . . 范 睢 、魏 牟 、田 文 、莊 周 、慎 到 、田 駢 、墨 翟 、 宋 鉼 、鄧 析 、惠 施 之 徒 也 (HSWZ 4). 如 何 , ‘how about it?’, is probably derived from 如 之 何 (cf. 18.3–⑮), with deletion of 之 . ㉑ The commentaries on the reading of 鮮 and 雖 make quite a difference in the understanding of this passage—can you follow their arguments? ㉒ Note that 安 here represents an adv, perhaps a variant of 焉 (cf. 8.2.5); it has the same meaning as the latter. ㉓ In 必 於 申 生 , the implicit P is 始 from the preceding question. The expression 其 為 人 也 often functions as absolute topic, introducing the description of personal traits; it may simply be translated as ‘He is a man who. . .’ ㉔ After a sequence of short conditional and consecutive clauses, 辱 之 近 行 should be construed as a hortatory utterance; the construction seems to be pseudoditransitive (cf. Box 11). ㉕ The plot is beginning to unfold: The duke is being talked into removing his sons from the capital in order to guard against the Rong—precisely the people that Duke Xian had attacked before (cf. ⑯). ㉖ 生 之 言 seems to be an exceptional case of ditransitive usage of a vi: Li Ji ‘fabricated ⟨slanderous⟩ talk against him’, thus throwing the state of Jin into turmoil.

260

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

21.3 a f e m m e fata l e

Map 2  Jin in the Chunqiu period Source: Su Chenshen 蘇振申 (ed.). Zhongguo lishi tushuo 中國歷史圖說 [A Pictorial General History of China]. vol. 4. Taipei: Shin Shin Publishing Co., 1978, 30.

261

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 18

Transmission of classical literature No Classical Chinese text has been transmitted in its original form. In fact, most texts of the classical period have not been transmitted at all. Tradition has it that in 213 bc the first emperor of Qin ordered a great part of ancient literature—historical records of states other than Qin, the Odes, the Documents, and the writings of the ‘Hundred Schools’—to be burnt (SJ 6). Even more disastrous, perhaps, was the looting of the imperial palaces of Qin in 206 bc, which destroyed the literary collection of the court. The subsequent history of the Chinese empire has seen many more such catastrophes: libraries plundered, burnt, or sunk in rivers. But these are just the spectacular cases. The vast majority of ancient texts probably vanished quietly and imperceptibly simply by not being copied. In an age of manuscripts written on perishable ­material—bamboo, wood, silk, and in post-­classical times paper—texts that were not regularly reproduced were doomed to dematerialize. The earliest Chinese bibliography we possess, the Hanshu ‘Treatise on Arts and Literature’ (‘Yiwenzhi’ 藝文志, HS 30), compiled around ad 90, gives a vivid impression of these losses. Based on a catalogue of the Han imperial collection, it lists a total of 653 works—but only about six per cent of these have come down to us in whole, and another six per cent in fragments. Nor is the quota any better for works not included in the Hanshu bibliography. Numerous, heretofore unknown texts that have come to light in recent archaeological discoveries (cf. Focus 19) have given us an impression of the amount and variety of texts that existed in ancient China. Many fewer texts would have come down to us were it not for the efforts of two remarkable personalities at the end of the Western Han period, Liu Xiang 劉向 (79–78 bc) and his son Liu Xin 劉歆 (46 bc–ad 23). Librarians at the Han court, they produced the first bibliographies of the imperial collection: the Bielu 別錄 and its abridgment, the Qilüe 七 略, which was the basis for the Hanshu ‘Yiwenzhi’. But the Lius did more than just compile bibliographies, they compiled many of the very books that are included therein. Examining the 13,000 juan in the imperial collection, they discarded duplicates, selected superior versions, and appar262

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 18 : t r a n s m i s s i o n o f c l a s s i c a l l i t e r at u r e

ently edited numerous texts for the first time (cf. 30.1). About a dozen of Liu Xiang’s editorial reports that have come down to us testify to his textual work. Collections like Zhanguo ce, Shuoyuan, or Lienü zhuan were first put together by the Lius, other texts, like Xunzi or Guanzi, were given their definite form by the two. It is no exaggeration to say that we owe a great part of classical literature to Liu Xiang and Liu Xin. However, this does not mean that this literature had reached its final form with the Lius. Rather, the precarious process of manuscript transmission continued for another millennium or so. Throughout the Chinese middle ages, classical texts were copied and re-­copied by hand, introducing further changes at every stage. This process of continuous alteration only gradually gave way to a more stable transmission in Tang and Song times. In 837, the Canonical Texts were engraved in stone ­steles—they have survived until today—and in the tenth century, they were printed for the first time. The bulk of classical literature, however, was only printed in subsequent centuries, and the earliest editions we possess date from the twelfth to sixteenth centuries; for example: Shiji c.1131–62 Zhuangzi c.1163–90 Han Feizi c.1165–73 Xunzi 1181 Lüshi chunqiu: c.1341–68 Mozi 1445 Shangjun shu: c.1522–67 Of course, these editions are rarissima that are hard to access. Therefore, scholars usually rely on modern editions that have the advantage of being readily available—and the disadvantage of having undergone yet more cycles of copying, editing, and, inevitably, alteration. It is well to bear in mind Derek Herforth’s admonition: All our texts are vulgates. Precisely for that reason, any conscientious grammarian of the language must realize that his analyses are based, faute de mieux, on this body of transmitted textual material, the victim for well over two millennia of contamination and corruption, unwitting and otherwise, at the hands of copyists, editors and, more recently, type-­setters and proofreaders. (Herforth 1985–87, 211) References: For the Hanshu ‘Yiwenzhi’, cf. Fölster 2016; for editions of classical texts, cf. Loewe 1993; for exemplary studies in textual criticism and transmission, cf. Simson 2006 and Roth 1992.

263

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 22

Laozi

22.1 Introduction264 22.2 The ineffable Dao (Lao 1) 265 22.3 Coincidentia oppositorum (Lao 2) 267

22.4 The sage’s order (Lao 3) 22.5 The ideal state (Lao 80)

269 270

22.1 Introduction The Laozi 老 子 , or Daodejing 道 德 經 , is perhaps the most mystified work of Classical Chinese literature. No other text has exuded more fascination for Western readers, and no other work has been translated nearly as often. There are more than 300 trans­lations into Western languages, which often display astonishing differences in their understanding of the text: they show how poorly the Laozi is understood. Its alleged author, variously called Li Er 李 耳, Lao Dan 老 聃 , or simply Laozi, supposedly was an elder contemporary of Confucius from the southern state of Chu. However, the lore surrounding his person is clearly fictitious, and the text that bears his name almost certainly does not predate the third century bc. The Laozi is a short text of about 5,000 characters. It consists of eighty-one short paragraphs, which are subsumed under two sections, the ‘Dao jing’ 道 經 and the ‘De jing’ 德 經 . They take the form of ­didactic sayings, often in rhyme, which revolve around concepts such as 道 , literally ‘the way’, metaphorically meaning the universal principle, 德 , the power that emanates from 道 , and 無 為 , ‘non-action’ which nonetheless accomplishes every­thing. These sayings contain no con­ text­ual information and mostly only a minimum of determinative elem­ents. This lack of specification gives the Laozi an air of ambiguity which makes it so attractive for modern readers: it has been read as a 264

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

22.2 THE INEFFABLE DAO

storehouse of timeless wisdom, a handbook for managers, a sex manual, a military treatise, a meditative text, or whatever its interpreters have looked for. Apparently, the people of Zhanguo and Han times also regarded the Laozi as more than a merely edifying book, since several manuscript copies have been found in noble tombs of the third and second centuries bc. Along with other philosophical, medical, and mantic texts, they were probably meant to assist the dead in the netherworld (cf. Focus 19). It is no mere coincidence that the Laozi became a founding text not only of philosophical Daoism but also of Daoist religion. Unsurprisingly, the Laozi has received numerous commentaries, the most important ones being that by an obscure Heshang gong 河 上 公 , which dates from some time between ad 300 and 500, and that by Wang Bi 王 弼 (226–49). The following excerpts are taken from the Xinyi Laozi duben 新 譯 老 子 讀 本 , a Taiwanese edition that provides pronunciation aids in bopomofo ㄅㄆㄇㄈ, the phonetic transcription used in Taiwan, a modern commentary, and a MSC translation (not reproduced here). Editions: Laozi jiaoshi 老 子 校 釋 , ed. by Zhu Qianzhi 朱 謙 之 . Beijing 1984; Gao Ming 高 明 (ed.). Boshu Laozi jiaozhu 帛 書 老 子 校 注 . Beijing 1996 (an edition of the Laozi silk MSS); Yu Peilin余 培 林 . Xinyi Laozi duben 新 譯 老 子 讀 本 . Taipei 2002. Translations: No attempt is made here to select among the countless available translations. Studies: Boltz 1984; Robinet 1991; Kohn and LaFargue 1998; Allan and Williams 2000; Wagner 2000; Shaugh­nessy 2005; Liu Xiaogan 2015, 25–126.

22.2  The ineffable Dao (Lao 1) The first paragraph introduces the main concept of the ‘Dao jing’, the Dao—‘Way, Principle’, or the like—in the kind of elliptical, enigmatic language that is characteristic for Laozi. Additionally, it presents a second key concept of early Chinese philosophy, that of ming, ‘names’. ㄉ ㄠˋ

265

常 cháng, vi, ‘be constant’ 妙 miào, vi, ‘be pro­foun­d’

ㄇ ㄥ

ㄉ ㄠˋ

ㄎ  ˇ ㄜ

徼 jiào, nc, ‘margin’

ㄎ  ˇ ㄜ

玄 xuán, vst, ‘be ob­scure’

ㄇ ㄥ





道2 dào, vn, ‘express’



ㄇ ㄥ ㄊ ㄢ









ㄨ ㄢ

ㄔ ㄤ

② ①





ㄇ ㄥ





ㄇ ㄥ



ㄔ ㄤ

ㄡ ˇ

ㄔ ㄤ

第一章  道可道,非 常道 ;名可名,非

ㄍ ㄨ

ㄕ ˇ







ㄇ ㄨ ˇ



常名 。無、名天地

ˇ ㄡ



之 始 ; 有、 名 萬 物

ㄔ ㄤ



ㄓ ˇ ㄜ

ㄇ ㄠ





ㄌ ˇ ㄤ

ㄍ ㄨ ㄢ ㄑ



ˇ ㄘ





ㄐ ㄠ

ˇ

之母 。故常無,欲





ㄨ ㄟ

ㄍ ㄨ ㄢ



ㄊ ㄨ ㄥ





ㄒ ㄩ ㄢ

ㄇ ㄥ







ˇ

以觀其妙;常有,欲



ㄔ ㄨ



ㄒ ㄩ ㄢ

ㄊ ㄨ ㄥ

以觀其徼 。此兩者,

ㄒ ㄩ ㄢ





ㄇ ㄣ



同 出 而 異 名, 同 謂





之 玄。 玄 之 又 玄 ,

眾妙之門 。 ㄓ ㄨ ㄥ ㄇ ㄠ

ㄉ ㄠˋ

ㄈ ㄟ

ㄈ ㄟ

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi





道可道非常道 第一個「道」字是名詞,指宇宙的本源,亦即創生天地萬物

解 老 》 曰 :「 道 者 萬 物 之 所 然 也。」 又 曰 :「 道 者 萬 的 總 原 理 或 原 動 力。《 韓 非 子 • 物之所以成也。」皆謂道是天地萬物的本源。第二個「道」字是動詞,講說的意思。

無、有,都是指的道。道, 「視



名可名非常名 第一個「名」字是名詞,指道的真相。第二個「名」字是動詞,稱謂的 意思。蓋老子之道,無狀無象,是形而上的,故不可名。只是為了解說方便起見,不

能沒有一個稱呼,所以勉強用一個「道」字來稱呼它。

之不見名曰夷,聽之不聞名曰希,搏之不得名曰微」 十 ( 四章 ,)並非具體的事物,所 以可稱為「無」。析言之, 「無」指「道之體」, 「有」指「道之用」。「體」必先於「用」,

妙,精微莫



所以「無」亦必先於「有」,四十章云:「天下萬物生於有,有生於無。」又自王弼以來,

古人多以「無名」、「有名」為讀,直至王安石才以「無」、「有」為讀。

測的意思。徼,陸德明曰:「邊也。」引申有「廣大無際」的意思。「妙」形容道之體

眾妙之門 指道。謂道為一切妙理與變化的門戶。



無 ) 「徼」形容道之用 有 ( 。)又這兩句古人多以「無欲」、「有欲」為句,而讀成「故 ( ; 常無欲,以觀其妙;常有欲,以觀其徼。」這樣的斷法,無論在文字上或意義上都說不通。

玄之又玄 是說幽微深遠到極點。



︻注釋︼

l e s s o n 22 : l ao z i

Comments ① This is one of the great first sentences in world literature. Its syntax is perplexing: 可 道 , evidently a modification to the first 道 , is not placed before but after the head; *可 道 之 道 (which, in fact, Wang Bi’s commentary says) would be the regular construction. Alternatively, the first three characters could be construed as a conditional clause, and the next three as the main clause, with deletion of the S.  For an impression of the variations in translation, compare the ­following, all by expert Sinologists: The Dao that can be trodden is not the enduring and unchanging Dao. (James Legge) La Voie vraiment Voie est autre qu’une voie constante. (J.J.L. Duyvendak) The Tao that may be called Tao is not the invariable Tao. (Derk Bodde) The way can be spoken of, / But it would not be the constant way. (D.C. Lau)

266

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

22.3 COINCIDENTIA OPPOSITORUM

The Way that can be “Way”-ed / Is not the constant Way. (A.C. Graham) Könnten wir weisen den Weg, es wäre kein ewiger Weg. (Günther Debon) To guide what can be guided is not constant guiding. (Chad Hansen)

② The same construction as before. The Dao’s name is also discussed in

Lao 32:  道 無 常 名 ,and Lao 25:  吾 不 知 其 名 ,字 之 曰 道 ,強 為 之 名 曰 大 。Observe how the text plays with the different name forms: the given name (名 ) of the Dao, which has been there for eternity, cannot be known; instead it first receives a courtesy name (字 ), a given name being reluctantly added afterwards. Combined, they read 大 道 , which actually appears several times in Laozi. ③ Observe how the punctuation of this sentence determines its interpretation; this is also discussed in the commentary. ④ Again, the placement of the comma is decisive for the analysis of the sentence; read the commentary and consider which punctuation makes more sense. ⑤ 玄 之 又 玄 is a partitive genitive, denoting a subset of a larger group, in this case a superlative: ‘the obscurest of obscure’. ⑥ 眾 here serves as mod indicating the plural number. 22.3  Coincidentia oppositorum (Lao 2) Whereas the first paragraph introduced some key concepts, the second displays the paradoxical reasoning that is characteristic of Daoist thought. The two sides of a distinc­tion inevitably refer to one another, in fact, they are one: coincidentia oppositorum.





ㄒ ㄤ

ㄅ ㄨ

ㄔ ㄥ

ㄔ ㄤ

ㄉ ㄨ ㄢ ˇ

ㄕ ㄢ





267

ˇ

ㄧ ㄧ ㄧ

ㄇ ㄟ ˇ

ㄨ ㄟ

ㄇ ㄟ ˇ

ㄙ ㄜ

ˇ









ㄕ ㄢ

ㄋ ㄢ



ㄏ ㄜ

ㄨ ㄟ



ㄧ ㄒ ㄤ















ㄧ ㄕ ㄥ

ㄕ ㄥ

ㄒ ㄚ ㄐ ㄝ





ㄕ ㄢ

ㄒ ㄤ

ㄊ ㄢ



ㄒ ㄤ ㄑ ㄥ



天 下 皆 知 美 之 為 美, 斯 惡 已。

ㄒ ㄚ

ㄡ ˇ



ㄍ ㄠ

ㄐ ㄝ

ㄍ ㄨ





皆 知 善 之 為 善, 斯 不 善 已

ㄒ ㄥ

,高下相傾,音聲相和,

ㄒ ㄤ ㄙ ㄨ ㄟ

ㄒ ㄤ

故 有 無 相 生, 難 易 相 成, 長 短

ㄏ ㄡ

相形

前後相隨。

既 有「 美 」 之 名,「 惡 」 醜 ( ) 之名則相對而生;既有「善」之名,「不善」

︻注釋︼

惡 ( 之 ) 名則相伴而出。美醜善惡之名既 有,天下必將求美而去醜,趨善而避惡,

微明》引此文並作「矣」。 •

由是紛爭迭起而詐偽滋生。「已」猶「矣」,

《文子



ㄑ ㄢ

有4 yǒu, nc,  ‘being’ 無4 wú, nc,  ‘non-being’ 形 xíng, nc, as vi,  ‘form’ 傾 qīng, vi,  ‘lean’ 音 yīn, nc,  ‘sound’ 和2 hè, vn,  ‘harmo­nize’ 隨2 suí, vn,  ‘follow’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

ㄕ ㄥ

是以



ˇ

ㄅ ㄨ ㄕ

ㄍ ㄨ ㄥ ㄔ ㄥ ㄦ ㄈ ㄨ ㄐ ㄩ

ㄈ ㄨ

ㄈ ㄨ ㄐ ㄩ





ˇ

ㄅ ㄨ ㄑ ㄩ



ㄔ ˇ ㄨ ㄨ



ㄨ ㄟ

ㄓ ㄕ

ㄒ ㄥ ㄅ ㄨ ㄢ ㄓ ㄐ ㄠ

ㄨ ㄢ ㄨ ㄗ ㄨ ㄛ ㄢ ㄦ ㄅ ㄨ







弗居,是以不去。



ㄨ ㄟ

ㄖ ㄣ





ㄕ ㄥ



ㄨ ㄟ





ˇ ㄡ

恃 shì, vn,  ‘rely on’





ㄅ ㄨ



聖人處無為 之事,行不言之教。萬物作焉而不辭

生而不有,為而不恃,功成而弗居。夫唯

是以 後面的文字,陳鼓應疑是錯簡。陳柱以為自





形顯的意義。王弼本原作「較」,畢沅曰:「古無『較』字,本文以『形』與

『傾』為韻,不應作『較』。



「萬物作焉而不辭」至章末,是由第五章錯來。高享以為「是以」以下八句與前

八句「文意截然不相聯」,乃後人合為一章,並疑「是以」二字為後人所加。

不辭 傅奕本、敦煌本、范應元本皆作「不

無為 是不恣意,不任己,一依自然而行的意思。無為並非不為事,只是如天地

創生萬物,順乎萬物生生之自然而已。

為始」,畢沅曰:「古始辭聲同,以此致異。」



l e s s o n 22 : l ao z i

Comments ⑦ 天下 obviously means ‘all the people’ in this context; compare French tout le monde. The clause 斯 惡 已 —note the Lu dialect word 斯 !—is strangely incomplete: either a S (but what could that be?) or a P is deleted. 已 here denotes the pt usually written 矣 (cf. 10.4–⑳). ⑧ For the use of 相 , cf. 7.5–㉕. ⑨ The commentary notes that the following sentences are from a ‘misplaced bamboo slip’ (錯 簡 ) which may have belonged to paragraph five of the Laozi. Such misplacement may occur when the strings that hold the bamboo slips tear, and it may greatly add to the obscurity of a text. However, since the silk MSS from Mawangdui have the same text (with slight variations), this supposition is not entirely convincing. ⑩ The 聖 人 is an ideal in Confucian, realist as well as Daoist texts, although the respective schools would disagree about his characteristic qualities. 無 為 , ‘doing nothing’ or ‘non-action’, is a key term in Laozi; it implies non-interference with the natural order of the world. ⑪ 辭 is puzzling; the commentary suggests 268

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

22.4 THE SAGE’S ORDER

that it be read as 始 : is the argument convincing? ⑫ A combination of two ptemp: 夫 , which introduces statements of general validity (cf. 18.3– ⑱) and 唯 , which here serves to emphasize the P. 22.4  The sage’s order (Lao 3) The third paragraph introduces the political thought of the Laozi, which is less than comforting: the text argues that the people should be disempowered and kept ignorant in order to establish a strong rulership.

ㄨ ㄨ ㄟ

ㄨ ㄅ ㄨ ㄓ





ㄈ ㄨ

ㄖ ㄨ ㄛ ㄑ

,強其骨。常使民無







ㄑ ㄤ





ㄍ ˇ ㄨ

ㄔ ㄤ

ˇ ㄕ



ㄇ ㄣ ㄨ

269

ㄧ ㄇ ㄣ



ㄧ ㄕ

,弱其志

ㄗ ㄜ

ㄒ ㄣ

ˇ ㄕ

ㄅ ㄨ ㄓ ㄥ

ㄅ ㄨ ㄍ ㄨ ㄟ ㄋ ㄢ ㄉ ㄜ ㄓ ㄏ ㄨ ㄛ

ˇ ㄕ ㄇ ㄣ



ㄨ ㄟ





ㄅ ㄨ ㄨ ㄟ ㄉ ㄠ

ㄅ ㄨ

ㄒ ㄢ



ˇ ㄝ

ㄒ ㄩ

ㄒ ㄢ

ㄎ ˇ ㄜ ㄩ

ˇ ㄕ ㄇ ㄣ





ㄨ ㄟ



ㄕ ㄤ

ㄒ ㄣ



ㄍ ㄢ ˇ



ㄅ ㄨ

ㄖ ㄣ





ㄓ ㄜ ˇ

ㄕ ㄥ





ˇ



ㄈ ㄨ



ㄅ ㄨ

, 使 民 不 爭 ; 不 貴 難 得 之 貨, 使 民 不 為 盜 ; 不 見 可 欲,使 民 心

ˇ ㄕ



ㄌ ㄨ ㄢ

。是以聖人之治,虛其心,實其腹

。使夫智者不敢為也。為無為,則無不治。

尚賢 河上公曰:「賢,謂世俗之賢。辯口明文,離道行權,去質為文也。不尚者,不





ㄅ ㄨ

不尚賢



不亂

知無欲

︻注釋︼

「見」與「現」同,



貴之以祿,不貴之以官。」按:賢,指賢者。尚賢,崇尚賢人。春秋時已有尚賢之風,如《論語·

子路》:「舉賢才。」《禮記·禮運》:「選賢與能。」故老子有反尚賢之言。



實其腹 使人民安飽,以免生巧偽。十二章曰:「是以聖人為

弱其志 志,即心志,是一切智力巧詐所產生之處,而智力巧詐是「聖人」治國的絆

顯現的意思。此句總上兩句而言。

腹不為目。」

腳石 十 ( 八章云:「智慧出,有大偽。」六十五章云:「民之難治,以其智多。」是其證明 。)所以 老子主張「弱其志」。 王弼曰:「守其真也。」按「使民無知無欲」,即使民「復歸於樸」 二 ( 十八章 。)









尚2 shàng, vi, ‘be superior’ 盜 dào, nc, ‘robber’ 虛 xū, vi, ‘be empty’ 腹 fù, nc, ‘belly’ 骨 gǔ, nc, ‘bone’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 22 : l ao z i

Comments ⑬ 尚 賢 was a tenet of the Mohists (cf. 24.1)—how does this tally with the image of Laozi as a repository of timeless wisdom? ⑭ Note the reading of 見 indicated in bopomofo: xiàn. This example shows how valuable the phonetic transcription is: it is a commentary in its own right that indicates which word is written by a given character, thereby fulfilling the first task in reading a Classical Chinese text (cf. p. xxv). In this case, it is a vi, used causatively, with可 欲 as O.  ⑮ These and the following parallel clauses all employ intransitive or stative verbs in a factitive sense. What does 其 refer to? ⑯ Note the parallels given in the commentary to explain this clause: it is all about making the people feeble-minded and submissive. ⑰ This sums it up: the people should be kept ignorant and devoid of free will. These aspects of Laozi are rarely acknowledged in the mainstream reception of Daoism. 22.5  The ideal state (Lao 80) This paragraph describes an ideal social order, in which the people are immobile, living in their separate villages without trade, writing, or interregional communication. Laozi advocates a primitive segmentary society, which, presumably, is easy to rule.

ㄨ ˇ ㄤ

ㄌ ㄞ

ㄌ ㄣ ㄍ ㄨ ㄛ ㄒ ㄤ ㄨ ㄤ





ㄙ ㄨ ㄟ

ˇ ㄡ

ˇ ㄕ

ㄓ ㄡ

ㄇ ㄣ



ㄈ ㄨ



ㄐ ㄝ



ㄕ ㄥ

ㄙ ㄨ ˇ ㄛ



ㄔ ㄥ

ㄩ ㄥ





ㄙ ㄨ ㄟ

ˇ ㄡ

ㄍ ㄢ

270

ㄐ ㄚ

ㄧ ㄧ

ˇ ㄕ





ㄩ ㄢ ˇ

ㄍ ㄨ ˇ ㄚ ㄇ ㄣ

ˇ ㄡ



寡 guǎ, vi,  ‘be few’ 什 shí, nc,  ‘squad (of ten soldiers)’ 伯2 bó, nc,  ‘pla­toon (of hundred soldiers)’ 器 qì, nc,  ‘weapon’



徙 xı ̌, vi,  ‘migrate’

ㄅ ㄛ

乘2 chéng, vn,  ‘mount’



陳2 chén, vn,  ‘deploy’







ㄒ ˇ





ㄧ ㄧ



ㄙ ㄨ

ㄙ ㄨ ˇ ㄛ ㄔ ㄣ

ㄒ ˇ ㄠ ㄍ ㄨ ㄛ

結 jié, vtr,  ‘tie (knots)’



甘 gān, vst,  ‘be sweet’

ㄅ ㄨ

樂3 yào, vn,  ‘delight in’

ㄩ ㄥ

ˇ ㄕ ㄇ ㄣ





ㄒ ㄤ





ㄅ ㄨ



ㄅ ㄨ

ㄌ ㄜ





ㄙ ˇ

ㄐ ㄩ



ˇ ㄙ



ㄌ ˇ ㄠ













ㄇ ㄣ

ㄈ ㄨ



ㄧ ㄧ



ㄨ ㄣ



ㄅ ㄥ

小 國 寡 民, 使 有 什 伯 之 器 而 不 用, 使 民 重

ㄒ ㄤ

ㄇ ㄟ ˇ

死而不遠徙。雖有舟輿 , 無所乘之;雖有甲 。使民復結繩而用之 。甘

ㄕ ㄥ





兵,無所陳之



,民至老死,不相往來。



ㄑ ㄩ ㄢ ˇ

其食,美其服,安其居,樂其俗 。鄰國相望,



雞犬之聲相聞

︻注釋︼ 俞樾曰:「什伯之器,乃兵器也。」 馬敘倫曰:

謂民淳事簡,可不用文字而治。此為老子反古

蘇轍曰:「內足,則外無所慕,故以其所有為美,

「『雖有舟輿』四句,古注文,誤入經文者也。」其說可供

參考。

思想。

以其所處為樂,而不復求也。」

㉑ ⑳



ㄓ ㄨ ㄥ

雞 jī, nc,  ‘chicken’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

22.5 THE IDEAL STATE

Comments ⑱ The paragraph provides no context at all: is this an ideal, a utopia, a prescription to be realized or a situation to be avoided? 什 伯 are military units, which explains the meaning of 器 . ⑲ The constructions 無 所 乘 之 and 無 所 陳 之 require closer analysis. In 所 陳 之 , 所 evidently substitutes for a PrP (cf. 12.2.2); the 所 phrase is a transformation of *陳 之 於 X, and it may be translated as ‘where ⟨they⟩ deploy them’. It follows that 無 所 陳 之 literally means ‘have no ⟨place⟩ where ⟨they⟩ deploy them’ > ‘deploy them no­where’, or, taking X as a time word, ‘never deploy them’, or even ‘deploy them under no circumstances’. 無 所 乘 之 is analogous. ⑳ 結 繩 , ‘to tie knots’, is supposed to refer to an ancient form of keeping records, similar to the Inka quippu. Reverting to it would mean to abandon writing. ㉑ In three of these parallel clauses, stative verbs are used in a putative sense; compare ⑮. ㉒ An absolute topic, in which the subordinated NP is logically the S of the following sentence (cf. 13.4 ㉕–㉖): it is the chickens and dogs, not their sounds, that hear each other.

Box 22  Deletion Deletion, the omission of superfluous words or constituents, is frequently encoun­tered in every­day speech. Consider sentences like ‘All right, I’ll be glad to’ (F. Scott Fitzgerald), ‘You don’t aim to answer my question, so I don’t yours’ (James Purdy) or fragmentary utterances like ‘Why?’, ‘Not really’, ‘When­ever’. In these and many other cases words that may be understood from the con­ text are omitted from a sentence. In Classical Chinese, deletion is very common in the case of the subject, if the latter has been previously mentioned. This regularly occurs when the S of a main clause is identical to the O (or a prepositional O) of the subordinate clause: 蒼梧之弟娶妻而美好。 吾問焉而不吾告 。

Cang Wus brother took a wife, and ⟨she⟩ was beautiful. (SY 3.21) I asked him, but ⟨he⟩ did not tell me. (Zhuang 19)

If adverbial expressions or subordinate clauses have the same O as the P or the main clause, respectively, it is regularly deleted: 余聞而弗忘。 眾聞則非之。

I have heard ⟨it⟩ and not forgotten it. (Zuo 10.3)

If the crowd hears ⟨it⟩, then it will condemn it. (Mo 17)

271

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 22 : l ao z i

Deletion of the Opr 之 is the rule after 不 and 以, and it often occurs after the prep 與 and 為: 孔子不見。 子貢以告孔子。 田成子與登臺… 為棄正妻。

Master Kong did not meet ⟨him⟩. (LY 17.1)

Zigong told Master Kong ⟨about it⟩. (HSWZ 9)

Tian Chengzi ascended the platform with ⟨him⟩. . .  (HFei 22) He abandoned his main wife for ⟨her⟩. (HFei 14)

Most irritating, perhaps, is the deletion of the P, which usually is the central part of the clause. Nevertheless, it may be omitted in cases where other constituents bear the main information: 學禮乎. . . 未也。 上醫醫國,其次疾人。 未知母之存否。 善不善本於義, 不於愛。

Have ⟨you⟩ studied the rites?…⟨I have⟩ not yet ⟨studied them⟩. (LY 16.13) The best doctors treat states, the second best ⟨treat⟩ patients. (GY 14.17)

I don’t even know whether my mother lives or doesn’t ⟨live⟩. (Zuo 7.2)

Good and bad are rooted in righteousness, ⟨they⟩ are not ⟨rooted⟩ in love. (Lü 19.4)

Deletion may occur with single parts of constituents as well. In NP, either the modi­fier or even the head may be omitted, if previously mentioned: 彼其髪短而心長。 晉大夫與楚孰賢。

His hair is short, but ⟨his⟩ ambitions are long. (Zuo 10.3) Who is more capable: Jin’s dignitaries or Chu’s ⟨dignitaries⟩? (Zuo 9.26)

By the same token, parts of other VP or PrP may be deleted: 臣聞以德和民, 不聞以亂。 景公問政孔子。

I have not heard ⟨of harmonizing the people⟩ through anarchy. (Zuo 1.4)

I have heard of harmonizing the people through virtue, Duke Jing asked Confucius ⟨about⟩ government. (SJ 47)

For more examples of the deletion of prep, which leads to pseudo-transitive constructions, cf. Box 11.

Finally, Classical Chinese allows for very succinct sentences that leave away pre­viously mentioned elements, making only new information explicit:

272

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

22.5 THE IDEAL STATE

大都,不過參國之一; 中,五之一; 小,九之一。

Big cities should not exceed one third of the capital; medium ⟨cities⟩ ⟨should not exceed⟩ one fifth ⟨of the capital⟩; small ⟨cities⟩ ⟨should not exceed⟩ one ninth ⟨of the capital⟩. (Zuo 1.1)

References: Unger 1994, 77ff; Yang Bojun and He Leshi 2003, 814–49.

273

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 19

Manuscripts The Shang and early Zhou periods may be regarded as an age of epigraphy, since little is known about texts being written on materials other than bones, bronze, or stone. However, since the fourth century bc, texts written in ink on perishable material—bamboo slips (MSC 竹簡 or 簡策), wooden slips (木牘) or silk sheets (帛書)—seem to have increased dramatically. The new writing material facilitated the production of significantly longer and more diverse texts than ever before: now, books were written. It is no accident that the classical age coincided with the first flowering of Chinese manuscript culture: the ‘contention between a hundred schools of thought’ (百家爭鳴) would have been unthinkable without a handy and portable medium of writing. In the case of bamboo manuscripts, the long, slender slips, which usually contained one line, were bound together with strings to form bundles (簡冊), as shown in Figure 10. The character 冊 is obviously the depiction of such a bundle. The fact that this character appears already on oracle bones has led scholars to believe that bamboo had been in use already in Shang times. While this would seem to be a plausible assumption, no such early specimens have been found, and—more importantly—the medium did not exert a noticeable impact on literary production before the classical age.

Just how multifaceted this manuscript culture was has only become clear in the past four decades, when archaeologists discovered ten thousands of manuscripts in ancient tombs, wells, and military fortifications. Whereas the greatest part of these manuscripts is bureaucratic in 274

Figure 10  Bamboo manuscript Source: Maurizio Scarpari. Das antike China. Cologne: Verlag Karl Müller GmbH, 2001, 45.

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 19 : ma n u s c r i p t s

nature, containing administrative, legal, or mili­tary communication, the others contain diverse genres of writing: medical, divinatory, astrological and prognostic texts, annalistic and historical writings, calendars, lists, letters, and philosophical texts. Many of these manuscripts were found in tombs, where they were buried together with various other valuables that were apparently meant to accompany the tomb occupant in the netherworld. The illustration in Figure 11 shows the Qin tomb M11 from Shuihudi 睡虎 地 (Hubei), in which the corpse is surrounded and literally covered by more than a thousand bamboo slips containing legal texts, a  chronicle, an almanac, and a manual for administrators. As the picture shows, such texts in tombs may be in considerable disarray due to the decay of their binding: these are not just cases of single misplaced bamboo slips, but of Figure 11  Shuihudi tomb entirely jumbled texts. Reassembling them in Source: Wenwu [Cultural Relics, an archaeological Journal] 6 (1973), 3 the right order is a daunting task. This is espefig. 4. cially so since the majority of the excavated manuscripts contain texts without transmitted counterparts. But even those excavated texts that correspond to received versions—among them Laozi, Lunyu, Yijing, parts of Zhanguo ce, Wenzi, Yili, and Liji, as well as several military classics—pose numerous difficulties. Especially the fact that ‘for the manuscripts the assumption of a conventionally accepted association of graph to word does not hold with anything like the degree of regularity it does for later texts’ (Boltz 1997, 255) presents serious problems. The case of Laozi is particularly well suited to exemplify the relation between transmitted texts and excavated manuscripts, since altogether five manuscripts of the text have been found: three, written on bamboo slips, were found in a tomb in the village of Guodian 郭店, Hubei, dated to about 300 bc. Another two, written on silk sheets, were found in a tomb in Mawangdui 馬王堆, near Changsha, dated to 168 bc (cf. the illustrations in Figure 12). These finds have revealed just how different 275

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 19 : ma n u s c r i p t s

the Laozi text was in classical times before its subsequent standardization. The script styles are very dissimilar from the kaishu of later texts, the orthography, especially the use of classifiers, is still far from later norms, the association of characters with words differs noticeably from what we know from transmitted texts, the syntax diverges in some places, indeed the entire structure of the text is reversed: the Laozi silk manuscripts both place the ‘De jing’ before the ‘Dao jing’. The following example provides a comparison of the respective versions of Laozi 2 (22.3): try to locate all the variant readings and consider what consequences they may have for our understanding of the text. Received version: 天下皆知美之為美,斯惡已。皆知善之為善, 斯不善已。故有無相生,難易相成,長短相較,高下相傾, 音聲相和,前後相隨。是以聖人處無為之事,行不言之教;萬物 作焉而不辭,生而不有,為而不恃,功成而弗居。夫唯弗居,是 以不去。 郭店老子甲: 天下皆智美之為美也,亞已;皆智善,此其不善 已。又亡之相生也,難惕之相成也,長耑之相型也,高下之相浧 也,音聖之相和也,先後之相隨也。是以聖人居亡為之事,行不 言之教。萬勿作而弗始也,為而弗志也,成而弗居。天唯弗居 也,是以弗去也。 馬王堆老子甲: 天下皆知美為美,惡已;皆知善,訾不善已。有 无之相生也,難易之相成也,長短之相刑也,高下之相盈也,意 聲之相和也,先後之相隋,恆也。是以聲人居无為之 事,行□□□□。□□□□□□也,為而弗志也,成功而弗居 也。夫唯居,是以弗去。 馬王堆老子乙: 天下皆知美之為美,亞已。皆知善,斯不善 矣。□□□□生也,難易之相成也,長短之相刑也,高下之相盈 也,音聲之相和也,先後之相隋,恆也。是以聖人居无為之 事,行不言之教。萬物昔而弗始,為而弗侍也,成功而弗居 也。夫唯弗居,是以弗去。

276

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 19 : ma n u s c r i p t s

Figure 12  Left: Silk ms from Mawangdui; Right: Bamboo ms from Guodian Sources: Left: Mawangdui Hanmu boshu Laozi. Beijing. Wenwu chubanshe, 1976; Right: Jingmen bowuguan [Jingmen Museum] (ed.). Guodian Chumu zhujian. Beijing: Wenwu chubanshe, 1998.

References: The past two decades have seen a flood of studies on ancient Chinese manuscripts. With the publication of the Handbook of Early Chinese Manuscripts by Behr, Meyer, and Kern forthcoming, Tsien Tsuen-­hsuin 2004, chs. 5 and 6, Giele 2010, and Wilkinson 2018, 1023–9, may serve as introductions; Chavannes 1905 and Wang Guowei 2014 are still edifying; Pian Yuqian and Duan Shu’an 2006 and Li Junming 2011 provide overviews of the field; Goldin 2021 is an up-­to-­date bibliography of Western publications; Handa wenku is a database containing many excavated MSS. Wei Desheng 2000 is a thorough study that shows just how different the language of MSS is from that of transmitted texts. Notable recent monographs include Shaughnessy 2006, Cook 2012, Meyer 2012, and Richter 2013.

277

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 23

Zhuangzi

23.1 Introduction278 23.2 The butterfly dream (Zhuang 2) 279 23.3 The wheelwright’s wisdom (Zhuang 13) 280

23.4 The fish’s joy (Zhuang 17) 23.5 Life and death (Zhuang 18)

282 283

23.1 Introduction Zhuangzi 莊子, a founding text of Daoism, is perhaps the most beloved piece of ancient Chinese literature among contemporary readers. A work of sparkling wit, linguistic finesse, and compelling narratives, it is a literary gem. Its fanciful anecdotes have often been contrasted to the prosaic moralizing discourse of the ‘Confucian’ school, and indeed, Confucius and his followers are often the butt of Zhuangzi’s humour. Little is known about its alleged author, Zhuang Zhou 莊周 (trad. 369–286 bc). A contemporary of Meng Ke’s, he is said to have come from Song, in present-­day Henan, where apparently he once held a minor office. But Zhuang Zhou certainly was no careerist: when the king of Chu offered him a post, so Zhuangzi has it, he replied that he would rather roll in the mud like a turtle. Zhuang Zhou is portrayed as a nonconformist who despised worldly success and ridiculed conventional knowledge—and yet, the body of philosophical literature associated with his name is substantial. The received version of the Zhuangzi was edited by Guo Xiang 郭象 (d. 312 ad), who condensed a work of originally fifty-­two chapters to thirty-­three chapters containing some 60,000 characters. Of these, only the first seven, the ‘Inner Chapters’, are considered as authentic (a tenet which has recently been challenged); chapters 8–22 are called ‘Outer’, and the remaining chapters 23–33, 278

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

23 . 2 t h e b u t t e r f ly d r e a m

‘Miscellaneous Chapters’. Whoever their authors were—‘primitivists’, ‘hedonists’, or ‘syncretists’—they were no lesser men than the author of the ‘Inner Chapters’. Their anecdotes, parables, and criticism of rival schools add to this rich repository of Daoist thought. The following excerpts follow Guo Qingfan’s Zhuangzi jishi 莊子集釋 with three commentaries: the principal commentary (注) by Guo Xiang, the sub-­commentary (疏) by Cheng Xuanying 成玄英 (seventh cent.), and the glosses of Lu Deming’s Jingdian shiwen. Editions: Zhuangzi jishi 莊子集釋, ed. by Guo Qingfan 郭慶藩. 4 vols. Beijing 1997. Translations: Wilhelm 1923; Watson 1968; Graham 1981; Mair 1994; Levi 2006. Studies: Chang Tsung-­Tung 1982; Cook 2003; Roth 2003; Coutinho 2004; Klein 2010; McCraw 2010.

23.2  The butterfly dream (Zhuang 2) The ‘butterfly dream’ is probably the best-known parable from Zhuangzi. The final section of Chapter two, 齊物論 (‘On regarding things as equal’), which is a core chapter of the work, it illustrates Zhuang Zhou’s scepticism toward all kinds of cognitive categories and differentiations: those between ‘this’ and ‘that’, life and death, or dream and reality. The Dao knows no such differentiations. 昔者莊周夢為胡蝶, 栩栩然胡蝶也 , 自喻適志與! ︻注︼自快得意,悅豫而行。



︻疏︼栩栩,忻暢貌也。喻,曉也。︻釋文︼ 《胡蝶》



故稱覺耳,未必非夢也。︻釋文︼《然覺》古孝反。《蘧



徐徒協反。 《栩》徐況羽反,喜貌。 不知周也。 俄然覺,則蘧蘧然周也 。︻注︼自周而言,

蘧 》 徐 音 渠, 又 其 慮 反。 不 知 周 之 夢 為 胡蝶與,胡蝶之夢為周與? 周與胡 蝶,則必有分矣 。︻注︼夫覺夢之分,無異於 此之謂物化 。︻注︼夫時不暫 死生之辯也。

停,而今不遂存,故昨日之夢,於今化矣。死生

之變,豈異於此,而勞心於其間哉! 方為此則不





279



夢 mèng, vn,  ‘dream’ 胡蝶 húdié, nc,  ‘butterfly’ 栩栩 xǔxǔ, vi,  ‘be blithe’ 喻/愉 yú, vi,  ‘be happy’ 適 shì, vn,  ‘satisfy’ 俄 é, adv,  ‘suddenly’ 覺 jiào, vn,  ‘wake up’ 蘧蘧 qúqú, adv, ‘jumping (from sleep)’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 23 : z h ua n g z i

知 彼,夢 為 胡 蝶

是也。取之於人,

則 一 生 之 中,今

不 知 後,麗 姬 是

也。而愚者竊竊

然自以為知生

之 可 樂,死 之 可

苦,未 聞 物 化 之

謂也。 ︻釋文︼《可

樂》音洛。



Comments ① While the ‘butterfly dream’ is the best-­known par­able in the Zhuangzi, the ‘butterfly’ itself, 胡蝶, is probably the best-­known two-­ syllable word in Classical Chinese (cf. Focus 9)—yet, Lu Deming (quoting an earlier commentary) felt it necessary to gloss the pronunciation of the second character. ② Another two-­syllable word of the reduplicative type. Such elusive expressions, which describe moods, feelings, or the like, are very frequent in Zhuangzi—and they are always a challenge for commentators. This clause is typical for Zhuangzi’s style: it may either be construed as an inversion of S and P (cf. 14.3) or as a regular verbal clause, in which case 胡蝶 functions as verbal P.  Consider the arguments for and against either interpretation. ③ Recall that 自 often substitutes for the O of the following verb; it regularly seems to function in this way with stative verbs expressing emotions: thus 自喻 means not simply ‘pleased by himself ’, but ‘pleased with himself ’. 適志 may be rendered as ‘to follow one’s desires’ or the like. Just like 乎, 與 not only marks questions but also open or suggestive sentences that end on a high pitch (cf. 7.2–①). ④ Note the reading of 覺 given in the commentary; despite its meaning in MSC 睡覺, it means ‘wake up’. Again, this is either an inversion of S and P or 周 is used as a verb. ⑤ For this construction, review 5.3.1–㉝. ⑥ Try to analyse this sentence carefully: what is the function of 則 (don’t let the punctuation mislead you)? Observe what liberties Guo Xiang’s commentary takes in interpreting this ­sentence. ⑦ An anteposed O. Now the commentator unfolds his own agenda in earnest. Note how Jingdian shiwen glosses 樂, which appears not in the main text but in Guo Xiang’s commentary: this is a commentary on a commentary. 23.3  The wheelwright’s wisdom (Zhuang 13) Most anecdotes in Zhuangzi do not feature Zhuang Zhou but other imaginary beings: strange creatures, the infirm and the outsiders, and 280

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

23 . 3 t h e w h e e lw r i g h t ’ s w i s d o m

often inferior artisans possessed of superior wisdom. In this story, a wheelwright expresses a core tenet of Zhuangzi: distrust of language and contempt for books. 桓公讀書於堂上。輪扁斲輪於堂下, 釋椎鑿而上,問桓公曰:「敢問,公 之所讀者 ︻校︼世德堂本者作為。何言邪?」 ︻疏︼桓公,齊桓公也。輪,車輪也。扁,匠人 名 也。 斲,雕 斫 也。 釋,放 也。公 曰 :「 聖 人 之 言 也。 」 曰 :「 聖 人 在 乎?」 公 曰 : 「已死矣。」︻疏︼答曰:「聖人雖死,厥教尚存焉。」曰:「然則君之所讀者,古人之糟 魄已夫!」 ︻疏︼ 夫( 酒) 滓曰糟,漬糟曰粕。夫醇酎比乎道德,糟粕方之仁義,已陳芻狗,曾 何足云!桓公曰: 「寡人讀書,輪人安得議乎! 有說則可,无說則死。」輪扁曰: 「臣也以臣之事觀之。 斲輪,徐則甘而不固,疾則苦而不入。不徐不疾,得 之於手而應於心,口不能言,有數存焉於其間 。 ︻釋文︼《甘》如字,又音酣。司馬云: 甘者,緩也。苦者,急也。《有數》李云:色注反,數,術也。臣不能以喻臣之子,臣之子 亦不能受之於臣,是以行年七十而老斲輪 。 ︻注︼此言物各有性,教學之無益也。古 之人與其不可傳也死矣, 然則君之所讀者,古人之糟魄已夫!」︻注︼當古之事,

已滅於古矣,雖或傳之,豈能使古在今哉!古不在今,今事已變,故絕學任性,與時變化而後至焉。

















桓 Huán, np, posth. name 讀 dú, vn, ‘read (aloud)’ 扁 Biǎn, np 斲 zhuó, vn, ‘carve’ 釋 shì, vtr, ‘put aside’ 椎 chuí, nc, ‘hammer’ 鑿 zuò, nc, ‘chisel’ 上2 shàng, vi, ‘ascend’ 糟 zāo, nc, ‘sediment’ 魄/粕 pò, nc, ‘dregs’ 議 yì, vn, ‘criticize’ 說2 shuō, vn, as nc, ­‘explanation’ 徐 xú, vi, ‘be slow’ 苦 kǔ, vst, ‘be unpleasant’ 應 yìng, vn, ‘react’ 數2 shù, nc, ‘skill’ 間 jiān, nloc, ‘midst’

Comments ⑧ Duke Huan of Qi (r.  685–643 bc) was the first of the Chunqiu ‘hegemons’ and the most powerful ruler of his time. Bian is one of many speaking names in Zhuangzi: it means ‘flat’ (the Kangxi zidian 康熙字典 specifies: 凡器物不圓者曰扁)—which is certainly a paradoxical name for a wheelwright. ⑨ For 敢問, cf. 4.4.1–㉑. Note the use of 邪 in a 281

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 23 : z h ua n g z i

­variable question (cf. 3.4.3), similar to MSC 呢. In the commentary, 校 introduces a text critical note (校勘記). ⑩ The conj 然則, literally ‘that being so, then’, introduces conclusions; in this sentence, it is being supplemented by the final 夫, ‘isn’t it’. 已 is a graphic variant of 矣. ⑪ For 安, cf. 21.3–㉒. The modal auxiliary 得, ‘get to (do)’, here means ‘dare to’ or the like; compare German ‘wie kommst du dazu?’ ⑫ For the expression 以 X 觀之, cf. 9.2.1–③. ⑬ For the function of 有 in 有數存焉, cf. 17.2–③. There is a definite redundancy in 存焉於其間: perhaps 於其間 was a gloss on the word 焉, which a later copyist erroneously integrated into the main text. Or is this the informal language of a wheelwright? The comment《甘》如字 means that the character 甘 is to be read like it usually is (not as a loan for another character). ⑭ For 是以, cf. 9.2.3–⑯. In 老斲, 老 serves as adv. ⑮ In this clause, all of 古之人與其不可傳 is the S, topicalized by 也: ‘the men of old and their untransmittable ⟨skills⟩’. 23.4  The fish’s joy (Zhuang 17) Another playful anecdote typical for the tone of Zhuangzi. The ‘sophist’ Huizi questions the possibility of cognition, and Zhuang Zhou, in turn, questions his question with an ingenious pun. Despite its ostensible disregard for language, Zhuangzi is a masterpiece of the Classical Chinese language.



莊子與惠子遊於濠梁之上 。 ︻疏︼濠是水名, 在淮南 鍾離郡,今見有莊子之墓。莊子曰:「儵 魚出遊從容,是魚之樂也。」︻疏︼儵魚,白







儵也。從容,放逸之貌也。︻釋文︼《儵魚》徐音條。

說文直留反。李音由,白魚也。 惠子曰:「子 非魚,安知魚之樂?」莊子曰:「子非我, 安知我不知魚之樂?」惠子曰:「我非子, 固 不 知 子 矣 ; 子 固 非 魚 也, 子 之 不 知 魚 之 樂, 全 矣。」 莊 子 曰 :「 請 循 其 本 。 ︻疏︼循,猶尋也。循本之義,列在下交。子曰『汝 安知魚樂』云者, 既已知吾知之而問 我, 我知之濠上也。」





282

濠 Háo, np, a river 梁2 liáng, nc, ‘bridge’ 儵魚 Tiáoyú, np, a fish 從容 cóngróng, vi, ‘be carefree’ 全 quán, vi, ‘be complete’ 循 xún, vn, ‘stick to’ 云 yún, pt, marks end of quotation

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

23 . 5 l i f e a n d d e at h

Comments ⑯ Huizi, or Hui Shi 惠施, a fourth-­century philosopher from Song, who is usually classified as a ‘sophist’ (名家), often appears in Zhuangzi disputing with Zhuang Zhou. ⑰ 從容 is another descriptive two-­syllable word. What does the commentary tell us about the fish in question? ⑱ 全, ‘complete’, may be understood as ‘entirely clear’. ⑲ The pt 云 is used to mark the end of quotations (which is here introduced by 曰), somewhat like the English ‘unquote’. The following 者 serves to topicalize the entire phrase: this is the origin of the conversation that Zhuangzi wanted to return to. ⑳ Two synonymous adv, 既已, emphasize the point: ‘you already knew very well’ or the like. ㉑ This is Zhuangzi’s punch line: he ridicules Huizi’s question by playing on the double meaning of the word 安. 23.5  Life and death (Zhuang 18) The story about the death of his wife is another example of Zhuangzi’s disbelief in conventional differentiations—life and death—and his defiance of social customs. In stark contrast to Confucian insistence on elaborate death rituals, he refuses to even mourn his wife, arguing that this would mean misunderstanding the nature of being.







283

莊 子 妻 死, 惠 子 弔 之, 莊 子 則 方 箕 踞 鼓 盆 而 歌 。︻疏︼箕踞者,垂兩腳如簸箕型也。盆,



瓦缶也。莊子知生死之不二,達哀樂之為一,是以

妻亡不哭,鼓盆而歌,垂腳箕踞,敖然自樂。 惠 子 曰 :「 與 人 居 長 子 老 身, 死 不 哭 亦 足 矣,又鼓盆而歌,不亦甚乎!」︻釋文︼《長 子》丁丈反。 莊子曰:「不然。是其始死 也, 我 獨 何 能 无 概 然! 察 其 始 而 本 无 生, 非 徒 无 生 也, 而 本 无 形, 非 徒 无 形 也而本无氣 。 雜乎芒芴之間,變而有氣, 氣變而有形,形變而有生,今又變而之死, 是相與為春秋冬夏四時行也 。



弔 diào, vn, ‘condole’ 箕 jī, nc, ‘winnowing basket’ 踞 jù, vn, ‘sit’ 盆 pén, nc, ‘pot’ 歌 gē, vn, ‘sing’ 概然 gàirán, vi, ‘be sad’ 徒2 tú, adv, ‘only’ 雜 zá vi, ‘be mixed’ 芒芴 mángwù, nc, ‘tohubohu’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

人 且 偃 然 寢 於 巨 室, 而 我 噭 噭 然 隨 而 哭 之, 自 以為不通乎命,故止也。」



︻釋文︼《无槪》古代反。司

馬云:感也。又音骨,哀亂貌。

所 以 誨 有 情 者, 將 令 推 至 理 以

遣累也。︻疏︼偃然,安息貌也。

巨 室, 謂 天 地 之 閒 也。 且 夫 息

我 以 死, 臥 於 天 地 之 閒, 譬 彼

炎 源, 何 得 隨 而 哀 慟! 自 覺 不

通天命,故止哭而鼓盆也。

︻注︼未明而概,已達而止,斯

l e s s o n 23 : z h ua n g z i

偃 yǎn, vi, ‘rest’ 巨 jù, vst, ‘be great’ 噭噭 jiàojiàorán, adv, ‘sobbingly’

Comments ㉒ Again, Huizi has a discussion with Zhuangzi, whom he catches lolling in a rather indecorous manner, with his legs splayed ‘like a winnowing basket’. ㉓ Huizi expresses dismay at Zhuangzi’s behaviour in a complex sentence; start your analysis by determining which clauses are subordinated and which are superordinated. How is 長 to be pronounced? ㉔ The rest of the text consists of Zhuangzi’s answer. For the analysis of 是其始死也, recall the uses of 其 (cf. Box 18). The vi 獨, used adverbially, may serve to emphasize an interrogative P, similar to German ‘bloß’ (compare also, loosely, ‘if only. . .’): ‘possibly’ is a fitting translation. ㉕ 氣, the primal ‘pneuma’ is the immaterial basis for all life and forms: this is a core assumption of ancient Chinese cosmology shared by many thinkers. It is in fire and water, plants, animals and humans (Xun 9), reaches up to heaven, down to the lowest depths (Guan 49) and makes up all physical substance (Meng 2A2). Zhuangzi’s notion that his wife in the beginning (始) was virtually nothing is reminiscent of Lao 1: 無,名天地之始 (cf. 22.2). ㉖ 芒芴 is a hapax legomenon, similar in meaning to tohu wa bohu in Genesis 1.2: the primaeval chaos. The pronominal adv 相, serving as O to a prep, precedes the latter (cf. 7.5–㉕). Can you analyse the NP following 為? ㉗ 命 here clearly means more than just an ‘order’; perhaps ‘fate’ or ‘destiny’ comes close to what is meant. In any case, it is something not to be actively sought but simply to be accepted by man.

284

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

23 . 5 l i f e a n d d e at h

Box 23  Bracket constructions As noted in Box 16, some CC words have corresponding fusion forms with a ‘pronominal’ final *-n, namely: 於 *ʔa ~ 焉 *ʔan, and 如 *na ~ 然 *nan. Since neither 於 nor 如 take 之 as an object, it is likely that *-n represents an Opr in the fusion words. Hence, 焉 means ‘in/on/from it’, and 然, ‘(be) like it’. The case of 曰 *wat ~ 云 *wən may be comparable, since 曰 is never followed by 之, and 云 means ‘say it/so’. Interestingly, all these pairs also lend themselves to bracket constructions, which are introduced by the word without *-n and closed by the word with *-n. 於此有人焉。 於文王取法焉。

Here is a man . . . (Meng 3B4)

⟨He⟩ derives his model from King Wen. (Mo 16)

In these constructions, where the PrP is placed in initial position, the function of 焉 is similar to that of object ‘dummies’, which take the regular position after the P when the O is exposed (cf. 13.3). The cases of 如/若. . . 然/焉, and 曰 . . . 云 are somewhat different, since they do not involve exposition, nor do the two words represent the same c­ onstituent. Here, the pleonasm seems to be motivated simply by rhetorical reasons. . . . 如日月之食焉。

. . . like an eclipse of the sun or the moon. (Zuo 7.12)

聖人若天然 . . . 若地然。

The sage is like heaven . . . and he is like earth . . . (Guan 37)

. . . 若擊鼓而求亡子焉。

. . . like beating a drum to search for a lost son. (Zhuang 13)

子曰『汝安知魚樂』云者。

You said: ‘How do you know the fishs’ joy?’ (Zhuang 17)

Finally, the bracket construction involving 於 and 乎 seems to be equally pleonastic, since both represent a prep (thus yielding a PrP with two prep). They are especially frequent with the prdem 是 and 此: 財用於是乎出。

Funds and resources emerge from there. (GY 1.3)

於是乎百官降物。

Thereupon, the officials submitted their things. (Zuo 10.17)

於此乎言之。

Now I will say it. (Zhuang 24)

References: Pulleyblank 1995, 82–3; Unger 2019, 709–11, 918–19, and 925–26.

285

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 20

Text structures Throughout this textbook, the importance of analysing clause structures has been repeatedly emphasized. It should be added that the ­analysis of text structures is no less important. In fact, as Michael Broschat has emphasized, ‘only when those structures, and thus their effects, are as clear as possible, can the reader presume to have a good understanding of the author’s intentions’ (Broschat 1985, 106–7). Just as the signal words of a clause may give us essential information about how to understand it (cf. Reviews 2 and 4), so the structural elements of a text may provide us with decisive clues about how it is to be read—even before we have trans­ lated a single word. When dealing with texts, we should proceed just like we do with clauses: not by rashly attempting translation but by first determin­ing their overall struc­ture. Text structures are the framework of any utterance, they establish the very precon­ditions for its meaning. There are many features that may be used as signposts for the analysis of text structures, including: • rhymes and/or parallelism indicate the interconnection of succes­ sive clauses; • conditional, consecutive, or final conjunctions point to the logical association of clauses; • repetition of names or other expressions suggest close connection between the respective clauses; • by the same token, the appearance of new names or themes may be a sign of disparate textual units; • beginnings and ends of direct speech (marked by 曰, 云, and, in many editions, by quotation marks) may demarcate textual units; • time words like 初, 昔, 今, or other temporal expressions may serve to parse a text; • equally, expressions like 夫, 故, 然則, and others in initial position mark the begin­ning of a new paragraph. For an example of how these signposts may be used, consider the open­ ing of Xun 23: 286

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 20 : t e x t s t r u c t u r e s

1 人之性惡,其善者偽也。 2   今人之性, 3    生而有好利焉, 4    生而有疾惡焉, 5     生而有耳目之欲,有好聲色焉, 6 7 8 9 10

順是,故爭奪生 而辭讓亡焉; 順是,故殘賊生而 忠信亡焉; 順是,故淫亂生 而禮義文理亡焉。

   然則從人之性,順人之情,    必出於爭奪, 合於犯分亂理, 而歸於暴。   故必將有師法之化,禮義之道,    然後出於辭讓, 合於文理, 而歸於治。 用此觀之,人之性惡明矣,其善者偽也。

The sentence in line 1, which expresses the topic under discussion, is repeated ver­batim in line 10: these two lines clearly frame the entire para­ graph. Line 2 begins with 今, which typically introduces a new paragraph in which a general statement is applied to a more specific case in point, thus indicating a deduction of sorts. Line 2 reiterates the theme from line 1, followed by three neatly parallel lines (3–5) with the structure: 生而有 X 焉,順是,故 Y 生而 Z 亡焉。 The last line is the most elaborate: following the law of increasing con­stituents (cf. Review 2), it concludes this complex sentence. In line 6 and 8, 然則 and 故 both indicate conclusions from the preceding obser­ vations. They are likewise followed by parallel lines with the structure: 出於 X,合於 Y,而歸於 Z。 Moreover, lines 7 to 9 reiterate the Y and Z terms from lines 3 to 5—爭奪, 辭讓, 禮義, and 文理—creating an elegant interlocking pattern, which weaves lines 3 to 9 together (interestingly, however, the key terms from line 4—殘賊 and 忠信—as well as 淫 from line 5 are not repeated). Line 10, then, expresses the quod erat demonstrandum, mark­ing the formal conclusion of the argument. ­Just by analysing its structural components, we have gained an understanding of how this text is to be read—now we can begin to trans­late. 287

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 20 : t e x t s t r u c t u r e s

Unfortunately, such basic text structures are rarely reflected in stud­ ies and trans­lations of Classical texts. This may be partly due to the fact that virtually all editions of Classical texts conceal these struc­tures by their layout (cf. the presentation of Xun 23 in 19.3). It is therefore highly recom­mended that students of Classical Chinese re-­format the texts they deal with before embarking on translation and interpretation. Electronic versions make it easy to rearrange lines in the above fashion, add different colours and other graphic features to mark the signposts of a text’s structures and thus make its logical shape visible. This exercise may reveal patterns that testify to the sophistication and coherence of a text—or it may exhibit gaps, breaks, and incon­sistencies that testify to its heterogeneous nature (cf. Focus 17). In any case, the analysis of text structures can be a fascinating eye-­opener and a key to precise, ­reasoned interpretations of Classical Chinese texts.

288

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 24

Mozi

24.1 Introduction289 24.2 Against Offensive Warfare (Mo 17) 290

24.3 Beware of ghosts (Mo 31)

294

24.1 Introduction The Mozi 墨子, Master Mo, is a collection of texts attributed to the school of Mo Di 墨翟 (c.479–381 bc). From the little we know, Mo Di, a native of Song, supposedly was a former adherent of the Confucians who left them to found a rival school with quite different tenets. In search of a new social order, they opposed all particularities of a lineage-­based society, proposing instead the principle of ‘universal solidarity’ (兼愛) that transcended kinship ties; they despised the ritual pomp cherished by Confucians, arguing ‘against music’ (非樂) and urging ‘thrift in consumption’ (節用) as well as ‘thrift in funerals’ (節葬). The Mohists wanted to pacify society by ‘rejecting offensive warfare’ (非攻)—some Mohists excelled as engineers of defensive devices—and implementing a strictly hierarchical order based on ‘elevating the worthy’ (尚賢) and compliance with superiors: a ‘state machine’ (V. Rubin) directed from the top of a political pyramid. Although the Mohists are described as a motley group of plebeians, they were just as tightly organized as the state they envisioned. Commanded by ‘grand masters’, they remained highly influential at some feudal courts in the fourth century bc, even after Mo Di’s death. In the third century bc, the Mohists are said to have split into several groups, eventually disappearing entirely from the political landscape. Mohism was subsequently all but ignored in imperial times, and it was not until the Qing scholars Bi Yuan 畢沅 (1730–97) and Sun Yirang 孫詒讓 (1848–1908) edited the Mozi that this school was appreciated as an important element of Chinese tradition. 289

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 24 : m o z i

The Mozi consists of seventy-­one chapters, eighteen of which are only known by their title, with very different contents. The core chapters (Nos. 8–37) contain the main teachings attributed to Mo Di, each discussed in three different versions (corresponding to the three schools?); furthermore, the collection contains ‘dialectical’ chapters which discuss questions of logic, ‘military’ chapters concerned mainly with defensive military technology, records of Mo Di’s speeches and deeds (Nos. 46–50), and miscellaneous other chapters. These parts are very heterogeneous in style—ranging from colloquial speech to highly abstract logical reasoning—and content; they were apparently composed in different times, and, certainly due to their unsteady transmission, display a host of textual corruptions and grammatical problems. The following excerpts are taken from the Mozi xiangu 墨子閒詁 by Sun Yirang. Editions: Mozi xiangu 墨子閒詁, ed. by Sun Yirang 孫詒讓. 2 vols. Beijing 2001, is a standard edition. Translations: Forke 1922; Watson 1963b; Schmidt-­Glintzer 1975a, 1975b, 1992; Johnston 2010; Knoblock and Riegel 2013. Studies: Graham 1978 is the classic treatment of the ‘dialectical’ sections; the military ­chapters are discussed by Yates 1979; Oriens Extremus 48 (2009) contains essays by Michael Nylan, Satō Masayuki, and Joachim Gentz on the Mozi and Mohist thought; essays on ­several chapters are assembled in Defoort and Standaert 2013.

24.2  Against Offensive Warfare (Mo 17) The following selection from Mozi 17 is the first and shortest of three sections ‘Against Offensive Warfare’. It is a good example for the style of argument in the core sections of the Mozi and for the kind of textual problems the editors had to grapple with. 非攻上

290

今有一人,入人園圃,



畢云:「說文云:『園所





以 樹 果。』『 種 菜 曰 圃。』」

竊 其 桃 李, 眾 聞 則 非 之, 上 為 政 者 得 則 罰 之。 此 何 也? 以 虧 人 自 利 也。 至 攘 人 犬 豕 雞 豚 者, 其 不 義 又 甚 入 人 園 圃 竊 桃 李。 是 何 故 也? 以 虧 人 愈多, 依下文當有「苟 ④



圃 pǔ, nc, ‘garden’ 竊 qiè, vn, ‘steal’ 李 lı ̌, nc, ‘plum’

虧 kuī, vi ‘harm’ 攘 ráng, vn, ‘rob’ 豚 tún, nc, ‘piglet’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

24.2 ag a i n s t o f f e n s i v e wa r fa r e

虧 人 愈 多 」 五 字。 其 不 仁 茲 甚 , 茲、 滋 古 今 字。 罪 益 厚。 至 入 人 欄 廄 , 欄, 即 闌 之 借 字。 廣



雅 釋 室 云 :「 欄, 牢 也。」

取 人 馬 牛 者, 其 不 仁義又甚攘人犬豕雞 豚 。 依 上 下 文, 此 句 疑 不 當有「仁」字。 此何故 也? 以 其 虧 人 愈 多。 苟 虧 人 愈 多, 其 不 仁 茲甚,罪益厚。 ⑧



欄 lán, nc, ‘corral’ 廄 jiù, nc, ‘stable’



Comments ① Chapters or sections with the same title are listed as 上 and 下—No. 1 and 2—or, in the case of three, as 上, 中, and 下. If there are more than three, the ‘Heavenly stems’ (cf. Focus 8) may be used to count them, as in Guanzi, 輕重甲 to 輕重庚. ② The expression 今有 introduces hypothetical statements; compare 今有殺人者 ‘If somebody killed a man’ (Meng 2B8). ③ Two complex sentences in which both clauses have the same O: in both cases, it is deleted in the subordinate clause. ④ Here, 以 is used as a verb—in which sense? The expression 至 . . . 者 may be rendered as ‘when it comes to. . . ’; the entire clause functions as topic for the following sentence. The adv 又 here serves to express the comparative of a vst. Note that the prep 於 is omitted after 甚: in later CC such pseudo-­transitive constructions become more and more common (cf. Box 11). ⑤ Note that the A, which serves as adverbial modifier, is placed after the P, because it is the core of the statement. We have already observed similar usage of ­temporal expressions (cf. 10.3 and Box 13), and it is also not uncommon with vst; compare GYang 7.6: 滋欲 殺之 甚. ⑥ The commentary c­ onjectures that the text is missing five characters in this place. While such text critical notes are very common in Qing editions, the editors mostly do not go so far as to emend the text: the doubtful reading is duly noted, but not corrected. ⑦ Here, the commentary explains 兹 as an archaic form of the character 滋. ⑧ Again, a character gloss: 欄 is explained as a loan character (假借) for 闌; in support of this, the corresponding entry from the Guangya, a third-­century encyclopaedic dictionary, is adduced. ⑨ Again, a text critical note, employing a similar argument as above: is the argument convincing?

291

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi





至殺不辜人也,扡其衣裘 ,畢云:「陸德明易



音義云:『褫,鄭本作『拕』,徒可反。』『扡』即『拕』



異文。」王云:「『也』即『扡』字之誤而衍者。」

取戈劍者,其不義又甚入人欄廄、取人馬 牛。此何故也?以其虧人愈多。苟虧人愈 多,其不仁茲甚矣,罪益厚。當此,天下 之 君 子 畢 云 :「 舊 脫 此 字, 據 後 文 增。」 皆 知 而非之,謂之不義。今至大為攻國,則弗 知 非 ,畢云:「知,一本作『之』。舊脫『非』字, 據 後 文 增。」 從 而 譽 之, 謂 之 義。 此 可 謂知義與不義之別乎? 可,舊本作「何」。畢云:

「一本作『可』,是。」今據正。

l e s s o n 24 : m o z i

辜 gū, nc, ‘guilt’ 扡/拖 tuō, vn, ‘tear off ’ 裘 qiú, nc, ‘fur coat’ 戈 gē, nc, ‘battle axe’ 譽 yù, vn, ‘praise’ 別 bié, vi as nc, ‘difference’

Comments ⑩ This is a fine example of the convoluted lines of argument that commentaries sometimes offer. The problem is to explain the character 扡 that is usually pronounced chĭ, meaning ‘split wood’—which evidently does not make sense here. The commentary first quotes Lu Deming’s (c.550–630) commentary to the Yijing (周易音義, part of his Jingdian shiwen 經典釋文), in which he notes that one edition of the Yijing, instead of the character 褫, has 拕. Then, the character 扡, which appears in the Mozi text, is identified as a variant of 拕: hence 扡 = 拕 = 褫, which means ‘tear off (clothes)’. Moreover, Wang Niansun 王念孫 (1744–1832) and Wang Yinzhi 王引之 (1766–1834) are quoted saying that the character 也 in the Mozi sentence is redundant (衍) and has slipped into the text because the following 扡 is graphically similar. ⑪ 當此 may be translated as ‘in the face of this’; Bi Yuan adds a text critical note, stating that the character 此 is missing (脫) in older editions. ⑫ Another text critical note: do you understand it? ⑬ 從而 is perhaps best rendered as ‘accordingly’.

292

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

24.2 ag a i n s t o f f e n s i v e wa r fa r e

殺一人謂之不義,必有一死罪矣。若以此說往, 殺十人十重不義, 必有 十死罪矣;殺百人百重不義,必有百死罪矣。當此,天下之君子皆知而非之, 謂 之 不 義。 今 至 大 為 不 義 攻 國, 則 弗 知 非 ,舊本「知」作「之」,下又衍「而」字。 畢云:「一本無『而』字,是。」從而譽之,謂之義。情不知其不義也 ,王云:「情、 誠通用。」 故書其言以遺後世。若知其不義也, 夫奚說書其不義以遺後世哉? 奚 說, 言 何 辭 以 解 說 也。 畢 云「 奚 說 猶 言 何 樂 」, 失 之。 今 有 人 於 此, 少 見 黑 曰 黑, 多 見 黑 曰 白, 則 以 此 人 不 知 白 黑 之 辯 矣 ; 少 嘗 苦 曰 苦, 多 嘗 苦 曰 甘, 則 必 以 此 人 為 不 知 甘 苦 之 辯 矣。 今 小 為 非, 則 知 而 非 之。 大 為 非 攻 國, 則 不知非 ,舊本「不知」下衍「而」字,今據王、蘇校刪。 從而譽之,謂之義 。畢云:「舊 此 可 謂 知 義 與 不 義 之 辯 乎? 舊本「可」上脫「此」字,

『 之 謂 』二 字 倒, 一 本 如 此。」

是以知天下之君子也,辯義與不義之亂也。



又「謂」誤「為」。畢云:「一本作『謂』,是。」案:道藏本「可」上有「此」字,「為」正作「謂」,

今據補正。季本「謂」亦不誤。







重2 chóng, vtr, ‘multiply’ 奚 xī, print, ‘what?’ 黑 hēi, vst, ‘be black’ 白 bái, vst, ‘be white’ 辯 biàn, vn as nc, ‘distinction’





⑲ ⑰





Comments ⑭ 往 is used metaphorically here: ‘if one proceeds/goes on with this reasoning. . . ’. Note how the following line of reasoning resembles the preceding sentences: how would you explain these repetitions? ⑮ In 十重不義, 重 is probably to be construed as P, which would make 不義 an O, hence a NP. It is not unusual for negated verbs to be treated as NP; cf. 是不為也,非不能也 (Meng 1A7). For the translation, one may choose an infinitive, an -ing participle (cf. 12.3), or simply a noun like, in this case, ‘injustice’. ⑯ The commentary suggests that the

293

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 24 : m o z i

c­ haracters 情 and 誠 are interchangeable (通用), so that 情 here means ‘truly’. Compare the OC pronunciations to check whether this is plaus­ ible. ⑰ 奚 is an etymologically related synonym of 何; just like the latter, it may be used adverbially, as in 奚說 (note how the commentators struggle with this expression!). What follows, should be construed as an object clause which is not formally nominalized (perhaps to avoid a second 其?). ⑱ Observe that the vst 少 and 多, although formally in adverbial position, are logically meant as modifiers to the respective O: ‘a little white’ and ‘much white’ (a hypallage, cf. 16.4–㉑). ⑲ This sentence seems ungrammatical; how would you emend it? (As always, it helps to read the entire text before tackling single sentences.) ⑳ The editor notes the redundant (衍) character 而 in old textual witnesses and adds that he deleted it (刪). ㉑ The commentary notes that the old witnesses invert (倒) the characters 謂之, writing 之謂 instead. ㉒ The commentary notes that the old witnesses wrongly (誤) have 為 instead of 謂, and it judges the reading 謂 in one witness as ‘right’ (是). Then the readings of the edition from the Daoist canon (道藏) are quoted—­ curiously, the Mozi was included in the Daoist canon! ㉓ 是以知 introduces the conclusion which is couched in quite entangled wording. A regularly constructed object clause would run *天下之君子之亂辯義 與不義也. Here, 也 is oddly employed to stress the preceding NP, and 亂 functions as the head of what must be construed as an elaborate NP starting with 辯. This construction, albeit comprehensible, brings us to the limits of syntactical analysis. Ulrich Unger, in an impious comment, has argued that some of the Mohists, hailing from the southern state of Yue, where a non-­Chinese language was spoken, could not speak proper CC. He detects traces of their imperfect language in passages of the Mozi which appear decidedly ungrammatical. Perhaps our translations, which naturally try to make sense of obscure constructions, conceal the fact that some parts of CC literature may not be very refined at all. Cf. Unger 1998, 163–4.

24.3  Beware of ghosts (Mo 31) One of the most perplexing aspects of Mohist thought is its insistence on the existence of ghosts. However, what may seem like a relic of archaic superstition is actually an integral part of the ‘state machine’: in the absence of effective law enforcement agencies, ghosts take over the function of police, judges, and executioners. 294

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

24.3 b e wa r e o f g h o s t s

計 jì, vn, ‘calculate’ 杜, Dù, np, a place

數3 shù, nc, ‘a number (of)’ 從2 zòng, nc, ‘follower’ 滿 mǎn, vi, ‘be full’ 素 sù, vi, ‘be plain’

朱 zhū, vi, ‘be vermilion’ 弓 gōng, nc, ‘bow’

冠 guān, nc, ‘cap’ 挾 xié, vn, ‘hold’

矢 shı̌, nc, ‘arrow’

射 shè, vn, ‘shoot (arrows)’ 折 zhé, vtr, ‘break’ 脊 jı̌, nc, ‘spine’

殪 yì, vi, ‘drop dead’

弢 tāo, nc, ‘bow case’

著 zhù, vtr, ‘write down’

今執無鬼者言曰:「夫天下之為聞見鬼神之物者,不可勝計也,亦孰為聞見鬼神有無 之物哉?」 子墨子言曰:「若以眾之所同見,與眾之所同聞,則若昔者杜伯是也。 周宣王殺其臣杜伯而不辜,杜伯曰:『吾君殺我而不辜,若以死者為無知則止矣;若 死而有知,不出三年,必使吾君知之。』 其三年,周宣王合諸侯而田於圃,田車數百乘, 從數千,人滿野。日中,杜伯乘白馬素車,朱衣冠 ,朱衣冠,蓋韋弁服也。周禮 司服「凡兵事, 韋弁服」,鄭注云:「韋弁,以韎韋為弁,又以為衣裳也。」韎、朱色近通稱。執朱弓,挾朱矢,追周宣王, 射之車上,中心折脊,殪車中,伏弢而死。 當是之時,周人從者莫不見,遠者莫不聞, 著在周之《春秋》。 國語 晉語「司馬侯謂悼公曰:羊舌肸習於春秋」,韋注云「春秋,紀人事之善惡,而

追 zhuī, vn, ‘pursue’







史通 六家篇、隋書 李德林傳,並引墨子云「吾見百國春秋」,蓋即此。史通又云「汲璅冢語記太丁時事,目為

管子 法法篇云「故春秋之記,臣有弒其君,子有殺其父者矣」。尹注云「春秋,即周公之凡例,而諸侯之國史也。」

以感勸其心。」公羊莊七年傳云「不脩春秋曰:雨星不及地尺而復」,何注云「謂史記也。古者謂史記為春秋」。

目以天時,謂之春秋,周史之法也。時孔子未作春秋」。又楚語「莊王使士亹傅太子申叔,時告之曰:教之春秋,



295

勝2 shēng, vn, as adv, ‘completely’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 24 : m o z i

夏 殷 春 秋, 又 有 晉 春



秋,記獻公十七年事。」

為君者以教其 臣, 為 父 者 以 警 其 子, 曰 :『 戒 之 慎 之! 凡 殺 不 辜者,其得不祥, 鬼 神 之 誅, 若 此 之 憯 遫 也! 以 若 書 之 說 觀 之, 則 鬼 神 之 有, 豈 可 疑哉? ㉘

戒 jiè, vi, ‘guard against’ 慎 shèn, vn, ‘be careful’ 誅 zhū, vtr as nc, ‘punishment’ 憯 cǎn, vst, ‘be painful’

Comments: ㉔ The idiomatic expression 不可勝v means something ‘cannot be fully done’, in this case: ‘cannot be fully counted’, hence ‘are countless’. While the preceding construction may be challenging, the following clause is decidedly ungrammatical (cf. note above). Johnston’s translation—‘but how many [of these] can truly say that they have seen such things?’ (Johnston 2010, 281)—makes it sound much more coherent than it is. ㉕ Mozi’s answer is similarly anacoluthic. Can you pinpoint the grammatical mistakes it contains? ㉖ Try to find out, using the reference works you have come to know, when this story took place and who 杜伯 was. For the construction of the first sentence, recall 15.3–⑬. ㉗ When ‘the sun was in the centre’, that is, at noon the demonic avenger appears on a plain carriage drawn by a white horse, the colour of death, and dressed in vermilion, the colour of military dress. The symbolic meaning of these colours befits the bloody revenge that follows. ㉘ The Chunqiu of Zhou, which is the alleged source of this story, is supposed to have been some kind of historical work (not to be confused with the Chunqiu of Lu). However that may be, tales of demons and other supernatural phenomena are also recounted as facts in the Shiji, and they remained a part of Chinese historiography for many centuries. ㉙ The final sentences clarify the function of history as life’s teacher: usually, (pseudo-)historical anecdotes are recounted in Classical Chinese literature not in order to ascertain facts but for didactic reasons. Moreover, the function of ghosts becomes clear: they are agents of law enforcement, as it were. Belief in them kept people in awe even if the state was too weak to effectively punish crimes.

296

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

24.3 b e wa r e o f g h o s t s

Box 24  Adverbs of degree We have already encountered quite a few adverbs of degree, which serve to express the comparative of stative verbs (for which, of course, there were no inflectional forms). The most common of these are: 愈, as in 愈多, ‘more’. 滋/茲, as in 茲甚, ‘worse’. 益, as in 益厚, ‘heavier’. 加, as in 加多, ‘more’. 又, as in 又甚, ‘more serious’. 彌, as in 彌高, ‘higher’. Note that 更, which has an analogous function in MSC, does not belong to these adv in CC.

Moreover, the superlative is expressed with 至, as in 至大, ‘biggest’, or, like in MSC, with 最, as in 最強, ‘strongest’.

297

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 21

Parallelism O death, where is thy sting? O grave, where is thy victory?

Paral­lelismus membrorum, parallelism of constituents, is well known as the most prominent rhetorical device in the Hebrew Bible. But arguably, its importance in Chi­nese literature goes far beyond anything known from the Biblical traditions. Parallel structures appear in the bone inscriptions of Shang times; they permeate the Book of Odes; they were a defining feature of Han fu poetry and the parallel prose of the Six Dynasties period; they reached their most sublime form in Tang poetry and were taken to staggering levels of complexity in the ‘eight-­legged’ examination essays of late imperi­al times. And, yes, they are ubiquitous in classical texts. In fact, parallelism is the most important com­pos­ ition­al principle in Classical Chinese literature: it contributes decisively to the meaning of texts and thus provides the key for their proper under­standing. All ‘philosophical’ works, no matter of what ‘school’, make use of parallel construc­tions to emphasize their points: 君子和而不同, 小人同而不和。 道可道,非常道, 名可名,非常名。 愚者闇於成事, 智者見於未萌。

Gentlemen harmonize without being alike, petty men are alike without harmonizing. (LY 13.23) The Way that can be expressed is not the ­constant Way; The name that can be named is not the constant name. (Lao 1) The stupid ignore even accomplished matters, The knowledgeable perceive things even before they develop. (SJS 1)

As these examples (and their translations) indicate, the features of the Old Chinese language—preponderance of monosyllabic words, word

298

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 21 : pa r a l l e l i s m

class flexibility, optional deter­mination, and the lack of inflection— make it more favourable to parallelism than inflect­ed or agglutinative lan­guages. But that is not all. More than just rhetorical embellish­ment or mnemonic device, parallelism has been identified as ‘a central structural principle of composition’ of Chinese literature (A. Plaks), indeed a structural principle of Chinese thought. It has been said to reflect correlative cosmology and the dualistic theory of Yin and Yang, parallel pairs expressing complementary aspects of a higher unity, the Dao. Indeed, the Chunqiu fanlu says just that: ‘The hundred things all have a fitting peer: to pair and fit them together, to match and join them is very well’ (CQFL 1). Parallelism, then, implies the joining of different elements. Although some words in coupled sentences are typically identical, paral­lelismus membrorum does not mean verbatim reiteration. Rather, it combines sameness of syntactic structures with differ­ence in meaning. To be sure, syntactic structures are not always as nicely parallel as in the above examples (cf. 15.2–① for an analysis). Take, for example, 令尹誅而楚姦不上聞,仲尼賞而魯民易降北 (HFei 49), where 不上聞 and 易降北 are syntactically quite different—but apparently ­similar enough to pass as parallel. Cf. also 德蕩乎名,知出乎爭 (Zhuang 26), discussed in 8.2.3.

In analogy to Biblical literature, at least three kinds of parallelismus membrorum may be distinguished in Classical Chinese texts: (1)  Synonymous parallelism, in which the same idea is expressed twice in different words; as in: 論至德者不和於俗, 成大功者不謀於眾。

He who is concerned about the highest virtue is not in harmony with popular ideas; he who accomplishes a great work does not take counsel with the multitude. (SJS 1, tr. Duyvendak)

(2)  Antithetic parallelism, in which the second clause is opposed in meaning to the first; as in: 始吾於人也,聽其言而信其行; 今吾於人也,聽其言而觀其行。

At first, I would listen to people’s words and trust in their conduct; now I will listen to people’s words and observe their conduct. (LY 5.11)

(3)  Synthetic parallelism, in which parallel clauses add up to an overarching state­ment; as in: 不登高山,不知天之高也;

Without having climbed a high mountain, one does not realize how high the heaven is;

299

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 21 : pa r a l l e l i s m

Without having faced a deep canyon, one does not realize how deep the earth is; 不聞先王之遺言,不知學問之 Without having heard the transmitted 大也。 words of the former kings, one does not realize how great knowledge is. (Xun 1) 不臨深谿,不知地之厚也;

This third kind, which combines complementary statements, is certainly the most pervasive in Classical Chinese literature. As the example shows, it is not limited to two lines but can conjoin three or more utterances. In this way, long complex sentences (cf. 15.2), indeed entire paragraphs may be structured by parallelism. Unfortunately, these are hardly apparent in traditional editions (see, for example, 19.2 with  the above Xunzi lines). It is a good exercise to copy these texts or re-­align them ­electronically in the above fashion in order to make such structures visible. Try doing this with the entire first chapter of the Xunzi and you will realize that it consists of parallel ­sentences throughout.

Identifying such parallel structures is essential for understanding Classical Chinese texts. As noted before (cf. Review 2), parallels may provide valuable clues for syntactic analysis. Since the last in a series of parallel clauses often is the most explicit, it may indicate the construal of the preceding clauses, as well. More­over, parallel structures may provide clues for textual criticism and editing. Take the following pas­sage from Han Feizi 49: 行仁義者非所譽 . . .

Practitioners of benevolence and righteousness should not be praised . . .

工文學者非所用 . . .

Experts in literary learning should not be employed . . .

The editor, Wang Xianshen, duly notes that the character 工 is missing in his copy-­text, but convincingly justifies his editorial decision with reference to the parallelism (cf. 28.3 for the commentary). When reading the terse texts of ancient China with their scarcity of determinative elements, their deletions and allusions, the analysis of parallel structures is a crucial tool, indeed the very basis for a correct understanding and trans­lation: ‘Only when those structures, and thus their effects, are as clear as possible, can the reader presume to have a good understanding of the author’s intentions’ (Broschat 1985, 106). Students of Classical Chinese ignore them at their peril.

300

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 21 : pa r a l l e l i s m

References: Plaks 1990 and Gentz 2007 give good overviews, Broschat 1985 provides incisive argu­ments for the importance of parallel structures; Tchang Tcheng-­ming 1937 de­scribes Shijing parallelisms; Cheng 1988 discusses cosmological origins; and Gentz 2006 analyses a specific case of parallelism in syllogisms.

301

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 25

Guanzi

25.1 Introduction302

25.2 Shepherding the People (Guan 1) 303

25.1 Introduction With 135,000 characters in twenty-­four juan and eighty-­six chapters, Guanzi is one of the largest collections of Classical Chinese literature. To simply call it a ‘book’ would be misleading. Its contents are so diverse that they cannot be regarded as a homogeneous composition, much less the work of a single man. Guan Zhong 管仲, or Guan Yiwu 管夷吾 (d. 645 bc), the celebrated minister of Duke Huan of Qi, to whom the book is attributed, certainly had nothing to do with its compilation. More likely, the collection grew at the hands of anonymous contributors over the course of the entire classical period (cf. the illustration in Focus 17). Major parts of it perhaps originated in the Jixia academy in Qi, where around 300 bc, a coterie of thinkers from different traditions assembled in the service of the Qi rulers. This would explain why Guan Zhong, a towering personality in Qi history, was chosen as a patron; and it would account for the multiplicity of topics and perspectives represented in the Guanzi. Its definite redaction, however, was only done by Liu Xiang 劉向 (79–78 bc) in about 26 bc, and the standard commentary was only written in Tang times by Fang Xuanling 房玄齡 (578–648). The collection contains political and military treatises, discussions of economy, agriculture, irrigation, education, and other topics, some chapters being commentaries on others. It has variously been attributed to the Daoist and realist ‘schools’, and it is the latter with which it is usually associated nowadays. Indeed, chapters like 重令, ‘Stressing Commands’ or 法禁, ‘Norms and Prohibitions’, convey a decidedly realist message: 302

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

25.2 s h e p h e r d i n g t h e p e o p l e

‘Formerly, the sage kings, ruling the people, did not value people’s extensive learning; rather, they wanted them to harmonise, blend in and obey commands’ (Guan 5.14). However, traces of the Yin/Yang school are also present, and Confucian influences are unmistakable in many parts of the text. The following excerpt is taken from the edition Guanzi jiaozhu 管子校注. Editions: Guanzi jiaoshi 管子校釋, ed. by Yan Changyao 顏昌嶢. Changsha 1996; Guanzi jiaozhu 管子校注, ed. by Li Xiangfeng 黎翔鳳. 3 vols. Beijing 2004. Translation: Rickett 1985–98. Studies: van der Loon 1952; Rickett 1987; McNeal 2003.

25.2  Shepherding the People (Guan 1) The opening chapter is considered a prime example of ‘Confucian’ thought in Guanzi. The ‘four bonds’ it endorses—ritual behaviour, righteousness, integrity, and a sense of shame—were core values of Chinese tradition, and they are being reintroduced today in the PRC. When Xi Jinping, speaking to students and teachers at Beijing University in May 2014, quoted the words ‘四維不張,國乃滅亡’, he trusted that his audience would recognize them as a quote from Guanzi—and some of them surely also understood what they implied, for another sentence from the same chapter runs: ‘四維張則君令行’.

四時所



管子校注卷第一 唐司空房玄齡註

牧民第一

凡 有 地 牧 民 者, 務 在 四 時,

以生成萬物也。張佩綸云:本書有四時篇, 「務

在四時」,即堯典之「敬授人時」,論語之「使

民 以 時 」。 翔 鳳 案 : 形 勢 解 云 :「 春 夏 生 長,

秋 冬 收 藏, 四 時 之 節 也。」 守 在 倉 廩。 食 者,

人 之 天 也。 國 多 財 則 遠 者 來, 地 辟 舉, 盡 也。 言 地 盡 闢 則 舉 則 民 留 處,

人 留 而 安 居 處 也。 戴 望 云 : 朱 東 光 本 作「 地

舉辟則可留處」,據尹注似亦作「地舉辟」。舉、







303

牧 mù, vtr, ‘shepherd’ 守 shǒu, vn as nc, ‘protection’ 舉2 jǔ, adv, ‘completely’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi



處 為 韻, 上 下 文 皆 協 韻, 此 不 宜 獨 異。 翔 鳳 案 :「 辟 舉 」

改為「舉辟」失韻,決非管書之舊。倉廩實則知禮節, 衣食足則知榮辱,上服度則六親固, 服,行也。

上行禮度則六親各得其所,故能感恩而結固之。張守節云:



六 親 謂 外 祖 父 母 一, 父 母 二, 姨 妹 三, 妻 兄 弟 之 子 四, 從

母之子五,女之子六也。王弼云:「父母兄弟妻子也。」四 維 張 則 君 令 行。 故 省 刑 之 要, 在 禁 文 巧 ;

文巧者,刑罰所由生。豬飼彥博云:文巧,淫工也。翔鳳案:

重 令 云 :「 工 以 雕 文 刻 鏤 相 穉, 女 以 美 衣 錦 繡 綦 組 相 穉。」

「穉」訓驕,相驕則作偽而犯刑,故禁文巧,兼以省刑也。

l e s s o n 25 : g ua n z i

節 jié, nc, ‘moderation’ 榮 róng, nc, ‘honour’ 維 wéi, nc, ‘bond’ 張 zhāng, vtr, ‘stretch’ 要 yào, nc, ‘essence’ 禁 jìn, vtr, ‘forbid’

Comments ① Below the title and chapter number we find the name of the commentator to whom the edition is attributed: Fang Xuanling, who held the highly prestigious post of ‘minister of works’ (司空) under the Tang. ② Note that 在, which appears several times in this paragraph, is hardly ever used as a prep in Zhanguo texts (for an exception, cf. 15.4.2–㉚); rather, it is to be construed as P, meaning ‘lie/reside in’, or even ‘rest on’ > ‘depend on’. The discourse of the ‘seasons’ is widespread in Classical Chinese texts: the ruler is supposed to respect the agricultural seasons and the work associated with them. ③ The commentary alludes to the common saying 民以食為天. As the context shows, it is not about culinary delights but about the satisfaction of basic needs. ④ 舉 is an adverb, which should be placed between S and P. It seems misplaced here—why? What else do you notice about these four clauses? What does the commentary discuss? ⑤ Look up the OC pronunciations of 實 and 節 in the glossary: what do you notice? ⑥ Again, look up the OC pronunciations of 張 and 行: the latter may come as a surprise.

304

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

25.2 s h e p h e r d i n g t h e p e o p l e

守國之度,在飾四維; ⑦

俞樾云:禮義廉恥非由修飾, 「飾」當讀為飭。「飭」與「飾」

古通用。郭沫若云:本篇乃有韻之文。而「守國之度在飾四維」, 「維」既不與「巧」韻, 「度」

復不與「維」韻,字當有誤。疑「度」當作「癸」,「揆」之省文。「揆」亦度也。「癸」則與「維」

為句中韻矣。翔鳳案:郭疑「度」為「癸」,非也。凡古音多存於現代方音中,如江、浙呼「母」



為「嬭」,江西稱「右邊」為「以邊」,安徽呼「牛」為「疑」,皆古音也。凡顧、段之古音、

皆周代雅言,乃官音,而尚有方音之存在,則顧、段不知也。順民之經 , 俞樾云:「順」 當讀為訓。「訓民之經」,言教訓其民之道也。古順,訓通用。在明鬼神,祗山川 ,鬼神、 山川皆有尊卑之序,故敬明之。敬宗廟,恭祖舊 。謂恭承先祖之舊法。不務天時則財 不生, 不務地利則倉廩不盈 。翔鳳案:「廟」、「舊」叶,「生」、「盈」叶。野蕪曠 則民乃菅 。 菅,當為姦。安井衡云:上下韻語,此獨不韻,非體也。「菅」當為「營」,

字之誤耳。營,猶貪也。翔鳳案:楊本誤字,多為漢隸之別體,實際不誤。唐皇甫誕碑「榮」

作「荣」,上半之二「火」作「卄」。則「菅」為「營」字無疑,安井說是也。



飾/飭 chì, vtr, ‘bring in order’ 祗 zhī, vn, ‘honour’ 經 jīng, nc, ‘guideline’ 盈 yíng, vi, ‘be full’ 蕪 wú, vi, ‘be overgrown (with weeds)’ 曠 kuàng, vi, ‘be desolate’ 菅/姦 jiān, vi, ‘be treacherous’

Comments ⑦ Whereas the entire passage from 國多財 onward is rhymed, 度 and 維 do not rhyme. This is intensely discussed in the commentaries: Guo Moruo (1892–1978) suggests that 度 should read 癸, which rhymes with 維. But the editor, Li Xiangfeng 黎翔鳳 ­disagrees; he argues that although 度 and 維 did not rhyme in standard OC, they may well have done so in OC dialects: a reminder that different dialects were spoken in ancient China which are mostly not reflected in our reconstructions of OC phonology (cf. Focus 26). Li supports this ­argument with examples from some southern Chinese dialects: do you understand why he chose them? ⑧ The commentary explains 順 as 訓, 305

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 25 : g ua n z i

since the two characters were interchangeable. Note that while the semantic classifier changes, the phonetic element remains stable (cf. also the case of 說 and 悅, 17.2–②): this is typical for graphic variation. For an analysis of the NP, see the commentary. ⑨ Observe the two synonymous conj in 則民乃菅: this construction, employed here for rhythmic purposes, is possible since 則 precedes and 乃 follows the S (cf. 15.4.2– ㉝). The character 菅 is taken to stand for 營: again, it is the semantic classifier that is unstable. 上無量則民乃妄。文巧不禁則民乃淫,不璋兩原則刑乃繁 。

璋, 當 為 章。 章, 明 也。 兩 原, 謂 妄 之 原, 上 無 量 也 ; 淫 之 原, 不

禁文巧也。能明此法者則刑簡。豬飼彥博云:「璋」當作「障」,塞也。

俞樾 云:「璋 」乃「墇 」字之誤。不 明 鬼 神 則 陋 民 不 悟 ,不悟鬼

神有尊卑之異也。豬飼彥博云:「悟」音與「神」不協,義亦難通,疑

校,効也。君無所尊,人



不祗山川則威令不聞 ,言能登封降禪,祗祀山川,



當作「信」。翔鳳案:「悟」讀霸。「神」讀煞,「神」、「煞」為古今字。「神」

與「悟」韻。

則威令遠聞。不敬宗廟則民乃上校 ,

何如璋云:其文每句諧聲如頌體,故名曰「國頌」。張佩綸

亦效之。不恭祖舊則孝悌不備。四維不張,國乃滅亡。

右國頌

云:頌、容通。廣雅 釋詁:「容,灋也。」此篇乃為國之法。





量 liàng, nc, ‘measure’ 妄 wàng, vi, ‘be reckless’ 璋/彰 zhāng, vst, ‘be clear’ 繁 fán, vst, ‘be numerous’ 悟 wù, vn, ‘realize’ 校2 /效 xiào, vn, ‘imitate’ 備2 bèi, vi, ‘be complete’ 頌 sòng, nc, ‘hymn’

Comments ⑩ Another didactic verse with perfect meter (note 則 and 乃, again) and rhyme. Such sayings were very likely meant to be memorized, hence the use of rhythm and verse as mnemonic devices. Even though literacy must have been fairly widespread in elite circles, the ubiquity of rhymes in classical texts suggests that oral transmission still played an important role in classical times—and, indeed, long thereafter. In the discussion of 璋 in the commentary, 章, 障, and 墇 are 306

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

25.2 s h e p h e r d i n g t h e p e o p l e

s­ uggested as explanations—in any case, the phonetic element remains the same. ⑪ Does 明鬼神 sound familiar? Note the philological acrobatics in this commentary: since 悟 and 神 do not rhyme, the former should be read 霸 and the latter 煞—and, voilà, they rhyme. ⑫ In 上校, the nloc serves as adverbial modifier to 校, expressing the place at which the action is directed; compare 左右望, ‘look left and right’ (Lü 21.1). ⑬ 右國頌 is an editorial remark: ‘the above [lit. “to the right”] are state hymns’. This is perplexing, since 頌 designates a genre of songs represented in the last section of the Shijing; the commentary suggests the following solution: 頌 = 容 = 灋 = 法, so the above is about the norms of a state. 國有四維。一維絕則傾,二維 絶則危,三維絕則覆,四維絕 則 滅。 傾 可 正 也, 危 可 安 也, 覆可起也,滅不可復錯也。何 謂四維?一曰禮,二曰義,三 曰 廉, 四 曰 恥。 禮 不 踰 節, 義不自進,廉不蔽惡,恥不從 故 不 踰 節 則 上 位 安, 枉。 不自進則民無巧詐,不蔽惡則 行自全,不從枉則邪事不生。

右四維





覆 fù, vtr, ‘overturn’ 錯 cù, vtr, ‘settle’ 廉 lián, nc, ‘integrity’ 進 jìn, vi, ‘advance’ 蔽 bì, vtr, ‘conceal’ 枉 wǎng, nc, ‘crookery’

Comments ⑭ The idea that moral qualities, not laws or institutions, are essential for the state is typically Confucian. The ‘Four Bonds’ have been revived in the twentieth century by dictators like Yuan Shikai and Chiang Kai-­shek in the name of a ‘Confucian’ order. Indeed, their ideas of law and order may not have been so far from those of Guanzi— witness the last sentences of this paragraph. ‘Shepherding the people’ is all about controlling the people. ⑮ These, too, are definitions: how are the P to be translated?

307

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 25 : g ua n z i

Box 25  The verb 有 有 is not only one of the most frequent verbs in Classical Chinese, it is also one of the most perplexing. First of all, there is not only one verb 有 but two which should be distinguished: the ‘possessive’ vn and the ‘existential’ vi. The latter, which is much less frequent, simply states that a subject ‘exists’: 法有,民安其次。

If there is law, the people will acquiesce in their position. (SJS 20)

In most cases, 有 represents the vn which expresses that someone ‘has’ something. The subject may well be a locative subject, as in: 宋人有耕者,

Among the people of Song, there was one who plowed.

田中有株。

In the middle of the field, there was a tree stump. (HFei 49)

有齊無秦,有秦無齊。

If Qi exists, Qin cannot, and if Qin exists, Qi cannot. (SJ 69)

These constructions may be compared to French sentences with ‘il y a’ (cf. 6.2.2–⑱). If the locative subject is omitted, we get statements very similar to those with an ‘existential’ 有: Nevertheless, this is not the same construction as in the first example: there, it is the subject that exists, whereas here it is the object. Finally, ‘possessive’ 有 may be used with an object and a VP, which must probably be construed as a complement. Consider the following clauses: 人有雞犬放…

If a man has chickens or dogs that become lost. . . (Meng 6A11)

爾有母遺…

You have a mother to provide for. . . (Zuo 1.1)

These clauses are similar to constructions with 使, where the object of 使 is logically the subject of the following VP, which functions as object complement (cf. 11.3–⑭). Compare 使人來 with the following: …未嘗有顯者來。

…there has never been anybody famous to come. (Meng 4B33)

While in this clause the adverbs 未嘗 make clear that 有 must be construed as a verb, the following constructions are less unambiguous: 有朋自遠方來…

When a friend comes from afar. . . (LY 1.1)

有聖人作…

A sage appeared. . . (HFei 49)

In these cases, two construals are possible: we either have the same construction as above or 有 does not represent a verb at all but a prind, functioning as modifier to the following noun.

308

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 22

Rhymes Rhymes are universal. Correspondence of sound between the endings of words features as a stylistic device in every language and in literary works of all ages. Especial­ly in oral cultures, they perform the important function of stabilizing hallowed texts that may not be altered. Ancient China is no excep­tion: rhymes already occur in bronze inscrip­tions of the eleventh century bc, they are employed systemati­cally in the odes of the Shijing (cf. Focus 11), in the Chuci, in the Yijing, and in large parts of the Laozi. Even military texts and the supposedly dry documents of the Shujing contain rhymed pas­sages where one might not expect them; inscriptions of the first emperor of Qin had con­sistent rhymes in every stanza; and the mythical ruler Yao is quoted in impeccable verse in Lunyu 20.1: 堯曰:咨!爾舜, 天之曆數在爾躬 。

Yao said: ‘Oh! you, Shun, *kuŋ

允執其中。

the Heaven-­determined order of ­succession now rests in your person. *truŋ Sincerely hold fast the due Mean.

四海困窮,

*guŋ

天祿永終。

*tuŋ

If there shall be distress and want within the four seas, the Heavenly revenue will come to a perpetual end. (tr. James Legge)

When king Wen of Zhou asked an advisor what it meant to care for the people, the latter equally responded in verse. Although some of these rhymes are less than perfect, they are clearly meant as such: 利之而勿害,

*gâts

Benefit them and don’t harm them,

成之勿敗,

*prâts

Complete them, don’t ruin them;

生之勿殺,

*srât

Let them live, don’t kill them,

與之勿奪,

*lôt

Give unto them, don’t take from them;

309

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 22 : r hym e s

樂之勿苦,

*khâʔ

Please them, don’t embitter them,

喜之勿怒。

*nâh

Make them happy, not angry. (SY 7)

Of course, not only rulers reportedly spoke in verse, but also folk sayings, children’s songs, mnemonic verses, and quite a few other passages in Classical Chinese literature are rhymed: philosophical texts like Zhuangzi and Huainan zi, historical texts, even military texts contain verse. Readers should always reckon with rhymes, especially since they are not always as easy to identify simply by the MSC pronunciation as in the example above. Often, rhymes are quite difficult to spot, as in the ­following case: 不出戶

*gâʔ

Without leaving through the door,

知天下,

*grâʔ

One understands the realm;

不闚牖

*luʔ

Without peeking through the window,

見天道。

*lûʔ

One perceives the ways of Heaven. (Lao 47)

Although the rhymes are no longer recognizable in Modern Standard Chinese, this piece of wisdom was in verse. It is no mere coincidence that these lines reappear in other classical works (Lü 17.2, HSWZ 3 and 5, HNan 9 and 12): they were meant to be memorized and repeated. Rhymes serve as mnemonic devices not only for the revered wisdom of antiquity but also for more ordinary texts like lists and didactic sayings. Such texts have left their traces in many passages of classical literature, for example in the first chapter of Guanzi: 國多財

*dzə̂

If the state has an abundance of wealth,

則遠者來,

*rə̂

people will come from afar;

地辟舉

*klaʔ

if the land has been opened for cultivation,

則民留處,

*k-­hlaʔ *dit

they will settle down.

則知禮節,

*tsît

they will know propriety and moderation;

衣食足

*tsok

when their clothing and food is adequate,

則知榮辱,

*nok

they will know [the distinction between] honor and shame.

倉廩實

When the granaries are full,

310

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 22 : r hym e s

上服度

*dâkh

則六親固,

*kâ(k)h

If the sovereign complies with the rules [regarding proper dress and expenditure], the six relationships will be secure;

四維張

*traŋ

if the four cardinal virtues prevail,

則君令行。

*grâŋ

the prince’s orders will be carried out.

故省刑之要,

*ʔiauh

在禁文巧。

*khrûh

Therefore the essential component in reducing punishments is to prohibit luxury and artfulness. (tr. W. Allyn Rickett)

This arrangement of the text reveals a pattern that the usual layout (cf. 25.2) tends to obscure: the entire passage consists of paired lines that rhyme, it is a mnemonic verse. Indeed, several lines are repeated verbatim in Chapters 71 and 80 of the same book: they are independent ­textual units which perhaps originated not in Guanzi but in some other context. Identifying rhymes is a worthwhile exercise. It brings some life to the mute texts that usually speak to us only in graphic form; and, just like identifying parallelism, it helps to analyse the structure of classical texts. In order to fully understand Classical Chinese texts it is necessary to recognize the rhymes they contain. References: Behr 1996 discusses the origins and early developments of Chinese rhyme; Unger 1994, 140ff, gives examples of rhymes in classical texts; Long Yuchun 1962/63 lists passages from twenty-four pre-­Qin texts; Tharsen 2015, though broader in scope, contains much material concerning rhymes; for rhymes in specific text corpora, see Ma Shinian 2011, 191–203, Cook 2011, and Weingarten 2016.

311

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 26

Military texts

26.1 Introduction312 26.2 Waging War (Sun 2) 313

26.3 Planning the Attack (Sun 3)

316

26.1 Introduction The Zhanguo period was a period of intense warfare that saw the development of huge infantry armies employed in large-­scale battles and long sieges, development of new weapons and tactics, professionalisation of officers, and systematic allocation of economic resources for military purposes. As a corollary of this development, a number of texts on military theory were written: Xunzi, Guanzi, Mozi, Huainan zi, and other classical works contain chapters dealing with warfare, and many texts were entirely devoted to military matters—the Hanshu ‘Yiwenzhi’, China’s oldest bibliography, notes fifty-­three titles in the category ‘military texts’ (兵書)! Only a few of these have been transmitted, namely: • Sunzi bingfa 孫子兵法, attributed to Sun Wu 孫武 (!) of Qi (sixth/ fifth cent. bc); • Wuzi bingfa 吳子兵法, attributed to Wu Qi 吳起 (c.440–381 bc); • Sima fa 司馬法, attributed to Sima Rangju 司馬穰苴 (sixth cent. bc); • Wei Liaozi 尉繚子, attributed to Wei Liao (fourth cent. bc); • Liutao 六韜, attributed to Lü Wang 呂望 (eleventh cent. bc); • Sanlüe 三略, also attributed to Lü Wang. The so-­called ‘Seven Writings on Warfare’ (武經七書) include the above as well as the Li Wei gong wendui 李衛公問對, a work from the Tang dynasty.

312

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

26.2 wag i n g wa r

Parts of these as well as other military texts—including the long lost Sun Bin bingfa 孫臏兵法, attributed to Sun Bin (fourth cent. bc)—have been excavated in 1972 from a Western Han tomb in Shandong, testifying to the prominence of these texts in classical times. As a genre, military texts are rather heterogeneous: they contain anecdotes, aphorisms, didactic dialogues, systematic treatises, philosophical essays, etc. The most influential of these writings is Sunzi bingfa, a book in thirteen chapters attributed to Sun Wu, a general from Qi who supposedly once served the king of Wu (hence it is also called 吳孫子兵法). The book gives a systematic exposition of diverse aspects of warfare: economic measures and logistic problems, the importance of terrain and timing, intelligence, sabotage, deception, and diverse stratagems that make use of the enemy’s weaknesses. Edited by no lesser personality than Cao Cao 曹操 (155–220) in the Later Han, it has been widely read and commented upon throughout the imperial age. In the twentieth century, although it is clearly written with regard to the massive infantry warfare of Zhanguo times, Sunzi bingfa became something like a bible for guerilla warfare; indeed, its stratagems have been adapted to business, politics, private life, and other spheres, making it the second most often translated Chinese text after the Laozi. The following excerpts are taken from the edition Shiyi jia zhu Sunzi jiaoli 十一家注孫子校理. Editions: Shiyi jia zhu Sunzi jiaoli 十一家注孫子校理, ed. by Yang Bing’an 楊丙安. Beijing 1999; Sun Bin bingfa jiaoli 孫臏兵法校理, ed. by Zhang Zhenze 張震澤. Beijing 2004; Wu Sunzi fawei 吳孫子發微, ed. by Li Ling 李零. Beijing 2000, including the newly excavated manuscript texts; Bingshu sizhong (Sunzi, Wei Liao zi, Wuzi, Sima fa) zhuzi suoyin 兵書四 種 (孫子, 尉繚子, 吳子, 司馬法) 逐字索引, ed. by D.C. Lau 劉殿爵. Hongkong 1992, with texts and concordances. Translations: Sawyer 1993 contains translations of all texts mentioned above; Ames 1993 and Lau and Ames 2003 both contain newly excavated manuscript texts. See also, in French, Levi 2000, and Tang Jialong 2004; and, in German, Klöpsch 2009. Studies: Weigand 1969; Yates 2000a, 2000b. Cf. also the introductory chapters in the above-­ mentioned translations.

26.2  Waging War (Sun 2) After an introductory chapter on the importance of warfare, Sunzi ­discusses the essentials of ‘Waging War’: not armaments or tactics, but logistics. The importance attached to the economic basis of war and the 313

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 26 : m i l i ta ry t e x t s

provisions for the troops reflects the historical background of the book: it was written for the needs of large-­scale warfare as developed in Zhanguo times, in which mass infantry armies were employed on protracted campaigns. Sunzi has no illusions about the risks and dangers of such total wars.

314

  曹操曰:欲戰,必先算其費,務因糧於敵也。

馳車千駟,革車千乘,帶甲十萬 ,〇李筌曰:馳車,戰車也;



作戰篇

孫子曰:凡用兵之法:

革車,輕車也;帶甲,步卒。車一兩,駕以駟馬,步卒七十人,計千駟之軍,帶甲七萬,馬四千匹。

千里饋糧 ,曹操曰:越境千里。則內外之費,賓客之用,膠漆之材,車甲之奉, 日費千金,然後十萬之師舉矣 。〇李筌曰:夫軍出於外,則帑藏竭於內。舉千金者,言多費也。



其用戰也勝,久則鈍兵挫銳,攻城則力屈,久

〇張預曰:日費千金,師久暴,則國用豈能給?若漢武帝窮征深討,久

千里之外贏糧,則二十人奉一人也。

暴師則國用不足。

〇 張 預 曰 : 兵 已 疲 矣, 力 已 困 矣, 財 已 匱 矣, 鄰 國 因 其 罷 弊,

而不解,及其國用空虛,乃下哀痛之詔是也。夫鈍兵挫銳,屈力殫貨,則諸侯乘其弊而起,

雖 有 智 者, 不 能 善 其 後 矣 。

起兵以襲之,則縱有智能之人,亦不能防其後患。若吳伐楚入郢,久而不歸,越兵遂入吳。當是時,

雖有伍員,孫武之徒,何嘗能為善謀於後乎? 故兵聞拙速,未睹巧之久也。 〇孟氏曰:雖拙,







馳 chí, vi, ‘speed’ 駟 sì, nc, ‘team of four horses’ 革 gé, nc, ‘armour’ 饋 kuì, vn, ‘supply’ 糧 liáng, nc, ‘provisions’ 費 fèi, nc as vtr, ‘expend’ 客 kè, nc, ‘client’ 膠 jiāo, nc, ‘glue’ 漆 qī, nc, ‘varnish’ 奉 fèng, nc, ‘service’ 鈍 dùn, vi, ‘be blunt’ 挫 cuò, vtr, ‘wear away’ 銳 duì, nc, ‘lance’ 屈2 jué, vtr, ‘wear out’ 殫 dān, vtr, ‘exhaust’ 弊 bì, nc, ‘weakness’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

26.2 wag i n g wa r

言初賦民而便取勝,不復歸國發兵也。始載糧,

有以速勝。夫兵久而國利者, 未之有也。 故不盡知用兵之害者,則不能盡知用 兵 之 利 也。 〇王皙曰:久而能勝,未免 於害 ;速, 則利斯盡也。善 用 兵 者, 役 不 再 籍, 糧 不 三 載。 曹 操 曰 : 籍, 猶 賦 也。

後 遂 因 食 於 敵, 還 兵 入 國, 不 復 以 糧 迎 之 也。

取用於國,因糧於敵,故軍食可足也。

杜 佑 曰 : 兵 甲 戰 具, 取 用 國 中, 糧 食 因 敵 也。

晉師館穀於楚是也。  . . . 故智將務食於敵, 食 敵 一 鍾, 當 吾 二 十 鍾。 杜 牧 曰 : 六

石四㪷為一鍾。一石,一百二十斤。





役 yì, nc, ‘conscript’ 籍 jí, nc as vn, ‘enlist’ 將2 jiàng, nc, ‘general’ 鍾 zhōng, nc, ‘bowl’

Comments ① 凡, ‘generally’, always used in initial position, introduces general rules, namely 法, which clearly does not mean ‘law’ in this sentence. ② Recall that measures are usually placed behind the phrases they refer to (Box 13). Each quadriga, manned by elite soldiers, was accompanied by a provision cart and, as the commentary informs us, by seventy infantrymen. If that is true, the ‘leather chariots’ would seem to have required less foot soldiers, yielding a total of 100,000 troops. In early Zhanguo times, this would have been about the maximum number of forces employed in battles; in the third century bc, armies reportedly grew to twice or three times this size. ③ 金 was a monetary unit equal to twenty ‘ounces’ (兩) of coins or gold. The number ‘1,000’, just like the ‘1,000 miles’ simply indicates a great amount; Sunzi’s point is: raising an army is expensive and burdensome. ④ Especially prolonged sieges—Sunzi deals with this at length in another passage—are onerous for an army, since they inevitably entail considerable losses and costs that could mean ruin to the state. ⑤ This was the geopolitical problem for all states in Zhanguo times: the other regional lords were always lurking, ready to pounce on any weakness (note the example given in the commentary, which involves Sun Wu himself!). Military leaders had to be prepared not just for a battle, but for the bellum

315

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 26 : m i l i ta ry t e x t s

omnium contra omnes. Here, 後 may be translated as ‘consequences’. ⑥ What do you notice about the usage of 利 in 國利? ⑦ 鍾, ‘bowl’, was a capacity measure used in Qi, roughly equivalent to 128 litres. Sunzi concludes that supplying troops from home is twenty times as costly as having them live off the enemy, in other words, he encourages ‘wise generals’ to plunder the land they passed through. Indeed, the intense wars of the time must have left few alternatives to solve the crucial problem of logistics. Although little is known about looting and ­ravaging troops, they must have been a scourge for the civilian population in Zhanguo times. 26.3  Planning the Attack (Sun 3) The third chapter in Sunzi contains the famous precept that it is best not to wage war. However, far from being inspired by humaneness or pacifism, this advice follows from the fact that in Zhanguo times wars were no longer raids in order to gain booty nor vendettas pursued to restore one’s honour; rather, their aim was to occupy a territory— and this was most advantageous if it was still intact and not ravaged by war.

316

謀攻篇

孫子曰:凡用兵之法:全 國 為 上, 破 國 次 之 ; 曹操曰:

興 師 深 入 長 驅, 距 其 城 郭, 絕 其 內

外,敵舉國來服為上。以兵擊破,敗

而得之,其次也。全 軍 為 上, 破 軍 次 之 ; 曹 操、 杜 牧 曰 : 司 馬 法曰:「一萬二千五百人為軍。」全 旅 為 上, 破 旅 次 之 ; 曹操曰: 五 百 人 為 旅。全 卒 為 上, 破 卒 次 之 ; 李 筌 曰 : 百 人 已 上 為 卒。 全伍為上,破伍次之。 是



破 pò, vtr, ‘destroy’ 旅 lǚ, nc, ‘battalion’ 伍 wǔ, nc, ‘squad’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

26.3 p l a n n i n g t h e at tac k

故百戰百勝,非善之善者也;不戰而屈人之兵,善之善者也 。杜 牧 曰 : 以 計 勝 敵。 . . .  故 善 用 兵 者, 屈 人 之 兵 而 非 戰 也, 拔 人 之 城 而 非 攻 也 , 〇 孟 氏 曰 : 言 以 威 刑 服 敵, 不 攻 而 取, 若 鄭 伯 肉 袒 以 迎 楚 莊 王 之 類。 毀 人 之 國 而 非 久 也。 〇 賈 林 曰 : 兵 不 可 久, 久 則 生 變。 但 毀 滅 其 國, 不 傷 殘於人,若武王伐殷,殷人稱為父母。 必以全爭於天下,故兵不頓而 故用兵之法:十則圍之 ,〇何氏曰:

利 可 全, 此 謀 攻 之 法 也。

圍者,四面合兵以圍城。而校量彼我兵勢,將才愚智、勇怯等,而我十倍勝於敵人,

是 以 十 對 一, 可 以 圍 之, 無 令 越 逸 也。五 則 攻 之 , 〇 杜 佑 曰 : 若 敵 并 兵 自 守, 不與我戰,彼一我五,乃可攻戰也。倍 則 分 之, 敵 則 能 戰 之, 少 則 能 逃 之, 〇張預曰:彼眾我寡,宜逃去之,勿與戰,是亦為將智勇等而兵利鈍均也。 不若則能避之。故小敵之堅,大敵之擒也 。〇李筌曰:小敵不量力而

堅戰者,必為大敵所擒也。





拔 bá, vtr, ‘capture’ 毀 huı̌, vtr, ‘destroy’ 倍 bèi, vst, ‘be double’ 逃 táo, vi, ‘evade’ 擒 qín, nc, ‘prey’



Comments ⑧ The highest aim in warfare is to keep the state and the military units intact: this surely applies to the own state and military as well as to the enemy’s. ⑨ Note the construction 而非 X 也, which, despite 非 and 也, must be construed as verbal; just like 者 in 善之善者, these words seem to be used purely for emphasis. ⑩ The last sentence, which repeats the title, seems to conclude the treatment of this topic; the following section, introduced by 故用兵之法, discusses an entirely new subject. ⑪ Just as there is no morphological distinction between ordinals and cardinals (cf. Box 3), the expression of multiples (‘ten times’, ‘five times’) can only be understood from the context. Note that 敵 in this sentence does not mean ‘enemy’.

317

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 26 : m i l i ta ry t e x t s

夫將者,國之輔也,輔周則國必強,輔隙則國必弱。故君之所以患於軍者三: 梅堯 臣 曰 : 患 君 之 所 不 知。不 知 軍 之 不 可 以 進, 而 謂 之 進 ; 不 知 軍 之 不 可 以 退, 而 謂 之 退, 是謂縻軍。不知三軍之事,而同三軍之政者,則軍士惑矣。 〇杜牧曰:「蓋謂禮度法令, 自有軍法從事,若使同於尋常治國之道,則軍士生惑矣。」不知三軍之權,而同三軍之任,則軍 士 疑 矣。 〇 何 氏 曰 : 不 知 用 兵 權 謀 之 人, 用 之 為 將, 則 軍 不 治 而 士 疑。三 軍 既 惑 且 疑, 則 諸 侯 之難至矣,是謂亂軍引勝。 〇杜牧曰:言我軍疑惑,自致擾亂,如引敵人使勝我也。 故知勝 有五:知可以戰與不可以戰者勝;識眾寡之用者勝 ,李筌曰:量力也。上下同欲者勝; 以 虞 待 不 虞 者 勝 ; 〇 陳 皞 曰 : 謂 先 為 不 可 勝 之 師, 待 敵 之 可 勝 也。 〇 何 氏 曰 : 春 秋 時, 城 濮 之 後,



晉無楚備,以敗於邲。邲之後,楚無晉備,以敗於鄢。自鄢已來,晉不失備,而加之以禮,重之以睦,是

以楚弗能加晉。將能而君不御者勝。 〇杜佑曰:王子曰「指授在君,決戰在將」也。此五者,知 故 曰 : 知 彼 知 己 者, 百 戰 不 殆 ; 王 皙 曰 : 殆, 危 也。 謂 校 盡 彼 我 之 情, 知

勝 之 道 也。

勝而後戰,則百戰不危。不知彼而知己,一勝一負;不知彼,不知己,每戰必殆。



輔 fǔ, nc, ‘bulwark’ 周2 zhōu, vn, ‘be all-­ embracing’ 隙 xì, nc as vn, ‘be divided’ 縻 mí, vn, ‘fetter’ 權 quán, nc, ‘authority’ 引 yı̌n, vtr, ‘incur’ 識 shì, vn, ‘know’ 負 fù, vi, ‘be defeated’





⑫ 夫 introduces a new paragraph and a change of topic, namely the

role of the general vis-­à-­vis the ruler. The passage is remarkable. It argues that the military should be autonomous and the ruler should not interfere, since he ‘does not know’ enough about military strategy. The ‘three armies’ are those that a feudal lord had under his command, perhaps specifically: infantry, cavalry, and chariot divisions. ⑬ The word 權 means ‘balance’, and hence ‘weight, power’ here perhaps still 318

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

26.3 p l a n n i n g t h e at tac k

implies the balance of power between the ‘three armies’. Observe that some key military terms—政, 權, also 勢 —were also key terms of politics: both realms apparently evolved in close conjunction in Zhanguo times. For 引勝, see the commentary. ⑭ 故 indicates a conclusion of the preceding argument; but how many of the five factors that contribute to victory have actually been discussed before? ⑮ Again, 故 introduces a conclusion, one of the most famous lines from Sunzi. But again, it does not fit the preceding text; rather, it seems to connect to the first part of the chapter. Like so many others, this text seems to consist of snippets from different contexts that were pieced together by an unknown editor.

Box 26  Weights and measures Just like dates, weights and measures are a thorny problem for historians of ancient China. Variations between regions and through time were significant, and even the Qin standardization did not introduce anything like the degree of unity we take for granted in modern times. Like in other ancient cultures, ancient Chinese measures were based on human bodily features—the width of a thumb (寸), the span of a hand (尺), the length of a double step (步)—, utensils—a dipper (斗), a beaker (角)—or other items of everyday life. Major weights and measures, along with conventional translations and approximate values (based on the Qin system), include the following: Weight:

Length:

銖 zhū, ‘scruple’ 兩 liǎng, ‘ounce’

= 24 銖

approx. 16 g

斤 jīn, ‘catty’

= 16 兩

approx.256 g

鈞 jūn

= 30 斤

approx. 7.7 kg

石 shí, ‘picul’

=4鈞

approx. 30.7 kg

分 fēn 寸 cùn, ‘inch’ 尺 chı ̌, ‘foot’

approx. 2.3 mm = 10 分

approx.2.3 cm

= 10 寸

approx. 23 cm

丈 zhàng, ‘rod’

= 10 尺

approx. 2.3 m

步 bù, ‘pace’

=6尺

approx. 1.38 m

= 300 步

approx. 415 m

里 lı ̌, ‘mile’ Area:

approx. 0.7 g

畝 mŭ 頃 qı ̌ng

approx. 194 m2 (465 m2 in Qin/Han) = 100 畝

319

approx. 19,400 m2

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 26 : m i l i ta ry t e x t s

Capacity:

升 shēng, ‘pint’

approx. 200 ml

斗 dǒu, ‘peck’

= 10 升

approx. 2 l

斛 hú, ‘bushel’

= 10 斗

approx. 20 l

鍾 zhōng, ‘bowl’

= 640 升

approx. 128 l

While most of these measures continued to be in use throughout imperial and even modern times, their values have changed greatly. Although quite a few of the above units may seem familiar, they should not be confused with their modern equivalents. Reference: Wilkinson 2018, 609–21.

320

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 23

Schools of thought In the third century bc, ‘hundred schools’ (百家) of thought are said to have contended with one another for the patronage of regional rulers. Indeed, the rivalry of different thinkers left their traces in many classical texts: the very first paragraphs of Mengzi are devoted to indirect ­criticism of other thinkers (cf. 18.3), which becomes even more outspoken in a later chapter: ‘Mr. Yang [Zhu teaches] egoism: that is rulerless; Mr. Mo [Di teaches] universal solidarity: that is fatherless. Being rulerless and fatherless: that is like wild beasts’ (Meng 3B9). Mozi contains a chapter ‘Against Confucians’ (非儒), whom he accuses of ‘brutalizing the people in the realm’ (Mo 39). The very label Mozi uses for his rivals, ru 儒, is derogatory. It probably meant ‘weakling’ or ‘coward’—perhaps not a far-­ fetched characterization of Confucius’ bookish followers. However, the term ru shed its derisive connotations and became the established Chinese label for the people we call ‘Confucians’. This latter term is, in fact, a European creation first used by the Jesuits in Ming times.

Xunzi even devotes an entire chapter ‘Against Twelve Masters’, most of which he blames for ‘cheating, deluding, and stupefying the masses.’ The twelve are: (1) Tuo Xiao 它囂 and Wei Mou 魏牟, (2) Chen Zhong 陳仲 and Shi Qiu 史鰌, (3) Mo Di 墨翟 and Song Xing 宋鈃, (4) Shen Dao 慎到 and Tian Pian 田駢, (5) Hui Shi 惠施 and Deng Xi 鄧析, (6) Zisi 子思 and Meng Ke 孟軻 (Xun 6). This list may come as a surprise: No Daoist or military theorist is mentioned, nor do personalities like Shang Yang, Guan Zhong, Gongsun Long, and others appear among the ‘twelve masters’. Instead, people like Tuo Xiao, Wei Mou, Chen Zhong, and Shi Qiu figure prominently on the list. Apparently, Xunzi and his contemporaries had quite a different idea of who was influential in his age than historians nowadays do. Zhuangzi, too, discusses thinkers in a similar way. The text singles out the following: 321

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 23 : s c h o o l s o f t h o u g h t

(1) (2) (3) (4) (5) (6) (7)

Mo Di 墨翟 and Qin Hua-­li 禽滑釐, Song Xing 宋鈃 and Yin Wen 尹文, Peng Meng 彭蒙, Tian Pian 田駢, and Shen Dao 慎到, Guan Yin 關尹 and Lao Dan 老聃, Zhuang Zhou 莊周 himself, Hui Shi 惠施, Huan Tuan 桓團 and Gong-­sun Long 公孫龍 (Zhuang 33).

Finally, the Han Feizi discusses only two groups ‘famous for their learning’ (顯學): the Mohists and the Ruists, whom he traces back to Mo Di and Kong Qiu, respectively. After the deaths of these thinkers, Han Feizi explains, the Ruists split into eight branches, and the Mohists into three (HFei 50). Zhanguo texts present quite different ideas about the importance and association of different thinkers. Apart from the Mohists, which seem to have been rather strictly organized, and perhaps the followers of Confucius, these thinkers were not conceived of in terms of ‘schools’. It was not until Han times that these thinkers were first systematically grouped in schools (jia). Sima Tan 司馬談 (d. 110 bc), the father of Sima Qian, allegedly wrote a treatise on the ‘Six schools’, in which he distinguishes the following: (1) (2) (3) (4) (5) (6)

Yin/Yang (陰陽) Confucians (儒) Mohists (墨者) Realists (法家) Logicians (名家) Daoists (道家) (SJ 130)

This was quite a different scheme than those proposed in older texts. Here, for the first time, thinkers other than Confucians and Mohists were grouped in ‘schools’. Sima Tan’s classification has stayed with us ever since. It became the basis of bibliographies (albeit somewhat expanded) and scholarly inquiries. Modern scholars have spoken of ‘Confucians’, ‘Daoists’, or ‘Legalists’ as a matter of course, and not until recently some have pointed out how problematic this is. The label ‘Legalism’ (for 法家) is especially misleading, not only because it suggests a ‘school’ of thought where there was none, but also because the term is used in an entirely different way in the social sciences (cf. the note at 27.1). The present book therefore uses the term ‘(political) realism’, instead.

322

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 23 : s c h o o l s o f t h o u g h t

These ‘schools’, conceived by an astrologer some 2,000 years ago are not only pre-­scholarly labels that should be replaced by scholarly cat­ egor­ies, but they also imply classifications that most likely were alien to Zhanguo thinkers. The texts themselves suggest that matters were much more intricate: They reveal surprising connections between ‘realists’ and ‘Daoists’, ‘Mohists’ and ‘realists’, ‘Mohists’ and ‘Daoists’, even ‘realists’ and ‘Confucians’ on the one hand and divisions between ‘Confucians’ like Xunzi and Mengzi, ‘realists’ like Shang Yang and Han Fei or ‘Daoists’ like Laozi and Zhuangzi on the other hand. Thinking in terms of ‘schools’ of thought may be convenient at first sight, but on closer inspection, the ‘hundred schools’ may turn out to be ‘hundred thinkers’. References: Smith 2003 recounts Sima Tan’s invention of ‘schools’; Csikszentmihalyi and Nylan 2003 offer a critique of the ‘myth of pre-­Han “schools” ’; Goldin 2011 specifically ­criticizes the label ‘Legalists’.

323

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 27

Shangjun shu

27.1 Introduction324

27.2 Changing the rules (SJS 1)

325

27.1 Introduction The Shangjun shu—Writings of the Lord of Shang—is one of the most influential and yet most neglected texts in all of Chinese tradition. It is attributed to the Machiavelli of ancient China, Shang Yang 商鞅 (c.390–338 bc), a nobleman from Wei 衛, who went into the service of Duke Xiao of Qin 秦孝公 (r. 361–338 bc). As his advisor and later chancellor, Shang Yang initiated the duke into the art of hardheaded government. Qin, a semi-‘barbarian’ state in the far west of the Chinese world, had started encroaching on the ‘Middle States’; it was beginning to build up formidable economic and military power that eventually en­abled Qin to vanquish all other states and establish an empire. Shang Yang is said to have been a key figure in this process: he changed the ‘customs of the Rong and Di barbarians’ (SJ 68) that supposedly prevailed in Qin and established a centralized regime dictated by raison d’état. The ­people were organized in groups of five or ten households that were mutually responsible, subject to severe penal laws and regular taxation. Chapters like ‘Agriculture and Warfare’ (農戰), ‘Unification of Words’ (壹言), ‘Implementation of Laws’ (錯法), or ‘Rewards and Punishments’ (賞刑) testify to the cool rationale of Shang Yang’s thought, which became the basis of Chinese political realism (法家). The people were to engage in agriculture and military service only, abstaining from idleness, luxury, travel—and most of all from ‘Odes and Documents, rites and music, goodness and cultivation, benevolence and integrity, sophistry and wisdom’ (SJS 3): the pillars of Confucian teaching are 324

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

27.2 c h a n g i n g t h e r u l e s

denounced in the Shangjun shu, they had no place in the logic of state power. It is no accident that Shang Yang was despised—he is said to have been torn apart by chariots, a victim of his own cruel laws—and that the Shangjun shu has been given short shrift by Confucian scholars. While apparently widely read in Qin and Han times, there was no standard edition or commentary before Qing times. Vicissitudes of transmission have led to the loss of two of the Shangjun shu’s twenty-­six chapters, the remaining twenty-­four are riddled with corruptions, omissions, and redundancies. Only a few of them—if at all—perhaps go back to Shang Yang himself, the others were likely added by later adherents of political realism. The traditional term ‘Legalism’ for fajia is a misnomer, which nevertheless persists in scholarly literature. The word 法 did not simply mean ‘law’ in Classical Chinese; it may also refer to ‘rules’ or ‘norms’ in a broad sense (Goldin 2011). Nor were texts like Shangjun shu or Han Feizi primarily concerned with legal questions; rather, they were all about politics (Vogelsang 2016). Therefore, the more adequate term ‘(political) realism’ is used here.

The neglect of the Shangjun shu is also revealed by the facts that the first critical edition, with extensive collation notes, was published by Yan Kejun 嚴可均 (1762–1843) only in 1793, and that Sinological scholarship for decades has all but ignored the text. The following excerpts are taken from the Shangjun shu zhuizhi 商君書錐指, edited by Jiang Lihong 蔣禮鴻 (1916–95). Editions: Shangjun shu zhuizhi 商君書錐指, ed. by Jiang Lihong 蔣禮鴻. Beijing 1996; Shangjun shu jiaoshu 商君書校疏, ed. by Zhang Jue 張覺. Beijing 2012 (includes much additional material). Translations: Duyvendak 1928, solitary for more than half a century, was complemented by Levi 1981, and, more recently, by Pines 2017, and Vogelsang 2017. Studies: Vandermeersch 1965 is a classic introduction to ‘Legalism’; cf. also Kroker 1950; Li Yu-­ning 1977; King 2015; Oriens Extremus 55 (2016) contains articles by Yuri Pines and Kai Vogelsang on the SJS; the Bochum Yearbook of East Asian Studies 44 (2021) assembles papers from a conference on the Shangjun shu.

27.2  Changing the rules (SJS 1) Having suffered defeats at the hands of Zhao, Han, and Wei, Duke Xian of Qin first initiated reforms in Qin, which his successor, Duke Xiao (r. 361–338 bc) continued in earnest. Shortly after he assumed power in 361 bc, he allegedly convened three advisors to discuss these reforms. This crucial discussion, which led to the implementation of Shang Yang’s reforms, is recounted at the very beginning of the Shangjun shu. 325

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

亟2 jí, adv, ‘urgently’ 顧 gù, vn, ‘consider’ 負2 fù, vn, ‘ignore’

驁 áo, vn, ‘slander’

闇 àn, vi, ‘be ignorant’ 萌 méng, vi, ‘sprout’ 郭偃 Guō Yǎn, np

君曰: 「代立不忘社稷,君之道也。錯法務民主張,臣之行也。

公孫鞅 Gōngsūn Yāng, np



孫詒讓曰:

王與肥義趙造論胡服章文與此多同,彼曰:『嗣立不忘先德,君之道也。錯質務明主長,臣之論也。』

今吾欲變法以治,更禮以教百姓, 恐天下之議我也。」公孫鞅曰: 「臣聞之, 疑行無成,疑事無功。君亟定變法之慮,殆無顧天下之議之也。 且夫有高 人 之 行 者, 固 見 負 於 世 ; 有 獨 知 之 慮 者, 必 見 驁 於 民。 語 曰 : 『 愚 者 闇 於成事, 闇猶蔽也。成事,既成之事。知者見於未萌。民不可與慮始,而可與樂成。』 嚴 萬 里 曰 :「 舊 本 無 而 字, 成 下 有 功 字。 今 依 史 記 增 刪。」 禮 鴻 案 : 舊 本 自 可 通, 不 必 改。郭 偃之法曰: 『論至德者不和於俗,成大功者不謀於眾。』 郭偃即卜偃,見國語 晉語。 ⑤







⑧ ⑥ ④

官卒;管仲始治也,桓公有武車。戒民之備也。」商、韓並稱道郭偃,以其變法也。

易齊,郭偃毋更晉,則桓、文不霸矣。凡人難變古者,憚易民之安也。…故郭偃之始治也,文公有

韋昭注:「郭偃,晉大夫卜偃也。」法,蓋謂言之可以為法者也。又案:韓非子 南面篇:「管仲毋



326



更 gēng, vtr, ‘change’

「錯法務民主張句義殊不可通。新序 善謀篇作『錯法務明主長』,是也,當據正。戰國 趙策 趙武靈

l e s s o n 27 : s h a n g j u n s h u

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

27.2 c h a n g i n g t h e r u l e s

Comments ① The ‘lord’ is Duke Xiao of Qin, he is speaking to Shang Yang and two Qin officers. ② Two parallel sentences, both with infinitive subjects (cf. 12.3). The first recalls the duke’s obligation to the state, represented metonymically by the ‘altars of soil and millet’. The second sentence is obviously corrupt; the parallels given in the commentary may help to make sense of it. ③ Again, two parallel sentences; note that the modal auxiliary 欲 is implied also in the second sentence. For the changing meaning of 百姓, cf. 2.4.3–㊳; it is not quite clear what is meant here. ④ Shang Yang’s original name was Gongsun Yang (Gongsun, ‘duke’s grandson’, was originally a title that later became a last name; in this case, it indicates a relation to the rulers of Wei). Later, he was given an estate in Shang, from which his current name is derived. ⑤ In this case, the pr 之 does not refer to something already mentioned but to what follows, it is not anaphoric but cataphoric. ⑥ The vst 殆, ‘be uncertain, risky, dangerous’, used as adv, indicates uncertain assumptions, ‘perhaps’ or ‘should’. ⑦ 且夫, a combination of the conj 且 and the ptemp 夫, introduces and emphasizes a new argument. For the function of 見, review 4.4.2. ⑧ 語 means an anonymous ‘saying’. Such orally transmitted maxims—sometimes also called 傳, ‘tradition’, or 諺, ‘proverb’—are often quoted in classical texts to support an argument. Occasionally they are associated with some personality, as below, which, however, should not be taken as reliable information: the attribution—just like those of entire texts—merely serves to give weight to the saying. ⑨ For an analysis of 民不可與慮始, recall that 可 passivizes the following verb: 可與, then, means ‘can be joined’. The subsequent VP may be construed as a C (compare constructions like 可使從政, LY 6.8). Hence, the clause may be translated as ‘The people cannot be joined to ponder beginnings’ > ‘One cannot ponder beginnings with the people’. Shang Yang believed that the people should be kept stupid, that is, apolitical. ⑩ Another saying, attributed to a certain Guo Yan (on whom see the commentary). 法 is a rather unusual name for such a saying; it is ­reminiscent of the expression 法言, ‘model words’, which appears in several classical texts (e.g. Zhuang 4: 故法言曰) and even as the title of a book by Yang Xiong 揚雄 (53 bc–ad 18).

327

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 27 : s h a n g j u n s h u

甘龍 Gān Lóng, np 易2 yì, vtr, ‘change’ 據 jù, vn, ‘rely on’

霸 bà, nc, ‘hegemon’

不肖 bùxiào, vst, ‘be unfit’ 拘 jū, vtr, ‘cling to’

法者,所以愛民也; 禮者,所以便事也 。商君之任刑,雖毒苦其民,固猶曰吾愛民也。 學者不察,見其書或有仁義字,便謂非所宜言, 必欲去之,若遇仇讎焉,夫豈知言者哉?是 以聖人苟可以彊國,不法其故;苟可以利民,不循其禮。」 孝公曰:「善!」 甘龍曰: 史記索隱曰:「孝公之臣,甘姓,名龍也。甘氏出春秋時甘昭公 子帶之後。」 「不 然。 臣 聞 之 : 聖 人 不 易 民 而 教, 知 者 不 變 法 而 治。 因 民 而 教 者, 不 勞 而 功 成 ; 據 法 而 治 者, 吏 習 而 民 安。 此 禮 記 所 謂 禮 從 宜, 使 從 俗 也。 然 守 常 之 道耳,非救弊之用也。今若變法,不循秦國之故,更禮以教民,臣恐天下之議 君。願孰察之。」 公孫鞅曰:「子之所言,世俗之言也。夫常人安於故習, 學者溺於所聞。 此兩者,所以居官而守法,非所與論於法之外也。 三 代不同禮而王 ,嚴萬里曰:「舊本作同道,史記作同禮。案此篇禮法並舉,作道訛,今改正。」 五霸不同法而霸。 故知者作法,而愚者制焉;賢者更禮,而不肖者拘焉。 拘禮之人,不足與言事;制法之人,不足與論變。 君無疑矣。」

吏 lì, nc, ‘government official’









⑰ ⑯



328

便 biàn, vi, ‘be useful’









OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

27.2 c h a n g i n g t h e r u l e s

Comments ⑪ This seems perplexing. What is meant, apparently, is that by laws the ruler cares for the people: in the sense of controlling them and protecting them from crime. ⑫ This is a crucial turn in the history of Chinese thought: the proposal not to follow ancient custom but instead change the rules must have been revolutionary in Zhanguo times. ⑬ The ruler, who has nothing more to say than ‘good!’—an elliptical clause—is a stereotype in many anecdotes; it is the advisors who do the thinking and arguing. ⑭ Gan Long, according to a Tang (!) commentary to the Shiji, was a member of the old Qin elite: it is no surprise that he puts forth a conservative argument. ⑮ These parallel sentences may both be construed as consisting of conditional and consecutive clauses, 者 functioning not as a pt but as a conj, ‘if ’. As such, it is sometimes used instead of 則 in later classical texts (e.g. SJ 87: 地廣 者粟多,國大者人眾,兵彊則士勇). ⑯ Interestingly, the ruler has to reckon with something like public opinion—here called 天下—that might criticize his decisions. He cannot simply decide but has to justify his decisions. Arguably, this fact is at the core of all political debates documented in classical literature. ⑰ The 學者 are the arch enemies of Shang Yang: the people should stay stupid, and learning only distracts from the basic activities of agriculture and warfare. ⑱ In the case of 所 以 (beware of the faux ami: it does not mean ‘therefore’ like in MSC!), 所 is not followed by a v but by a prep, the O of which it represents; it means ‘that by which’, or, in this sentence, ‘those by/with whom’. The construction 所與 is to be analysed in the same way: ‘those with whom’. ⑲ The ‘Five Hegemons’ were the leaders of the central states in the Chunqiu period, when the power of the Zhou kings had waned; the title contrasts with that of ‘king’. ⑳ Parallel sentences in which 賢者 and 不肖者 are contrasted; both seem to have been technical terms for the evaluation of personnel. ㉑ For the analysis of 不足與言事 and 不足與論變, cf. ⑨.

329

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 27 : s h a n g j u n s h u

伏羲 Fúxī, np

神農 Shénnóng, np 黃帝 Huángdì, np

殷 Yīn, np, another name for the Shang dynasty 反 fǎn, vn, ‘oppose’

法。」顏師古注:「復,因也。」禮鴻案:復讀如复。說文:「复,行故道也。」何禮之循?





怒疑讀為孥。書 甘誓:「不用命,戮於社。予則



及至文、武,各當時而立法,因事而制禮; 禮法以時而定,制令各順其宜,兵 甲器備各便其用。臣故曰:『治世不一道,便國不必法古。』 禮鴻案:不必法古義最完 密。云法古者,正對杜摯法古無過語而發;云不必者,古固可以法,可以不法,惟便於事耳。湯、武 之王也,不脩古而興, 嚴萬里曰:「諸本及史記作循古,今據司馬貞索隱改。」王時潤曰:「脩、 循二字隸書形近,故易致誤。此當依諸本作循為是。」夏 殷之滅也,不易禮而亡。然則反古者 未必可非,循禮者未足多是也。 君無疑矣。」孝公曰:「善!」

孥戮汝。」蓋古者誅而有孥,起於夏后氏與?禮鴻案:商君行刑重其輕者,是誅而怒矣。則以時勢又變也。

神農教而不誅,黃帝、堯、舜誅而不怒。

伏羲、

杜 摯 曰 : 近 人 朱 師 轍 曰 :「 杜 摯 與 王 稽 攻 趙, 見 國 策。」 「 臣 聞 之 : 利 不 百, 不 變 法 ; 功 不十,不易器。臣聞法古無過,循禮無邪。君其圖之。」 公孫鞅曰:「前世不同 教,何故之法?帝王不相復 ,漢書 武帝紀元朔六年夏六月詔曰:「朕聞五帝不相復禮,三代不同

是2 shì, vst, ‘be right’









330

杜摯 Dù Zhì, np

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

27.2 c h a n g i n g t h e r u l e s

Comments ㉒ A second conservative advisor speaks up, but all he has to offer are stereotypes he has ‘heard’: this is perhaps what Shang Yang meant by 學者溺於所聞. ㉓ Shang Yang makes clear that his political philosophy is grounded in a critical theory of history. He holds that since things have changed fundamentally over the ages, and—contrary to what Mengzi and others claim—one cannot learn from history, much less emulate historical patterns. These patterns, he argues, have constantly changed, so: 何故之法?. . . 何禮之循?For the syntax of these rhetorical questions, cf. 14.2–⑭. ㉔ Ironically, Shang Yang refers to history—he adduces the ‘Three Sovereigns’, mythical founders of Chinese culture, and two mythical thearchs—to make his point: the only thing one can learn from history is that one cannot learn from history. ㉕ For the use of 及, cf. 16.3. Wen and Wu were the posthumous names of the first two Zhou kings. Note the effect of 各 (as opposed to 皆) in this sentence: Wen and Wu each established laws in their own specific way; cf. 7.3–⑫ ㉖ Tang and Wu were the founders of the Shang and Zhou dynasties, respectively: they indeed ‘ruled as kings’. For the analysis of this clause, note the parallel to the next one. ㉗ This is a good example for the parallelism of 可 and 足 (cf. 4.4.2); observe that and 是 are both verbs. ㉘ ‘Good!’ Shang Yang’s arguments, of course, convinced the duke, who proceeded to implement ambitious reforms in Qin.

Box 27  Conceptual history The knowledge of syntactical rules, lexical properties, scholarly commentaries and relevant dictionaries enables us to translate most Classical Chinese utterances with great confidence. Yet, even the most accomplished translations often leave us puzzled: ‘Is it not a pleasure, having learned something, to try it out at due intervals? Is it not a joy to have like-­minded friends come from afar?’ (LY 1, tr. D.C. Lau) ‘The Way that can be “Way”-ed / Is not the constant Way.’ (Lao 1, tr. A.C. Graham ) ‘Force gives birth to strength; strength gives birth to awesomeness; awesomeness gives birth to virtue; virtue is born of force.’ (SJS 5, tr. Y. Pines)

Such utterances are shrouded by an aura of mystery which makes them all the more impressive—but what do they actually mean? Evidently, mere translation does not suffice to make clear what the Chinese classics were talking about. In order to really understand what they mean, we need to clarify the concepts of ancient China: What, exactly, is a ‘friend’ in the Lunyu, and why is

331

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 27 : s h a n g j u n s h u

‘learning’ so important? What is the ‘Way’ in Laozi? What is meant by ‘strength’, ‘awesomeness’, and ‘virtue’ in the Shangjun shu? This is where the real ­interpretive work starts: with conceptual history. Just like we make sense of words in the context of a sentence, we need to make sense of concepts within discursive frameworks that evolve with social change. We need to appreciate, for example, that in Confucius’ times the emergence of a trans-­local elite society first necessitated formal learning, and that only the expansion of social contexts made it thinkable that ‘friends’ could come from ‘afar’. We need to explore how increasing social complexity gave rise to the concept of an ineffable ‘Way’ that underlay the diversity of being. And we need to analyse how the terms ‘strength’, ‘awesomeness’, and ‘virtue’ functioned within the creed of political realism, how they related to new concepts of rulership, power, state, etc. This task is much more formidable than that of linguistic analysis, and it has hardly been tackled by Sinological scholarship. But unless we gain a clear understanding of the concepts, in which the ancient Chinese described their world, their sayings will always remain shrouded by an aura of mystery. References: The potential of Conceptual history is still largely untapped in Chinese studies. For a brief introduction, cf. Vogelsang 2012; for case studies concerning ancient China, cf. Schwermann 2011; Vogelsang 2010, 2016; and Crone 2020.

332

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 24

Multiple transmissions As discussed in Focus 17, virtually all Classical Chinese books are composite texts that often appear as a pastiche of smaller textual units pieced together by anonymous contributors. Not only breaks and inconsistencies within single books testify to this process, but also the fact that many of the smaller textual units appear in different books. Thus Zuozhuan and Guoyu share some seventy anecdotes (cf. 20.3), many of which also appear in Shiji. Lüshi chunqiu, Shuoyuan, and other texts are veritable collections of anecdotes with considerable overlap. However, they do not simply repeat identical anecdotes. Rather, such multiple transmissions often display significant differences: sometimes anecdotes are placed in different contexts, sometimes elements are added, exchanged, or attributed to different characters. The Shangjun shu text discussed in 27.2 provides a striking example of this, since it has parallels in Xinxu, Shiji, and Zhanguo ce. Especially the latter’s version is remarkable: SJS 1

ZGC 19.4

孝公平畫,公孫鞅、甘龍、 杜摯三大夫御於君,慮世事 之變,討正法之本,求使民 之道。 君曰:「代立不忘社稷,君之道 也。錯法務民主長,臣之行也。

武靈王平晝間居,肥義侍坐, 曰:「王慮世者之變,權甲兵 之用,念簡、襄之跡,計胡、狄之 利乎?」

今吾欲變法以治,更禮以教百 姓,恐天下之議我也。」

今吾欲繼襄主之業,啟胡、翟之 鄉,而卒世不見也。敵弱者,用力 少而功多,可以無盡百姓之勞,而 享往古之勳。

王曰:「嗣立不忘先德,君之道 也;錯質務明主之長,臣之論 也。是以賢君靜而有道民便事之 教,東有明聲先世之功。為人臣 者,窮有弟長辭讓之節通有補民益 主之業。此兩者,君臣之分也。

333

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 24 : m u lt i p l e t r a n s m i s s i o n s

夫有高世之功者,必負遺俗之 累;有獨知之慮者,必被庶人之 恐。今吾將胡服騎射以教百姓,而 世必議寡人矣。」 公孫鞅曰:「臣聞之,疑行無成, 疑事無功。君亟定變法之慮, 殆無顧天下之議之也。且夫有高人 之行者,固見負於世;有獨知之慮 者,必見驁於民。

肥義曰:「臣聞之,疑事無功,疑 行無名。今王即定負遺俗之慮,殆 毋顧天下之議矣。

語曰:『愚者暗於成事,智者見於 未萌。民不可與慮始,可與樂成 功。』郭偃之法曰:『論至德者 不和於俗,成大功者不謀於眾。 』法者,所以愛民也;禮者,所 以便事也。是以聖人苟可以強 國,不法其故;苟可以利民,不 循於禮。」孝公曰:「善!」 . . .

夫論至德者,不和於俗;成大 功者,不謀於眾。昔舜舞有 苗,而禹袒入裸國,非以養欲而樂 志也,欲以論德而要功也。愚者闇 於成事,智者見於未萌,王其遂 行之。」

孝公曰:「善。吾聞窮巷多恡, 曲學多辨。愚者之笑,智者哀 焉;狂夫之樂,賢者喪焉。拘世 以議,寡人不之疑矣。」於是遂 出墾草令。

王曰:「寡人非疑胡服也,吾恐天 下笑之。狂夫之樂,知者哀焉;愚 者之笑,賢者戚焉。世有順我 者,則胡服之功未可知也。雖敺世 以笑我,胡地中山吾必有之。」

The Zhanguo ce relates an entirely different event, the decision of king Wuling of Zhao to introduce ‘barbarian’ trousers to his country in 307 bc. But details of the setting and many rhetorical elements of the discussion—marked in grey above—are strikingly similar to the discussion recorded in Shangjun shu which is said to have taken place more than half a century earlier in Qin. Duke Xiao begins the discussion with almost the same words as King Wuling, and Gongsun Yang gives the same advice as Fei Yi, quoting exactly the same proverbs as the latter (slight differences in wording are signs of oral transmission). Such cases of parallel transmission are puzzling: did the compilers of the Shangjun shu copy from the Zhanguo ce or vice versa? And there is yet another possibility, namely that these parallels derive from an entirely different collection, a ‘text behind the texts’, which has not come down to us. In fact, Hans Stumpfeldt has detected quite a few cases in which snippets of travel accounts, mnemonic verses, anecdote collections, 334

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 24 : m u lt i p l e t r a n s m i s s i o n s

or didactic texts are re-­embedded in classical texts. Originally meant to be memorized, such texts were widely known and freely used as commonplaces in the compilation of texts. These observations have far-­reaching consequences for the understanding of Classical Chinese literature. They show us how little we know about its composition and the original sitz im leben of its elements; and they show us how careful we should be in using these texts as sources of history. Much of what they tell us is not historical at all but simply ­commonplace. References: Stumpfeldt 2002, 2007, and 2010 are illuminating case studies; cf. also Fischer 2009 and Vogelsang 2011 for Shizi and Zuozhuan, respectively; for examples of parallels between Lunyu and excavated texts, cf. Csikszentmihalyi and Kim 2014, esp. 159–62.

335

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 28

Han Feizi

28.1 Introduction336 28.2 Past and present (HFei 49) 337

28.3 Public and private interest (HFei 49)

340

28.1 Introduction Han Feizi 韓非子 is a collection of treatises, anecdotes, and other writings attributed to Han Fei (c.280–233 bc). Han Fei was a student of Xun Kuang (cf. 19.1), but certainly no ‘Confucian’. Rather, he was a realist, and his works are widely regarded as the summa of political realism (fajia) in ancient China. While earlier thinkers like Shen Buhai or Shang Yang, so the text states, were ‘not altogether good’ regarding the theory of state power (HFei 43), Han Feizi supposedly combines all their strengths. Biographically, Han Fei has much in common with Shang Yang to whom the Shangjun shu is attributed: like the latter, he was the scion of a ruling house, Han 韓, who took up service in Qin but fell victim to an intrigue and the harsh laws of Qin. But quite unlike the Shangjun shu, Han Feizi has always been widely read and debated. Its first known commentary (now lost) was written in the fifth/sixth century, others followed in Tang times. The text’s popularity may be due to its sophisticated style, its witty anecdotes, and the sheer wealth of its subject matter. Besides treatises on political theory, typically illustrated by anecdotes, Han Feizi also contains models of social history as well as discussions of rhetoric, the rise and fall of states, and of diverse philosophical schools. The overlap with realist texts like Shangjun shu and Shenzi is significant. But Confucian teachings are also discussed, and, notably, two of Han Feizi’s sections (解老 and 喻老) are specifically devoted to commentaries on Laozi: a remarkable indication 336

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

28.2 pa s t a n d p r e s e n t

of the connections between these two schools of thought, which at first sight seem so unrelated. Just like Mozi, Han Feizi was also included in the Daoist canon. The received text, containing fifty-­five chapters in twenty juan, is considered by most scholars to derive (with the exception of just a few chapters) from Han Fei’s own hand; it may thus be dated to the second half of the third century bc. Indeed, it provides a vivid image of political and philosophical debates at the dawn of the imperial age: not only a summa of realist theory, but a high point of Chinese thought. Editions: Han Feizi jishi 韓非子集釋, ed. by Chen Qiyou 陳奇猷. Taipei 1982; Han Feizi jijie 韓非子集解, ed. by Wang Xianshen 王先慎. Beijing 1998. Translations: Liao Wenkui 1959; Watson 1964; Mögling 1994; Levi 1999. Note also the me­ticu­ lous translation by Harbsmeier in his Thesaurus Linguae Sericae. Studies: Wu Geng 1978 analyses Han Feizi’s political theory; Lundahl 1992 focuses on the structure of the text; Ma Shinian 2011 studies its genesis; Zheng Liangshu 1993 and Goldin 2013 offer comprehensive studies on different aspects of the Han Feizi.

28.2  Past and present (HFei 49) Chapter 49, ‘Five Vermin’ (五蠹), which contains one of the earliest descriptions of the evolution of political institutions, is a remarkable document of historical thought. Describing a development from ‘high antiquity’ through ‘middle antiquity’ to ‘recent antiquity’, it clearly expresses a linear—not cyclical—view of history. This view of history was new in Zhanguo times; similar accounts of social history also occur in Shangjun shu, Zhuangzi, Guanzi, Lüshi chunqiu, and other classical texts. But Han Feizi most clearly spells out the consequences that follow from this view: namely that, since circumstances change over time, one cannot take history as a model.

④ ③





337

五蠹第四十九

上 古 之 世, 人 民 少 而 禽 獸 眾, 人 民 不 勝 禽 獸 先慎曰:御覽 蟲蛇; ○ 七 十 八 引「 眾 」 作「 多 」。有 聖 人 作, 搆 木 為 巢, 以 避 羣 害, 而 民 悅 之, 使 王 天 下, 號 之 曰 有

先慎曰:各本「號」 巢氏。 ○ 下 無「 之 」 字, 御 覽 有, 依 下

文 當 有, 今 據 補。民 食 果 蓏



蠹 dù, nc, ‘vermin’ 蟲 chóng, nc, ‘insect’ 蛇 shé, nc, ‘snake’ 搆 gòu, vtr, ‘build’ 巢 cháo, nc, ‘nest’ 號 hào, vtr, ‘name’ 蓏 luǒ, nc, ‘fruit’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 28 : h a n f e i z i

蜯 蛤, 腥 臊 惡 臭 而 傷 害 腹 胃, 民 多 疾病;有聖人作,鑽燧取火,以化腥臊, 而民說之,使王天下,號之曰燧人氏。 中古之世,天下大水,而鯀、禹決瀆。 近 古 之 世, 桀、 紂 暴 亂, 而 湯、 武 征 伐。 今 有 搆 木 鑽 燧 於 夏 后 氏 之 世 者, 必為 鯀、 禹 笑矣 ; 有決瀆於殷、 周 之 世 者, 必 為 湯、 武 笑 矣。 然 則 今 有 美 堯、 舜、 湯、 武、 禹 之 道 於 當 今 之 世 者, 必 為 新 聖 笑 矣。 是 以 聖 人 不 期 脩 古, 不 法 常 可, 論 世 之 事, 因為之備。 ⑦











蜯 bàng, nc, ‘oyster’ 蛤 gé, nc, ‘mussel’ 腥 xīng, vi, ‘be rancid’ 臊 sāo, vi, ‘putrid’ 胃 wèi, nc, ‘stomach’ 鑽 zuān, vn, ‘drill’ 燧 suì, nc, ‘kindler’ 鯀 Gǔn, np 瀆 dú, nc, ‘drainage ditch’



Comments ① The concept of ‘high antiquity’—in contrast to ‘middle antiquity’ and ‘recent antiquity’—is remarkable: it implies a view of history not as a revolving, but as an evolving process in which truly new structures emerge while the old become obsolete. ② For the analysis of 有, cf. Box 25. Note that the commentary quotes the Taiping yulan 太平御覽, an encyclopedia from Song times. Many parts of classical texts, often no longer extant in the received editions, have been transmitted through quotations in encyclopedias. ③ 羣, a nc meaning ‘group, crowd’, is occasionally used as a modifier indicating the plural number (cf. Box 28). ④ The O after 使 is deleted; it is not 王. ⑤ ‘Mr. Nest-­owner’, just like the following ‘Mr. Fire drill’, is a speaking name and, of course, fictional. ⑥ The relation of two coordinated verbs as P with two coordinated nouns as O may be elegantly translated as v1 n1 v2 n2: ‘hurt their bellies and harmed their stomachs’ (rather than ‘hurt and harmed their bellies and stomachs’); cf. 4.2.2–⑪. ⑦ Gun was the mythical father of the equally mythical Yu. According to legend, he was appointed by Yao to drain the floods in northern China, but failed. The task was then transferred to his son, Yu, who gloriously succeeded. ⑧ Jie and Zhou were the reputedly cruel last rulers of the Xia and Shang, respectively, who were deposed by Tang and king Wu, the reputedly virtuous founders of the following dynasties: 湯放桀,武王伐紂, as Meng 1B15 puts it. ⑨ For this construction, recall 11.2–⑨–⑩. ⑩ This is directed against 338

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

28.2 pa s t a n d p r e s e n t

all rival philosophical schools, especially the Confucians and the Mohists, who turn to (pseudo-)historical figures to support their teachings. Like the Shangjun shu, Han Feizi takes the idea of historical change to its extreme: the only thing one can learn from history is that one cannot learn from history. ⑪ 常可 should be construed as a NP. ⑫ 因, originally a vn, ‘follow’, is also used as a conj 因, ‘consequently, thereupon’, especially in later classical texts. 宋人有耕者,田中有株,兔走觸株,折頸而死; 因 釋 其 耒 而 守 株, 冀 復 得 兔, 兔 不 可 復 得, 先慎曰:藝文類聚引「笑」上有「所」 而身為宋國笑 。 ○ 字。 今 欲 以 先 王 之 政, 治 當 世 之 民, 皆 守 株 盧文弨曰:「古」下似當分段。丈 之類也。 古者 ○ 夫 不 耕, 草 木 之 實 足 食 也 ; 婦 人 不 織, 禽 獸 之皮足衣也。 不事力而養足,人民少而財有 餘,故民不爭。是以厚賞不行,重罰不用,而 民自治。 今人有五子不為多, 子又有五子, 大 父 未 死 而 有 二 十 五 孫。 是 以 人 民 眾 而 貨 財 寡, 事 力 勞 而 供 養 薄, 故 民 爭 ; 雖 倍 賞 累 罰 而不免於亂。















株 zhū, nc, ‘tree trunk’ 兔 tù, nc, ‘rabbit’

頸 jı ň g, nc, ‘neck’ 耒 lěi, nc, ‘plough’

冀 jì, vn, ‘hope’ 類 lèi, nc, ‘category’ 丈夫 zhàngfū, nc, ‘man’ 織 zhī, vn, ‘weave’ 皮 pí, nc, ‘hide’ 大父 dàfù, nc, ‘grandfather’ 孫 sūn, nc, ‘grandchild’ 供 gōng, nc, ‘supply’ 累 lěi, vtr, ‘pile up’

⑬ 冀, unlike its synonyms 欲 or 願, is not used as a modal verb but

mostly followed by an object clause or, as in this case, an object infinitive.

⑭ Again, the same construction with 為 as above (cf. ⑨); note the variant

given in the commentary, which makes the analysis even clearer. A modern Chinese rendering of the entire anecdote is given on page xxiii. ⑮ This is the moral of the story: one cannot—like Confucians and other nostalgics—use history as a guide to the present. The reference of 皆 is  not quite clear; the preceding clause is probably to be understood as nominalized. ⑯ This description of ‘antiquity’—which actually matches the Paleolithic!—must have been purely speculative; nonetheless, it offers 339

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 28 : h a n f e i z i

a perceptive theory of the evolution of social institutions. ⑰ Rewards and punishments were the two main instruments for ordering society that realist thinkers advocated. Yet, their ultimate goal was to make these instruments superfluous. ⑱ Again, the clause preceding the cop 為 should be construed as nominalized, i.e. as a subject clause. ⑲ This is a perceptive description of Han Fei’s age: the Warring States period indeed experienced sharp population growth and increase in warfare or, as the text puts it, ‘anarchy’. This, according to Han Feizi, is the underlying reason for the evolution of statecraft. 28.3  Public and private interest (HFei 49) Another passage in the same chapter discusses the basic difference between Confucian and realist values: While the former stress family values and obedience to the parents, the latter insist on loyalty to the state and its laws. As usual, the text begins with a short treatise followed by an anecdote to illustrate the point. 儒 以 文 亂 法, 俠 以 武 犯 禁, 而 人 主 兼 禮 之, 此 所以亂也。 夫離法者罪,而諸先生以文學取; 犯 禁 者 誅, 而 羣 俠 以 私 劍 養。 故 法 之 所 非, 君 之 所 取 ; 吏 之 所 誅, 上 之 所 養 也。 法 趣 上 下 太 田 方 曰 : 趣, 當 作 取。 法, 謂「 法 之 所 四 相 反 也, ○

非 也 」。 取, 謂「 君 之 所 取 也 」。 上, 謂「 上 之 所 養 也 」。 下,

謂「吏之所誅也」。而無所定,雖有十黃帝不能治也。 王 渭 曰 : 句 絕。譽 之 則 害 故 行 仁 義 者 非 所 譽, ○ 王 渭 曰 : 為 一 句, 下 文「 非 所 用 」 句 絕。工 文 學 功; ○

先慎曰:乾道本「文」上無「工」字。有「工」 者非所用, ○ 字是。上文「行仁義者非所譽」,與「工文學者非所用」句法

一律,明此不當少一字。用之則亂法。









340

儒 rú, nc, ‘Confucian’ 俠 xiá, nc, ‘knight’ 諸2 zhū, prind, ‘all’ 先生 xiānshēng, nc, ‘tutor’ 私 sī, vi, ‘be private’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

28.3 p u b l i c a n d p r i vat e i n t e r e s t

Comments: ⑳ The text singles out two opponents of the rule by law: the ‘Confucians’ (on this appellation, cf. Focus 23) who prefer etiquette over laws, and the ‘knights’ who follow their own code of honour, violating state prohibitions (of blood feud, presumably). Though disparaged by Han Feizi, these swordsmen became immortalized in the Shiji chapter on ‘assassins’ (SJ 86) and, much later, in Wuxia literature and films. For the use of 兼, cf. Box 30. ㉑ Two nicely parallel sentences involving passive constructions; even 取, actually a vn, is being used in the passive voice: apparently, stylistic criteria here overrule lexical properties. On the pluralizing words 諸 and 羣, cf. Box 28. ㉒ The construction 法趣上 下四相反也 is tricky: the commentary by Tai Tianfang (not originally in the Han Feizi jijie) clarifies the matter. ㉓ Recall that 所 phrases sometimes express what the S should do (cf. 12.2.1–⑦). Note the pains the commentator takes to  parse the text: this is precisely what 章句 commentaries do (cf. Focus 14). 楚之有直躬,其父竊羊而謁之吏。 令尹曰:「殺之!」 以 為 直 於 君 而 曲 於 父, 報 而 罪 之。 以 是 觀 之, 夫 君 之 直臣,父之暴子也。 魯人從君戰,三戰三北。仲尼問 其故,對曰:「吾有老父,身死,莫之養也。」仲尼以 為孝,舉而上之。 以是觀之,夫父之孝子,君之背臣 也。故令尹誅而楚姦不上聞,仲尼賞而魯民易降北。 上 下 之 利 若 是 其 異 也, 而 人 主 兼 舉 匹 夫 之 行, 而 求 致 社稷之福,必不幾矣。 古者蒼頡之作書也,自環者謂

盧文弨曰:說文引作「自營為厶」,「營」、 之私,背私謂之公 , ○ 「環」本通用。「私」當作「厶」,下同。顧廣圻曰:說文又云:「公,

从八,从厶。八,猶背也。」引此曰「背厶為公」。公私之相背也,

乃蒼頡固以知之矣。 ㉙











341

直躬 Zhígōng, np 羊 yáng, nc, ‘sheep’ 令尹 lìngyı̌n, nc, ‘governor’ 北2/背 bèi, vn, ‘retreat’ 仲尼 Zhòngní, np 降2 xiáng, vi, ‘surrender’ 匹夫 pı̌fū, nc, ‘ordinary man’ 致 zhì, vn ‘bring about’ 幾 jī, vn, ‘come close to’ 蒼頡 Cāng Jié, np 環 huán, nc as vn, ‘revolve (around)’ 公2 gōng, vi, ‘ be public-­ spirited’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 28 : h a n f e i z i

㉔ Two anecdotes follow to illustrate the incompatibility of family val-

ues and loyalty to the state. Note the speaking name: ‘Zhigong from Chu’ also happens to be ‘zhi’, honest. The 之 in this clause seems redundant. ㉕ 令尹 was the highest official post in Chu, equivalent to a chancellor. His judgment of the case is quite remarkable. ㉖ Interestingly, Confucius—here called by his courtesy name, 仲尼 (cf. Focus 5)—is associated not with this but with the preceding anecdote in LY 13.18. For 莫之養, recall that 之 moves into medial position in negated sentences (cf. 5.2.3). ㉗ Han Feizi is not consistent in the use of verbs. What do you notice about 令尹誅 compared to 犯禁者誅, above? Moreover, usage of the vn 背 as mod is unusual. ㉘ Constructions like 若是其異, ‘as different as that’, are not infrequent (cf. Meng 1A7: 若是其甚, SY 2: 若是其 貧, etc.); for the analysis, cf. Box 18. ㉙ Finally, Han Feizi sums up the point: the fundamental opposition between private and public interest. The text invokes Cang Jie, the legendary creator of Chinese writing, offering a witty play on the characters 私 and 公. In order to fully understand it, see the commentary. The antagonism of 私 and 公 is a central topic in political realism. It is also explicitly dealt with in the Lüshi chunqiu (cf. 29.2), which, incidentally, offers another version of the above story with quite a different drift (Lü 11.4): 楚有直躬者,其父竊羊而謁之上,上執而將誅之。直躬者請 代之。將誅矣,告吏曰:「父竊羊而謁之,不亦信乎?父誅而 代之,不亦孝乎?信且孝而誅之,國將有不誅者乎?」荊 王聞之,乃不誅也。孔子聞之曰:「異哉直躬之為信也,一父而 載取名焉。」故直躬之信,不若無信。

Box 28  Pluralizing words In the absence of morphological distinctions, Classical Chinese makes use of several strategies to express the plural number. Coordinations of synonyms such as 朋友 may serve this function (cf. 2.2); distributives such as 皆, 莫, 孰, or 各 may indicate that the subject is in the plural (cf. 7.3); and adverbs like 兼, 盡, 悉, and others may signalize plural objects (cf. Box 30). Moreover, various modifiers may express the plural. Numerals are the most obvious example:

342

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

28.3 p u b l i c a n d p r i vat e i n t e r e s t

三人行。

Three men walk. (LY 7.22)

百工之所為

what all the artisans produce (Meng 3A4)

萬物與我

the myriad things and me (Zhuang 2)

As noted before (cf. Box 3), the numerals ‘hundred’ and ‘ten thousand’ should not be taken to indicate exact quantities; they simply mean ‘all’ of the things in question. In this respect, they are similar to other words which function as modifiers, expressing an indefinite plurality, such as 諸, 眾, 羣, and 數: 國之諸市

all markets in the capital (Zuo 10.3)

眾妙之門

the gate to all profundities (Lao 1)

羣俠

the multitude of knights (HFei 49)

數年

several years (LY 7.17)

Observe that 凡 is not a case in point, since it serves as A, not as mod (cf. 13.2–④).

343

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 25

Synonyms ‘Diction is choice of words’, the Sinologist Edward Schafer once wrote, ‘Why this word and not that word? It is necessary, then, to know exactly what the medieval poet’s words mean, as distinguished from approximate synonyms he might have used instead’ (Schafer 1990–91: 41). In other words, in order to really understand what is being said, we also need to know what is not being said. This applies not only to medieval poetry but also to classical texts. In fact, the ancient Chinese themselves were keenly aware of this matter, as some of the earliest exegetic works, the Gongyang zhuan and Guliang zhuan show. Written as commentaries to the Chunqiu (cf. Focus 10), they dissect virtually every entry in the Chunqiu by discussing why it chose this specific expression and not another: Why does it say ‘the king’s first month’? To emphasize the unification of government. What does ‘together with’ express? That there was an inner intention. Why does it not say ‘killed’? Because Duan had followers and troops. Duan was a younger brother, but the text does not call him that, he was a prince, but the text does not call him that: in order to blame him. Why does it not mention the day? In order to express detachment. Why does it give a day? Because it considers it dangerous.

While the details of these interpretations are debatable, their underlying rationale is not: The choice of words is crucial to the message of the text. Words obtain their meaning not by themselves but from their difference to other words. This obviously includes their antonyms (recall Lao 2 in 22.3), but also their synonyms, that is those words that mean nearly the same as the given word. Again, the ancient Chinese have given due attention to these. The oldest Chinese glossary, the Erya 爾雅, lists ­hundreds of synonyms in entries like the following: 初,哉,首,基,肇,祖,元,胎,俶,落,權,輿,始也。 林,烝,天,帝,皇,王,后,辟,公,侯,君也。 怡,懌,悅,欣,衎,喜,愉,豫,愷,康,妉,般,樂也。 命,令,禧,畛,祈,請,謁,訊,誥,告也。 卬,吾,台,予,朕,身,甫,余,言,我也。 

344

(Erya 1)

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 25 : s yn o nym s

Of course, none of these synonyms are exactly equal in denotation, ­connotation, and lexical properties: for example, there is an important gradation within the group of 帝, 皇, 王, 公, 侯, etc., 初, 元, 始, and the others in that group are used in different contexts, and the different properties of the pronouns 吾, 予, 我, and others have been discussed in Box 5. In fact, there are hardly any two Classical Chinese words that mean exactly the same: ‘synonyms’ are, paradoxically, not synonymous. This is exactly why the choice between them is so important. For example, 境 and 邊 both denote ‘borders’, but whereas the former means state borders, the latter means borders of cities or other smaller units within a state. Dictionaries list both 朋 and 友 as ‘friend’, but the former seems to denote ‘colleague’, while the latter implies mutual affection. The terms 惡, 姦, 邪, and others all express different kinds of moral depravity; 死, 崩, 薨, 卒, 祿, and 殪 are clearly distinct ways of saying ‘die’; just as 殺, 弒, 賊, 戮, 殘, and 滅 denote different ways of killing. These words are not used arbitrarily, their selection makes a great ­difference. Cf. the notes at 死 [78.2] and 殺 [79.6] in the glossary. While the glossary in this textbook refers to synonyms and antonyms in quite a few cases, sometimes discussing their nuances in notes, it is not systematic and far from complete in this respect.

Even the choice of ‘synonymous’ function words matters. When Mengzi answers King Xiang’s somewhat coarse question ‘天下惡乎定’, he replies: 定于一, choosing the dignified preposition 于 over the more common 於 or 乎 (cf. 8.2). This choice among synonyms is not incidental but a central part of Mengzi’s statement: it expresses subtle criticism of the king’s brusqueness. Similarly, it makes a difference whether an author chooses the formal conjunction 及 instead of the usual 與, or the antiquated preposition 用 instead of the current 以. Sometimes the choice between synonyms simply gives nuances to an utterance, sometimes it makes all the difference. But it is never trivial. That is why students of Classical Chinese should take synonyms very seriously. We will never capture the precise meaning of Classical Chinese utterances as long as we ignore what could have been said, but was not, with other words. References: The most comprehensive resource, discussing more than 4,000 words in over 1,400 synonym groups, is Wang Fengyang 1993; Hong Chengyu 2009 is good for quick reference. For the impressive case of death terminology, cf. Harbsmeier 1995, and Bao Tingyi 2006, which lists more than 10,000 words (from all periods of Chinese history) referring to death; Harbsmeier’s Thesaurus Linguae Sericae is by far the best resource for synonym groups in any Western language.

345

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 29

Lüshi chunqiu

29.1 Introduction346 29.2 Eliminate Self-­Interest (Lü 1.5) 347

29.3 Promote Farming (Lü 26.3)

351

29.1 Introduction The Lüshi chunqiu 呂氏春秋, Spring and Autumn of Mr. Lü, is a massive anthology allegedly compiled by a group of scholars under the direction of Lü Buwei 呂不韋, the chancellor of Qin, in 239 bc. Despite its title, it is not an annalistic text like the Chunqiu of Lu; nor is it purely ­historical, anecdotal, or philosophical. The Lüshi chunqiu defies classification. It consists of three discrete parts, namely twelve juan of ‘records’ (紀), each corresponding to a month of the year, eight juan of ‘surveys’ (覽), and six juan of ‘discussions’ (論). The twelve ‘records’ all start with a description of cosmological, ­natural, and ritual events within the given month. Similar calendrical treatises, called ‘monthly ordinances’ (月令) are also preserved in ­several other ancient texts: Liji, Huainan zi, and the Yi Zhoushu. They contain astronomical observations as well as notes on corresponding cultural matters, accounts of plant and animal life, the agricultural cycle, rituals, sacrifices and, importantly, the duties of the ruler. These ‘monthly ordinances’ are precursors of the Chinese agricultural calendar (農曆) and Chinese almanacs (萬年曆) that determine propitious times for travel, marriage, investments, and other activities. But the descriptions—or: prescriptions?—of the ‘monthly ordinances’ are even broader in scope. They correlate all activities of man—especially the emperor— with the course of nature, thus integrating state and society with the universe. 346

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

29.2 e l i m i n at e s e l f - i n t e r e s t

While the attention to calendrical matters may explain the term ‘chunqiu’ in the book title, there is much more to the Lüshi chunqiu. Often drawing on other texts, the Lüshi chunqiu assembles philosophical treatises, historical anecdotes, agricultural and other technical writings, as well as a wealth of miscellaneous material. An ambitious attempt to bring together the diverse traditions of Zhanguo times in a unified framework, its ultimate subject matter is the art of rulership. The Lüshi chunqiu has been called an encyclopedia of knowledge for ancient China: a universal guide for the universal empire that Qin aspired to. Indeed, its patron considered it a complete work. Lü Buwei is said to have publicly suspended a copy from the gate of the Qin capital Xianyang, offering 1,000 pieces of gold to anyone who could add or delete even a single character to improve the text. No one came forward. The reason for this reluctance was certainly not the literary perfection of the Lüshi chunqiu (in fact, recent scholarship has suggested that it is actually a fragmentary work: only the ‘records’ were probably finalized by Lü Buwei) but rather the position of its patron. Lü Buwei was not only the second-­most powerful man in Qin—an anecdote even insinuates that King Zheng, the future first emperor, was his illegitimate child—but also a representative of the hard-­nosed politics and unforgiving penal system once introduced by Shang Yang (cf. 27.1). Indeed, the text clearly has an inclination for political realism. It shares many themes and snippets with texts like Shangjun shu, Han Feizi, or Guanzi: they indicate that the dominant political creed at the dawn of the Chinese empire was not Confucianism or Daoism but political realism. The following selections are taken from Chen Qiyou’s 陳奇猷 Lüshi chunqiu jiaoshi 呂氏春秋校釋. Editions: Lüshi chunqiu zhushu 呂氏春秋注疏, ed. by Wang Liqi 王利器. 4 vols. Chengdu 2002; Lüshi chunqiu jishi 呂氏春秋集釋, ed. by Xu Weiyu 許維遹. Beijing 2009; Lüshi chunqiu jiaoshi 呂氏春秋校釋, ed. by Chen Qiyou 陳奇猷. 2 vols. Shanghai 1984. Translations: Wilhelm 1928; Kamenarovic 1998; Knoblock and Riegel 2000. Studies: Carson 1980 deals with negatives, restrictives and other syntactic features, whereas Sellmann 2002 studies calendrical and temporal matters in the Lüshi chunqiu. Shorter studies include Kalinowski 1980; Lau 1991, 1992b; Riegel 1995; and Cook 2002.

29.2  Eliminate Self-­Interest (Lü 1.5) The most urgent task for the emerging states of the Zhanguo period was to suppress the power of local aristocrats. The self-­interest of the noble houses, which ‘strive to serve their own advantage, but do not strive to 347

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 29 : lü s h i c h u n q i u

enrich the state’ (HFei 6), was highly detrimental to central rule. Therefore, the condemnation of ‘self-­interest’ (私) and the praise of ‘public interest’ (公) became a major theme of political realism. The term 私 now took on a meaning close to ‘corruption’, in short: ‘Self-­ interest plunges the empire into anarchy’ (Guan 37). Unsurprisingly, this theme also figures prominently in the first juan of the Lüshi chunqiu: the fourth chapter is entitled ‘Cherish public interest’ (貴公), and the fifth, ‘Eliminate self-­interest’ (去私).



去私

天無私覆也,地無私載也,日月無私燭也,四時無私 行也, 行其德而萬物得遂長焉。黃帝言曰:「聲禁重, 色禁重,衣禁重,香禁重,味禁重,室禁重。」

奇 猷 案 : 篇 中 稱 頌 墨 者 大 公 無 私。 考 墨 子 之 學, 以 為 虧 人 自 利

畢沅曰:舊校云:「行,一作為」。



為眾亂之源(詳墨子 兼愛、非攻),故以去私為法,則此篇出於墨家

者流之手,可以斷言也。

蘇時學曰:「行其德而萬物得遂長焉」下,忽接「黃帝言曰:聲禁重,

色禁重,衣禁重,香禁重,味禁重,室禁重」,此數語與前後文義並

不 相 蒙, 通 篇 亦 無 此 意, 蓋 必 重 己 篇 內 所 引, 而 後 人 轉 寫 錯 誤, 溷

入此篇者。







覆2 fù, vtr, ‘cover’ 燭 zhú, nc, as vn, ‘shine on’ 香 xiāng, nc, ‘fragrance’ 味 wèi, nc, ‘flavour’



Comments ① The commentary associates the chapter title with the teachings of Mozi. Do you recognize the passage referred to? Is the association convincing? ② The text begins with commonplace sayings; compare the following:

348

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

29.2 e l i m i n at e s e l f - i n t e r e s t

天無私覆,地無私載,天地豈私貧我哉。 (Zhuang 6) 聖人若天然,無私覆也;若地然,無私載也。 (Guan 37) 天公平而無私,故美惡莫不覆。地公平而無私,故小大莫不載。(Guan 64) 天地無不覆載,日月無不照明。(Wen 2) 孔子曰:天無私覆,地無私載,日月無私照。(Li 40)

The images of heaven, earth, as well as sun and moon that shine on everything was associated with a sovereign who reigns over all under heaven: it also appears on stone inscriptions of the first emperor of Qin. ③ Now the ‘Yellow Thearch’ is invoked, a mythical ruler who gained prominence in Zhanguo texts (cf. Focus 7). Of course, such quotations are not authentic; more likely, they were invented ad hoc in order to bolster a speaker’s argument. 堯 有 子 十 人, 不 與 其 子 而 授 舜 ; 舜 有 子 九 人, 不 與 其 子 而 授 禹 ; 至 公 也。 晉 平 公 問 於 祁 黃 羊 曰 :「 南 陽 無 令, 其 誰 可 而 為 之?」 祁 黃 羊 對 曰 :「 解 狐 可。」 平公曰:「解狐非子之讎邪?」對曰:「君問可,非問 臣之讎也。」平公曰:「善。」遂用之。國人稱善焉。居 有 間, 平 公 又 問 祁 黃 羊 曰 :「 國 無 尉, 其 誰 可 而 為 之?」 對 曰 :「 午 可。」 平 公 曰 :「 午 非 子 之 子 邪?」 對曰:「君問可,非問臣之子也。」平公曰:「善。」又 遂用之。國人稱善焉。孔子聞之曰:「善哉!祁黃羊之 論也,外舉不避讎,內舉不避子。」 祁黃羊可謂公矣。

高 注 : 國 語 曰 :「 舜 有 商 均 」, 此 曰「 九 子 」, 不 知 出 於 何 書 也。

高注:南陽,晉山陽 河北之邑,今河內 溫陽 樊州之屬皆是也。令,

⑤ ④







349



祁黃羊 Qí Huángyáng, np 南陽 Nányáng, np 解狐 Xiè Hú, np 讎 chóu, nc, ‘enemy’ 尉 wèi, nc, ‘commandant’ 午 Wǔ, np

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

君也。而,能。為,

治。 王 念 孫 曰 :

義, 此「 而 」 字 不

可訓為「能」。「而」

猶 以 也。 言 誰 可 以

為 之 也( 誰 可 以 為

之, 猶 言 誰 能 為

之 ; 若 云 誰 可

高注:間,

能 為 之, 則 不 辭

矣)。

頃也。



「而」「能」古雖同

l e s s o n 29 : lü s h i c h u n q i u

④ More convenient legends—note the critical commentary—to illus-

trate the virtue of being ‘public-­spirited’ as opposed to mean ‘self-­ interest’. Note that, at least in these examples, ‘self-­interest’ does not refer to individuals but to kinship groups: Yao and Shun refrained from benefitting their sons, not themselves. ⑤ Following (pseudo-) historical sequence, this last example refers to a ruler of Chunqiu times (which, however, does not make it any more authentic than the preceding ones). Two points may be noted about the duke’s question: firstly, the fact that it is not a rhetorical question despite the use of 誰 as S, and secondly the unusual subordination of 可 by 而. How do the commentators explain the latter? ⑥ 居, followed by expressions of time, means ‘occupy, take up (some time)’; just like simple 有間, it may be used as a temporal adjunct at the beginning of a sentence (cf. 10.2) in the sense of ‘after a while’. ⑦ Like in other texts (cf. 20.3), Confucius is invoked to pronounce the moral of a story—no matter whether the historical Confucius would have endorsed it. 墨者有鉅子腹䵍,居秦,其子殺人, 秦惠王曰:「先生之年長矣,非有 它子也,寡人已令吏弗誅矣,先生 之以此聽寡人也。」 腹䵍對曰: 「 墨 者 之 法 曰 :『 殺 人 者 死, 傷 人 者 刑 』, 此 所 以 禁 殺 傷 人 也。 夫 禁 殺傷人者,天下之大義也。王雖為 之賜,而令吏弗誅,腹䵍不可不行 墨者之法。」不許惠王,而遂殺之。 忍所私以行 子, 人 之 所 私 也, 大義,鉅子可謂公矣。庖人調和而 弗敢食,故可以為庖。若使庖人調 ⑨



350

墨 mò, nc, as np

鉅子 jùzı ̌, nc, ‘grandmaster’ 腹䵍 Fù Tūn, np 庖 páo, nc, ‘kitchen’ 調 tiáo, vi, ‘harmonize’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

29.3 p r o m ot e fa r m i n g

和 而 食 之, 則 不 可 以 為 庖 矣。 王 伯 之 君 亦 然, 誅 暴 而 不 私, 以 封 天 下 之 賢 者,故 可 以 為 王 伯;若使王伯之君誅暴 而 私 之, 則 亦 不 可 以 為 王伯矣。

高 注 : 私, 愛 也。



范 耕 研 曰 : 按 : 之、 其 古

通。

范 耕 研 曰 : 私, 謂 私 其 國 土

於 己 而 不 以 封 賢 者。 此 譏 始 皇

滅六國行郡縣為自私。







⑧ Next, an anecdote involving a Mohist ‘grandmaster’—a title within the Mohist group—is adduced (for the association with Mozi, cf. ①).

The last clause is problematic; note how the commentary tries to make sense of it. ⑨ Fu Tun executes his own son despite the pardon given by the ruler. Perhaps Confucius would have approved of that, too; recall LY 4.11: 君子懷刑,小人懷惠 (17.5–⑳). ⑩ An allegory caps the chapter: Just as the cook does not eat the food he prepares, so the ruler does not, for his own benefit, deplete the state he rules. In contrast to the prebendal lords of earlier times who actually owned their fiefs, the new-­style rulers of Zhanguo times only represented their states. But note the commentary that gives an entirely different interpretation. 29.3  Promote Farming (Lü 26.3) In an agricultural society like Ancient China, farming was of fundamental importance. This a point that all thinkers of ancient China agreed on. The Hanshu ‘Yiwenzhi’ lists nine works specifically devoted to agriculture, and the last four chapters of the Lüshi chunqiu also deal with farming. As the following excerpt shows, it did not promote ­farming simply a means of securing the necessities of life but also for political reasons.

351

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 29 : lü s h i c h u n q i u

上農

古先聖王之所以導其民者,先務於農。民農非徒為地利也,貴其志也。 民農則樸,樸則易用,易用則邊境安,主位尊。 民農則重,重則少私 義, 少 私 義 則 公 法 立, 力 專 一。 民 農 則 其 產 復, 其 產 復 則 重 徙, 重 徙則死處而無二慮。

夏緯瑛呂氏春秋上農等四篇校釋後記曰:呂氏春秋的上農等四篇,大致取材於后稷農書。

后稷農書既為呂氏春秋所采,它的原來著作年份,應該更早一些;但也只能視為是戰國時

譚戒甫曰:

奇猷案:商君書 農戰云:「聖人知治國之要,故令民歸心於農。歸心於農

奇猷案:「力專一」, 謂專力於農事也。



候的東西。

則民樸而可正也, 紛紛而易使也。」

季 冬 記「 水 澤 復 」, 高 注 :「 復 亦 盛 也 」。 亦 作「 腹 」, 故 月 令 作「 水 澤 腹 堅 」, 鄭 注「 腹,

厚也」。蓋復、腹、複,皆可訓厚。









樸 pǔ , vi, ‘be simple’ 邊 biān, nc, ‘border’ 專 zhuān, vn, ‘concentrate on’ 產 chǎn, vtr, as nc, ‘(agricultural) produce’







Comments ⑪ The commentary speculates that these chapters derive from writings by Hou Ji, a legendary ancestor of the Zhou who supposedly taught the people how to cultivate the land. ⑫ Political realists not only wanted the people to farm ‘for the benefits of the soil’ but they ‘valued their commitment’ to it. Tilling the land kept the people simple and easy to rule: it was ultimately an instrument of state control (cf. the SJS quotation given in the commentary). ⑬ 重, literally ‘heavy’, may perhaps best be rendered as ‘lethargic’ or ‘apathetic’ in this context. No matter whether they are awarded the epithet ‘simple’ (above), ‘stupid’

352

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

29.3 p r o m ot e fa r m i n g

(cf. 27.2), or ‘lethargic’, the rationale stays the same: Since such peasants hardly entertain a particular sense of justice, the ruler can easily impose his laws and unite. ⑭ For an explanation of 復, cf. the commentary. For 重徙, compare Lao 80: 使民重死而不遠徙 (22.5); the goal is the same in both texts: to keep the people immobile.

舍 本 而 事 末 則 不 令, 不 令 則 不 可 以 守, 不可以戰。 民舍本而事末則其產約,其產 約則輕遷徙,輕遷徙,則國家有患,皆有遠 志, 無 有 居 心。 民 舍 本 而 事 末 則 好 智, 好智則多詐,多詐則巧法令,以是為非,以 非為是。

高 注 : 令, 善。 孫 詒 讓 曰 :「 不 令 」, 謂 不 受 令 也。

奇猷案:「遠志」,

夏緯瑛曰:高注固非,孫氏以「不令」作「不受令」解,

高注:戰,攻。



與上「易用」反正相對,於文義可通,但憑空加一「受」

字,似亦未安。

高注:巧,讀如「巧智」之巧。

謂 遠 徙 他 處 之 志, 與 上 文「 重 徙 」 義 相 反, 則「 居 心 」

猶言留住其地之心。











末mò, nc, ‘nonessentials’ 遷 qiān, vi, ‘migrate’





⑮ Note the different interpretations for 令. It seems to represent the vtr ‘cause’, used in the passive voice: ‘be unmoved, unruly’. ⑯ As military terms, 守 and 戰 are antonyms, meaning ‘defend’ and ‘attack’, respectively. ⑰ This is the worst case scenario for political realists: that the people ‘take migration lightly’, become itinerant and ‘have no desire to settle’. ⑱ Another thing political realists abhor is knowledge. Cleverness is associated with deceit, which will inevitably lead to disorder: ‘The

353

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 29 : lü s h i c h u n q i u

people are hard to rule because they know too much’, as Lao 65 has it. But the root of all evil is the neglect of farming, for ‘if the fields lie fallow, falsehood will thrive among the people’ (SJS 6).

Box 29  Transcriptions of Chinese To English speakers, the umlaut in Lüshi chunqiu may seem irritating. As students of Chinese know, this is not the only problem with the Hanyu pinyin system of transcription: the ‘i’ in shi is pronounced differently than in li, the ‘u’ in chun differently than in qiu, the ‘a’ in tan differently than in yan—and the ‘x’ will forever remain a mystery. Although it has become the international standard, pinyin is not an ideal system. Nor is it the only attempt to transcribe the sounds of Chinese. Ever since the Jesuits started studying Chinese in Ming times, scholars have proposed different systems of transcription. Before Hanyu pinyin was introduced in 1958, there were literally dozens of different systems. Most of them have fallen into oblivion, but some are still used by Sinologists and publishers. Until a few decades ago, the most widespread system was not pinyin but that propagated by Thomas Wade and Herbert Giles in the ninteenth century. In Wade-­ Giles transcription, which is, among others, still used in the Cambridge History of China (without the umlaut, incidentally), Lüshi chunqiu reads Lü-­shih ch’un-­ch’iu; and the vowels in t’an and yen are nicely differentiated. Instead of ‘x’, Wade-­Giles uses ‘hs’, as in hsiao or hsieh, a reminiscence of the old distinction between dentals (*s-) and velars (*h-) in Middle Chinese (cf. Focus 2). In France, one still has to reckon with the system of the École française d’Extrême-­Orient (E.F.E.O.) created by Séraphin Couvreur in 1902 with its representation of glottal stops, as in ngan (pinyin: an), and its differentiation between dentals and velars, as in kiang vs. tsiang (pinyin both: jiang). In Taiwan, the zhuyin fuhao system (注音符號, also called bopomofo ㄅㄆㄇㄈ) from 1918, which is similar to the Japanese katakana system, is still much in use: especially in schools and as a reading aid next to Chinese characters. Moreover, the Gwoyeu Romatzyh proposed by the linguist Zhao Yuanren 趙 元任 in 1926, with its ‘tonal’ spelling is still used by some; other Sinologists, like Ulrich Unger, have devised their own systems. The following samples may serve to give an impression of the range of possible transcriptions (without tone marks):

354

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

29.3 p r o m ot e fa r m i n g













pinyin

jiang

xue

si

qing

rou

zi

Wade-­Giles

chiang

hsüeh

ssŭ

ch’ing

jou

tzŭ

E.F.E.O.

kiang

hiue

sseu

ts’ing

jeou

tseu

zhuyin

ㄐㄧㄤ

ㄒㄩㄝ



ㄑㄧㄥ

ㄖㄡ



Gwoyeu

jiang

shyue

syh

chiing

row

tzyy

Unger

kiang

hioh



ts’ing

juh

tsï

Of course, students of Classical Chinese do not need to actively use these transcriptions; but they should be prepared to read them. References: Wilkinson 2018, 61–4, gives a succinct historical overview; the monument to Wade-­ Giles transcription is Herbert Giles’ Chinese-­English Dictionary (1892); the E.F.E.O. system was first presented in the Bulletin de l'École française d'Extrême-­Orient 2 (1902); Gwoyeu Romatzyh is used in the Zhongwen da cidian (cf. Focus 16); for Unger’s system, cf. Unger 1985.

355

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 30

Zhanguo ce

30.1 Introduction356

30.2 Feng Xuan trades debt for gratitude (ZGC 11.1)

357

30.1 Introduction The image of the Warring States period—its battles and diplomacy, chivalry and ruthlessness, its brilliance and its foolishness—is perhaps best captured in the Zhanguo ce, the Stratagems of the Warring States. This collection of anecdotes was edited by Liu Xiang (cf. Focus 18) at the end of the Western Han. In his report to the throne—a rare glimpse of  early editorial practice—Liu writes that he had found numerous ­writings in the palace, which were in disorder and had been mixed up. Moreover, there were eight bundles divided up by states in a basket . . . Your servant, Xiang, thereupon adhered to those divided up by states and brought them into an approximate chronological order. I separately sorted those which did not fit into the order, to supplement the others. Having discarded redundant parts, I obtained 33 bundles. The original characters are often wrong, leaving out half of the character: thus 趙 is rendered 肖, 齊 is rendered 立, and many other characters are like this. The texts in the palace were originally called ‘Stratagems of the States’ [國策], ‘Affairs of the States’ [國事], ‘Shortcomings and Advantages’ [短長], ‘Discourses on Affairs’ [事語], ‘Advantageous Writings’ [長書], or ‘Refined Writings’ [修書], respectively. Your servant, Xiang, believes that since in the period of the Warring States, roaming knights would assist the states that employed them and devise stratagems and plans for them, it is suitable to call them ‘Stratagems of the Warring States’. (ZGC, ‘Liu Xiang shulu’)

This description tallies well with the transmitted text, which consists of thirty-­three juan, arranged by eleven different states of Zhanguo times: Eastern and Western Zhou (juan 1–2), Qin (3–7), Qi (8–13), Chu (14–17), Zhao (18–21), Wei (22–25), Han (26–28), Yan (29–31), Song (32), and Zhongshan (33). Altogether, it contains about 500 anecdotes, mostly 356

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

30.2 f e n g x ua n t r a d e s d e bt f o r g r at i t u d e

­ iscussions between rulers and/or their advisers, covering the period d from 490 bc to 221 bc. Witty and eloquent, these dialogues are certainly not to be taken as historical sources but as models of political debate. Quite a few pieces of the Zhanguo ce have become classics either in their own right or as the source of idiomatic sayings (成語); and some of its protagonists, like the brilliant rhetoricians and strategists Su Qin and Zhang Yi, have become immortalized in Chinese cultural memory. A  silk ­manuscript containing twenty-­seven pieces, eleven of which ­correspond to anecdotes in the Zhanguo ce, was found in 1973 in a Han dynasty tomb at Mawangdui, near Changsha (Hunan): it testifies to the currency of such material in early Imperial China. The following excerpts are taken from the edition Zhanguo ce, Shanghai 1985. Editions: Zhanguo ce 戰國策, ed. by Liu Xiang 劉向. 3 vols. Shanghai 1985. Translations: Crump 1996 and Bonsall 2005 are still the only translations into Western languages. Studies: Crump 1960; He Jin 2001; Friedrich 2009.

30.2  Feng Xuan trades debt for gratitude (ZGC 11.1)





357

齊人有馮諼者



鮑本「屬」,「囑」同。

鮑本草,不精也。具,饌具。



齊人有馮諼者,貧乏不能自存, 使人屬 孟嘗君,願寄食門下。 孟 嘗 君 曰 :「 客 何 好?」 曰 : 「客無好也。」曰:「客何能?」 曰 :「 客 無 能 也。」 孟 嘗 君 笑 而 受 之 曰 :「 諾。」 左 右 以 君 賤之也,食以草具。

The ‘Lord of Mengchang’, whose name was Tian Wen 田文, was a grandson of king Wei of Qi (r. 356–320). Famous for his huge retinue of some three thousand people, he figures in over a dozen anecdotes of the Zhanguo ce. One of the most popular of these is the story of the unassuming but clever Feng Xuan, perhaps the best-­known of the lord of Mengchang’s retainers. 馮諼 Féng Xuān, np 乏 fá, vi, ‘be destitute’ 屬 zhǔ vn, ‘entrust to’ 寄 jì, vi, ‘sojourn’ 門下 ménxià, nc, ‘retinue’ 諾 nuò, vn, ‘agree’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

倚 yı,̌ vn, ‘lean on’ 鋏 jiá, nc, ‘sword’

比2 bì, vn, ‘comply with’ 駕 jià, vn, ‘harness’ 揭 jiē, vtr, ‘raise’

家/姑2 gū, nc, ‘father’s sister’ 貪tān, vn, ‘be greedy’

給gěi, vi, ‘be well-­supplied’

④ ③



鮑本補曰:以下文例之,疑當有「鋏」字。



鮑本待我以客。

下客食菜。

姚本一本「客」上有「魚」字。 鮑本補曰:列士傳,孟嘗君廚有三列,上客食肉,中客食魚,

④ ③

鮑本補曰:吳士韻補,家,叶工乎反。

⑥ ⑤

358

柱 zhù, nc, ‘pillar’

居有頃,倚柱彈其劍 ,歌曰:「長鋏歸來乎!食無魚。」左右以告。孟嘗君曰: 「食之,比門下之客 。」居有頃,復彈其鋏,歌曰:「長鋏歸來乎!出無車。」 左右皆笑之,以告。孟嘗君曰:「為之駕,比門下之車客。」於是乘其車,揭 其劍,過其友曰:「孟嘗君客我 。」後有頃,復彈其劍鋏,歌曰:「長鋏歸 來乎!無以為家 。」左右皆惡之,以為貪而不知足。孟嘗君問:「馮公有親 乎?」對曰:「有老母。」孟嘗君使人給其食用,無使乏。於是馮諼不復歌。

l e s s o n 30 : z h a n g u o c e

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

30.2 f e n g x ua n t r a d e s d e bt f o r g r at i t u d e

Comments ① The courts of nobles like the lord of Mengchang also served a social function, offering destitute men a chance for employment. ② Although the commentaries do not discuss this, the sentence 左右以君賤之也 seems incorrect. Do you see the problem, and can you explain it? ③ 有頃, ‘for a while’, is another example of a temporal adjunct which is placed behind the VP it refers to (cf. 10.3 and Box 13). The following passage is a variation of the theme that songs ‘express one’s intention’ (cf. Focus 11)—and that a good ruler (or, in this case, patron) takes them seriously. ④ The commentary, in a typical sys­tem­ atiz­ing effort, proposes that there were three different classes of re­tainers, distinguished by their diet. More likely, the word ‘fish’ was simply ­chosen for the rhyme: 乎 *ɦâ—魚 *ŋa. 來 apparently functions as a hortatory pt in this clause, meaning ‘come on’ rather than ‘come’. ⑤ Next, Feng Xuan asks for a carriage, so he can brag to his old comrades. Does this line of the song rhyme? ⑥ The commentary tells us that 家 is pronounced gū. Does this make sense (consider the rhymes)? If so, how is the sentence to be understood? Note the neat conclusion that follows after this last wish is granted: the paragraph began by describing Feng Xuan as 貧乏, and it ends on 無使乏. 後 孟 嘗 君 出 記, 問 門 下 諸 客 :「 誰 習 計 會 , 能 為文收責 於薛者乎?」馮諼署曰:「能。」孟嘗 君 怪 之, 曰 :「 此 誰 也?」 左 右 曰 :「 乃 歌 夫 長 鋏歸來者也。」孟嘗君笑曰:「客果有能也 ,吾 負之,未嘗見也。」請而見之,謝曰:「文倦於事, 憒 於 憂, 而 性 懧 愚, 沉 於 國 家 之 事, 開 罪 於 先 生 。先生不羞,乃有意欲為收責於薛乎?」馮諼曰: 「願之。」於是約車治裝,載券契 而行,辭曰:「責 畢 收, 以 何 市 而 反?」 孟 嘗 君 曰 :「 視 吾 家 所 寡 有者。」

鮑本計會,會,總合也。正曰:會,古外反。













359

記 jì, vn as nc, ‘notification’ 會 kuài, nc, ‘account’ 責 zhài, nc, ‘debt’ 薛 Xuē, np, a place 署 shǔ, vn, ‘sign up’ 怪 guài, vst, ‘be strange’ 謝 xiè, vn, ‘apologize’ 倦 juàn, vi, ‘be worn out’ 憒 kuì, vi, ‘be troubled’ 懧/懧 nuò, vi, ‘be weak’ 沉 chén, vi, ‘be immersed’ 開 kāi, vtr, ‘begin’ 羞 xiū, vn, ‘be ashamed’ 裝 zhuāng, nc, ‘luggage’ 券 quàn, nc, ‘contract’ 契 qì, nc, ‘tally’ 畢 bì, vi as adv, ‘all’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 30 : z h a n g u o c e



鮑本「責」, 「債」同。集韻,

逋財也。



鮑 本 言 果, 則 孟 嘗 固 意 其

能也。



鮑本得罪於煖,自我啟之。

鮑 本 券, 亦 契。 契 別 書 之,

以刀判其旁。



⑦ The nloc 後 here appears in initial position, indicating a leap in time (cf. 10.2, note). The Lord of Mengchang does not seem to employ specialized accountants but rather assigns the job on an ad hoc basis. ⑧ Xue, a place in Qi, was the lord’s fief. Do you notice anything unusual about the question? ⑨ The Lord of Mengchang seems to have expected this turn of events, as the use of 果 suggests (cf. 4.3–⑯). ⑩ The Lord of Mengchang continues his courteous treatment of Feng Xuan with a highly self-­deprecating speech. He ‘asks’ the latter for an audience, humbly calls himself by his given name, profusely apologizes for his negligence, calls Feng Xuan ‘sir’, and admits to his ‘guilt’ (開罪 is a variant of 得罪, cf. 4.4.1, note). In a hierarchical society like that of ancient China, this must have been extremely unusual. ⑪ The debt tallies that Feng Xuan takes with him are wooden tablets, which were split into halves, each party keeping one. Upon collection of the debts, the halves were matched.

驅 而 之 薛, 使 吏 召 諸 民 當 償 者, 悉 來 合 券。 券 徧 合, 起 矯 命 以 責 賜諸民,因燒其券,民稱萬歲 。

鮑本祝孟嘗也。

長 驅 到 齊, 晨 而 求 見。 孟 嘗 君 怪 其 疾 也, 衣 冠 而 見 之, 曰 :「 責 畢 收 乎? 來 何 疾 也!」 曰 :「 收 畢 矣。」「 以 何 市 而 反 ?」 馮 諼 曰 :「 君 云『 視 吾 家 所 寡 有 者 』。 臣 竊 計, 君 宮 中 積 珍 寶, 狗 馬 實 外 廄, 美 人 充 下 陳。 君 家 所 寡 有 ⑬





360

驅 qū, vi, ‘gallop’ 償 cháng, vn, ‘pay back’ 悉 xī, vtr as adv, ‘all’ 徧/遍 biàn, vi as adv, ‘all’ 矯 jiǎo, vn, ‘falsify’ 燒 shāo, vi, ‘burn’ 到 dào, vn, ‘arrive at’ 冠2 guàn, vn, ‘put on a cap’ 積 jī, vn, ‘accumulate’ 珍 zhēn, vst, ‘be precious’ 狗 gǒu, nc, ‘dog’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

鮑本孟嘗問也。



者以義耳!竊以為君市義。」孟嘗君曰: 「市義柰何?」 曰:「今君有區區之薛, 不拊愛子其民,因而賈利之。臣竊矯君 命,以責賜諸民,因燒其券,民稱萬歲。 乃 臣 所 以 為 君 巿 義 也。」 孟 嘗 君 不 說, 曰 :「 諾, 先 生 休 矣!」 後 期 年, 齊 王 謂孟嘗君曰:「寡人不敢以先王之臣為 臣。」 孟嘗君就國於薛,未至百里,民 扶老攜幼,迎君道中。孟嘗君顧謂馮諼: 「先生所為文市義者,乃今日見之。」

30.2 f e n g x ua n t r a d e s d e bt f o r g r at i t u d e





充 chōng, vi, ‘be full’ 耳2 ěr, ptemp, ‘only’ 柰/奈 nài, vn, ‘deal with’ 區區 qūqū, vi, ‘be tiny’ 拊 fǔ, vn, ‘comfort’ 賈 gǔ, nc, ‘merchant’ 期2/朞 jī, vi, ‘be entire’ 攜 xié, vn, ‘take by the hand’ 迎 yìng, vn, ‘welcome’



⑫ Like many classical anecdotes, this one centres around an ambiguous

expression (‘something which is scarce in my house’): Feng Xuan takes the opportunity to distort his lord’s command and not collect anything material at all. ⑬ Another comical scene: Feng Xuan wakes his lord up at the crack of dawn only to tell him he burned his tallies. Note the ­construction 來何疾也, which may be construed as a case of post-­ modification (cf. Box 13). ⑭ There are some oddities in Feng Xuan’s language: in both 以義耳 and 竊以為君市義, the prep 以 seems redundant. Moreover, it is by no means clear what Feng Xuan means by 義. Surely, the term cannot mean ‘righteousness’ here, since cancelling the debts was an act of pure generosity. The precise meaning of many ancient Chinese concepts and their changes over time are still largely unexplored. 柰何 is a late classical variant of 如之何 or 若之何 (cf. 18.3–⑮). ⑮ This scene must be set in 319 bc, when king Wei of Qi, lord Menchang’s grandfather, had died and his successor, king Xuan of Qi, acceded the throne. Since he did not grant lord Mengchang an office, the latter had to leave the capital and retire to his fief. ⑯ In Xue, lord Mengchang finds a safe haven thanks to Feng Xuan’s stratagem. However, this is not the end of the story. Feng Xuan informs his patron that ‘a clever rabbit has three burrows’ (狡兔有三窟)—a well-­known idiomatic saying until today—and goes on to restore him to the position of minister in Qi.

361

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 30 : z h a n g u o c e

Box 30  Quantifying objects As discussed in 7.3, distributive adverbs serve to indicate which parts of the subject—‘all’, ‘some’, etc.—relate to the predicate. Similarly, there are several CC words that may be used adverbially to indicate that ‘all’ objects are affected by the action expressed by the P. These include 兼, 遍, 周, 盡, and 悉, which regularly function as adverbs: 臣願悉言所聞。

I wish to tell everything I have heard. (HFei 1)

晉侯兼享之。

The marquis of Jin entertained them all. (Zuo 9.26)

繆公乃求酒,遍飲之。 Duke Mu ordered ale and gave them all to drink. (HSWZ 10) 晉文公…周流天下。 盡食其肉。

Duke Wen of Jin . . . wandered all over the realm. (SY 6) ⟨He⟩ ate all his flesh. (HSWZ 7)

Despite this common function, these words are not all adverbs but belong to different word classes. Hence, unlike distributives, they do not constitute a coherent group. Reference: Harbsmeier 1981: 49–77.

362

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 26

Dialects The concept of ‘Classical Chinese’, a scholarly construction of the nineteenth century, is actually a bold simplification of the linguistic diversity that existed in ancient China. In the preceding lessons, we have already encoutered some traces of this diversity. Recall the somewhat peculiar speech of the wheelwright in Zhuang 13 (23.3) or of the retainer Feng Xuan (30.2), which may be explained as sociolects, i.e. forms of speech limited to socially defined groups. Recall also words like 斯 (instead of 則 or 此) or the conjunction 如 (instead of 若), which are characteristic of Lu texts like Lunyu and Mengzi: these are indicators of dialects, i.e. forms of speech used exclusively by certain regional communities of speakers. Finally, consider Ulrich Unger’s suggestion that some of the Mohists, who came from Yue, could not speak proper Classical Chinese (cf. 24.2 note): this may be a case of ‘pidgin’ Chinese spoken by people whose mother tongue was not Chinese. Such cases may have been quite common. After all, the ‘middle states’ and the Yi, Man, Rong, and Di peoples surrounding them had ‘mutually unintelligible languages and different predilections’ so that they needed translators—which, incidentally, were called differently in every region: ‘in the east they are called “transmitters” [寄], in the south “symbolizers” [象], in the west “Di experts” [狄鞮], and in the north “translators” [譯]’ (Li 5) or, according to another account, ‘men of tongues’ (舌人, GY 2.7). But these ‘barbarians’ were not the only ones to depend on translators when dealing with the Chinese. The people of Chu, Wu, and Yue, who became part of the Chinese interstate system in Chunqiu times, also spoke non-­Chinese languages: ‘Qi and Wu have different customs, and their languages are mutually unintelligible’ (Lü 23.3). The languages of these polities south of the Yangzi were probably closer to Thai languages than to Old Chinese. Consider the song of some boatsmen from Yue, one of the few specimens of Yue language that have been transmitted, which goes like this:

363

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 26 : d i a l e c t s

濫兮抃草濫予昌枑 澤予昌州州𩜱州焉 乎秦胥胥縵予乎昭 澶秦踰滲惿隨河湖 (SY 11) The Chinese writing system makes this song look deceptively familiar. But the sound and meaning of the words behind these characters escape us: this is certainly not Classical Chinese! Even a Chu dignitary, upon hearing this song, had to ask for a translation into Chu speech. These were not just different dialects but genuinely different, ‘mutually unintelligible languages’. But even within the core of the Chinese realm there were significant dialectal differences. It was a commonplace that ‘every hundred Chinese miles, there will be different habits, every thousand Chinese miles, there will be divergent customs’ (Yan 3.18)—and surely different idioms. While not being ‘mutually unintelligible’, these dialects could certainly pose problems. For example, when the army of Qin encountered the men of Wei at the Huanghe, they needed ‘men from the east, who can talk to the leaders’ of Wei in order to communicate with them (Zuo 6.13). And in the following anecdote a faux ami caused some confusion even among close neighbours: In Zheng, the people call an untreated piece of jade ‘pu’, and in Zhou, they call a rat which has not yet been cured, ‘pu’. When a man from Zhou, carrying uncured rats, passed by a merchant from Zheng, he asked the latter whether he would like to buy some ‘pu’. The merchant from Zheng replied that he did—but when the man from Zhou showed him his ‘pu’ and he saw that they were rats, he declined them. (ZGC 5.12)

While hardly anything is known about phonological or syntactical characteristics of these dialects, we possess comprehensive evidence on their lexicon. Some 2,300 dialect words were systematically compiled by Yang Xiong 揚雄 (53 bc–ad 18) at the end of the classical period in China’s first dialect dictionary, the Fangyan 方言. A typical entry reads as follows: 嫁、逝、徂、適,往也。 自家而出謂之嫁, 由女而出為嫁也。

*krâh, *dats, *dzâ, and *lhek mean *waŋʔ, ‘to go’. To depart from the family is called *krâh, ‘marry’; for a girl to depart is to marry.

364

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 26 : d i a l e c t s

逝,秦晉語也。徂,齊語也。 *dats is the language of Qin/Jin; *dzâ is the language of Qi; 適,宋魯語也。往,凡語也。 *lhek is the language of Song/Lu; *waŋʔ is common language. (Fangyan 1) Note that the reconstructions given are of OC. They represent the way Yang Xiong and his peers may have pronounced these words, not the way they were pronounced in the ­respective dialects.

Besides distinguishing between two dozen dialect areas like Qin/Jin, Song/Lu, and Qi, the Fangyan regularly refers to a ‘common language’. Indeed, such a ‘common language’ may have evolved as early as Chunqiu times. With growing interregional communication, there seems to have developed a lingua franca of sorts, a ‘polite speech’ (yayan 雅言) spoken by the elites of Zhou, Zheng, Chen, Song, Lu, Qi, and other states of the north China plain. Confucius, so we are informed in LY 7.18, used ‘polite speech’ when talking about the ‘Odes’, the ‘Documents’, and the conduct of rituals (which would seem to imply that otherwise he did not speak ‘polite speech’ but Lu dialect). This elite common language is probably responsible for the relative linguistic homogeneity of Classical Chinese texts. Yet, this should not obscure the fact that ‘Classical Chinese’ is not simply ‘polite speech’ but a mixtum compositum, containing elements of diverse sociolects, dialects, and even non-­Chinese languages. References: Serruys 1959 is a classic study of the Fangyan; Hua Xuecheng 2007 has comprehensive material on dialects and their study; for two specific dialects, cf. Wang Qiming 1998 and Zhao Tong 2006.

365

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 31

Shiji

31.1 Introduction366

31.2 Biography of Confucius (SJ 47)

368

31.1 Introduction The Shiji 史記, written by Sima Qian 司馬遷 (c.145–86 bc), is the texte fondateur of Chinese historical writing. In fact, the Shiji became eponymous for history itself. Its original title was Taishi gong shu 太史公書, Writings of His Excellency, the Grand Astrologer, which was Sima Qian’s office at the Han court. Accordingly, the title that later gained currency, Shiji, means Records of the Astrologer. The word shi 史 eventually took on the meaning ‘historical writing, history’ because this grand astrologer happened to write China’s greatest work of history. To be precise, work on the Shiji was begun by Sima Qian’s father, Sima Tan 司馬談 (d. 110 bc). Sima Qian continued his father’s enterprise under the most adverse circumstances, having suffered the punishment of castration at the behest of emperor Wu for speaking in defence of a general who had surrendered to the Xiongnu. After Sima Qian’s death, further parts were added by a certain Chu Shaosun 褚少孫 and others. Containing half a million characters in 130 juan, the Shiji is a vast compendium of Chinese history from the legendary Yellow Thearch down to Sima Qian’s own times. In compiling this universal history, Sima Qian drew on older historical writings such as Shangshu, Zuozhuan, Guoyu, Zhanguo ce, as well as the Classical philosophical writings and archival material available at the Han court. Weaving these sources together, Sima Qian arranged his material in five sections:

366

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

31.1 i n t r o d u c t i o n

本紀, ‘basic annals’ (twelve juan): the histories of successive ruling dynasties from the mythical thearchs to the Qin; 表, ‘tables’ (ten juan): chronological concordances of rulers and events in pre-­imperial regional states as well as Han aristocratic families; 書, ‘writings’ (eight juan): treatises on ritual, music, the calendar, sacrifices, waterways, economy, and other important subjects; 世家, ‘hereditary houses’ (thirty juan): mainly the histories of the regional states in pre-­imperial times; 傳,‘traditions’ (seventy juan): mainly biographies of eminent ­personalities in pre-­imperial and imperial times as well as descriptions of hermits, assassins, foreign peoples, and other groups. The last chapter is an autobiographical postface by Sima Qian.

This arrangement, slightly altered, became the model for dynastic histories throughout the next two millennia. Moreover, the Shiji became the prime resource for the history of ancient China: its wide scope and literary brilliance have continued to fascinate historians until today. However, it should be borne in mind that the Shiji is not an historical source, but an account that was created decades, centuries, even millennia after the events it narrates. Nor should it be mistaken for an ob­ject­ ive, critical history in the modern sense. This was not Sima Qian’s intention. Rather, he bestowed praise and blame on historical figures not only through explicit judgments, but through subtle wording and rearrangement of information throughout the work. Nor did he hesitate to include myths (like the Yellow Thearch), miracles (like emperor Gaozu’s mother being impregnated by a dragon), and other un­believ­ able accounts in his work that, by modern standards, would certainly not qualify as ‘history’. The following excerpts are taken from the Zhonghua shuju edition (Beijing 1997). Editions: Shiji 史記. 10 vols. Beijing 1997, is the standard edition. Shiki kaichū kōshō kōhō 史記會注考證校補, ed. by Takigawa Kametarō 瀧川龜太郎 and Mizusawa Toshitada 水澤利忠. 2 vols. repr. Shanghai 1995, contains a wealth of commentaries. Translations: Chavannes 1895–1905 is a classic translation of juan 1–47. Watson 1993 contains juan 5–12, 15–20, 28–30, 48–59, 68, 71–3, 79, 84–104, 106–27, 129. Nienhauser 1994– will, upon completion, be the first full translation in a Western language. Studies: The corpus of studies on the Shiji is enormous. Hardy 1999; Nienhauser 2011; van Ess 2014; and van Ess, Lomová, and Schaab-­Hanke 2015 are just a few recent titles.

367

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 31 : s h i j i

31.2  Biography of Confucius (SJ 47) There is not a shred of contemporary evidence that testifies to the life of Confucius, China’s most eminent thinker. All the information we have about him derives from the Lunyu and other Zhanguo texts— and e­ specially from the Shiji, which contains the earliest biography of the ‘master’. Strangely, it is not classified among the biographies but among the ‘hereditary houses’, which section is otherwise reserved for nobles. Perhaps this is an expression of Sima Qian’s admiration for the sage; however, his biography does not read like a success story at all.





368

孔子世家第十七

孔子生魯 昌平鄉陬邑。 其先宋人也,曰孔防叔。



索隱 陬是邑名,昌平,鄉號。 孔子居魯之鄒邑 昌平鄉之闕里也 正義

括地志云:「故鄒城在兗州 泗水縣東南六十里 輿地志云鄒城西界闕里有

尼丘山。」

索隱 家語:「孔子,宋微子之後。 宋襄公生弗父何,以讓弟厲公。 弗

父何生宋父周,周生世子勝,勝生正考父,考父生孔父嘉,五世親盡,別



為公族,姓孔氏 。」

防 叔 生 伯 夏, 伯 夏 生 叔 梁 紇。 紇 與 顏 氏 女 野 合 而 生 孔 子, 禱於尼丘得孔子。 魯襄公二十二年而孔子生。 生而首





昌平 Chāngpíng, np, a place 陬 Zōu, np, a place 孔防叔 Kǒng Fángshū, np 叔梁紇 Shūliáng Hé, np 顏 Yán, np, a family name 禱 dǎo, vn, ‘pray’ 丘 qiū, nc, ‘hill’ 襄 Xiāng, np, posth. name

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

31.2 b i o g r a p hy o f co n f u c i u s

上 圩 頂, 故 因 名 曰 丘 云。 字 仲 尼, 姓 孔 子為兒嬉戲,常陳俎豆, 孔氏。  . . . 設禮容。



索隱 家語云「梁紇娶魯之施氏,生九女 其妾

生孟皮,孟皮病足,乃求婚於顏氏 徵在,從父命

為 婚 」。 其 文 甚 明。今 此 云「 野 合 」者,蓋 謂 梁

紇 老 而 徵 在 少, 非 當 壯 室 初 笄 之 禮, 故 云 野 合,

謂不合禮儀。

索 隱 公 羊 傳「 襄 公 二 十 一 年 十 有 一 月 庚 子,

孔 子 生 」。 今 以 為 二 十 二 年, 蓋 以 周 正 十 一 月 屬

明年,故誤也。





圩 yú, vi, ‘be indented’

頂 dı ̌ng, nc, ‘crown (of the head)’ 嬉 xī, vn, ‘amuse oneself ’ 戲 xì, vn, ‘play’ 俎 zǔ, nc, ‘small altar’ 豆 dòu, nc, ‘sacrificial bowl’ 設 shè, vtr, ‘put on’ 容 róng, nc, ‘expression’

Comments ① Confucius’ birthplace was the town of Zou in Changping county, Lu. Observe the omission of the prep after the P, which is accordingly used in a pseudo-­transitive way (cf. Box 11): this is not uncommon for the language of the Shiji. ② Confucius’ family allegedly belonged to the nobility of Song: his earliest documented ancestor who carries the name Kong, Kongfu Jia 孔父嘉, a grand marshal (大司馬) of Song, was murdered in 710 bc, whereupon his family fled to Lu. His name is a combination of courtesy name (孔父) and personal name (嘉); the last name Kong was then derived from the courtesy name, a process that was quite common in ancient China. Confucius’ father Shuliang He, mentioned below, also went by his courtesy name and personal name. ③ There has been much discussion about the meaning of 野合: ‘wild union’ or ‘united in the wilds’? The commentary gives a politically correct interpretation. ④ This is the third time the text tells us that Confucius was born! Just like the redundancies in Genesis 1, where heaven, earth, man, and all creatures are created twice, these repetitions may indicate different sources (cf. Focus 17). ⑤ Note how the reason given for Confucius’ name—the bump on his head—does not fit well with the preceding story of his mother’s prayer: the latter would suggest that the name of the hill—尼丘—inspired Confucius’ personal name as well as courtesy name (cf. Focus 5). Perhaps two traditions have been conflated in this passage.

369

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 31 : s h i j i

孔子年十七,魯大夫孟釐子病且死, 誡其嗣懿子曰:「孔丘,聖人 之後,滅於宋。 其祖弗父何始有宋而嗣讓厲公。及正考父佐戴、武、 宣公,三命茲益恭,故鼎銘云: 『一命而僂,再命而傴,三命而俯, 循牆而走,亦莫敢余侮。饘於是,粥於是,以餬余口。』 其恭如是。 吾聞聖人之後,雖不當世,必有達者。今孔丘年少好禮,其達者歟? 吾即沒,若必師之。」

索隱 昭公七年左傳云「孟僖子病不能相禮,乃講學之,及其將死,召大夫」云云。

索解 杜預曰:「孔子六世祖孔父嘉為宋 華督所殺,其子奔魯也。」

索解 杜預曰:「三命,上卿也。考父廟之鼎。」

索解 杜預曰:「於是鼎中為饘粥。饘粥,餬屬。言至儉也。」

⑨ ⑧ ⑦ ⑥









孟釐子 Mèng Lízı ̌ , np 懿子 Yìzı ̌, np

厲 Lì, np, posth. name 正考父 Zhèng Kǎofù, np 佐 zuŏ, vn, ‘assist’ 戴 Dài, np, posth. name 鼎 dı ̌ng, nc, ‘tripod’

銘 míng, nc, ‘inscription’

僂 lü ,̌ vi, ‘stoop’

傴 yǔ, vi, ‘bow down’ 俯 fǔ, vi, ‘bow the head’ 饘 zhān, nc, ‘gruel’ 粥 zhōu, nc, ‘congee’ 餬 hú, nc as vn, ‘fill’ 沒 mò, vi, ‘die’



Comments ⑥ In Confucius’ times, three cadet families of the ducal house—the Meng 孟, Shusun 叔孫, and Ji 季 (collectively known as San Huan 三桓)—wielded great political power in Lu. Meng Lizi (?–524 bc) was the scion of the Meng family. ⑦ Kongfu Jia, Confucius’ forebear, was murdered in 710 bc by a high minister in Song, who also ­kidnapped

370

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

31.2 b i o g r a p hy o f co n f u c i u s

his wife. Thereupon the Kong family fled to Lu. ⑧ According to this statement, Confucius’ ancestors belonged to the high nobility of Song: Fu Fuhe relinquished succession to the Song throne, and Zheng Kaofu was a minister to three Song rulers in the eighth century bc. The duplication of adverbs of degree in 茲益 seems redundant. ⑨ This ‘tripod inscription’ is highly unusual; nothing even remotely resembling it is known from Western Zhou or Chunqiu times. Note the rhymes: 僂 loɁ – 傴 *ɁoɁ – 俯 *poʔ – 走 *tsôɁ – 侮 *moʔ – 口 *khôʔ. ⑩ With this instruction began the Meng family’s patronage of Confucius. Perhaps it was this patronage that gained him access to the court of Lu. 孔 子 貧 且 賤。 及 長, 嘗 為 季 氏 料量平;嘗為司職吏 史, 而 畜 蕃 息。 由 是 為 司 空。 已 而 去 魯, 斥 乎 齊, 逐 乎 宋、 衛, 困於陳 蔡之閒,於是反魯。 孔 子 長 九 尺 有 六 寸, 人 皆 謂 之 「長人」而異之。魯復善待,由 是反魯。 . . . 

索隱 有本作「委吏」。按: 趙岐曰「委

吏,主委積倉庫之吏」。









料 liào, vtr, ‘measure’ 司職 sīzhí, nc, ‘supervisor of pastures’ 蕃 fán, vi, ‘flourish’ 司空 sīkōng, nc, ‘minister of public works’ 斥 chì, vtr, ‘reject’ 逐 zhú, vtr, ‘expel’ 困 kùn, vi, ‘be in difficulty’ 尺 chı̌, nc, ‘foot’ 寸 cùn, nc, ‘inch’

Comments ⑪ This is an intriguing sentence: was Confucius a registrar to the Ji family (cf. ⑥) or simply a ‘commissioner’, 委吏, as the commentary suggests? The latter is supported by Meng 5B5: 孔子嘗為委吏 矣,曰,會計當而已矣. ⑫ After a few more government posts follows the story of Confucius’ travels and failures. The events are narrated in greater detail later in the chapter. ⑬ Quite an abrupt change of subject, followed by a repetition of what had been said before: again, these are telltale signs of a textual pastiche. How tall was Confucius according to this account?

371

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 31 : s h i j i

定 公 十 四 年, 孔 子 年 五 十 六, 由大司寇行攝相事, 有喜色。 門 人 曰 :「 聞 君 子 禍 至 不 懼, 福 至 不 喜。」 孔 子 曰 :「 有 是 言 也。不曰『樂其以貴下人』乎?」 於是誅魯大夫亂政者少正卯。 與 聞 國 政 三 月, 粥 羔 豚 者 弗 飾賈;男女行者別於塗;塗 不拾遺;四方之客至乎邑者不 求有司,皆予之以歸。

索隱 家語作「皆如歸」。



⑯ ⑮





定 Dìng, np, posth. name 攝相 shèxiàng, nc, ‘deputy chancellor’ 少正卯 Shǎozhèng Mǎo, np 粥2 /鬻 yù, vtr, ‘sell’ 羔 gāo, nc, ‘lamb’ 飾2 shì, vtr, ‘raise (prices)’ 賈2 /價 jià, nc, ‘price’ 塗 tú, nc, ‘street’ 拾 shí, vn, ‘pick up’ 有司 yǒusī, nc, ‘officer’

Comments ⑭ Duke Ding of Lu reigned 509–409 bc, thus his fourteenth year was 496 bc. If Confucius was born in 551 bc, he would have been fifty-­four or fifty-­five years old according to European reckoning—but fifty-­six according to traditional Chinese reckoning, since the Chinese counted current years, not elapsed years: babies were considered one year old at birth, and they turned one year older with every new calendar year (not on their birthday!). The high offices mentioned here were actually the patrimony of noble families; it would seem doubtful that Confucius ever held them. ⑮ This utterance by Confucius is one of many that are not recorded in the Lunyu. It seems like an intrusion into the text, since the following sentence seems to connect to the preceding statement about Confucius’ position as minister of justice. ⑯ This is a notorious episode from the life of the sage: Confucius had Shaozheng Mao executed, allegedly for five detestable acts, among them ‘false teachings’ (Xun 28). One tradition has it that Shaozheng Mao had repeatedly won over many of Confucius’ disciples. . . ⑰ This is Confucius at the pinnacle of his career: he actually ‘participated’ (與) in the government. The praise for his administration became a commonplace in later literature. For the use of 弗, cf. the note at 5.2.3. A legal manuscript unearthed in 1975 may shed some light on the last two clauses: it tells us that s­ trangers, upon arriving in a city, had to report to the officials and produce their credentials before engaging in trade (Hulsewé 1985,  174). Under Confucius’ administration, apparently, they were not inconvenienced by the authorities.

372

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

31.2 b i o g r a p hy o f co n f u c i u s

齊人聞而懼,曰:「孔子為政必霸,霸則吾地近焉,我 之為先并矣。盍致地焉?」 黎鉏曰:「請先嘗沮之; 沮之而不可則致地,庸遲乎!」 於是選齊國中女子好 者 八 十 人, 皆 衣 文 衣 而 舞 康 樂, 文 馬 三 十 駟, 遺 魯 君。 陳女樂文馬於魯城南高門外。季桓子微服往觀再三, 將受,乃語魯君為周道游, 往觀終日,怠於政事。子 路 曰 :「 夫 子 可 以 行 矣。」 孔 子 曰 :「 魯 今 且 郊, 如 致 膰乎大夫, 則吾猶可以止。」桓子卒受齊女樂,三日 不聽政;郊,又不致膰俎於大夫。孔子遂行。

索隱 謂請魯君為周偏道路游行,因出觀齊之女樂。

集解 王肅曰:「膰,祭肉。」

㉒ ㉑













并 bìng, vtr, ‘annex’ 黎鉏 Lí Chú, np 沮 jǔ, vtr, ‘obstruct’ 文2 wèn, vtr, ‘decorate’ 庸2 yōng, adv, ‘how?’ 遲 chí, vst, ‘be late’ 選 xuǎn, vn, ‘choose’ 舞 wǔ, vn, ‘dance’ 康樂 Kānglè, np, a dance 季桓子 Jì Huánzı ̌, np

微2 wéi, vst, ‘be inconspicuous’ 怠 dài, vi, ‘be negligent’ 郊 jiāo, nc as vn, ‘sacrifice (in the outskirts)’ 膰 fán, nc, ‘sacrificial meat’

Comments ⑱ Qi was Lu’s neighbour to the north; according to HFei 10.31, it was Duke Jing of Qi (齊景公, 547–490 bc) himself, who was worried about Confucius’ successful administration. For the meaning of 霸, cf. 27.2–⑲. For 盍, cf. 9.3–㉜. ⑲ Li Chu was a high minister of Qi. 庸, used in rhetorical questions, is synonymous with 豈: ‘how could it be…?’ Like the latter, it is often used in conjunction with 乎. ⑳ The clause 皆衣文衣而舞康樂 is perplexing, since it is placed, like a relative clause, between the two O of 選; moreover, 皆 usually has its place between the S and the P. Is this a gloss that slipped into the text? The ‘beautiful woman stratagem’ (美人計), as it was later called, was a commonplace in Classical texts: Duke Mu of Qin is said to have corrupted the king of the Rong by sending him female musicians; Duke Xian of Jin, planning to attack Yu and Guo, first sent their rulers female musicians; the king of Chu did likewise with the ruler of Zou, etc. Perhaps this part of the story should be treated as a literary device rather than historical fact. ㉑ Ji Huanzi was the head of the Ji family (cf. ⑥) that had, 373

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 31 : s h i j i

by the end of the sixth century bc, effectively taken over power from the dukes of Lu. 微服, quite literally, means ‘plain clothes’, i.e. ‘incognito’. 語 yù, ‘tell’, can take two O: the person whom one tells, and what one tells. ㉒ Confucius is quoted in LY 9.18 as having said 吾未見好德如好色者也: this would have fit the occasion well. For Zilu, cf. 11.3–⑳. ㉓ According to this account, Confucius left Lu for two reasons: firstly, the female musicians that distracted the minister (compare 聽政 and 聞國政, above: both mean ‘attend to the government’); secondly, the fact that sacrificial meat was not parcelled out to the dignitaries (including himself). Interestingly, Meng 6B6 only mentions the meat, but not the women: 孔子為魯司寇,不用,從而祭,燔肉不至,不稅冕而行. This may be another indication that the Shiji narrative is a composite text, conflating (at least) two stories.

374

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 27

Textual criticism No Classical Chinese text has been transmitted in its original form. Instead of dealing with ancient manuscripts, students of Chinese culture usually turn to readily available modern editions, neatly arranged, punctuated, and accompanied by extensive scholarly commentaries. The convenience of these editions may dim our awareness of how far removed they are from what classical texts have looked like more than 2,000 years ago. The difference must have been dramatic. Not only were they originally written in scriptio continua (cf. Box 7), without commentaries, and in quite a different orthography (cf. Focus 12), but the texts themselves differed from what we read now. When classical texts were copied and re-­copied by hand, changes were introduced at every stage: characters were miswritten, omitted, duplicated, added, or transposed; entire lines were skipped (or bamboo slips misplaced); glosses were copied into the text; entire chapters were lost, misplaced, or jumbled. Add intentional alterations such as the replacement of taboo characters (cf. Box 20), deletion of objectionable content, or editorial emendations and one can imagine how dramatic the cumulative effect of these changes may be. Nor have these textual alterations been remedied by scholarly editions. For all their philological acumen, Chinese scholars of the Qing and later periods usually simply adhered to one received version of a text as their setting copy. Unlike European editors, they do not seem to have regarded textual changes as corruptions. Therefore, ‘rather than taking textual decisions on doubtful points, trying to restore them to an original form, [they] would usually leave the traditional reading in possession of the text and explain it’ (Vogelsang 2002b: 530): we have seen examples of this practice in Focus 14–⑪ and 24.2–⑥. As a result, even our best editions of classical texts reproduce historical distortions. They are not critical editions in the full sense of the word, hence they cannot be trusted for serious scholarly research. In order to obtain a reliable text, it is necessary to reverse the process of

375

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 27 : t e x t ua l c r i t i c i s m

textual alteration, as it were, and try to restore the text as nearly as possible to its original form. This is the task of textual criticism. Textual criticism, which has been most fully developed in Biblical studies, proceeds from the examination of all extant witnesses— manuscripts and print editions—of the text. This examination will likely yield quite a few variant readings in different witnesses; but how to decide which one is right? The answer is not by simply counting the witnesses that have a given reading but by evaluating them. In order to do so, their genetic relationship must be determined. This method is called stemmatic analysis. Take, for example, a passage of the Zuozhuan, which is testified in four manuscripts: one formerly held in the Kanazawa bunko, Japan (金澤本), one Japanese copy of the Tang compilation Qunshu zhiyao (羣書治要), another one from the collections of Li Mingnan and Nakamura Fusetsu (李鳴南/中村), and a Dunhuang scroll in the Pelliot collection in Paris (P.3729); moreover, the text is ­testified in the Tang stele inscriptions (石經, cf. Focus 18) and, of course, in many print editions. Through analysis of their textual relation, the stemma in Figure 13 may be reconstructed. Chinese print editions

P.3729

Figure 13  Stemma codicorum Source:  Adapted from Vogelsang 2007, 979.

As the stemma shows, all Chinese print editions derive from the tang stele inscriptions, whereas the Japanese edition 左氏會箋 stems from the Kanazawa manuscript, and the other three manuscripts derive from an unknown witness, here called 乙本. It also shows that the different editions are by no means equally valuable as witnesses. For example, even if dozens of Chinese print editions have the same reading, whereas 376

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 27 : t e x t ua l c r i t i c i s m

P.3729 and 羣書治要 have a different one, this does not mean that the former is to be preferred by dint of sheer quantity. Rather, the two readings are equal in value, because they represent two independent lines of transmission. By the same token, the combined testimony of 羣書治要, 李鳴南/中村, and P.3729 cannot overrule a variant reading in the 左氏會箋, because the latter represents an independent branch of transmission. In fact, the Chinese as well as the Japanese print editions can be discarded as witnesses altogether, since the editions they derive from—the Tang stele text and the Kanazawa manuscript, respectively—are still extant. The unknown witness 乙本, in turn, may be reconstructed by critical collation of the three manuscripts that derive from it. The underlying rationale is very straightforward: wherever all three coincide, this must be the text of 乙本; wherever two coincide and one differs, the latter is probably wrong, whereas the two others represent the text of 乙本; only in those rare cases where all three witnesses differ, the editor will have to make a conjecture to reconstruct the text of 乙本. Having thus established 乙本, we can proceed to reconstruct an even earlier witness, here called 甲本, by collating the Tang stele inscription, 乙本, and the Kanazawa manuscript by the same method. A text established in this way, then, rightly deserves the epithet ‘critical’. To be sure, this is an arduous process that may involve many vicissitudes; and even in an ideal case, the final result will be a version of the text from the Six Dynasties period—still about a millennium removed from a presumed ‘original’ of classical times. Only in rare cases do excavated manuscripts afford a glimpse at earlier stages of certain texts (cf. Focus 19). The business of textual criticism may not seem rewarding, and that is perhaps why it has been all but neglected in Chinese studies. Nevertheless, it must be done. For as long as we do not have critical ­editions of classical texts, all historical, linguistic, or philosophical studies based on these texts—including the present book—must remain ­tentative. References: For the method of textual criticism, cf. Maas 1958 and Timpanaro 2005. The only critical edition of a CC text is Thompson 1979, but cf. the meticulous studies by Simson 2002 and 2006. For the Zuozhuan example, cf. Vogelsang 2007; for the problems of textual transmission and scholarship in China, cf. Boltz 1995 and Vogelsang 2002b.

377

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 32

Yijing

32.1 Introduction 32.2 The Great Tradition I (Yi 7)

32.3 The Great Tradition II (Yi 8)

378 380

385

32.1 Introduction The Yijing 易經, or Canon of Changes, is the most venerable of the Five Canonical Texts (cf. Focus 1). Its origins are shrouded by myths. The mythical ruler Fuxi 伏羲 is said to have invented eight trigrams (八卦), that is combinations of three broken or unbroken lines that represented natural phenomena or states of being. They are: ☰



乾 Qian 兌 Dui



離 Li



震 Zhen



巽 Xun



坎 Kan



艮 Gen



坤 Kun

The origin of the trigram names is unclear. They seem to be quite old, since at least half of them are not part of the CC vocabulary: the characters 乾, 兌, 艮, and 坤 are only used to write trigram (or hexagram) names.

Tradition has it that the trigrams were later combined to form sixty-­ four hexagrams (卦)—permutations of six broken or solid lines. These hexagrams were used in milfoil divination, in which they represented a numerical result obtained by drawing milfoil stalks from a stack. But they probably also symbolized magical powers: totem animals or clan emblems that evoked the power of the ancestors. ‘The highest mani­fest­ ation of the world’s hidden powers resides in the hexagrams’ (極天下之 賾者,存乎卦, Yi 7). In the Yijing, every hexagram is accompanied by its name and two short comments: a ‘hexagram statement’ (卦辭) that characterizes the hexagram in a few words, and ‘line statements’ (爻辭), often rhymed, specifying the mantic implications of every line for itself. Thus the entry for the first hexagram reads: 378

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

32.1 i n t r o d u c t i o n

䷀乾。元亨利貞。

qian/active

初九。潛龍勿用。

Base (9):

九二。見龍在田, 利見大人。

(9) 2:

九三。君子終日乾乾,

(9) 3:

夕惕若,厲, 无咎。 九四。或躍在淵,无咎。 (9) 4: 九五。飛龍在天,

(9) 5:

利見大人。 上九。亢龍有悔。

Top (9):

用九。見群龍无首,吉。 All (9):

Supreme offering. Favourable augury. A dragon lies beneath the lake. No action take. Lo, on the field a dragon bides. To meet with great men well betides. A prince is active all day long and after dark still stays alert. DANGEROUS. NO MISFORTUNE. Sometimes it leaps above the tide. NO MISFORTUNE. A dragon through the heavens glides. To meet with great men well betides. A dragon soaring away. There will be trouble. See dragons without heads. AUSPICIOUS.

Translation: Rutt 2002. Richard Rutt’s typographic differentiation between the oracle utterance, the indication or injunction that follows, and the PROGNOSTIC WORDS has been retained.

These comments are traditionally attributed to King Wen (cf. Focus 7). Although this is no longer credible, the cryptic nature and arcane language of these comments suggest that they belong to the oldest part of the book. Hence they together with the hexagrams themselves are sometimes called Zhouyi 周易. Over time, the Zhouyi was supplemented by ten distinct commentaries, called the ‘ten wings’ (十翼 shiyi), that have been attributed to Confucius. Although this attribution, too, is highly doubtful, the ‘ten wings’, largely written in Classical Chinese, are obviously younger than the Zhouyi, and they certainly added a reflective quality to the text. They not only expanded the Zhouyi to become a much larger text now called the Yijing but also changed the character of the text from a div­in­ation manual, steeped in magical beliefs, to a philosophical work that contains reflections on ethical, political, and cosmological issues, correlative 379

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 32 : yi j i n g

thought, and essentials of the Yin/Yang theory. As such, the Yijing has occupied the minds of Chinese scholars for well over two millennia. The excavation of two Yijing manuscripts from Han dynasty tombs as well as the appearance of another bamboo manuscript al­leged­ly from the state of Chu testify to the importance of the text in Classical times; and ever since it was declared a canonical text under emperor Wu (武帝, r. 140–87 bc), the Yijing has been studied and memorized by virtually every scholar in imperial China. By the eighteenth century, they had produced a staggering 1,800 book-­length commentaries to this text; the commentary by Zhu Xi was mandatory reading for examination candidates for almost seven centuries. All the while, the Yijing has continued to be used as a divination manual throughout Chinese history. Its hexagrams and outlines of its cosmology are annually reprinted in Chinese popular almanacs; its binary system of representation has excited Leibniz in the eighteenth century as well as modern computer scientists around the globe; and it has inspired countless adaptations to all situations in life: the Yijing is applied to love and marriage, business, managing crisis situations, psychoanalysis, life counselling, and much more. After more than 2,000 years, the Yijing has lost nothing of its magic. The following passages are taken from the Zhouyi zhengyi 周易正義 by Kong Yingda 孔穎達 (574–648), re-­edited in Ruan Yuan’s Shisan jing zhushu. Editions: Standard editions include Zhouyi zhengyi 周易正義, ed. by Kong Yingda 孔穎達, in Shisan jing zhushu 十三經注疏, ed. by Ruan Yuan 阮元. repr. Shanghai 2009; Zhouyi benyi 周易本義, ed. by Tanaka Keitarō. Taipei 1975; and Zhouyi jijie zuanshu 周易集解纂疏, ed. by Li Daoping 李道平. Beijing 1994. The Mawangdui silk MS was published as Mawangdui boshu Zhouyi jingzhuan jiaodu 馬王堆帛書周易經傳校讀, ed. by Zhang Zhenglang 張政烺. Beijing 2008. Translations: Wilhelm 1924 set the standard for the twentieth century. More recent translations are Rutt 2002 and Schilling 2009. For the Mawangdui MS, cf. Hertzer 1996 and Shaughnessy 1996. Studies: Waley 1933; Wilhelm 1972; Shaughnessy 1983, 2014; Smith 2012.

32.2  The Great Tradition I (Yi 7) The ‘Great Tradition’ (大傳), or ‘Tradition of Appended Statements’ (繫辭傳), the two parts of which constitute the fifth and sixth ‘wings’ of the Yijing, is one of the most important treatises of Chinese philosophy. 380

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

32.2 t h e g r e at t r a d i t i o n i

Its numerological theories; the idea of Yin and Yang as complementary forces, emanating from an undifferentiated unity, that pervade all things; and the correlation between the sixty-­four hexagrams and the structure of the universe have profoundly influenced Chinese thought over two millennia. This cosmological theory dictated political decisions since Han times, it informed the understanding of Daoist literature in the middle ages, served as a metaphysical basis for neo-­Confucian philosophy since Song times, and it has pervaded popular beliefs up to the present day. In a way, the structure of the text itself mirrors the principle that the manifold things are all interwoven, since it is a patchwork of disparate snippets that have been knitted into one system by the unknown editors of the text.





381

周易 繫辭上第七 ︻疏︼正義曰:謂之「繫辭」者,凡有二義,論字取繫屬之義。



聖人繫屬此辭於爻卦之下,故此篇第六章云:「繫辭焉以斷其吉凶。」

天尊地卑, 乾坤定矣。︻疏︼正義曰:天以剛陽而尊,地以柔陰而卑,則乾坤 之體安定矣。卑高以陳,貴賤位矣。 天尊地卑之義既列,則涉乎萬物,貴賤之 位明矣。動靜有常,剛柔斷矣。︻疏︼正義曰:天陽為動,地陰為靜,各有常度, 則剛柔斷定矣。 方以類聚,物以羣分,吉凶生矣。 方有類,物有羣,則有 同有異,有聚有分也。順其所同,則吉;乖其所趣,則凶,故吉凶生矣。在天成象, 在地成形,變化見矣。︻疏︼正義曰:「象」謂懸象,日月星辰也。「形」謂山川 草木也。 是故剛柔相摩, 八卦相盪。︻疏︼正義曰:剛則陽爻也。柔則陰爻也。 剛柔兩體,是陰陽二爻,相雜而成八卦,遞相推盪。鼓之以雷霆,潤之以風雨。 日月運行,一寒一暑。 乾道成男,坤道成女。乾知大始,坤作成物。 乾以易知,坤以簡能。︻疏︼「鼓之以雷霆」至「簡能」。 ○ 正義曰:八卦既 ④

繫 xì, vi, ‘be attached (to)’ 辭2 cí, nc, ‘statement’ 卑 bēi, vst, ‘be low’ 乾 qián, np, name of a trigram and a hexagram 坤 kūn, np, name of a trigram and a hexagram 剛 gāng, vst, ‘be hard’ 柔 róu, vst, ‘be soft’ 斷 duàn, vtr, ‘separate’ 吉 jí, vst as nc, ‘auspiciousness’ 凶 xiōng, vst as nc, ‘inauspiciousness’ 象 xiàng, nc, ‘image’ 摩mó, vn, ‘rub’ 卦 guà, nc, ‘trigram’ 雷 léi, nc, ‘thunder’ 霆 tíng, nc, ‘lightning’ 潤 rùn, vi, ‘be moist’ 運 yùn, vi as adv, ‘in circles’ 暑 shǔ, vst as nc, ‘heat’ 簡2 jiǎn, vst, ‘be simple’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

業 yè, nc, ‘task’

推 tuī, vn, ‘replace’

爻 yáo, nc, ‘line (of a trigram or hexagram)’ 玩 wán, vtr, ‘ponder’ 占 zhān, vn as nc, ‘divination’

相推盪,各有功之所用也。又鼓動之以震雷離電,滋潤之以巽風坎雨,或離日坎月,運動而行,一節為寒,一節為暑。易則易 知,簡則易從。易知則有親,易從則有功。 順萬物之情,故曰有親。通天下之志,故曰有功。︻疏︼正義曰:此二句, 論聖人法此乾坤易簡,則有所益也。有親則可久, 有功則可大。可久則賢人之德,可大則賢人之業。︻疏︼正義曰: 行天地之道,總天地之功,唯聖人能。然今云賢人者,聖人則隱迹藏用,事在无境。 易簡而天下之理得矣。天下之理得, 而成位乎其中矣。︻疏︼正義曰:成位況立象,言聖人極易簡之善,則能通天下之理,故能成立卦象於天地之中,言並天地也。 聖人設卦觀象,繫辭焉而明吉凶,剛柔相推而生變化。 繫辭所以明吉凶,剛柔相推所以明變化也。吉凶者, 存乎人事也。變化者,存乎運行也。是故吉凶者,失得之象也。 由有失得,故吉凶生。︻疏︼正義曰:此下四句經,總明 諸象不同之事,辭之吉者是得之象;辭之凶者是失之象。悔吝者,憂虞之象也。變化者,進退之象也。剛柔者, 晝夜之象也。六爻之動,三極之道也。 三極,三材也。 是故君子所居而安者,易之序也。所樂而玩者, 爻之辭也。是故君子居則觀其象而玩其辭,動則觀其變而玩其占。是以自天祐之,吉无不利。︻疏︼正義曰:







愛樂而耽玩也。卦之與爻,皆有其辭,但爻有變化,取象既多,以知得失。故君子尤所愛樂,所以特云「爻之辭」也。



382

吝 lìn, vn as nc, ‘distress’



「所樂而玩者,爻之辭也」者,言君子愛樂而習玩者,是六爻之辭也。辭有吉凶悔吝,見善則思齊其事,見惡則懼而自改,所以

l e s s o n 32 : yi j i n g

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

32.2 t h e g r e at t r a d i t i o n i

Comments ① The chapter opens with three parallel sentences that correlate the forces of the universe with the symbols of the Yijing. Heaven and earth, the poles of the universe, are associated with the first two hexagrams 乾 and 坤, from which all others originate (note that 乾 and 坤 are also associated with penis and vagina, respectively). Then the high and low beings are associated with social strata as well as with the upper and lower trigrams (貴賤) of the hexagrams; movement and stillness are associated with hard/unbroken and soft/broken lines (剛柔), respectively. The correlations are: Heaven 天 Earth 地

Qian 乾 high 高 upper movement 動 lines 貴 Kun 坤 low 卑 lower stillness 靜 lines 賤

hard lines 剛 soft lines 柔

Although the sentences may appear like conditional/consecutive constructions, this is not the way the Xici zhuan argues. It expresses correlation rather than causation: ‘just as . . . so’, rather than ‘if . . . then’. The parallelismus membrorum is the syntactic counterpart to correlative thought. ② Here, two other main categories of the Yijing are introduced: ‘auspiciousness vs. inauspiciousness’ (吉凶) and ‘change and transformation’ (變化). Both are interrelated, since Yijing divination does not simply yield a hexagram signifying an auspicious or inauspicious state, but also the tendency in which it will change and transform. The system of the Yijing implies perpetual change. ③ 是故 opens a new paragraph, which introduces the mechanics of change according to the Yijing: the unbroken and broken lines ‘rub against’ one another, and the trigrams replace one another, so that ever new symbols emerge, representing so many moments in change. The following forces of nature have been associated with trigrams: ‘thunder’ with 震 Zhen, ‘lightning’ with 離 Li, ‘wind’ with 巽 Xun, and ‘rain’ with 坎 Kan. Thus the trigrams, while forms in their own right, equally represent natural phenomena. ④ 乾 is associated with the male, 坤 with the female. 乾 ‘knows the great beginnings’, which the commentary takes as ‘Heavenly Yang energy’ (天陽之

383

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 32 : yi j i n g

氣), i.e. semen. Just like ‘Adam knew Eve his wife, and she conceived’ (Genesis 4.1), this sentence may refer to sexual reproduction; consequently 坤 ‘brings about the completed things’, and the lineage is founded. On a different level, the two represent the realm of knowledge vs. action, ­perhaps also the realm of politics, in which 乾 is the ruler and 坤 the subject (compare SJS 1: 愚者闇於成事,知者見於未萌,民不可與慮 始,而可與樂成; cf. 27.1). ⑤ Again, the tone of the text seems to change: in the following lines causal reasoning, indicated by 則, appears for the first time. The structure of the argument is as follows: 易 → 易知      簡→ 易從    易知 → 有親    易從 → 有功        有親 → 可久     有功 → 可大            可久 → 賢人之德   可大 → 賢人之業 The double cascade begins with ‘being easy’ (易 does not mean ‘change’ in this case) and ends with the ‘worthy man’. Observe how the commentary tries to explain why the ruler (聖人) is not mentioned here. ⑥ Now the text has moved from the cosmological to the political sphere: from 天 to 天下, the realm ‘under Heaven’, and its ‘pattern’. If the ‘pattern’ is grasped (note the passive usage of the vn 得), the social ranks (cf. ①) are completed (or: perfection is established?). ⑦ In this section, the ‘sages’ are introduced, who created the trigrams (according to tradition: Fuxi) and their explanations (according to tradition: king Wen). They made the world and its fortunes legible. ⑧ Here, the relation of the Yijing ­concepts and the world is detailed. The hexagrams reflect success and failure, worries and anxiety, advance and retreat, day and night; to sum up: ‘the movement of the six lines is the principle of the three ultimate potencies’, heaven, earth, and man. ⑨ The 君子 is King Wen, in particular; but more generally, the term refers to all members of the elite (cf. 17.5). It was their duty to accord with cosmic harmony as reflected in the Yijing, hence they should ‘dwell on’, ‘contemplate’, and ‘ponder’ its symbols—and then ‘Heaven will assist them’ (which quotes a line statement from hexagram 14, 大有). The use of the prep 自 is puzzling.

384

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

32.3 t h e g r e at t r a d i t i o n i i

32.3  The Great Tradition II (Yi 8) The second part of the ‘Great Tradition’ adds mythical details to the theory of the Yijing symbols. Its account of the creation and uses of the hexagrams amounts to a veritable history of Chinese civilization from a hunter-­gatherer society to the beginnings of agriculture and commerce as well as government, writing, and navigation. Just like Han Feizi (cf. 28.2), it reflects a view of history that is not based on a cyclical notion of time, in which rulers come and go, events occur and re-­occur without any real change, but describes the movement from an emphatically different past to genuinely new stages of development. The Xici zhuan thus introduces a new concept of ‘change’ which is absent from the older parts of the Canon of Changes. 古者包犧氏之王天下也,仰則觀 象于天,俯則觀法于地,觀鳥獸 之 文, 與 地 之 宜, 聖 人 之 作 易, 无

大不極,无微不究。大則取象天地,細則

觀鳥獸之文,與地之宜也。︻疏︼正義曰:

「自此」至「取諸離」。此一節明包犧法天

地 造 作 八 卦, 法 離 卦 而 為 罔 罟 也。近 取 諸身,遠取諸物,于是始作八卦, 以 通 神 明 之 德, 以 類 萬 物 之 情。 作 結 繩 而 為 罔 罟, 以 佃 以 漁, 蓋取諸離。︻疏︼正義曰:「近取諸身」,

者若耳目鼻口之屬是也。「遠取諸物」者,



385

包犧 Bāoxī np, a mythical ruler 仰 yǎng, vn, ‘look up’ 罔2 wǎng, nc, ‘net’ 罟 gǔ, nc, ‘(fishing) net’ 佃 tián, vi, ‘hunt’ 漁 yú, vn, ‘fish’ 蓋 gài, adv, ‘probably’ 離 Lí, np, a hexagram

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 32 : yi j i n g

大壯 Dàzhuàng, np, a hexagram

宇 yǔ, nc, ‘eaves’

葬 zàng, vtr, ‘bury’

薪 xīn, nc, ‘brushwood’

樹 shù, nc as vn, ‘plant with trees’ 大過 Dàguò, np, a hexagram

夬 Guài, np, a hexagram 耜 sì, nc, ‘spade’

益 Yì, np, a hexagram

噬嗑 Shìhé, np, a hexagram 窮 qióng, vi, ‘be exhausted’ 垂 chuí, vi, ‘hang down’ 裳 cháng, nc, ‘skirt’

若雷風山澤之類是也。舉遠近則萬事在其中矣。 包犧氏沒,神農氏作,斲木為耜,揉木為耒,耒耨之利,以教天下, 蓋取諸益。 制器致豐,以益萬物。日中為市,致天下之民,聚天下之貨,交易而退,各得其所,蓋取諸噬嗑。 噬嗑, 合也。巿人之所聚,異方之所合,設法以合物,噬嗑之義也。 神農氏沒,黃帝、堯、舜氏作,通其變,使民不倦;神 而化之,使民宜之。易窮則變,變則通,通則久。︻疏︼正義曰:「神而化之,使民宜之」者,言所以「通其變」者,欲使

穴 xué, nc as adv, ‘in caves’



義也。

上古穴居而野處,後世聖人易之以宮室,上棟下宇,以待風雨,蓋取諸大壯。宮室壯大於穴居,故制為宮室, ⋯

得宜也。是以自天祐之,吉无不利。黃帝、堯、舜垂衣裳而天下治,蓋取諸乾、坤。 垂衣裳以辨貴賤,乾尊坤卑之

神理微妙而變化之,使民各得其宜。若黃帝已上,衣鳥獸之皮,其後人多獸少,事或窮乏。故以絲麻布帛而制衣裳,是神而變化,使民



是不封也。不種樹以標其處,是不樹也。「喪期无數」者,哀除則止,无日月限數也。



百官以治,萬民以察,蓋取諸夬。 夬,決也。書契所以決斷萬事也。

上古結繩而治,後世聖人易之以書契,

非是後物以替前物,故不云「上古」也。 古之葬者厚衣之以薪,葬之中野,不封不樹,喪期无數,後世聖人易之 以棺椁,蓋取諸大過。︻疏︼正義曰:厚衣之以薪,葬之中野,猶在穴居結繩之後,故直云「古」也。「不封不樹」者,不積土為墳,

取諸大壯也。︻疏︼正義曰:巳前不云「上古」,已下三事,或言「上古」,或言「古」,與上不同者,巳前未造此器之前,更无餘物之用,



⑮ ⑯

386

棟 dòng, nc, ‘ridgepole’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

32.3 t h e g r e at t r a d i t i o n i i

Comments ⑩ 包犧 is another name of Fuxi, the mythical figure associated with the beginnings of a hunter and gatherer society (yet he is said to have ‘ruled as a king’, as if there had been a unified realm ‘under Heaven’). He is credited with the introduction of hunting and fishing as well as with the creation of the eight trigrams. Observing natural phenomena, so the Xici zhuan tells us, Fuxi recreated the ‘images’, ‘models’, ‘patterns’, and ‘properties’ in the trigrams. They ‘penetrate’ the forces of the universe, literally the ‘power of spiritual brightness’, ‘brightness’ being a metaphor for the all-­encompassing capacity of this power. ⑪ ‘Tying knots’ not only serves to make nets but also as a method of information storage (cf. 22.5–⑳); this will become clear a few lines later. The method of making nets, the text tells us, was taken from the hexagram Li—which, indeed, looks like a net: ䷝. ⑫ Fuxi was succeeded by Shennong. With him, a new stage of social evolution begins. As his name, ‘Divine Farmer’, suggests, this mythical figure is supposed to have introduced farming to the Chinese (inspired by the hexagram Yi). Interestingly, the text associates this with the emergence of trade and interregional markets: a network that connects the local segments of peasant society. Note that all these ‘innovations’ are said to be derived from the hexagrams (see the commentaries for the conceptual link): the writing had been on the wall, as it were, since Fuxi’s times. ⑬ The next evolutionary stage is represented by the Huangdi, Yao, and Shun. Their contribution to social development was the introduction of government. They were the first to exercise power by ‘making’ or ‘compelling’ (使) the people to do things. Note the emphasis on ‘change’ (變) in these lines: this was a central theme for political realism, as well (cf. 27.1). However, Huangdi, Yao, and Shun apparently did not rule by law and bureaucracy but ‘let their garments hang loose’: they were charismatic rulers, and therefore, ‘Heaven will assist them’ (cf. ⑨). ⑭ After an enumeration of further inventions—boats, oars, beasts of burden, gates, pestles and mortars—‘to benefit all under Heaven’, which is left out here, a new paragraph (or a new snippet?) begins. It reiterates the process of civilization from a different perspective. Now, the invention of dwellings is discussed: this theme as well as the term ‘high antiquity’ (上古) are reminiscent of HFei 49 (cf. 28.2) and Mo 6: 古之民未知為宮室時,就陵 阜而居,穴而處. Note how the commentary tries to make sense of the new terminology in this paragraph. ⑮ From the houses of the living, the text now turns to the abode of the dead: this is a central issue for Confucian thinkers. Perhaps the hexagram Daguo, which is invoked in 387

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 32 : yi j i n g

this context, should accordingly be read as the ‘Great passing’. ⑯ Here is a history of Chinese government in a nutshell: from ‘tying knots’ to written documents and a bureaucracy that administers a ‘myriad of people’. This is no longer charismatic government but the system of the empire. Whereas the Xici zhuan depicts this history as the unfolding of a natural order—Fuxi had prefigured it all in the hexagrams!—other authors have interpreted it as the decline from a golden age. The ­admiration for the inventions of the sages on the one hand and the sentimental longing for the simple, unspoiled society of pre-­historic times on the other would remain common themes in Chinese literature until the twentieth century.

388

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Lesson 33

Ritual texts

33.1 Introduction 33.2 The Cycle of Ritual (Li 9)

33.3 The Great Learning (Li 42)

389 390

394

33.1 Introduction Ritual was a central component of ‘Confucian’ teachings throughout the Classical period. ‘All students talked a lot about ritual’ (SJ 121), and the topic was treated at length in works like Lunyu (cf. 17.4) and Xunzi. However, it was only in the early Han period that three veritable compendia appeared, which systematically described an authoritative ritual order. Attributed to the Duke of Zhou (周公, eleventh cent. bc) himself, these three ritual texts all attained canonical status by Tang times. Despite their claim to antiquity, these compilations were no doubt first written and edited in Classical times. Perhaps they were a reaction to the increased codification of law in Zhanguo and Qin times, affirming the credo of an elite that defended its own ritual norms against state-­ imposed law: ‘Ritual does not reach down to the common people, and penal law does not reach up to the dignitaries’ (禮不下庶人,刑不上 大夫), as the very first chapter of the Liji states. Indeed, the earliest ritual text, the Yili 儀禮, Ceremonials and Rituals, was known in Han times, inter alia, as Shili 士禮, Rituals for Officers, and this is exactly what it describes: the seventeen chapters of the received version bear titles like ‘Capping Rites for Officers’ (士冠禮), ‘Nuptial Rites for Officers’ (士昏禮), ‘Mourning Rites for Officers’ (士喪 禮), and the like. The Yili, was the Canon of Ritual (禮經) in Han times; when Han texts spoke of Li, they meant the Yili, and it was this text that was officially engraved in stone in 175 ad. 389

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 33 : r i t ua l t e x t s

The Zhouli 周禮, Zhou Rituals, also called Zhouguan 周官, Zhou Offices, is quite a different text: an idealized—and demonstrably unhistorical—description of the Zhou governmental structure, it provides systematic descriptions of hundreds of offices that allegedly existed in the heyday of Zhou rule, specifying titles, ranks, and duties in considerable detail. Yet this description that claims to hark back to antiquity was first mentioned only in Han times; and even then it seems to have remained all but unknown until Wang Mang (王莽, c.45 bc–ad 23), who put an end to the Western Han, used it to legitimize his Xin 新 dynasty. The least systematic, and most intriguing of the ritual texts is certainly the Liji 禮記. A huge compilation of diverse material, it is attributed to a certain Dai Sheng 戴聖 (first cent. bc). While his uncle, Dai De 戴德, is credited with a work in eighty-­five chapters, the Da Dai Liji 大戴禮記 (in reality a post-­Classical text), Dai Sheng is said to have compiled the canonical Liji (or Xiao Dai Liji 小戴禮記) in forty-­nine chapters. The work covers a wide range of topics, including long lists of ritual terms, theoretical treatises on ritual, detailed prescriptions for sacrifices, seasonal observances, rules of conduct within a household, sayings by and anecdotes about Confucius and his disciples, as well as treatises on music (thought to be at least part of a lost Canon of Music, 樂經) and learning. Zheng Xuan (鄭玄, 127–200) wrote authoritative commentaries to all three ritual texts. The following sections are taken from the edition Liji zhengyi 禮記正 義 by Kong Yingda 孔穎達 (574–648), which is part of Ruan Yuan’s Shisan jing zhushu. Editions: All three texts are included in the Shisan jing zhushu 十三經注疏, ed. by Ruan Yuan 阮元. repr. Shanghai 2009. Other scholarly editions include Zhouli zhengyi 周禮正義, ed. by Sun Yirang 孫詒讓. 14 vols. Beijing 2000; Liji jijie 禮記集解, ed. by Sun Xidan 孫希旦. 3 vols. Beijing 1989; and Liji xunzuan 禮記訓纂, ed. by Zhu Bin 朱彬. 2 vols. Beijing 1996. Translations: The Yili was translated by Couvreur 1916 and Steele 1917; the Zhouli by Biot 1851; and the Liji by Legge 1885 and Wilhelm 1930. Studies: Karlgren 1931; Broman 1961; Nylan 2001. 168–201; Elman and Kern 2009; Liu Yukai and Habberstad 2014.

33.2  The Cycle of Ritual (Li 9) The ‘Liyun’ is one of the most influential chapters of the Liji. Especially its short descriptions of idealized social orders, attributed to Confucius, 390

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

33.2 t h e c yc l e o f r i t ua l

enjoyed a remarkable revival in the twentieth century. They were invoked, among others, by the reformer Kang Youwei (1858–1927), by the first president of the Republic of China, Sun Yatsen (1866–1925), and by the CCP leader Deng Xiaoping (1904–97). Until today, the ideal of xiaokang,‘moderate prosperity’, remains an official ideal of the Communist party. The ‘Liyun’ is a textbook example for the enduring appeal of ­classical texts—and for their persistent misinterpretation.







391

禮運第九



陸曰:「鄭云禮運者,以其記五帝三王相變易及陰陽轉旋之道。」

昔者仲尼與於蜡賓, 蜡者,索也。歲十二月合聚萬物而索饗之,亦祭宗廟,時孔子仕魯,在助祭之中。 與音預。蜡,仕嫁反,祭名。事 畢, 出 游 於 觀 之 上, 喟 然 而 嘆。 觀,闕也。孔子見魯君於祭禮 ○ 有不備,於此又睹象魏舊章之處,感而嘆之。仲尼之嘆,蓋嘆魯也。言偃在側,曰:「君子何 嘆?」 言偃,孔子弟子子游。 孔子曰:「大道之行也,與三代之英,丘未之逮也,而有 「大道之行也, 志焉。」 大道,謂五帝時也。英,俊選之尤者。逮,及也,言不及見。志,謂識古文。 禪,善面反。 天下為公,選賢與能,講信修睦。 公猶共也。禪位授聖,不家之。睦,親也。 ○ 故 人 不 獨 親 其 親, 不 獨 子 其 子。 使 老 有 所 終, 壯 有 所 用, 幼 有 所 長, 矜 寡 孤 獨 廢 長,丁丈反。矜,古頑反。匱,其魏反。男 有 分, 分,猶職也。 疾者,皆有所養。無匱乏也。 ○ 女 有 歸。 貨 惡 其 棄 於 地 也, 不 必 藏 於 己, 力 惡 其 不 出 於 身 也, 不 必 為 己。 是 故 謀 閉而不興,盜竊亂賊而不作,故外戶而不閉,是謂大同。」 同,猶和也,平也。 ︻疏︼「大 ④

蜡 zhà, np, a sacrifice 觀2 guàn, vn as nc, ‘watchtower’ 喟 kuì, adv, ‘moan’ 言偃 Yán Yǎn, np 側 cè, nc, ‘side’ 英 yīng, nc, ‘heyday’ 逮 dài, vn, ‘reach’ 志/識2 zhì, vn as nc, ‘document’ 講 jiǎng, vn, ‘clarify’ 睦 mù, vi as nc, ‘concord’ 矜/鰥 guān, nc, ‘widower’ 孤 gū, vi, ‘be orphaned’ 廢 fèi, vtr, ‘abandon’ 匱 kuì, vi, ‘be lacking’ 藏 cáng, vn, ‘store’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

篤 dǔ, vi, ‘be loyal’

池 chí, nc, ‘moat’

紀 jì, nc, ‘leading principle’ 謹 jı̌n, vn, ‘be careful’ 考 kǎo, vn, ‘complete’

康 kāng, vst as nc, ‘tranquility’

殃 yāng, nc, ‘calamity’

「天下為公」,謂天子位也。為公,謂揖讓而授聖德,不私傳子孫,即廢朱均而用舜禹是也。 ○

勇 yoˇng, nc, ‘bravery’

「故人不獨親其親不獨子其子」者,君既無私,言信行睦,故人法之,而不獨親己親,不獨子已子。 ○

道」至「大同」

郭 guō, nc, ‘outer city wall’





猶則也。如有不由此者,在勢者去,眾以為殃,是謂小康。 」 康,安也。大道之人以禮,於忠信



為薄,言小安者失之,則賊亂將作矣。 ︻疏︼「今大」至「小康」 ○ 正義曰:前明五帝已竟,此明三代俊英 之事。孔子生及三代之末,故稱今也。隱,去也。干戈攻伐,各私其親,是大道去也。



392

溝 gōu, nc, ‘ditch’

「今大道既隱,天下為家。 傳位於子。各親其親,各子其子,貨力為己。 大人世及以為禮, 城郭溝池以為固。 亂賊繁多,為此以服之也。大人,諸侯也。禮義以為紀,以正君臣,以篤父子, 以 睦 兄 弟, 以 和 夫 婦, 以 設 制 度, 以 立 田 里, 以 賢 勇 知, 以 功 為 己。 故 謀 用 是 作, 知音智。禹、 而兵由此起。 以其違大道敦朴之本也。教令之稠,其弊則然。老子曰「法令滋章,盜賊多有。」 ○ 湯、文、武、成王、周公,由此其選也。 由,用也,能用禮義以成治。 此六君子者,未有 不謹於禮者也。以著其義,以考其信,著有過,刑仁、講讓,示民有常。 考,成也。刑,

l e s s o n 33 : r i t ua l t e x t s

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

33.2 t h e c yc l e o f r i t ua l

Comments: ① The commentary by Lu Deming quotes the Han commentary by Zheng Xuan for an interpretation of the chapter title. The chapter itself does not contain the word 運, ‘cycle, revolution’, at all. ② Confucius, serving as an official in Lu, seems to have been dissatisfied with a sacrificial ceremony: this is a well-­known motif (cf. 31.2–㉓). According to Shiji, Confucius’ disciple Yan Yan (note that names are not marked in this edition) was forty-­five years his junior: this scene must be set late in the master’s life. ③ In the twentieth century, the sentence 天下為公 was adopted by Sun Yatsen to describe his ideal of a republic: in Sun’s interpretation, the expression referred to people’s rights. But the author of the ‘Liyun’ certainly did not have anything like a republic in mind; rather, the passage deals with the difference between patrimonial rule, which centres on familial ties (親其親), and central rule, which replaces the principle of kinship by that of ability (選賢與能). Note the proximity to Mozi’s doctrine of ‘elevating the worthy’ (cf. 24.1). ④ These are some specific blessings of the ‘great principle’: everyone has his or  her place in a social system, and everyone is taken care of. Caring for the needy is also an important theme in Xunzi and Mengzi: 老而無妻曰鰥,老而無夫曰寡,老而無子曰獨,幼而無父 曰孤,此四者,天下之窮民而無告者。文王發政施仁,必先斯四 者 (Meng 1B5). ⑤ The essence of 大同 seems to be security, brought about by an extension of familial trust and devotion to the entire people. (Note that the commentary equates 同 with 和: do you agree?) The concept was revived in the nineteenth/twentieth century by Kang Youwei in his Datong shu 大同書, a utopia in which he envisages a global society characterized by unity and equality. In a less ambitious meaning, the term made it into the national anthem of the Republic of China, still sung in Taiwan today. Its opening lines are: 三民主義,吾 黨所宗 / 以建民國,以進大同. ⑥ This describes a decadent age with dynastic rulership and kinship groups as the basic units of society. This

393

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 33 : r i t ua l t e x t s

social structure, the text argues, leads to selfishness, distrust, rivalry, and violence. ⑦ This is where ‘ritual’ comes into play: as a remedy for a society that has lost its unity. This line  of thought is strikingly similar to Laozi, where the origin of Confucian virtues is described in much the same way: 大道廢,有仁義;智慧出,有大偽;六親不和,有孝 慈;國家昏亂,有忠臣 (Lao 18). ⑧ The great rulers of the ‘Three Dynasties’ (cf. Focus 7), living in a decadent age, needed to enact ritual in order to demonstrate virtue and steadiness. The construction 由此其 選也 is not clear, since an O is not expressed. ⑨ Another vague construction. But the drift is clear: whoever does not conform will be expelled. A society of ‘moderate prosperity’ is no longer ordered by mutual trust but by political power. This certainly remains true in the twenty-­first century, when xiaokang shehui was declared an official goal of the CCP. 33.3  The Great Learning (Li 42) The ‘Great Learning’ (大學), a short text traditionally attributed to Confucius’ disciple Zeng Can 曾參 (cf. 17.3–⑧), deals—quite literally— with ‘higher education’. Whereas xiaoxue 小學 probably meant the acquisition of reading skills, the ‘Great Learning’ deals with personal cultivation, development of charismatic virtue, ordering the family and, ultimately, pacifying the empire. This direct line from personal virtue to political order has made it a centrepiece of neo-­Confucian thought. As one of the Four Books (四書), it was elevated to canonical status by Zhu Xi 朱熹 (1130–1200) and became required reading for millions of students in late Imperial China. 大學第四十二

394

︻疏︼正義曰:案鄭目錄

云:「各曰大學者,以其記



博學,可以為政也。 」

大學之道,在明明德, 在親民,在止於至善。 知 止 而 后 有 定, 定 而 后 能 靜, 靜 而 后 能 安, 安 而 后 能 慮, 慮 而 后 能 得。 物 有 本 末, 事 有 終 始, 知 所 先 後, 則 近 道 矣。「 明 ⑩

靜 jìng, vi, ‘be tranquil’ 近2 jìn, vn, ‘come close’

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

33.3 t h e g r e at l e a r n i n g



格 gé, vtr, ‘investigate’ 謙 qiān, vi, ‘be humble’

大,舊音泰,劉 ○





明德」,謂顯明其至德也。止,猶自處也。得,謂得事之宜也。





直帶反。近,附近之近。 古 之 欲 明 明 德 於 天 下 者, 先 治 其 國。 欲 治 其國者,先齊其家。欲齊其家者,先脩其身。欲脩其身者,先正 其心。欲正其心者,先誠其意。欲誠其意者,先致其知。 知,謂知



善惡吉凶之所終始也。 ○ 其知如字,徐音智,下「致知」同。致知在格物。 格, 來 也。 物, 猶 事 也。 其 知 於 善 深 則 來 善 物, 其 知 於 惡 深 則 來 惡 物, 言 事 緣 人 所 好

來 也。 物 格 而 后 知 至, 知 至 而 后 意 誠, 意 誠 而 后 心 正, 心 正 而 后身脩,身脩而后家齊,家齊而后國治,國治而后天下平。 自 天 子 以 至 於 庶 人,壹 是 皆 以 脩 身 為 本, 其 本 亂 而 末 治 者 否 矣。 其所厚者薄,而其所薄者厚,未之有也。 此謂知本,此謂知之 至 也。 壹 是, 專 行 是 也。所 謂 誠 其 意 者, 毋 自 欺 也, 如 惡 惡 臭, 如 好好色,此之謂自謙。 故君子必慎其獨也。



齊2 qí, vtr, ‘regulate’

Comments ⑩ The text completely lacks contextual information. It is not clear who is addressed, and what the communicative situation is. Observe the three parallel clauses with infinitive O.  至, used as mod, means ‘the ultimate/highest’. ⑪ On the consecutive conj 而后, cf. 16.3. ⑫ Note that 先後 must be construed as causative, since preceded by 所. Moreover, 近 should be understood as a vn (what would it mean if it were a vst?). ⑬ This passage suggests that the text is addressed to a ruler: only he would be able to ‘order his state’. The ‘Daxue’ argues that the order of the state depends not on laws or institutions, but on the personal cultivation of the ruler (and the people), which in turn depends on ‘understanding’ and on ‘investigating things’. Although this latter expression—格物—was used in the early twentieth century as a word for modern science, it certainly means nothing of the sort in this text. 395

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

l e s s o n 33 : r i t ua l t e x t s

⑭  Here, the ambition of this treatise is clearly expressed: the empire pacified. ⑮ For this usage of the prep 至於, cf. 21.2–⑦; 以 seems to

resume the preceding NP, adding emphasis to the statement: ‘from the Son of Heaven all the way down to. . .’ Compare constructions like 自此 以往 (Zhuang 2), ‘from this onward’, 從今以往二年 (Guan 18), ‘two years from now’, or 自生民以來 ‘since mankind was born’ (Meng 2A2). The punctuation of this sentence is debatable: 壹是 can certainly not be construed as S (since 是 as prdem cannot be modified); rather, both words may be connected as P to the preceding, 是 representing a vst: ‘…uniformly holds true’. The next clause starts with 皆. ⑯ Again, 厚 and 薄 must be construed as causative because of 所. Recall that 所 phrases sometimes indicate an obligation (cf. 12.2.1–⑦), hence: ‘what one should treat as important’. ⑰ ‘To humble oneself ’ may be somewhere close to what 自謙 means: cheating oneself results from being overly self-­ important. ⑱ This is the highest stage of self-­cultivation: the gentleman is heedful even of what he does when he is alone. Or perhaps: ‘The noble man must take care of what is singular to him’, as Nylan 2001, 170, suggests? The ‘Daxue’ remains highly enigmatic: this is precisely what has made it so fascinating for more than two millennia.

396

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Focus 28

Translation While throughout this book the focus has been on syntactic analysis of Classical Chinese utterances, questions of translation have only been mentioned in passing. In fact, the two are quite distinct operations. Whereas syntactic analysis usually yields one, and only one, correct result, translation may generate many acceptable English renderings (reconsider, for example, the translations of Lao 1 given in 22.2–①). Precise syntactic analysis, bound by the structure of the source language, is the precondition for translation; but it does not predetermine the ­latter, insofar as translation is bound by the possibilities of the target language. A translation must, first and foremost, yield a correct English sentence. To be sure, it may be useful to mirror the original utterance in a word-­for-­word ‘crib’ translation as a first step; for example: 王使人問疾 > ‘king send man ask disease’ (Meng 2B2). But this also makes clear how much more it takes to arrive at a real translation: ‘The king sent someone to ask about the disease’.

In the case of Classical Chinese, translations often must deviate from the original to add certain elements that the original does not contain (recall the principle of optional determination, Box 8). This includes tense, number, deleted constituents, etc.: ⟨He is⟩ a good man. (Meng 7B25) 善人也。 所與立 ⟨those⟩ with whom ⟨he⟩ stands (LY 10.3) 孔子不見。 Master Kong did not meet ⟨him⟩. (LY 17.1) 男女居室 。 Men ⟨and⟩ women inhabit ⟨the same⟩ house. (Meng 5A2) 不好假道人。 ⟨He⟩ does not like to borrow methods ⟨from⟩ others. (Xun 9)

Just as these supplementary features add necessary precision to the translation, so the choice of translation words must often be more specific than the Chinese original. Recall, for example, the wide semantic range of the word 為 (cf. Box 14) of the conjunction 而 (cf. 7.6), or of the preposition 於 (cf. 8.2): on every occurrence of these words, a translator will have to decide which specific rendering best fits the given context.

397

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 28 : t r a n s l at i o n

Evidently, ‘concordant’ translations of CC texts, in which given words are consistently ­rendered by the same translation word, are well-­nigh impossible. Even basic concepts like 德, 法, 義, 治, and many others have such a wide range of meanings that they cannot be translated the same way on every occurrence.

Correctness is merely the minimum requirement for a translation. In order for it to be not only correct but also good, further deviations from Classical Chinese wording are often necessary. In some cases, the word order within phrases may be reversed for an idiomatic rendering: 往來 飲食

come and go eat and drink

Sometimes active constructions (especially such without an overt subject) are more elegantly translated in the passive voice: 在陳絕糧。 不違農時… …為天下戮。

In Chen, the provisions were exhausted. (LY 15.2) If the farming seasons are not interfered with. . . (Meng 1A3) …was disgraced in the realm. (Lü 9.2)

Complex nominal modifications should be translated as relative clauses or otherwise in English: 仲子所居之室, 伯夷之所築與。

Was the house in which Zhongzi lives built by Boyi? (Meng 3B10)

Note also how the question necessitates the anteposition of the auxiliary verb ‘was’ in English, which makes it impossible to retain the parallel structure of the original. Moreover, absolute topics may often be smoothly integrated into the sentence rather than left isolated at the beginning: When Yu governed the realm, 禹之治天下, 使民心變。 ⟨he⟩ changed the people’s hearts. (Zhuang 14) 其為人也,好善。 He is a man who loves what is good. (6B13)

Clauses with two coordinated nouns as S or O and two coordinated verbs as P may be more elegantly translated in a crisscross pattern: 君臣恐懼。 …傷害腹胃。

The ruler was afraid, and the ministers scared. (ZGC 11.1) …hurt their bellies and harmed their stomachs. (HFei 49)

398

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

f o c u s 28 : t r a n s l at i o n

Many more examples could be adduced for ways in which translations deviate from the original. A translation can never be an exact rendering but only an approximation of the original. Among the wide range of aspects of an utterance—its syntactic structure, terminology, modality, rhetorical devices, register, dialectal elements, rhymes, rhythm, parallel structures, intertextuality, etc.—only a few can usually be conveyed to the target language. The translator has to make a choice. Take the following passage from Lao 2: 天下  皆知美之為美,斯惡已。       皆知善之為善,斯不善已。         故  有無相生, 難易相成, 長短相較, 高下相傾, 音聲相和, 前後相隨。 A translator would have to consider, among other things, (1) the overall structure with two parallel sentences in the beginning and another six parallel clauses; (2) the elliptical constructions 斯惡已 and 斯不善已; (3) the Lu dialect word 斯; (4) the force of 已; (5) the rhythm of four syl­lables per line in the next six clauses; (6) the (imperfect) rhyme structure; and (7) the philosophical meaning expressed in the passage. Note that the syntactical analysis of the entire passage (except, perhaps, for the deletion in the beginning) is unambiguous—and yet it is not difficult to imagine that it may give rise to an almost infinite number of different translations. Syntactical analysis, as it was expounded in this book, is just a first step in the study of Classical Chinese literature. It clearly determines the limits of interpretation—and opens the way for an abundance of ­possible translations. References: Bellos 2011 is a wonderful introduction to translation; for translation from Chinese (not only CC), cf. Wilkinson 2018, 51–60, with many references.

399

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Glossary This glossary is primarily devised as a tool for using the present textbook. As such it includes all Chinese characters that occur in the example sentences and texts as well as the words they represent in these passages, supplemented by further entries that would seem essential for a basic core vocabulary of Classical Chinese. However, no claim to completeness is made. Proper names are included only in exceptional cases. Since in principle every Classical Chinese word could serve as proper name, it makes little sense to note them all. Therefore, only the names of regional polities and very eminent figures such as Yao or Shun are given entries.

The entries are arranged according to the tradi­tional order of 214 semantic classifiers and the number of additional strokes. Homographs—i.e. cases of one character writing different words—are discussed separately. The entries are structured as follows: • the character (in some cases followed by alternative form, marked by a slash); • the MSC pronunciation in pinyin (in some cases followed by a literary pronunciation [cf. Box 6], marked by a slash); • the OC phonological reconstruction, marked by an asterisk; The OC pronunciations follow Schuessler (2007, 2009); wherever Schuessler offers two alternative re­constructions, the one closer in accord with Baxter and Sagart 2014 has been chosen.

• For words, which may be represented by different characters, the entry for an ortho­graphic variant will simply refer to the main entry; The main entry, in principle, is to be found under the least ambiguous character (e.g. 歟 as opposed to 與, which can represent several words); note that this may not ­necessarily be the oldest or most frequently used form.

• the word class; Since no other dictionary of Classical Chinese supplies word classes, all word classes have been deter­mined specifically for this glossary. Especially the verbal sub­classes are unique to this glossary; it may therefore be useful to review their criteria of classification (cf. 1.2.3 and Box 12): • stative verbs are characterized by intransitive, putative (and factitive), and/or comparative usage (with 於 or an adverb); used as modifiers, they may be understood like adjectives; • intransitive verbs are characterized by intransitive and factitive (but not putative) usage; as modifiers, they must be understood as present participles;

401

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

• transitive verbs are characterized by transitive and passive usage (with or without 於); used as modifiers, they must be understood as past participles; • neutral verbs are characterized by transitive and intransitive (with or without 於) usage; ditransitive usage is possible (alternatively expressed by a single O and a PrP); usage as modifier is highly unusual.

• English translations in quotation marks; • citations taken from the textbook and from additional sources (indicated in parenthe­ses); in some cases, descriptive citations will follow directly after the trans­lation; most citations, however, serve to illustrate the u ­ sage of words, different usages being marked by boldface letters; boldface ­numerals separate usages usually associated with different word classes, e.g. verbal vs. nominal usage; The citations also serve to account for the word classes: for example, entries for stative verbs will usual­ly give a intransitive, b causative (possibly putative), c comparative, and d modifier usage. Punctuation is largely restricted to the comma, which may take the place of other punctuation marks found in the original edition; full stops at the end of clauses are omitted.

• in some cases, run-­on entries of phrasal usages, separated by a dash; • occasionally, notes discussing the usage of words as well as cross-­ references to syno­nyms, antonyms, or alternative characters. For the abbreviations used, cf. the list of abbreviations.

402

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

List of 214 Classifiers 1 stroke 1 2 3 4 5 6

一 丨 丶 丿 乙 亅

2 strokes 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

二 亠 人 儿 入 八 冂 冖 冫 几 凵 刀 力 勹 匕 匚 匸 十 卜 卩 厂 厶 又

3 strokes 3 0 口 31 囗 32 土

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

士 夂 夊 夕 大 女 子 宀 寸 小 尢 尸 屮 山 川 工 已 巾 干 幺 广 廴 廾 弋 弓 彐 彡 彳

4 strokes 6 1 62 63 64 65 66 67 68

心 戈 戶 手 支 攴 文 斗

6 9 70 7 1 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94

斤 方 无 日 曰 月 木 欠 止 歹 殳 毋 比 毛 氏 气 水 火 爪 父 爻 爿 片 牙 牛 犬

5 strokes 95 96 97 98 99 100 101 102 103 104

403

玄 玉 瓜 瓦 甘 生 用 田 疋 疒

105 106 107 108 109 110 111 112 113 114 115 116 117

癶 白 皮 皿 目 矛 矢 石 示 禸 禾 穴 立

6 strokes 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140

竹 米 糸 缶 网 羊 羽 老 而 耒 耳 聿 肉 臣 自 至 臼 舌 舛 舟 艮 色 艸

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

141 142 143 144 145 146

虍 虫 血 行 衣 西

7 strokes 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163

見 角 言 谷 豆 豕 豸 貝 赤 走 足 身 車 辛 辰 辵 邑

1 64 酉 165 釆 166 里

8 strokes 1 67 168 169 170 171 172 173 174 175

金 長 門 阜 隶 隹 雨 靑 非

9 strokes 176 177 178 179 180 181 182 183 184

面 革 韋 韭 音 頁 風 飛 食

185 首 186 香

10 strokes 187 188 189 190 191 192 193 194

馬 骨 高 髟 鬥 鬯 鬲 鬼

11 strokes

2 02 黍 203 黑 204 黹

13 strokes 2 05 206 207 208

黽 鼎 鼓 鼠

14 strokes 209 鼻 210 齊

魚 鳥 鹵 鹿 麥 麻

15 strokes

12 strokes

17 strokes

195 196 197 198 199 200

201 黃

404

211 齒

16 strokes 212 龍 213 龜

214 龠

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 1  一 0

一/壹  yī, *ʔit, vi, 1 ‘be one, first, single, whole’, a itr, of­ten emph., ‘one and the same’: 黃黑馬一也 (GLong 2); 所道則異,而所 繩則一也 (SJS 7); b caus, ‘unite, unify’: 用 兵之道,務在一賞 (SJS 6); 教誨之,調 一之 (Xun 16); ‘have one’: 齊人有一妻一 妾而處室者 (Meng 4B33); 晏子一狐裘三 十年 (Li 4); c mod, often emph., ‘a single’: 我 何愛一牛 (Meng 1A7); ‘one and the same’; 江河之水非一源之水也 (Mo 1); ‘the whole’: 一國慕之 (Meng 4A6); 是一國皆 善也 (SJS 18); 2 adv, ‘once, once and for all’: 願比死者壹洒之 (Meng 1A5); 3 nc, ‘one’: 二與一為三 (Zhuang 2); ‘unity, a whole’: 萬物與我為一 (Zhuang 2); ‘the first’: 何謂 四維,一曰禮 (Guan 1). Cf. Box 3. 1

丁  dīng, *têŋ, ntemp, 4th of the ‘Heavenly stems’: 秋八月丁卯,大事于大廟 (Zuo 6.2). Cf. Focus 8.

七  qī, *tshit, vi, ‘be seven’, a itr: 強國刑 七賞三 (SJS 4); b mod: 七教,父子兄弟 夫婦君臣長幼朋友賓客 (Li 5). Cf. Box 3. 2

三  sān, *sə̂m, v , 1 ‘be three’, a itr: 天地 i

與人,三而成德 (CQFL 24); b caus: 三 其本,若為食 (Guan 78); c mod: 三人行 (LY 7.22); 2 nc, ‘three’: 二與一為三 (Zhuang 2). Cf. 參 [28.9] and Box 3.

三  sàn, *sə̂ms, adv [< preceding], ‘thrice’: 季文子三思而後行 (LY 5.20); 禹 八年於外,三過其門而不入 (Meng 3A4); 三折肱知為良醫 (Zuo 11.13). Cf. 參 [28.9] and Box 3.

上  shàng/shǎng, *daŋʔ, vi, ‘ascend, rise’, a itr: 釋椎鑿而上 (Zhuang 13); 上於東階 則先右足,上於西階則先左足 (Li 1); b pseudo-­tr: 師曠至,履而上堂 (SY 18); 上山則見虎,下澤則見蛇 (Yan 2.10); c

caus, ‘raise, exalt, promote’: 上農除末, 黔首是富 (SJ 6); 上賢以崇德 (Li 5); syn. 尚 [42.5]. Ant. 下 [1.2].

上  shàng, *daŋh, nloc, 1 ‘top, upper part’, a head: 莊子與惠子遊於濠梁之上 (Zhuang 17); usually best translated as ‘on top of, above’: 我知之濠上也 (Zhuang 17); ‘up in/on’: 王坐於堂上 (Meng 1A7); fig. ‘the foremost, best’: 君子義以為上 (LY 17.20); 生而知之者,上也 (LY 16.9); b mod: 後舉而加諸上位 (Meng 5B6); in dates, ‘the first’: 十月上辛 (Zuo 12.13); ant. 下 [1.2]; cf. also 中 [2.3]; 2 adv, ‘upward’: 不敬宗廟則民乃上校 (Guan 1); 3 nc, ‘supe­rior, ruler’: 夫上之所為,民之歸也 (Zuo 10.21); 上好禮,則民莫敢不敬 (LY 13.4); ant. 下 [1.2], 民 [83.1]. 上帝  Shàngdì,

*daŋh-­ têh, np, ‘High God’, thought to be the highest ancestor of the Shang 商 dynasty: 紂 . . . 不肯事上帝 (Mo 27).

下  xià, *grâʔ, nloc, 1 ‘bottom, lower part’, a head: 邑于岐山之下居焉 (Meng 1B15); usually to be trans­lated as ‘below, underneath, at the bottom of’: 從山下望木 者,十仞之木若箸 (Xun 21); b mod, ‘low, sub­ordi­nate’: 下官之冀遷者 (SJS 3); in dates, ‘the last’: 十二月下辛 (GLiang 13.1); 2 nc ‘subordi­nate’: 為上則不能愛下 (Xun 5); 所惡於上,毋以使下,所惡於 下,毋以事上 (Li 42). Ant. 上 [1.2]; 高 [189.0].

下  xià, *grâh, vi, ‘descend, go down; submit (to), surrender’, a itr: 揖讓而 升,下而飲 (LY 3.7); 水無有不下 (Meng 6A2); 原人聞之乃下 (Lü 19.6); b pseudo­tr: 讓賢而下之,臣之忠也 (Lü 14.6); 樂 其以貴下人 (SJ 47); c caus, ‘send down, sub­ordinate’: 今父之親子也,親賢而下 無能 (Li 32). Ant. 上 [1.2].

丈夫  zhàngfū, *draŋʔ-­pa, nc, ‘(adult) man, husband’: 有賤丈夫焉 (Meng 2B10); 丈夫不耕 . . . 婦人不織 (HFei 49). Ant. 婦人 [38.8].

405

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

3

不  bù, *pə, adv, ‘not’, negates VP: 君子 不憂不懼 (LY 12.4). Cf. 否 [30.4], 無 [86.8], 勿 [20.2], 弗 [57.2], 非 [175.0]. 不肖  bùxiào,

*pə-­siauh, vst, ‘be ­ nworthy, unfit, in­capable’, a category for u judging people, a itr: 不肖而不肯事賢 (Xun 5); b comp: 嬰最不肖,故宜使楚耳 (SY 12); c mod: 不肖者使不肖主 (SY 12). Ant. 賢 [154.8].

丑  chǒu, *rhuʔ/nhruʔ, n

, 2nd of the ‘Earthly branches’: 乙丑,趙穿殺靈公於 桃園 (Zuo 7.2). Cf. Focus 8. temp

4

丙  bı̌ng,

*praŋʔ, ntemp, 3rd of the ‘Heavenly stems’: 以丙丁日為大赤龍一 (CQFL 74). Cf. Focus 8.

且  qiě, *tshaʔ, conj, 1 ‘and, moreover’, a

coordi­nating VP: 不義而富且貴 (LY 7.16); 孔子貧且賤 (SJ 47); b coordinating clauses: 事不可必也,且政在焉 (Zuo 10.25); cf. 15.2; syn. 又 [29.0], 亦 [8.4]; cf. 而 [126.0]; 2 adv, ‘be about to’: 我且賢之 用,能之使 (HFei 33); syn. 將 [41.8].

丘  qiū, *khwə, n , 1 ‘hill, mountain’: 黃 c

帝 . . . 登乎崑崙之丘 (Zhuang 12); 禱於 尼丘得孔子 (SJ 47); syn. 山 [46.0], 阜 [170.0]; 2 np, given name: 是魯孔丘與 (LY 18.6); cf. Focus 5.

世  shì, *lhats, nc, ‘generation; era, age’: 故書其言以遺後世 (Mo 17); 治當世之 民(HFei 49). 7

並  bìng, *bêŋʔ, adv, ‘side by side, jointly, to­gether’: 天地與我並生 (Zhuang 2). Classifier 2  丨 3

中  zhōng, *truŋ, n

loc, ‘the middle, average’, a head: 武王與紂戰於牧野之中

(SJS 17); 其劍自舟中墜於水 (Lü 15.8); b mod: 中世上賢而說仁 (SJS 7); 今居中 國,去人倫,無君子,如之何其可也 (Meng 1B11). Cf. 上 [1.2], 下 [1.2], 中國 [31.8].

中  zhòng, *truŋh, vn,‘hit (a target); match’, a tr: 祝聃射王,中肩 (Zuo 2.5); 其曲中規 (Xun 1); 言不中法者,不聽也,行不中 法者,不高也,事不中法者,不為也 (SJS 23); b itr: 秦無韓魏之規,則禍必中 於趙矣 (SJ 69). 中庸  zhōngyōng, * truŋ-­loŋ, vi, ‘be regular, proper, im­partial’, a itr: 君子中 庸,小人反中庸 (Li 31); b mod: 中庸民 不待政而化 (Xun 9). Classifier 3  丶 2

丸  wán, *wân, nc, ‘pellet, bullet; ball; egg’: 而觀其辟丸也 (Zuo 7.2). 4

主  zhǔ, *toʔ, nc, 1 ‘principal, lord, ruler, master’: 國日安,主日尊,天下日富, 此所謂吉主也 (Lü 24.6); 狗固吠非其主 也 (ZGC 13.5); 2 vi, ‘be principal/most important’, a caus: 主忠信, 無友不如己 者 (LY 1.8); 吾欲南伐,何主 (Guan 20); b mod: 主車則造父為御 (Lie 3). Note: 主 is the most general word for ‘master’, applicable to any­body from a dog owner to a feudal lord; 君 [30.4] more specifically means the ruler of a polity; 王 [96.0], originally the title of the Zhou kings, came to mean an exemplary ruler (just like 聖人 [128.7] often refers to an ideal ruler), and it was appro­pri­ated as a title by rulers of Zhanguo states; 霸 [173.13], by contrast, means the ‘hegemon’ who ruled by force. 主  zhǔ, *toʔ, vn, ‘rule (over), control, be in charge (of)’, a tr: 驪姬請使申生主曲沃 (GY 7.3); 君人民,主社稷,治國家 (Mo 8); b itr: 今夫齊,長主而自用也 (SJ 69).

406

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 4  丿 1

乃  nǎi, *nə̂ʔ, conj 1 a consecutive, ‘then’:

原人聞之乃下 (Lü 19.6); 數奪民時,大 饑乃來 (Lü 26.3); b temporal, ‘there­ upon’: 武王又再拜稽首,乃出 (SJ 4); cf. 15.4.2–㉝; cf. 16.3; 2 ptemp, 使白馬乃馬也 (GLong 1.2).

2

久  jiǔ,

*kwəʔ, vi, ‘last a long time’: 久,彌異時也 (Mo 40); a itr: 天下之無 道也久矣 (LY 3.24); 久矣,吾不復夢見 周公 (LY 7.5); esp. ‘stay a long time’: 久於 齊,非我志也 (Meng 2B14); b caus: 久 師則兵敝矣 (SJ 69).

3

之  zhī, *tə, prper, ‘him, her, it, them’, a as  Od: 父能生之,不能養之,母能食 之,不能教誨之 (Xun 19); b as Oi: 予之 法制,告之訓典 (Zuo 5.23); cf. 5.1–3 and Box 5; c as 'dummy' O= for ex­posed O: 夏 禮吾能言之 (LY 3.9); for anteposed O: 子 是之學 (Meng 3A4); cf. 14.2. 之  zhī, *tə, pt [] lose’: 司馬以 吾故亡其良子 (Zuo 10.21); ant. 存 [39.3]; c mod: 亡人無以爲寶 (Li 42); 2 adv, used to negate the verb 有: 諸侯之相 攻國者亡有 (Mo 14); 故盜賊亾有 (Mozi 4.14); syn. 無 [86.8]; ant. 有 [74.2]; cf. 忘 [61.3].

4

亥  hài, *gə̂ʔ, n

temp, twelfth of the ‘Earthly branches’: 六月丁亥,公假于大廟 (Li 25). Cf. Focus 8.

交  jiāo, *krâu, v , 1 ‘mingle, associate n

with, ex­change’, a itr: 與朋友交而不信乎 (LY 1.4); b tr: 士無賢不肖皆謙而禮交之 (SJ 77); 2 adv, ‘mutually, one another’: 上 下交征利,而國危矣 (Meng 1A1); 周鄭 交惡 (Zuo 1.3).

亦  yì, *jak, conj, 1 a coordinating, ‘also,

as well, likewise’, referring to the S: 舜人 也,我亦人也 (Meng 4B28); referring to the P: 知者不失人,亦不失言 (LY 15.8); syn. 且 [1.4]; b adversative, ‘nevertheless, nonetheless’: 寡人雖死,亦無悔焉 (Zuo 1.3); 雖不識義,亦不阿惑 (GY 7.4); cf. 16.2; 2 adv, ‘though, after all; sure­ly’: 學而 時習之,不亦說乎 (LY 1.1); 王何必曰 利,亦有仁義而已矣 (Meng 1A1).

享  xiǎng, *haŋ, v

n [< preceding], ‘feast, enter­tain; offer, sacrifice’: 享,獻也 (Xun 27); 珍,享,獻也 (Erya 1); 五官致 貢,曰享 (Li 2); a tr: 請享靈公 (Zuo 11.13); b tr2: 定王享之肴烝 (GY 2.7); c tr with PrP: 璧琮九寸,諸侯以享天子 (ZL 80); d itr: 賓服者享 (Xun 18); 古者先 王日祭,月享,時類,歲祀 (GY 18.2).

Classifier 9  人 0

人  rén, *nin, n , 1 ‘person, man’, meanc

ing a hu­man being of either sex: 人有氣有 生有知,亦且有義,故最為天下貴也 (Xun 9); 人之所以為人者,非特以二 足而無毛也,以其有辨也 (Xun 5); ‘s.o. else, others’: 人之所教,我亦教之 (Lao 42); 人皆有兄弟,我獨無 (LY 12.5); 不 戰而屈人之兵 (Sun 3); 2 adv, ‘like a man’: 豕人立 (Zuo 3.8).

1

亾  var. of 亡 [8.1]. 2

介  jiè,

*krêts, nc, ‘armour; armoured soldier’: 既而與為公介 (Zuo 7.2).

今  jīn, *kə̂m, n

temp, ‘now, currently’, a ini: 今老矣,無能為也已 (Zuo 5.30); often in a modal sense, in­troducing a new point: 今 公子蘭 . . . 必將為君 (Zuo 7.3); frequently in combination with 夫: 今夫水,搏而躍 之,可使過顙 (Meng 6A2); b mod: 昔者 疾,今日愈 (Meng 2B11); c head: 象刑殆 非生於治古,並起於亂今也 (Xun 18). Ant. 昔 [72.4], 古 [30.2]; cf. 夫 [37.1].

仁  rén, *nin, nc, 1 ‘humaneness, benevolence, kindness, compassion’: 小所以事

409

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

大,信也,大所以保小,仁也 (Zuo 12.7); 仁之實,事親是也 (Meng 4A27); 仁愛也,故親 (Xun 27); 溫潤而澤,仁 也 (Xun 30); 心兼愛人謂之仁,反仁為 戾 (XS 8.3); 2 vst, ‘be humane, benevolent’, a itr: 雍也仁而不佞 (LY 5.5); 仁者能仁 於人,而不能使人仁 (SJS 18); b caus, ‘treat as humane’: 愛之而弗仁 (Meng 7A45). Note: Within the teachings of Con­fucius as described in the Lunyu, 仁 is the highest moral virtue. Among the various explanations the master gives for the concept, the most succinct is: 愛人 (LY 12.2).

什  shí, *gip, nc, ‘group of ten; squad (of ten sol­diers)’: 使有什伯之器而不用 (Lao 80). 仇  chóu, *du, var. of 讎 [149.16]. 3

代  dài, *lə̂kh, vn, 1 ‘replace, supersede, succeed’, a tr: 若子死,將誰使代子 (HFei 22); b itr: 子之代乎父,弟之代乎 兄 (Xun 23); 2 adv, ‘successive­ly, one after another’: 暴君代作 (Meng 3B9); 3 nc, ‘line of succession, dynasty’: 三代之得天 下也以仁(Meng 4A3). Dist. 伐 [9.4]. 他/佗/它  tā/tuō, *lhâi, prdem, ‘other, anoth­er’, mod: 他日又獨立 (LY 16.13); 佗 邑唯命 (Zuo 1.1).

令  líng, *reŋ, vi, ‘be good, fine’, only mod: 子有令聞,而美其室 (Zuo 9.15); 巧言令色,鮮矣仁 (LY 1.3). 令  líng,

*reŋ, vtr, ‘cause, induce’, a tr with C: 道者,令民與上同意也 (Sun 1); 天帝令我長百獸 (XX 2.4); 臣能令君勝 (SJ 65); b pass: 民無不令矣 (Lü 19.6); c pass with C: 故民可令農戰,可令游 宦,可令學問 (SJS 23). Syn. 使 [9.6].

令  lìng, *reŋ(h), vn, 1 ‘order, command’, a tr: 不從其所令,從其所行 (Li 33); b itr: 上令而民知所以應 (SJS 5); 乃令於國 曰 (SY 3); 2 nc, ‘command; commander’:

惟令是視 (Guan 15); 南陽無令,其誰可 而為之 (Lü 1.5). Syn. 命 [30.5].

令尹  lìngyı̌n, *reŋ(h)-winʔ, nc, ‘governor’, a title in 楚: 令尹子文,三仕為令 尹,無喜色 (LY 5.19); 令尹誅而楚姦不 上聞 (HFei 49).

仞  rèn, *nəns, nc, unit of length, approx. eight feet: 從山下望木者,十仞之木若 箸 (Xun 21). 以  yı̌ , *ləʔ, prep, 1 ‘with, by means of ’:

以肱擊之 (Zuo 8.2); with anteposed O: 政 以治民,刑以正邪 (Zuo 1.11); used to emphasize an Od: 以金賜其母 (HSWZ 9); 授之以政 (LY 13.5); cf. 9.2.2; used to emphasize a C: 有文在掌曰友,遂以名 之 (SJ 33); from the basic meaning several special meanings derive: ‘accord­ing to’: 以 道觀之,物无貴賤 (Zhuang 17); ‘be­cause of ’: 司馬以吾故亡其良子 (Zuo 10.21); ‘thanks to’: 以君靈不死 (Zuo 10.14); ‘at (a time)’: 以十月入 (Zuo 3.16); cf. 9.2; syn. 用 [101.0]; 2 conj, a subordinating VP, ‘and (thus)’: 古之民樸以厚,今之民巧以偽 (SJS 7); 君子淡以親,小人甘以絕 (Zhuang 20); 其聲和以柔 (Li 19); b subordi­nating clauses, ‘in order to, so as to’: 終夜不竊,以思 (LY 15.31); cf. 16.5; 3 vn, ‘take, use’, only tr: 虛而待之,彼自以之 (HFei 8); 釋法而以知,背功而以譽 (SJS 23); ‘be due to’: 我之不共,魯故之以 (Zuo 10.13); 此何也,以虧人自利也 (Mo 17); 強秦之所以不敢加兵於趙 者,徒以吾兩人在也 (SJ 81).

4

伐  fá, *bat, v , 1 ‘attack, strike’, 凡師,有 n

鐘鼓曰伐,無曰侵,輕曰襲 (Zuo 3.29); a tr: 昔文公與秦伐鄭 (Zuo 9.14); b itr: 湯 武征伐 (HFei 49); 且欲東伐 (SJ 5); cf. note at 征 [60.5]; 2 nc, ‘feat, ex­ploit’: 乃可以威民 而懼戎,且旌君伐 (GY 7.6). Dist. 代 [9.3].

伐  fá, *bat, v , ‘cut, fell (trees)’, a tr: 禁 tr

止伐木 (Li 6); b pass: 直木先伐,甘井 先竭 (Zhuang 20). Dist. 代 [9.3].

410

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

伏  fú, *bək, v , ‘crouch, duck; hide; lie in i

am­bush’, a itr: 冬伏而夏游 (Xun 26); b caus: 晉侯飲趙盾酒,伏甲,將攻之 (Zuo 7.2); c mod: 見寢石以為伏虎也 (Xun 21). Ant. 起 [156.3].

任  rén, *nəm, vn, ‘carry, bear; sustain, endure’, only tr: 故不大其棟,不能任 重,任重莫若棟,任國莫若德 (Wen 11); 任大惡三,行將安入 (GY 9.4). 任  rèn,

*nəms, [< preceding] nc, 1 ‘duty, burden, responsibility’: 任重而道 遠 (LY 8.7); 故國者,重任也 (Xun 11); 2 vn, ‘trust (in); entrust, ap­point’, a itr: 信則人任焉 (LY 17.5); b tr: 尚明也 者,舉賢而任之 (HFei 32); 任之以政 而殺之 (SY 13); c tr2: 而后舉益,任之 政 (SJ 2).

伍  wǔ, *nâʔ, nc, 1 ‘squad (of five)’: 合其 卒伍,置其有司 (ZL 59); 2 vi, ‘be a squad’, caus: 取我田疇而伍之 (Zuo 9.30).

6.2); cf. Focus 5; ‘earl’, third-­highest of the five aris­tocratic ranks of the Zhou: 王者之 制祿爵,公侯伯子男,凡五等 (Li 5); cf. Box 15.

伯/佰  bó, *prâk, n , ‘group of

hunc dred; pla­toon (of hundred soldiers)’: 使 有什伯之器而不用 (Lao 80).

何  hé, *gâi, print, ‘what, why’: a typically as P: 若是者何也(SJ 6); b as O: 天亦何欲 何惡 (Mo 26); c as S in rhetorical questions: 何非君土 (Zuo 10.7); d as C: 匡人 其如予何 (LY 9.5); e mod, ‘what kind of ’: 是何鳥也 (Lü 18.2). Cf. 3.4.3; 5.3.2; syn. 奚 [37.7], 胡 [130.5].

位  wèi, *wrəts, nc, 1 ‘position, rank’: 故 不踰節則上位安 (Guan 1); cf. 即位 [26.5]; 2 vn, ‘hold/assume a position (in)’, a itr: 一虛一滿,不位乎其形 (Zhuang 17); 卑高以陳,貴賤位矣 (Yi 7); b tr: 故 堯舜之位天下也 . . . 為天下位天下也 (SJS 14).

yú, *la, pr , ‘I’, a head: 余不 休  xiū, *hu, v , ‘rest’, a itr: 霜始降,則百 余/予  女忍殺 (Zuo 10.1); b mod, ‘my’: 啟予 per

i

工休 (Li 6); b caus: 休卒徒 (Zhuang 29).

足,啟予手 (LY 8.3). Cf. Box 5.

仰  yǎng, *ŋaŋʔ, v , ‘look up (at)’, a itr: 作  zuò, 見橋竦焉實而仰 (SY 3.6); 仰則觀象於 n

天,俯則觀法於地 (Yi 8); b tr: 仰之如 日月,敬之如神明 (Zuo 9.14); 仰天而 歎 (SY 3). Ant. 俯 [9.8].

仲  zhòng, *druŋh, nc, ‘medius’, the middle one in a row of brothers, used in courtesy names: 髙辛氏有才子八人,以伯 仲叔季爲序(Zuo 6.18); 孔子…字仲 尼,姓孔氏 (SJ 47). Cf. Focus 5.

*tsâk, vi, ‘arise, emerge; be active’, a itr: 是故聖人作,為禮以教人 (Li 1); 日出而作 (Zhuang 28); b caus, ‘make, build, produce, com­pose’: 聖人作 樂 (Li 19); c mod: 以此食作夫五萬 (SJS 15).

佐  zuŏ, *tsâih, [< 左, 48.2] v , ‘assist, n

support’, a tr: 及正考父佐戴武宣公,三 命茲益恭 (SJ 47); b itr: 今蘧伯玉為 相,史鰌佐焉 (SY 12).

伏羲  Fúxī, *bəkh-­hŋai, n , a mythical 佑/右/祐  yòu, *wəh, [< 右, 30.2] ruler: 伏羲神農教而不誅 (SJS 1); 及 燧 p

人伏羲始為天下 (Zhuang 16). Cf. Focus 7 and 32.3–⑩.

5

伯  bó, *prâk, nc, ‘elder, senior’, the eldest in a row of brothers, used in courtesy names: 髙辛氏有才子八人,以伯仲叔 季爲序(Zuo 6.18); 仲弓問子桑伯子 (LY

vn, 1 ‘assist, help’, a tr: 是以自天祐之 (Yi 7); 乃命于帝庭,敷佑四方 (SJ 33); b itr: 良臣將死,天命不祐 (Zuo 10.1); 良臣 不生,天命不祐 (GY 14.17); 2 nc, ‘assis­ tance; assistant’: 其右提彌明知之 (Zuo 7.2).

佚/逸/失  yì, *lit, vi [< 失, 37.2], 1 ‘escape, break free; be free, relaxed, idle,

411

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

unemployed; unbound, unrestrained’: a itr: 馬逸不能止 (GY 11.9); 頃公用是佚 而不反 (GYang 8.2); 故居不隱者思不 遠,身不佚者志不廣 (Xun 28); b caus: 乃逸楚囚 (Zuo 8.16); 勞我以生,佚我 以老,息我以死 (Zhuang 6); c mod: 舜 無佚民,造父無佚馬 (HSWZ 2); 2 nc, ‘idleness’: 以近待遠,以佚待勞 (Sun 7). Ant. 勞 [19.10].

conj, condition­al, ‘if, sup­posing (that)’: 使 白馬乃馬也 (GLong 1.2); cf. 15.4.1; final ‘so that’: 屏王之耳目,使不聰明 (Zuo 10.27); cf. 16.5.

依  yī, *ʔəi, v , ‘rely on, lean on’, a itr: 依 n

於庭牆而哭 (Zuo 11.4); b tr: 輔車相 依,唇亡齒寒 (Zuo 5.5); 鬼神非人實 親,惟德是依 (Zuo 5.5).

佃/田  tián/diàn, *lîns, v , ‘hunt; cul- 供  gōng, *koŋ, var. of 共 [12.4]. i

tivate, till’, a itr: 宣子田於首山 (Zuo 7.2); 夢哭泣者,旦而田獵 (Zhuang 2); 作結 繩而為罔罟,以佃以漁 (Yi 8); b mod: 田車數百乘 (Mo 31); syn. 獵 [94.15].

佗  tā/tuō, *lhâi, var. of 他 [9.3]. 6

來  lái, *rə̂, v , 1 ‘come’, a itr: 不與陰陽俱 i

往來 (CQFL 16); b caus, ‘attract’: 遠人不 服,則脩文德以來之 (LY 16.1); 故欲來 民者,先起其利 (Guan 64); cf. 致 [133.4]; c mod: 請輕之,以待來年 (Meng 3B8); ant. 去 [28.3], 往 [60.5]; 2 ptemp, hortatory: 盍歸乎來 (Meng 4A13); 長鋏歸來乎 (ZGC 11.1).

侍  shì, *dəh, vn, ‘be present, participate’, a itr: 顏淵季路侍 (LY 5.26); b tr: 侍 食於君,君祭,先飯 (LY 10.18); 閔子 侍側,誾誾如也 (LY 11.13); 臣侍君 宴,過三爵,非禮也 (Zuo 7.2).

使  shì , *srəh, n , ‘envoy, ambassador’: c

7

保  bǎo,

*pûʔ, vtr, ‘protect, safeguard, defend’, a tr: 乃先保南里以待之 (Zuo 12.27); 保民而王 (Meng 1A7); b pass: 惡 於宋而保於我 (Zuo 3.12).

便  biàn, *bens, vi, ‘be useful, favourable, bene­ficial’, a itr: 作為衣服帶履,便 於身 (Mo 6); b caus: 禮者,所以便事也 (SJS 1); 審法制,明賞罰,便器用 (WLiao 4); c mod: 自謂便人 (Li 32). 俄  é, *ŋâi, adv, ‘abruptly, suddenly’: 俄 然覺,則蘧蘧然周也 (Zhuang 2).

侯  hóu, *gô, n , 1 a title, ‘marquis’: 王者 c

之制祿爵,公侯伯子男,凡五等 (Li 5); 吾之不遇魯侯,天也 (Meng 1B16); 2 vi, a itr: ‘be marquis’: 始南皮章武侯先帝不 侯 (SJ 57); b caus: ‘ennoble, make mar­ quis’: 及臣即位乃侯之 (SJ 57); 百里之 君,而封侯其臣 (SJS 17). Cf. Box 15 and 諸侯 [149.8].

兵交, 使在其間可也 (Zuo 8.9); 夢天 使與己蘭 (Zuo 7.3).

侵  qīn,

使  shı̌, *srəʔ, vn, 1 ‘send, employ; make s.o. do sth., compel, cause to’, a tr: 君使 臣,臣事君 (LY 3.19); b tr with C: 王使 人來 (Meng 2A2); c itr with C: 使盡之 (Zuo 7.2); 齊侯使請戰 (Zuo 8.2); syn. 令 [9.3], 命[30.5]; cf. 史 [30.2], 吏 [30.3]; 2

俗  sú, *s-­lok, nc, ‘custom, common practice, tra­dition, usage’: 夫論至德者,不和 於俗,成大功者,不謀於眾 (ZGC 19.4).

使  shı̌, *srəʔ, vi [< preceding], 1 ‘be/ serve as an envoy’, a itr: 使於四方,不辱 君命 (LY 13.5); 今臣使於秦,而三日不 見 (ZGC 18.12); b pseudo-­tr: 晏子使 魯,見昭公 (Yan 4.12).

*tshəm, vtr, ‘invade, raid, encroach, vio­late’, 凡師有鍾鼓曰伐, 無 曰侵, 輕曰襲 (Zuo 3.29); a tr: 夏,鄭 人侵許 (Zuo 3.29); 以為侵官之害甚於 寒 (HFei 7); b pass: 非侵於諸侯,必劫 於百姓 (SJS 25); 故上失其道則邊侵於 敵 (Lü 3.3). Cf. note at 征 [60.5].

侮  wǔ, *moʔ, v , ‘scorn, insult, slight’, a tr

tr: 亦莫敢余侮 (SJ 47); b pass: 恭則不侮 (LY 17.5); 禁而不止則刑罰侮 (Guan 16).

412

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

信  xìn, *sin, n , 1 ‘trust; trustworthiness, c

relia­bility, faithfulness’: 小所以事大,信 也 (Zuo 9.8); 將者智信仁勇嚴也 (Sun 1); 期果言當謂之信,反信為慢 (XS 8.3); 2 vst, ‘be sincere, trust­worthy, truthful’, a itr: 與朋友交而不信乎 (LY 1.4); b caus, ‘[consider trustworthy >] trust’: 吾斯之未 能信 (LY 5.6); ‘make trustworthy’; 設法度 以齊民,信賞罰以盡民能 (HFei 48); ant. 疑 [103.9]; 3 adv, ‘truly, indeed’: 水信 無分於東西 (Meng 6A2). Syn. 誠 [149.6].

俎  zǔ, *tsraʔ, n , ‘small altar’: 孔子為 c

兒嬉戲,常陳俎豆,設禮容 (SJ 47).

俠  xiá, *gêp, nc, ‘knight, swordsman’: 俠以武犯禁 (HFei 49); 古布衣之俠,靡 得而聞已 (SJ 124). 8

倍  bèi, *bə̂ʔ, v , ‘be multiple, (x-)times; st

double’, a itr: 四時事備,而民功百倍矣 (Guan 53); b caus, ‘multiply, augment’: 雖 倍賞累罰而不免於亂 (HFei 49); 此皆十 倍其國之眾,而未能食其地也 (Mo 19); c comp: 其難萬倍於和氏之璧 (XX 5); 諸侯之地五倍於秦 (SJ 69); d pseudo­tr: 賦粟倍他日 (Meng 4A14).

倉  cāng, *tshâŋ, n , ‘storage, granary’:

也 (Meng 5A2); 眾異不得相蔽以亂其倫 也 (Xun 21). Note: The 五倫, ‘five relations’, between father and son, ruler and minister, husband and wife, old and young, comrades and friends, respectively, as described in Meng 3A4, came to be regarded as a basic structural principle of Chinese society.

修/脩  xiū,

*siu, vst, ‘be cultivated, refined’, a itr: 田野不修,人民不安 (SY  1); 行忍情性,然後能脩 . . . 公脩而 才,可謂小儒矣 (Xun 8); b caus, ‘cul­ti­ vate, take care of, practice’: 欲脩其身者, 先正其心 (Li 42); 湯武之王也,不脩古 而興 (SJS 1).

倚  yı̌, *ʔaiʔ, v , ‘incline, lean against/ n

on; rely on’, a tr: 倚柱彈其劍 (ZGC 11.1); 曾點倚其門而歌 (Li 4); b itr: 在輿則見 其倚於衡也 (LY 15.6); 中立而不倚 (Li 31); c mod (!), ‘crooked’: 父母之喪,居倚 廬 (Li 22); 飾邪說,文姦言,為倚事 (Xun 4).

倦  juàn, *gons, vi, ‘be weary, tired, worn out’, a itr: 學而不厭,誨人不倦 (LY 7.2); 文倦於事,憒於憂 (ZGC 11.1); b caus: 去則追之,以倦其師 (Wu 2); syn. 厭 [27.12].

c

不務地利則倉廩不盈 (Guan 1).

倒  dǎo,

*tâuʔ/tâuh, vtr, ‘turn, reverse, invert’, a tr: 倒戟以禦公徒而免之 (Zuo 7.2); b pass: 小斂之衣,祭服不倒 (Li 22); c mod: 君為倒君,臣為亂臣 (Guan 31).

俯  fǔ, *poʔ, vi, ‘look down; bow down, bow the head’, a itr: 見梓勃焉實而俯 (SY 3.6); b caus: 行而俯項,非擊戾也 (Xun 2). Ant. 仰 [9.4]. 俱  jù/jū, *ko, adv, ‘both, all; together, jointly, likewise’: 父母俱存 (Meng 7A20); 雖與之俱學, 弗若之矣 (Meng 6A9); 孔子墨子俱道堯舜 (HFei 50); 其俱是 也,其俱非也邪 (Zhuang 2).

倫  lún, *run, n , ‘kind, class, category; c

natural relations’: 男女居室,人之大倫

9

偝/背/北  bèi, *bə̂kh, v , ‘turn one's n

back on, turn away/a­ gainst, oppose; retreat’, a itr: 鄉則不若,偝則謾之 (Xun 5); 三戰三北 (HFei 49); b tr: 新與晉盟而 背之,無乃不可乎 (Zuo 8.15); 沛公不敢 背項王也 (SJ 7); c mod (!): 夫父之孝 子,君之背臣也 (HFei 49). Cf. 負[154.2].

側  cè, *tsrɘk, nc, ‘side’: 子食於有喪者 之側 (LY 7.9); 遁逃反側之民 (Xun 9).

假  jiǎ, *krâʔ, vn, ‘borrow; arrogate; make use of ’, a tr: 假輿馬者,非利足也 (Xun 1); 晉侯復假道於虞 (Zuo 5.5); 周 公旦假為天子七年 (HFei 37); b itr: 君 子 . . . 善假於物也 (Xun 1); c pseudo-­tr2: 不好假道人 (Xun 9).

413

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

假  jiǎ, *krâʔ, v , ‘lend, bestow’, a tr : 彼 n

2

不假我道 (HFei 10); 唯器與名,不可以 假人 (Zuo 8.2); 假人元龜 (SJ 3); b tr with PrP: 若以假人,與人政也 (Zuo 8.2); 忠 臣不借人以力,貞女不假人以色 (LNZ 4.11).

偽  wèi, *ŋwaih, vtr, ‘contrive, forge, fake; pre­ tend to’, 可學而能,可事而成 之在人者謂之偽 (Xun 17); a tr: 我偽逃 楚 (Zuo 8.16); 辭,偽有疾 (Zuo 11.9); b pass: 其善者偽也 (Xun 23).

偃  yǎn, *ʔanʔ, vi, 1 ‘lie down, rest’, a itr: 草上之風,必偃 (LY 12.19); b caus, ‘put to rest’: 反而定三革, 偃五兵 (Xun 8); 2 adv, ‘relaxedly, peace­fully’: 人且偃然寢於 巨室 (Zhuang 18); 偃然案兵無動 (Xun 9). 10

備  bèi, *brəks, n , 1 ‘provision, equipc

ment; preparations’: 兵甲器備各便其用 (SJS 1); 2 vn, ‘provide (for), prepare, take pre­cautions (against)’, a tr: 行善而備敗 (GY 1.3); 譬之若修兵者,以備寇也 (Lü 1.2); b itr: 夫姦必知則備 (HFei 46).

備  bèi, *brəks, v , 1 ‘be complete’, a itr: i

萬物皆備於我矣 (Meng 7A4); 不恭祖舊 則孝悌不備 (Guan 1); b caus: 故求備 物,不求豐大 (GY 18.2). Syn. 具 [12.6].

傑  jié, *gat, n , ‘hero’: 英傑至 (Xun 9).

c

賢不肖不雜則

11

傳  chuán, *dron, v , ‘relay, transfer, pass tr

on, transmit’, a tr: 堯欲傳天下於舜 (HFei 34); b pass: 名號不傳乎千世 (XX 5.1); c pseudo-­tr2: 今先聖人為書,而傳之後世 (SJS 26); 夏亡,傳此器殷,殷亡,又 傳此器周 (SJ 4). Cf. 8.2.4.

僂  lǚ, *rô/roʔ, v , ‘stoop, hunch down; i

subdue’, a itr: 一命而僂 (SJ 47); b caus: 僂 身而下聲 (Yan 1.22).

傾  qīng,

*khweŋ, vi, ‘lean, incline towards’, a itr: 心枝則無知,傾則不精

(Xun 21); b caus: ‘make inclined, topple, tumble, tilt’: 高下相傾 (Lao 2); 傾耳而聽 之,不可得而聞也 (Li 29).

傷  shāng, *lhaŋ, v , ‘hurt, harm, injure’: tr

a tr: 矢人唯恐不傷人,函人唯恐傷人 (Meng 2A7); b pass: 郤克傷於矢 (Zuo 8.2); c mod: 川壅而潰,傷人必多 (GY 1.3).

傴  yǔ, *ʔoʔ, vi, ‘bend; bow down; be hunch­backed’, a itr: 再命而傴 (SJ 47); b mod: 傴巫跛擊之事也 (Xun 9). 傲/驁  ào,

*ŋâuh, vn, ‘be arrogant, haughty, contemp­tuous’, a itr: 故君子不 傲不隱不瞽,謹順其身 (Xun 1); 生而 尊者驕,生而富者傲 (SY 3); syn. 驕 [187.12]; b tr, ‘treat with contempt’: 士驁 祿爵者,固輕其主 (Lü 15.3); 夫不忍一 世之傷,而驁萬世之患 (Zhuang 26); syn. 輕 [159.7], 慢 [61.11].

12

僨  fèn, *pəns, v , 1 ‘overturn, fall’, a itr: i

鄭伯 之車僨于濟 (Zuo 1.3); b caus: 此謂 一言僨事,一人定國 (Li 42); 2 nc, ‘downfall’: 重僨可疾 (GY 7.6).

13

儆  jı̌ng, *kreŋʔ, var. of 警 [149.13]. 億  yì, *ʔək, v , ‘be (one) hundred thoui

sand’: 方一里者為田九百畝 . . . 方百里 者,為方十里者百,為田九十億畝 (Li 5). Cf. Box 3.

價/賈  jià, *krâh, n , ‘price’: 粥羔豚 者弗飾賈 (SJ 47).

c

14

儒  rú, *no, nc, ‘scholar, classicist, ritualist; Con­fucian’: 世之顯學,儒墨也,儒 之所至,孔丘 也 (HFei 50); 儒,以道得民 (ZL 3); 夫儒者 以六藝為法 (SJ 130). Note: The word 儒 is related to 懦 [61.14], ‘tim­id’, 弱 [57.7], ‘weak’, and 柔 [75.5], ‘soft’: it was originally a pejorative name for scholars, especially

414

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

the fol­ lowers of Confucius, meaning something like ‘wimp’ or ‘weakling’. Later, it became the com­mon label for classical scholars or, as Western scholars call them, ‘Confucians’. Cf. Focus 23.

15

優  yōu, *ʔu, n , ‘entertainer, comedian’: c

公之優曰施,通於驪姬 (GY 7.6).

償  cháng, *daŋ(h), vn, ‘recompense, pay back; replace’, a itr: 使吏召諸民當償者 (ZGC 11.1); b tr: 羣臣請相夫人以償馬 (Zuo 11.3); 賣妻子不足償之(ZGC 10.9). Classifier 10  儿 2

元  yuán, *ŋon, nc, ‘head’, a head: 狄人 歸其元,面如生 (Zuo 5.33); b mod, ‘chief, first, primary, principal’: 元惡不待 教而誅 (Xun 9); 二世元年七月(SJ 48). 3

兄  xiōng, *hwraŋ, n , ‘older brother’: 其 c

兄自外至 (Meng 3B10); 四海之內皆兄弟 也 (LY 12.5). Syn. 昆 [72.4]; ant. 第 [57.4].

4

先  xiān, *sə̂n, v , 1 ‘be first, earlier’, a itr: i

年先矣,而無經緯本末以期年耆 者,是非先也 (Zhuang 27); b caus, ‘put first, give priority to’: 仁者先難而後 獲,可謂仁矣 (LY 6.22); 知所先後 (Li 42); c mod, ‘pre­ceding, former’: 是先王之 道 (Xun 19), cf. 前[18.7]; 2 adv: 吾為子先 行 (ZGC 14.3); 欲治其國者,先齊其家 (Li 42); 3 nc, ‘forebear’: 其先宋人也 (SJ 47). Ant. 後 [60.6].

先  xiàn, *sə̂ns, [< preceding] v , 1 ‘pren

cede, come before; outstrip’, a tr: 疾行先 長者謂之不弟 (Meng 6B2); 禹不先 鯀,湯不先契,文武不先不窋 (Zuo 6.2); b itr: 天先乎地,君先乎臣 (Li 11); 2 prep, ‘before’, 先天地生而不為久

(Zhuang 6); 衛侯先期入 (Zuo 5.28); 先 事慮事,先患慮患 (Xun 27).

先生  xiānshēng, *sə̂n-­sreŋ, nc, ‘elder; gentle­man; tu­tor, teacher’: 諸先生以文 學取 (HFei 49); as form of address, ‘sir, master’: 先生之年長矣 (Meng 6B4); 先 生之年長矣 (Lü 1.5).

充  chōng, *thuŋ, v , ‘be abundant, full, i

complete’, a itr: 君之倉廩實,府庫充 (Meng 1B12); 珍怪異物,充於後宮 (YTL 15); b caus: 美人充下陳 (ZGC 11.1); 苟能充之,足以保四海 (Meng 2A6).

5

克  kè, *khə̂k, vn, ‘defeat, subdue, overcome’, a tr: 克己復禮為仁 (LY 12.1); 武王 克商,遷九鼎于雒邑 (Zuo 2.2); 克之於 牧野 (Lü 5.5); b itr: 夫以彊取,不義而 克,必以為道 (Zuo 10.1); 強大有 道,不戰而克 (Wen 3). 免  miǎn, *mranʔ, v , ‘avoid, escape’, a i

itr: 吾不免是懼 (Zuo 9.22); 事之以珠 玉,不得免焉 (Meng 1B15); b caus, ‘rescue, release’: 倒戟以禦公徒,而免之 (Zuo 7.2). Dist. 兔 [10.6].

6

兒  ér, *ŋe, nc, ‘son, boy’: 老人兒啼 (SJ 119). Cf. 子 [39.0]. 兔  tù, *lhâh, n , ‘rabbit’: c

(HFei 49). Dist. 免 [10.5].

兔走,觸株

Classifier 11  入 0

入  rù, *nəp, vi, ‘go into, enter’, a itr, usually im­plying the O ‘door, home’: 農夫蚤出 暮入 (Mo 32); 入以事其父兄, 出以事 其長上 (Meng 1A5); 生人立焉,死人入 焉 (SY 2); b caus: 輔重耳入之於晉 (HFei 10); 君子之聞道,入之於耳,藏之於 心 (HSWZ 9); c pseudo-­tr: 主人入門而 右,客入門而左 (Li 1); d pseudo-­tr2: 有 餘 粟 者 入 之 倉  . . . 有 餘 錢 者 入 之 府

415

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(HFei 10); 吾欲輔重耳而入之晉 (HFei 10). Ant. 出 [17.3].

2

內  nèi, *nûts, n

loc, 1 ‘(the) inside’, a head: 河內凶則移其民於河東 (Meng 1A3); 故 君子務脩其內,而讓之於外 (Xun 8); 外無敵國之患,內無亂臣之憂 (HFei 14); b mod: 君多內寵 (Zuo 2.11); 我以鄭 為內臣 (Zuo 5.7); 2 adv: 外舉不避 讎,內舉不避子 (Lü 1.5); 無以好惡內傷 其身 (Zhuang 5); 3 vi, ‘be internal, inside’, a itr: 弗知內矣,知之外矣(Zhuang 22); b caus: 故內其國而外諸夏,內諸夏而 外夷狄 (CQFL 6). Ant. 外 [36.2].

4

全  quán, *dzon, v , ‘be complete, whole, i

intact’, a itr: 子固非魚也,子之不知魚 之樂,全矣 (Zhuang 17); b caus: 善用兵 者持五殺以應,故能全其勝 (HNan 15); c mod: 全國為上,破國次之 (Sun 3).

6

兩  liǎng, *raŋʔ, v , 1 ‘be

two, double, i both’, a itr: 心未嘗不兩也 (Xun 21); b caus, ‘double, make/have two’: 一國而兩 君,一國不可理也,一家而兩父,一 家不可理也 (Guan 23); 行有進退,而 力不能兩也 (Guan 53); 2 adv, ‘twice; both, as a pair’: 冬與夏不能兩刑,草與 稼不能兩成 (Lü 24.5); 力不能兩護,故 棄妾之子 (LNZ 5.6); 今君高拱而兩有 之 (SJ 69). Cf. Box 3 and note at 二 [7.0].

Classifier 12  八 0

八  bā, *prêt, vi, ‘be eight’, a itr: 律中太 蔟,其數八 (HNan 5); b caus: 太史八之 (GY 1); c mod: 八卦定吉凶 (Yi 7). Cf. Box 3. 2

公  gōng,

*kôŋ, vi, 1 ‘all-­embracing, public-­spirit­ed; impersonal, impartial,

fair’: 兼覆無私謂之公,反公為私 (XS 8.3); a itr: 君身善則不公矣 (Guan 30); b mod: 匹夫有私便,人主有公利 (HFei 47); 少私義則公法立 (Lü 26.3); 2 nc, ‘public interest, public spirit’: 公私之分 明,則小人不疾賢 (SJS 14); 不以私害 公 (LNZ 5.6). Ant. 私 [115.2].

公  gōng, *klôŋ, nc, ‘duke’, highest of the five aris­tocratic ranks of the Zhou: 王者之 制祿爵,公侯伯子男,凡五等 (Li 5); 周公何人也 (Meng 2B9); with posth. names: 鄭武公莊公為平王卿士 (Zuo 1.3). Cf. Box 15.

公子  gōngzı̌, *klôŋ-­tsəʔ, nc, '(ducal) prince': 今公子蘭 . . . 必將為君 (Zuo 7.3). Note: 公子 was the official title of a ruler’s male offspring, only the heir apparent being called 太子 [37.1] 六  liù/lù, *ruk, v , ‘be six’, a itr: 定得田 i

方十里者六 (CQFL 28); b caus: 五而六 之,九而十之,不可為數 (Guan 51). Cf. Box 3.

4

共/供  gōng, *koŋ, v , 1 ‘provide, pretr

sent (tribute)’, a tr: 我之不共,魯故之以 (Zuo 10.13); b pass: 王祭不共,無以縮 酒 (Zuo 5.4); 2 nc, ‘sup­ply, provision’: 事 力勞而供養薄 (HFei 49).

共  gòng, *goŋh, vn, 1 ‘join in, share, have in com­ mon’, a tr: 吾與汝共之 (Zhuang 6); 魏王與龍陽君共船而釣 (ZGC 25.25); b itr: 願車馬,衣輕裘,與 朋友共 (LY 5.26); 2 adv, ‘all, jointly, together’: 萬人操弓共射一招 (Lü 1.2); 少事長, 賤事貴, 共帥時 (Li 12). 5

兵  bīng, *praŋ, nc, 1 ‘weapon; soldier, troops’: 兵者不祥之器 (Lao 31); 雖有甲 兵,無所陳之 (Lao 80); 凡所為有兵 者,為爭奪也 (Xun 15); 國大者人 眾,兵彊則士勇 (SJ 87); 2 vn, ‘attack (with weapons)’, tr: 士兵之 (Zuo 11.10).

416

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

6

典  diǎn,

*tə̂nʔ, nc, ‘statute, norm; codex’: 予之法制,告之訓典 (Zuo 5.23).

其  qí, *gə, prper, 1 ‘his, her, its, their’, a mod: 勿奪其時 (Meng 1A3); 今我非 其主也 (Lü 10.5); ‘proper, appropriate, right’: 天下不一,諸侯俗反,則天王 非其人也 (Xun 10); 義,非其門而由 之,非義也 (Xun 27); b in nominal­ized clauses: 夫子何以知其將見殺 (Meng 7B29); 2 pt, used to subordi­nate prdem: 彼其道遠而險 (Zhuang 20); 上下之利 若是其異也 (HFei 49). Cf. Box 5 and Box 18.

其  qí, *gə, adv, ‘possibly, probably, surely’, a in verbal clauses: 虢公其奔 (Zuo 5.5); typically with final 乎: 吳其亡 乎 (SY 9); 堯舜其猶病諸 (LY 6.30); with final 與: 其晏子之謂歟 (Yan 2.21); with final 也: 吾其靜也 (GY 7.4); with final 矣: 微管仲,吾其被髮左衽矣 (LY 14.16); hortatory, ‘should, had better’: 子 其圖之 (Zuo 8.16); in rhetorical questions: 其誰不知 (Zuo 5.32); 其何以行之 哉 (LY 2.22); cf. 豈 [151.3], 庸 [53.8]; b in nominal clauses: 無為而治者,其舜也 與(LY 15.5); 其田成氏乎 (HFei 34). Cf. 7.2 and Box 18. 具  jù, *goh, v , 1 ‘be complete, together’, i

a itr: 禮固自孔子時而其經不具 (SJ 121); 其功大者其樂備, 其治辯者其禮具 (Li 19); b caus, ‘put together, arrange, equip, prepare’ 繕甲兵,具卒乘 (Zuo 1.1); 居則具一日之積,行則備一夕之 衛 (Zuo 5.33); 古者諸侯出疆, 必具官 以從 (SJ 47); c mod, ‘complete, consummate’: 今由與求也,可謂具臣矣 (LY 11.24); 2 adv, ‘completely, all’: 良乃入,具 告沛公 (SJ 7); syn. 備 [9.10]; 3 nc, ‘equipment, utensil, instrument’: 法令者治之 具, 而非制治清濁之源也 (SJ 122).

8

兼  jiān, *kêm(s), v , 1 ‘unite, combine, tr

em­brace’, a tr: 孔子兼之 (Meng 2A3); esp.

‘seize, conquer’: 秦兼天下 (ZGC 10.12); b pass: 技兼於事,事兼於義,義兼於 德,德兼於道,道兼於天 (Zhuang 12); 陽兼於陰,陰兼於陽,夫兼於妻,妻 兼於夫,父兼於子,子兼於父,君兼 於臣,臣兼於君 (CQFL 53); 殷兼於 周,陳亡於楚 (SY 8); 2 adv, ‘all, universally, comprehensively’: 若使天下兼相 愛 (Mo 14); 兼利天下 (Xun 9); cf. Box 30. Ant. 別 [18.5].

14

冀  jì, *krəih, vn, ‘hope’, a tr: 冀復得兔 (HFei 49); 冀人之以己為公也 (Xun 8); b itr: 吾冀而朝夕修我 (GY 5.13). Classifier 13  冂 4

再  zài, *tsə̂h, v , 1 ‘be two’, a itr: 如是 i

者再三 (HFei 32); b caus, ‘double’: 事再 其本,民無饘者賣其子 (Guan 78); 一 之謂甚,其可再乎 (Zuo 5.5); 2 adv, ‘twice’: 公再拜稽首(Li 4). Cf. note at 二 [7.0].

Classifier 14  冖 7

冠  guān, *kôn, n , ‘cap’: 為人子, 父母 c

存, 冠衣不純素 (Li 1); 君子正其衣冠 (LY 20).

冠  guàn, *kôns, vn, ‘wear/put on a cap’, a itr: 許子冠乎,曰冠 (Meng 3A4); 衣冠 而見之 (ZGC 11.1); b tr: 衣威王之服,冠 其冠,帶其劍 (Lü 9.3). 冠  guàn, *kôns, v , ‘confer a cap’, as a tr

ceremony mark­ing the entry into adulthood, a tr: 將冠子而未及期日 (Li 7); 大 功之末,可以冠子,可以嫁子 (Li 21); b pass: 男子二十冠而字 (Li 1); 丈夫之冠 也,父命之 (Meng 3B2); c mod: 冠子不 言,髮子不笞 (HSWZ 7).

417

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 15  冫

3

3

冬  dōng, *tûŋ, n

temp, ‘winter’: 冬與夏不 能兩刑,草與稼不能兩成 (Lü 24.5); 天 有四時,春秋冬夏 (Li 29); 冬者,天之 威也 (CQFL 79).

4

冰  bīng, *prəŋ, nc, ‘ice’: 冰,水為之而 寒於水 (Xun 1). 8

凍  dòng, *tôŋh, v , ‘freeze’, a itr: 父母凍 i

餓 (Meng 1A5); 水始冰,地始凍 (Li 6); b caus: 比其反也,則凍餒其妻子 (Meng 1B6); c mod: 國有凍人,人主不 裘 (XS 6.1).

出  chū, *k-­hlut, vn, ‘go out, exit, emerge, depart (from); proceed (from), take (sth.) as point of de­parture’, a itr: 明日,王出 (HFei 31); 季孟自南門入,出自東門 (Zuo 11.6); 順人之情,必出於争奪 (Xun 23); b tr: 出門如見大賓 (LY 12.2); 不出戶而知天下 (HSWZ 3); ‘outdo, outvie’: 賢者以相出為道 (SJS 7). Ant. 入 [11.0]. 出  chū/chuì, *k-­hluts, v

tr [< preceding], ‘bring/send/cast out, expel; hand over’, a tr: 孟子惡敗而出妻 (Xun 21); 彼如出 之,可以得荊 (HFei 23); b mod: 夫無逆 出妻之喪 (GLiang 8.9).

Classifier 18  刀 2

Classifier 16  几

分  fēn,

1

凡  fán, *bam, pt, ‘generally, whenever, in every case’, mostly ini, a with S: 凡有血 氣者,莫不尊親 (Li 31); b with exp O: 凡我國之忠信之士,我將賞貴之,不 忠信之士,我將罪賤之 (Mo 13); c with absolute topic: 凡人臣之事君也,多以 主所好事君 (SJS 14); d with num, ‘altogether’: 凡三百有六十 (Yi 7); 凡火攻有 五 (Sun 12).

Classifier 17  凵

*pən, vtr, ‘divide; distinguish, differenti­ate’, a tr: 韓魏趙三子分晉 (HFei 44); 人性之無分於善不善也,猶水之 無分於東西也 (Meng 6A2); b pass: 澶漫 為樂,摘僻為禮,而天下始分矣 (Zhuang 9); c mod: 輯其分族 (Zuo 11.4).

分  fèn, *bəns, vn, 1 ‘apportion, allot, distribute (to)’, a tr: 分人以財謂之惠 (Meng 3A4); 分貧振窮, 長孤幼 (Zuo 10.14); b tr2: 今君分之土而官之 (GY 7.7); 分之都 城 (4.1); 2 nc ‘part, divi­sion; differ­ence; share, lot; position’: 禮樂則脩,分義則 明 (Xun 16); 周與胡蝶,則必有分矣 (Zhuang 2); 男有分,女有歸 (Li 9). 4

2

列  liè, *rat, n , ‘rank, hierarchy; order’, a 凶  xiōng, *hoŋ, v , 1 ‘be bad, disastrous, head: 宋人既成列 (Zuo 5.22); b mod c

st

inaus­pi­cious; suffer calamity, be in misery’, a itr: 河內凶,則移其民於河東 (Meng 1A3); b comp: 天下之物,莫凶於 雞毒 (HNan 9); c mod: 吉事尚左,凶事 尚右 (Lao 31); 2 nc, ‘inauspicious­ness; calamity’: 八卦定吉凶,吉凶生大業 (Yi 7); 以是為非,以吉為凶 (HSWZ 4). Ant. 吉 [30.3].

‘ranked’: 使公卿至於列士獻詩 (GY 1.3).

刑  xíng, *gə̂ŋ, n , 1 ‘punishment’: 政以 c

治民,刑以正邪 (Zuo 1.11); 夫刑者所以 禁邪也 (SJS 6); 2 vtr, ‘punish’, a tr: 及陷乎 罪,然後從而刑之,是罔民也 (Meng 3A3); b pass: 殺人者死,傷人者刑 (Xun 18); c mod: 國無刑民 (SJS 17). Note: 刑

418

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

denotes physical punishment, esp. any of the ‘five punish­ments’: tattooing the face (墨), cutting off the nose (劓), cutting off the feet (刖), castration/sterilization (宮), and capital punish­ment (大辟). Besides physical punishment, penal­ ties (罰) could include fines such as money, ar­mours, or other goods.

beneficial/of use (to), benefit’, a itr: 利於 民不利於君 (SY 1); b tr: 上可而利 天,中可而利鬼,下可而利人 (Mo 10); 亦將有以利吾國乎 (Meng 1A1); ‘benefit from’ (!): 仁者安仁,知者利仁 (LY 4.2); 因而賈利之 (ZGC 11.1). Ant. 害 [40.7].

6 刑  xíng, *gə̂ŋ, var. of 形 [59.4]. 刖  yuè, *ŋwat, v , ‘cut off feet’, as a pun- 刺  cì, *tshek, v , ‘prick, stab’, a tr: 魏瑩 n

tr

ishment, a tr: 王以和為誑,而刖其左足 (HFei 13); b pass: 衛國之法,竊駕君車 者罪刖 (HFei 12); c mod: 為母之故,忘 其刖罪 (HFei 12). Cf. note at 刑 [18.4].

5

別  bié, *prat, v , 1 ‘separate, differentin

ate, di­vide’, only tr: 夫禮者所以 . . . 別同 異,明是非也 (Li 1); 別物上下 (Mo 19); 故聖人聞其聲則別其清濁 (CQFL 20). Ant. 兼 [12.8].

怒,將使人刺之 (Zhuang 25); b itr: 今 有利劍於此,以刺則不中 (Lü 8.3).

制  zhì, *tats, v , 1 ‘decide, determine, tr

con­trol’: a tr: 各當時而立法,因事而制 禮 (SJS 1); b pass: 通者常制人,窮者常 制於人 (Xun 4); 知者作法,而愚者制 焉 (SJS 1); c pseudo-tr: 制法之人,不足 與論變 (SJS 1); 2 nc, ‘decision, regulation, rule, institution’: 列國之卿,當小國之 君,固周制也 (Zuo 10.23); 今日之 事,我為制(Lü 16.6).

別  bié, *brat, v [< preceding], 1 ‘be sep- 到  dào, *tâuh, v , ‘reach, arrive at’, a tr: n

i

arated, different’, a itr: 不敬,何以別乎 (LY 2.7); 水別於他水 (Guan 57); 男女行 者別於塗 (SJ 47); b mod: 唯天子之 喪,有別姓而哭 (Li 3); ant. 兼 [12.8]; 2 nc, ‘separation, division; differ­ence’: 父子 有親,君臣有義,夫婦有別,長幼有 序,朋友有信 (Meng 3A4); 此可謂知義 與不義之別乎 (Mo 17).

初  chū, *tshra, n

, 1 ‘beginning, origin; for­merly, at first’: 無以反其性情而復其 初 (Zhuang 16); 夫禮之初,始諸飲食 (Li 9); typically ini, referring to an unspec­ ified point in the past: 初,鄭武公娶于申 (Zuo 1.1); 初楚子將以商臣為大子 (Zuo 6.1); cf. 20.2–①; 2 adv, ‘first, initially’: 夫秦 之初滅諸侯,天下之心未定 (SJ 88). temp

利  lì, *rits, v , ‘be sharp’, a itr: 金就礪則 i

利 (Xun 1); b caus, ‘sharpen’: 工欲善其 事,必先利其器(LY 15.10); c mod: 今有 利劍於此 (Lü 8.3). Ant. 鈍 [167.4].

利  lì, *rits, nc, 1 ‘profit, benefit’: 君子喻 於義,小人喻於利 (LY 4.16); 2 vn, ‘be

長驅到齊 (ZGC 11.1); b itr: 民到于今稱 之 (LY 16.12). Syn. 至 [133.0].

券  quàn, *khwans, n , ‘contract, agreec

ment’: 使無券契之責 (Guan 81); 悉來合 券,券遍合 (ZGC 11.1). Syn. 契 [37.6].

7

前  qián, *dzên, nloc, 1 ‘(the) front; before, earlier’, a head: 不敢以陳於王前 (Meng 2B2); b mod: 前世不同教 (SJS 1); 2 vi, ‘go/step forward’, a itr: 孔子下車而 前 (Zhuang 29); b caus: 夜行者掩目而 前其手 (HNan 17). Ant. 後 [60.6]; cf. 先 [10.4]. 則  zé, *tsə̂k, conj, 1 ‘then’, logical or temporal: 氣衰則生物不遂 (Li 19); 多男子 而授之職,則何懼之有 (Zhuang 12); 我 非殺戎王則不敢入邑 (SJ 6); 於是至則 圍王離 (SJ 7) cf. 15.4.2; 2 ptemp, contrastive, a topicalizing a S: 鳥則擇木 (Zuo 12.11); b topicalizing an O: 子女玉帛,則君有之 (Zuo 5.23); cf. 13.2 and 者 [125.4].

419

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

8

剛  gāng, *kâŋ, vst, ‘be hard, unyielding’, a itr: 金柔錫柔,合兩柔則為剛 (Lü 25.2); 淮南王為人剛 (SJ 101); b caus: 強 者必剛鬬其意 (SJS 11); 故為天者務剛其 氣,為君者務堅其政 (CQFL 78); c comp: 土乃益剛 (Guan 57). Syn. 堅 [32.8]; ant. 柔 [75.5]. 刳  kū, *khwâ, v , ‘cut up, slit open; distr

embowel’, a tr: 刳木為舟,剡木為楫 (Yi 8); 燕王必當殺子,刳子腹及子之腸矣 (ZGC 31.2); 通質,刳白馬而盟 (SJ 69); b pass: 諫士刳囚 (LNZ 7.2); c mod: 令其 罪若此刳羊矣 (SY 12).

秦 (ZGC 4.1); c itr: 吾王不豫,吾何以 助 (Meng 1B4); 2 nc, ‘help, assistance’: 身 不肖而令行者,得助於眾也 (HFei 40).

7

勃  bó, *bə̂t, vi, ‘spread out, flourish’, itr: 見梓勃焉實而俯 (SY 3.6).

勉  miǎn, *mranʔ, vi, ‘exert oneself, strive’, a itr: 君子者,勉於糾人者也 (Guan 35); 夫勉於仁者不亦難乎 (Li 32); b caus, ‘encourage, stimulate’: 勉之以慶 賞 (Xun 9); 上不加勉,而民自盡 (Guan 4); c pseudo-­tr: 子之君將行仁政,選擇 而使子,子必勉之 (Meng 3A3).

劍  jiàn, *kams, n , ‘sword’: 其劍自舟中

勇  yǒng, *loŋʔ, nc, ‘bravery, courage, audacity’: 折而不撓,勇也 (Xun 30); 將 者智信仁勇嚴也 (Sun 1); 兵彊則士勇 (SJ 87); 持節不恐謂之勇,反勇為怯 (XS 8.3).

Classifier 19  力

9

13

墜於水 (Lü 15.8).

c

動  dòng, *dôŋʔ, v , ‘act, move, stir’, a itr:

0

i

力  lì, *rək, nc, 1 ‘strength, force, effort’: 其馬力竭矣 (Zhuang 19); 2 vn, ‘make an effort, exert one­self ’; a itr: 其從者肅而 寬,忠而能力 (Zuo 5.23); b tr: 民貧則力 富 (SJS 20). 3

功  gōng, *kôŋ, nc, ‘success, achievement, merit’: 巍巍乎其有成功也 (LY 8.19).

加  jiā, *krâi, vn, 1 ‘add, increase; bestow; im­pose’, a tr: 我不欲人之加諸我 也,吾亦欲無加諸人 (LY 5.13); b itr: 功 名之極也,不可以加矣 (Xun 14); 2 adv, ‘increasingly, more’, for comp of vst: 鄰國 之民不加少,寡人之民不加多 (Meng 1A3). Cf. Box 24. 5

助  zhù, *dzrah, vn, 1 ‘help, aid, assist’: a tr: 齊人以衛師助之 (Zuo 2.10); b tr with C: 予助苗長矣 (Meng 2A2); 齊助楚攻

知者動,仁者靜 (LY 6.23); 鄉人譙之弗 為動 (HFei 49); b caus: 是動天下之兵也 (Meng 1B11); c mod: 山林,其動物宜毛 物 (ZL 18). Ant. 靜 [174.8].

務  wù, *moh, v , 1 ‘strive for, exert onen

self ’, a tr: 故糟糠不飽者不務梁肉 (HFei 49); b itr: 莫若射,故聖王務焉 (Li 46); 天下方務於合從連衡 (SJ 74); c aux, ‘strive to (do sth.)’: 陽虎務取之,我務守 之 (HFei 33); cf. 4.4.1; 2 nc, ‘effort; concern; task’: 凡有地牧民者,務在四時 (Guan 1); 唯幾也,故能成天下之務 (Yi 7).

10

勞  lào, *râuh, v , ‘reward, recompense’, n

a tr: 勞農勸民 (Lü 4.1); 周公拂其首,勞 而食之 (SY 3.6); b itr: 上卿逆於境,晉 侯郊勞 (GY 1.14).

勞  láo, *râu, vi, ‘toil, labour, work (to exhaus­tion)’, a itr: 飢者弗食,勞者弗息 (Meng 1B11); 勞而欲休,此人之情性也 (Xun 23); b caus: 君子信而後勞其民 (LY 19.10); 或勞心,或勞力,勞心者治

420

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

人,勞力者治於人 (Meng 3A4). Ant. 息 [61.6], 佚 [9.5].

勝  shēng, *lhəŋ, vn, 1 ‘succeed, be up to (a task)’; a tr: 王喜,以為能勝其任也 (Meng 1B9); b itr: 執圭,鞠躬如也,如不勝 (LY 10.4); 心暇者勝 (Lü 17.3); 2 adv,‘successfully, completely, fully’: 不違農時,穀不可勝食 也 (Meng 1A3); 夫天下之為聞見鬼神之 物者,不可勝計也 (Mo 31); cf. 24.3–㉔; 3 nc, ‘success, victory’: 勝久則鈍兵挫銳 (Sun 2); 故能全其勝 (HNan 15). 勝  shèng, *lhəŋh, vn [< preceding], ‘exceed, de­feat, overcome, prevail’, a tr: 人 民不勝禽獸蟲蛇 (HFei 49); 故欲勝人 者,必先自勝 (Lü 3.3); b itr: 王者之 兵,勝而不驕 (SJS 10). Ant. 敗 [66.7], 負 [154.2]. 11

勢  shì, *nhets, nc, 1 ‘circumstances, condition; in­fluence; power’: 處勢而驕下者,庸 主之所易也 (HFei 36); 聖王就其勢,因 其便,不失其性 (SY 18); 2 vn, ‘condition, bring about’, tr: 故夫名分定,勢治之道 也;名分不定,勢亂之道也 (SJS 26). 14

勳  xūn, *hwən, n , ‘merit’: (Zuo 5.5).

c

勳在王室

18

勸  quàn, *khwans, v , ‘recommend, urge, tr

en­courage’, a tr: 故兄勸其弟,父勸其子 (XS 3.2); b tr with C: 勸齊伐燕 (Meng 2B8); c pass: 故賞不用而民勸 (Xun 12); 勸於善 言而學 (Mo 48); d pseudo-­tr: 若是則田野 大闢,而農夫勸其事矣 (Guan 81); 省婦 使以勸蠶事 (Li 6).

Classifier 21  匕 2

化  huà, *hŋrôih, v , 1 ‘change, transform’, i

a itr: 化而為鳥,其名為鵬 (Zhuang 1); b caus: 鑽燧取火,以化腥臊 (HFei 49); c mod: 神莫大於化道 (Xun 1); 2 nc, ‘change, trans­formation, metamor­phosis’: 此之謂 物化 (Zhuang 2).

3

北  běi/bò, *pə̂k, nloc, 1 ‘north’: 東西南 北,莫敢不服 (ZGC 11.5); 2 vi, ‘be/move north’, itr: 南之人不得北,北之人不得 南 (Mo 47); 日北,則景長多寒 (ZL 18). 北  bèi, *bə̂kh, var. of 偝 [9.9]. Classifier 22  匚 4

匡  Kuāng, *khwaŋ, np, a place: 匡人其 如予何 (LY 9.5). 12

匱  kuì, *grus, vi, ‘be lacking, deficient, inade­quate’, a itr: 遠鄉皆至,則財不匱 (Li 6); 無匱乏也 (Li 9); 賞匱而姦益多 (SJS 8); b caus: 國貧則上匱賞 (SJS 6). Syn. 乏 [4.4]. Classifier 23  匸 2

匹夫  pı̌fū, *phit-­pa, nc, ‘simple man, ordinary man’: 匹夫不可奪志也 (LY 9.26); 匹夫有私便,人主有公利 (HFei 47).

Classifier 20  勹 2

勿  wù,

some­times used like simple 毋: 王欲行王 政,則勿毀之矣 (Meng 1B5). Cf. 5.2.3 and Box 16.

*mət, fus [< 毋之], 1 ‘do not . . . him/her/it/them’: 君其勿許 (Zuo 5.7); 勿奪其時 (Meng 1A3); 2 adv, ‘do not’,

9

區區  qūqū, *kho-­kho, vi, 1 ‘be tiny, insignifi­cant, trifling’, a itr: 宋國區區

421

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Zuo 9.17); b mod: 武靈王以區區趙服單 于 (SJ 20); 2 adv: 區區然皆擅一國之利 (SJS 14); 區區焉相樂也 (Lü 26.2).

Classifier 24  十 0

十  shí, *gip, vi, ‘be ten’, a itr: 功不 十,不易器(SJS 1); b caus: 五而六 之,九而十之,不可為數 (Guan 51); 所伐不當其禍十之 (SY 16); c mod: 從山 下望木者,十仞之木若箸 (Xun 21). Cf. Box 3. 1

卑  bēi, *pe, vst, 1 ‘be low, mean, humble’, a itr: 天尊地卑 (Yi 7); 臣有寵矣,然而 臣卑 (HFei 36); 桓幼而貴,隱長而卑 (GYang 1.1); b caus, ‘demean, belittle, degrade’: 卑宮室而盡力乎溝洫 (LY 8.21); 卑其志意,大其園圃高其臺 (Xun 25); 夫禮者,自卑而尊人 (Li 1); c comp: 故主上愈卑,私門益尊 (HFei 11); d mod: 德行成於身而遠古,卑人也 (Guan 16); 2 nc, ‘the low, despicable’: 將使 卑踰尊,疏踰戚 (Meng 1B7); 卑高以 陳,貴賤位矣 (Yi 7). Ant. 尊[41.9], 貴 [154.5], 高 [189.0]. 7

a itr: 曲禮三千,其致一也 (Li 10); b caus: 人十能之己千之 (Li 31); c mod: 不 遠千里 (Meng 1A1); 2 nc, ‘a thousand’: 萬 取千焉 . . . 不為不多矣 (Meng 1A1). Cf. Box 3.

南  nán, *nə̂m, nloc, 1 ‘south’, a head: 東 西南北,莫敢不服 (ZGC 11.5); b mod: 季孟自南門入,出自東門 (Zuo 11.6); 2 vi, ‘be/move south’, itr: 南之人不得 北,北之人不得南 (Mo 47); 日南,則 景短多暑 (ZL 18); 3 adv: 雍也可使南面 (LY 6.1).

2

10

千  qiān, *tshîn, v , 1 ‘be (one) thousand’, i

午  wǔ, *ŋâʔ, ntemp, seventh of the ‘Earthly branches’: 四月壬午,孝文太后 崩 (SJ 11). Cf. Focus 8. 3

半  bàn, *pâns, n , ‘half ’: 食之,舍其半 c

(Zuo 7.2).

博  bó, *pâk, v , 1 ‘be broad’, a itr: 夫殺之 i

人,為利人也博矣 (Mo 19); b caus: 博 我以文,約我以禮 (LY 9.11); ant. 淺 [85.8]; 2 adv: 君子博學於文 (LY 6.27).

Classifier 25  卜 3

6

卒  zú, *tsût, n , ‘soldier; batallion’: 繕甲 占  zhān, c

兵,具卒乘 (Zuo 1.1); 休卒徒 (Zhuang 29).

卒  zú, *tsut, vi, 1 ‘end, finish; die’, a itr: 九月衛穆公卒 (Zuo 8.2); b caus: 此不卒 者也,何以卒之,以其來會葬,我卒 之也 (GLiang 6.3); cf. note at 死 [78.2]; 2 adv, ‘in the end, finally’: 桓子卒受齊女 樂 (SJ 47). Ant. 始 [38.5].

卒  cù, *tshût, adv, ‘abruptly, brusquely, sudden­ly’: 卒然問 (Meng 1A6); 卒然相 遇 (Wu 5).

*tem, vn, 1 ‘prognosticate, prophesy, di­vine’, a tr: 臣以見占隱,以往 察來 (Wu 1); 命大史釁龜策,占兆審卦 吉凶 (Li 6); b itr: 善為易者不占 (Xun 27); 衛孫良夫帥師及齊師占於新築 (GLiang 8.2); 2 nc, ‘prognostication, divination’: 動 則觀其變,而玩其占 (Yi 7).

6

卦  guà,

*kwrêh, nc, ‘(Yijing) trigram; hexagram’: 是故剛柔相摩,八卦相盪 (Yi 7); 聖人設卦觀象 (Yi 7); 極天下之賾 者,存乎卦 (Yi 8). Cf. 32.1.

422

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 26  卩 3

卯  mǎo, *mrûʔ, ntemp, fourth of the ‘Earthly branch­es’: 秋八月丁卯,大事于 大廟 (Zuo 6.2). Cf. Focus 8.

所厚者薄 (Li 42); 有知其厚厚而薄薄 (CQFL 1); c comp: 德莫厚焉,刑莫威焉 (Zuo 5.15); d mod: 或以厚葬久喪以為仁 也 (Mo 25). Ant. 薄 [140.13].

8

原  yuán, *ŋwan, n , 1 ‘plain, flat land’: c

4

危  wēi/wéi, *ŋoi, vst, 1 ‘be precipitous, steep; endan­gered, precarious’, a itr: 國危 矣 (Zuo 4.30); b caus., ‘regard as dangerous, dread’: 戰,民之所危也 (SJS 6); ‘endanger’: 愛臣太親,必危其身 (HFei 4); c mod: 登高山,履危石 (Zhuang 21); 2 nc, ‘danger, risk’: 明主 . . . 安天下之危者 也 (Guan 64). Syn. 殆 [78.5]; ant. 安 [40.3]. 5

即/卽  jí, *tsit, v , 1 ‘approach, arrive’,

猶其有原隰衍沃也 (GY 1.3); 2 var. of 源 [85.10].

12

厭  yàn, *ʔems, vn, 1 ‘be satiated; fed up (with)’; a itr: 非盡天下之地,臣海內之 王者,其意不厭 (SJ 86); b tr: 夫子時然 後言, 人不厭其言,樂然後笑, 人不 厭其笑,義然後取, 個不厭其取 (LY 14.13); 帶利劍, 厭飲食, 財貨有餘, 是 謂盜竽 (Lao 53); syn. 倦 [9.8] 2 nc, ‘satisfaction’: 姜氏何厭之有 (Zuo 1.1).

n

tr: 望之儼然,即之也溫 (LY 19.9); 將即 席,容毋怍 (Li 1); esp. ‘accede to the throne, assume an office’: 及莊公即 位,為之請制 (Zuo 1.1); 獻公即位 (SJ 5); 2 conj, ‘if, when’, a ini: 即戎寇至,傳 鼓相告 (Lü 22.3); b med: 吾即沒,若必 師之 (SJ 47).

7

卽  jí, *tsit, var. of 即 [26.5].

Classifier 28  厶 3

去  qǔ, *khaʔ, v , ‘remove, eliminate, get n

rid of ’, a tr: 今居中國,去人倫,無君 子,如之何其可也 (Meng 1B11); 能去此 取彼,則人主道備矣 (Guan 52); b itr: 必不得已而去,於斯三者何先 (LY 12.7). Ant. 取 [29.6], 致 [133.4], 存 [39.3].

去  qù, *khah, v , 1 ‘go away, leave’, a itr: n

9

卿  qīng, *khraŋ, n , ‘minister’, highest c

rank cate­gory in the pre-­imperial bureaucracy: 孟子為卿於齊 (Meng 2B6). Cf. note at 士 [33.0].

Classifier 27  厂

9

7

厚  hòu, *gôʔ, v , ‘be thick; rich, subst

聖人之行不同也,或遠,或近,或 去,或不去 (Meng 5A7); b tr: 治國去 之,亂國就之 (Zhuang 4); 管仲去魯而 入齊,魯弱而齊強 (HSWZ 4); 2 prep, ‘far from’: 顧爭於戎狄,去王業遠矣 (ZGC 3.7); 天道之與人道也,相去遠矣 (Zhuang 11). Ant. 就 [43.9], 來 [9.6].

stantial, heavy’, a itr: 不臨深谿,不知地 之厚也 (Xun 1); 其流長矣,其溫厚矣 (Xun 4); b caus, ‘enrich’: 家務相益,不 務厚國 (HFei 6); ‘treat as impor­tant’: 其

參  cān, *tshə̂m, v , 1 ‘be three, threefold’, i

a itr: 天子者,與天地參 (Li 26); 三王之 德,參於天地 (Li 29); 太宗與社稷於子 者參 (XS 10.4); b mod: 大都不過參國之 一 (Zuo 1.1); 2 adv: 君子博學而日參省 乎己 (Xun 1). Note: 參 is re­lated to 三

423

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

[1.2], but more specifically refers to a unit of three, a trinity.

友  yǒu, *wəʔ, n , 1 ‘friend’: 與朋友交而

Classifier 29  又

c

0

又/有  yòu, *wəh, adv [< 有], 1 ‘also, and’, a esp. with stative verbs, ‘more’: 殆有 甚焉 (Meng 1A7); 玄之又玄,眾妙之門 (Lao 1); cf. Box 24; with num: 吾十有五 而志于學 (LY 2.4); cf. Box 3; b with nominal P: 困而學之,又其次也 (LY 16.9); 2 conj, ‘also, fur­thermore, ad­ditionally, again’: 彼其道遠而險,又有江山 (Zhuang 20); 雖有槁曓,不復挺者,輮使之然也 (Xun 1); 又嘗為孔子門人 (Yan 7.11). Syn. 亦 [8.4], 且 [1.4].

2

反  fǎn, *panʔ, vi, a itr, ‘return, go back to’: 吾不反矣 (Zuo 8.15); 沈於樂者反於 憂 (Lü 20.5); b caus, ‘make return, call back’: 王如改諸,則必反予 (Meng 2B12); 故先王反之於農戰 (SJS 3); c pseudo-­tr: 反齊滕之路 (Meng 2B6); 晉 文侯反國 (Lü 24.4).

反  fǎn, *panʔ, vn, 1 ‘oppose, turn against, rebel’, a tr: 子不反親,臣不逆 主,先王之通誼也 (ZGC 19.4); 反古者 未可必非 (SJS 1); b itr: 蜀侯煇,相壯 反,秦使甘茂定蜀 (SJ 71); 2 adv ‘in turn, in­stead, on the con­trary’: 今釋之,而反 抱兄之子,何也 (LNZ 5.6); 3 nc, ‘opposite’: 害者,利之反也 (HFei 46). 及  jí, *gəp, v , 1, ‘arrive (at), reach’, a tr: n

it mainly coordi­ nates names, often im­plying that the second is less important than the first.

軍且及之,棄其所抱 (LNZ 5.6); b itr: 田氏弗先,禍及矣 (SJ 46); 2 prep, ‘upon, at (a time)’: 及其長也,無不知敬 其兄也 (Meng 7A15); 3 conj, a temporal, ‘when, as soon as’: 及至文武,各當時而 立法,因事而制禮 (SJS 1); cf. 16.3; syn. 比 [81.0], 至 [133.0]; b coordinating, ‘as well as, and’: 生莊公及共叔段 (Zuo 1.1). Note: As a coordinating conj, 及 is an older, more formal synonym of 與 [134.7];

不信乎 (LY 1.4); 2 vn, ‘be friends with; befriend, help’, a tr: 無友不如己者 (LY 1.8); 鄉田同井,出入相友,守望相 助,疾病相扶持 (Meng 3A3); also between siblings: 兄敬愛弟謂之友 (XS 8.3); b itr: 左儒友於杜伯,皆臣周宣王 (SY 4). Syn. 睦 [109.8]; cf. note at 朋 [74.4].

6

取/趣  qǔ, *tshoʔ, v , 1 ‘choose, select; n

take, seize, conquer’, a tr: 鄭人取貨於印 氏 (Zuo 9.26); 非擇而取之 (Meng 1B14); 耳目鼻口不知所取去 (HNan 14); cf. 8.2.4; b itr: 輕取於其民而重致於其君 (Guan 14); 興兵而伐,必取 (SJS 3); c pass (!): 諸先生以文學取 (HFei 49); 2 nc, ‘choice, prefer­ence’: 法趣上下四相反也 (HFei 49); 孰計之,然後定其欲惡取舍 (Xun 12). Ant. 去 [28.3], 致 [133.4], 舍 [135.2].

受  shòu, *duʔ, v , ‘accept, receive; bear’: n

a tr: 趙宣子…為法受惡 (Zuo 7.2); b itr: 季桓子微服往觀再三,將受 (SJ 49).

叔  shū/shú, *nhuk, n , third-­born in a c

group of brothers, used in courtesy names: 髙辛氏有才子八人,以伯仲叔季爲 序(Zuo 6.18); cf. Focus 5; ‘brother in law’: 父之兄妻為世母,父之弟妻為叔母 (Erya 4).

7

叛/畔  pàn,

*bâns, vn, ‘transgress; rebel, sepa­rate from, de­fect’, a tr: 君子博 學於文,約之以禮,亦可以弗畔矣夫 (LY 6.27); 吾聞忠不畔上,勇不畏死 (HSWZ 9); b itr: 燕人畔 (Meng 2B9).

8

叟  sǒu, *sôʔ, n , form

of address, ‘old c man’: 叟不遠千里而來 (Meng 1A1).

424

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

14

叡  ruì, *los/lots, vn, ‘be intelligent, sharp-­sighted, sharp-­witted’, itr: 齧缺之 為人也,聰明叡知 (Zhuang 12); 耳目聰 明,心意叡智 (Lü 26.6). Syn. 知 [111.3]. Classifier 30  口 0

口  kǒu, *khôʔ, nc, ‘mouth, opening’: 口 不能言(Zhuang 13); 防民之口,甚於防 川 (GY 1.3); 寧為雞口,無為牛後 (SJ 69). 2

古  gǔ, *kâʔ, n

temp, ‘olden times; classical antiqui­ty’, a ini, with 者: 古者天子守在四 夷 (Zuo 10.23); b mod: 古聖人也 (Meng 2B9); mostly with 之: 古之道也 (LY 3.16); 古之人皆然 (LY 14.40); 董狐古之良史 也 (Zuo 7.2); c head: 象刑殆非生於治 古,並起於亂今也 (Xun 18). Ant. 今 [9.2]; cf. 昔 [72.4]. Note: 古 does not mean ‘past times’ but implies timeless qualities; it points to an enduring, nor­ mative tra­dition. It per­haps derives from an etymon ‘firm, durable’, which relates it to words such as 固, ‘firm, solid’, 故, ‘tradition’, and others in the same phonetic series.

可  kě, *khâiʔ, v , ‘be possible, permisst

sible, ac­ceptable, adequate (to be done)’, a itr: 劌曰未可 . . . 劌曰可矣 (Zuo 3.10); b caus: 始皇可其議 (SJ 87); 胡亥可其書 (SJ 87); c aux for pass: 魯可取乎 (Zuo 4.1); d sometimes short for 可以: 士何如 斯可謂之達矣 (LY 12.20); 正亦可為國 乎 . . . 正可以圉盜乎 (ZGC 27.21). Cf. 4.4.2 and note at 可以.

可以  kěyı̌, *khâiʔ-­ləʔ, vn, ‘can’, only as aux: 不可以教民 (Guan 5); 詩,可以 興,可以觀,可以群,可以怨 (LY 17.9); 夫子可以行矣 (SJ 47). Cf. 4.4.1. Note: Whereas simple 可 passivizes the

following verb, 可以 precedes active verbs. However, 可以is occasionally abbreviated to 可, which is then followed by an active verb.

史  shı̌, *srəʔ, n , [< 使, 9.6], 1 ‘commisc

sioner, scribe; registrar’: 史為書,瞽為 詩 (Zuo 9.14); 天子言,則史書之 (Lü 18.2); 孔子 . . . 嘗為季氏史,料量平 (SJ 47); 2 vi, ‘be pedantic, formalistic’, itr: 質 勝文則野,文勝質則史 (LY 6.16). Dist. 吏 [30.3]. Note: 史 was an office in the proto-­bu­reaucracy of the Zhou and other regional polities associated with a wide range of diplomatic, mili­ tary, political, scribal, archival, and other duties. They were, quite literally, ‘com­missioners’ that could be appointed to different tasks; cf. 使 [9.6].

司  sī, *sə, vn, ‘supervise, oversee’, only tr: 左右有局,各司其局 (Li 1); 黃帝後 土之所司者,萬二千里 (HNan 5). A number of official titles are form­ed with 司 (cf. note at 官 [40.5]): - 司空  sīkōng, *sə-­khôŋ, n , ‘min­ c

ister of public works’: 由是為司空 (SJ 47).

-司寇  sīkòu, *sə-­khôh, nc, ‘minister of justice’: 孔子為魯司寇 (Meng 6B6). -司馬  sīmǎ,

*sə-­mrâʔ, nc, ‘mar­ shal’: 司馬以吾故亡其良子 (Zuo 10.21).

-司職  sī­zhí, *sə-­tək, n , ‘supervisor c

of pastures’: 嘗為司職吏而畜蕃息 (SJ 47).

- 有司  yǒusī, *wəʔ-­sə, nc, ‘of­fi ce-­ holder, officer’: 四方之客至乎邑者不求 有司 (SJ 47).

右  yòu, *wəʔ/h, nloc, 1 ‘right (side)’: 孔子之周,觀於太廟右陛之前,有金 人焉 (SY 10); 2 vi, ‘be/turn to the right’, a itr: 主人入門而右 (Li 1); b caus: 左之 不左,右之不右 (HFei 34). Ant. 左 [48.2]. 右  yòu, *wəh, var. of 佑 [9.5].

425

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

召  zhào,

*drauh, vn, ‘call (for), summon; invite’, a tr: 王召士季 (GY 2.7); 使 吏召諸民當償者(ZGC 11.1); b itr: 佛肸 召,子欲往 (LY 17.6).

3

吊/弔  diào,

*tiâuk/tiâukh, vn, ‘condole, com­fort, pay a visit of condolence’, a itr: 出吊於東郭氏 (Meng 2B2); b tr: 誅其 君而弔其民 (Meng 3B5).

各  gè, *kâk, adv [< 舉], ‘each; respec-

tively’: 自王以下,各有父兄子弟 (Zuo 9.14). Cf. 7.3.

合  hé, *gə̂p, vi, ‘unite, combine; join, bring to­gether; match’, a itr: 紇與顏氏女 野合而生孔子 (SJ 47); 心合於道,說合 於心,辭合於說 (Xun 22); b caus: 於是 乎合其州鄉朋友婚姻 (GY 18.2); 合天下 而君之 (Xun 10).

后  hòu, *gôʔ/h, nc, ‘sovereign, lord, queen’: 褒姒者,童妾之女,周幽王之 后也 (LNZ 7.3). 后  hòu, *ɦôʔ, var. of後 [60.6]. 吏  lì, *rəh, n , ‘government

official; officer’: 吏習而民安 (SJS 1); 其父竊羊而 謁之吏 (HFei 49). Dist. 史 [30.2]; cf. also 使 [9.6], 委吏 [38.5], and note at 官 [40.5]. c

名  míng,

*meŋ, nc, 1 ‘(given) name; reputation, fame’: 德蕩乎名,知出乎爭 (Zhuang 4); ant. 實[40.11]; 2 vtr, ‘name, call (by the given name)’, a tr: 異哉君之 名子也 (Zuo 2.2); 宋人有學者,三年反 而名其母 (ZGC 24.4); b tr with C: 寶玉 而名之曰石 (XX 5); c pass, with C: 故因 名曰丘云 (SJ 47). Note: 名 specifically means the ‘given name’ of a person as opposed to the ‘courtesy name’, 字 [39.3]; 號 [141.7] refers to an informal appel­ lation. Cf. Focus 5.

同  tóng, *dôŋ, vi, 1 ‘be the same, equal’, a itr: 君子和而不同,小人同而不和 (LY 13.23); 聖人之行不同也 (Meng 5A7); b caus ‘treat as/make equal; share, have in

common’: 生而同聲,長而異俗 (Xun 1); 同則同之,異則異之 (Xun 22); 此與 彼同類 (Mo 45); c mod: 君取於吳為同 姓 (LY 7.31); 2 adv, ‘equally, together’: 上 下與天地同流 (Meng 7A13); 3 nc, ‘coequal, peer’: 無恃寵,無違同 (Zuo 11.4). Ant. 異 [102.6].

吉  jí, *kit, vst, 1 ‘be auspicious, fortunate’, a itr: 史蘇占之,曰不吉 (Zuo 5.15); 卜之以守龜於宗祧,吉矣 (Zuo 12.23); 章其有常,吉哉 (SJ 2); b mod: 喪事先 遠日,吉事先近日 (Li 1); 擇吉日 (Li 6); 2 nc, ‘auspiciousness’: 八卦定吉凶,吉 凶生大業 (Yi 7); 元者吉之始也 (SY 12). Ant. 凶 [17.2]. 4

否  fǒu, *pəʔ, adv, ‘(or) not’, implies an aforemen­tioned P: 未知母之存否 (Zuo 7.2); 其本亂而末治者否矣 (Li 42).

告  gào, *kûkh, v , ‘inform; announce, n

tell’, a tr: 反以告商子 (SY 3.6); b tr2: 予之 法制,告之訓典 (Zuo 5.23); c itr: 反以 告乎商子 (SY 3.6).

君  jūn, *kwən, nc, 1 ‘lord; ruler’: 君 者,善群也 (Xun 9); 君之所以為君 者,賞罰以為君 (Guan 31);

2 vn, ‘be/act as a ruler’, a itr: 晉靈公不君 (Zuo 7.2); b tr: 今王公大人之君人 民 . . . 欲脩保而勿失 (Mo 8); 合天下而 君之 (Xun 10). Cf. note at 主 [3.4]; ant. 臣 [131.0]; dist. 君子.

君子  jūnzı̌, *kwən-­tsəʔ, nc, 1 ‘gentleman’: 君子和而不同,小人同而不和 (LY 13.23); 君子義以為上 (LY 17.20); 君 子者,禮義之始也 . . . 天地之參也,萬 物之摠也,民之父母也 (Xun 9); 以仁 為恩,以義為理,以禮為行,以樂為 和,薰然慈仁,謂之君子 (Zhuang 33); 2 vst, ‘be/act gentlemanly’, itr: 人不知而 不慍,不亦君子乎 (LY 1.1); 穆公知其 君子也 (SY 2). Ant. 小人 [42.0]; dist. 君.

吞  tūn, *thə̂n, vn, ‘swallow’, only tr: 江 海之魚吞舟 (SY 12); 乃吞藥自殺 (SJ 88);

426

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

fig. ‘conquer, annex’: 今秦王欲吞天下 (SJ 87).

吾  wú, *ŋâ, prper, ‘I; we’, a S: 吾楚人也 (ZGC 7.5); b mod: 吾王不遊,吾何以 休 (Meng 1B1); c O, med: 吾問焉而不吾 告 (Zhuang 19); 君不吾聽 (SY 1). Cf. Box 5.

吳  Wú, *ŋwâ, np, a polity: 此人不 死,必為吳患 (SJ 31). Note: Wu was a polity in the lower Yangzi re­gion, with its capital in the region of modern Wu­xi. Although its rulers belong to the royal clan of Ji姬 and trace their ancestry to Taibo太伯, a son of Zhou Taiwang太王, it was not part of the Central Polities and considered ‘barbarian’. Wu was vanquished by Yue越in 473 bc. 吝  lìn,

*m-­rəns, vn, 1 ‘be niggardly, stingy; jeal­ ous; resentful’: 積多不能分 人,而厚自養,謂之吝 (Yan 4.23); a itr: 使驕且吝,其餘不足觀也已 (LY 8.11); 時年歲凶,則民吝且惡 (Mo 5); 吝於財 者失所親 (Guan 1); b tr: 故輕棄之,非 所惜吝 (SJ 41); syn. 愛 [61.9]; 2 nc, ‘distress’: 悔吝者,懮虞之象也 (Yi 7).

吠  fèi, *ba(t)s, vn, ‘bark’, a itr: 衣狗裘者 當犬吠 (SY 11); 犬吠不驚,命曰金城 (SY 16); b tr: 其狗不知而吠之 (HFei 23); c mod (!): 北海則有走馬吠犬焉 (Xun 9). Cf. note at 鳴 [196.3]. 5

和  hé, *(g)wâi, vn, 1 ‘be harmonious, peaceful, in accord’, a itr: 夫妻不和,家 室大凶 (SY 10.7); 君子和而不同,小人 同而不和 (LY 13.23); 西周欲和于楚韓 (ZGC 1.4); b tr: 遽和東周 (ZGC 1.27); 2 nc, ‘har­mony’: 禮之用,和為貴 (LY 1.12); 剛柔得適謂之和,反和為乖 (XS 8.3). Syn. 睦 [109.8]. 和  hè,

*(g)wâih, vn, [< preceding], ‘harmo­nize, attune’, tr: 音聲相和 (Lao 2); 臣聞以德和民,不聞以亂 (Zuo 1.4); 齊 王和其顏色 (ZGC 10.4). Syn. 調 [149.8].

呼  hū, *hâ(h), vn, ‘call (someone), call (out)’, a itr: 順風而呼,聲非加疾也 (Xun 1); b tr: 王呼之曰,余不食三日矣 (GY 19.3).

命  mìng, *mreŋh, nc, 1 ‘order, mandate’: 棄君之命,不信 (Zuo 7.2); 敢不唯 命是聽 (Zuo 9.28); ‘fate, life span’: 死生 有命,富貴在天 (LY 12.5); 莫之致而至 者,命也 (Meng 5A6); 羣生皆得其命 (Xun 9); 2 vtr, ‘order; appoint; name’, a tr: 馬者,所以命形也,白者,所以命色 也,命色者,非命形也 (GLong 1.2); b tr with C: 大叔命西鄙北鄙貳於己 (Zuo 1.1); c pass: 三卿皆命於天子 (Li 5); syn. 令 [9.3], 使 [9.6]; cf. 名 [30.3]. 味  wèi, *məs, nc, ‘taste, flavour’: 三月 不知肉味 (LY 7.14); 目好色,耳好 聲,鼻好香,口好味 (Wen 3).

周  zhōu, *tiu, v , 1 ‘be all-­embracing, n

compre­hen­sive; encompass, embrace; encircle, surround’, a itr: 君子周而不 比,小人比而不周 (LY 2.14); 其藏之也 周,其用之也遍 (Zuo 10.4); b tr: 齊師 敗績,逐之,三周華不注 (Zuo 8.2); 周 其君子,不失成功,周其小人,不失 成命 (Guan 20); 2 adv, ‘in circles; all around, everywhere; all’: 周麾而呼 曰,君登矣 (Zuo 1.11); 晉文公出亡,周 流天下 (SY 6); cf. Box 30.

周  Zhōu, *tiu, n , a a dynasty: 天降禍 p

于周 (Zuo 10.32); b given name: 昔者莊 周夢為胡蝶 (Zhuang 2); 2 vi, ‘be (like) Zhou’, itr: 俄然覺,則蘧蘧然周也 (Zhuang 2). Note: The Zhou dynasty was found­ed in the mid-­eleventh century bc, when king Wu vanquished the Shang. Residing in the capital of Hao (near modern Xi’an), the Zhou rulers en­feoffed vas­ sals and relatives to rule over the terri­ tories of the North China plain. Although Zhou’s power greatly diminished when the capital was invaded in 771 bc and the Zhou kings fled east­ward to present-­day Luo­yang, they remained rulers of ‘All un­der Heaven’ at least in name, and

427

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

their founding fathers—king Wen, king Wu, and the duke of Zhou—became paragons of orderly ruler. In Zhanguo times, Zhou split up into two polities, which were finally conquered by Qin in 256 bc.

6

哀  āi, *ʔə̂i, n , 1 ‘sorrow’: 不勝其哀,故 c

哭也 (SY 6); 2 vst, ‘be sorrowful, mournful’, a itr: 鳥之將死也其嗚也哀 (LY 8.4); b caus: 人主胡可以不務哀士 (Lü 15.4).

哉  zāi, *tsə̂, ptemp, a with verbal P: 為仁 由己,而由人乎哉 (LY 12.1); b often with inverted P: 久矣哉,由之行詐也 (LY 9.12); 君子哉商子也 (SY 3.6); c with nom P: 是豈水之性哉 (Meng 6A2); cf. 14.3 and Review 4. 7

哭  kū, *khôk, v , ‘wail, sob’, a itr: 哭而過 n

市 (Zuo 6.18); b tr: 顏淵死,子哭之慟 (LY 11.10); 我噭噭然隨而哭之 (Zhuang 18); c mod (!): 哭日不歌(Li 1).

員  yuán, *wen, var. of 圓 [30.10].

問  wèn, *məns, v

n [< 聞 wèn], 1 ‘ask’, a tr: 吾將問之 (LY 7.15); b tr with PrP: 齊景 公問政於孔子 (LY 12.11); c itr: 公問於眾 仲曰 (Zuo 1.4); 夫民愚,不懷知而問 (SJS 7); cf. 8.2.4; d pseudo-­tr: 荊宣王問群 臣曰 (ZGC 14.3); 2 nc, ‘question’: 今者吾 忘吾答,因失吾問 (Zhuang 23). Ant. 答 [118.6].

啗  dàn,

*lə̂mʔ, vn, ‘feed, give to eat; tempt, lure’, a itr: 往說秦將,啗以利 (SJ 8); a tr: 主孟啗我 (GY 8.1); b tr with PrP: 嘗啗我以餘桃 (HFei 12); 淫之以色,啗 之以利 (LT 2). Syn. 食 [184.0].

9

喟  kuì, *khus, adv, ‘moan, groan’: 顏淵 喟然歎 (LY 9.11); 簡公喟焉太息 (SY 9).

喪  sāng, *sâŋ, n , ‘funeral; mourning, c

sorrow’: 夫三年之喪,天下之通喪也 (LY 17.21).

喪  sàng, *sâŋh, v

8

商  shāng, *lhaŋ, nc, 1 ‘merchant, trader’: 所謂四民者,士農工商也 (SY 7); 商賈 皆欲藏於王之市 (Meng 1A7); syn. 賈 [154.6]; 2 np, a dynasty: 夏道不亡,商德 不作;商德不亡,周德不作 (SY 1). Note: The Shang dynasty was allegedly founded by Tang 湯, who in the sixteenth century bc over­threw the depraved last Xia ruler and estab­lished his capital in Hao 亳, in modern Henan. Succeed­ing rulers supposedly moved the capital several times. 唯/惟  wéi, *wi, pt

唯  wěi, *wiʔ, interj, ‘yes’: 曾子曰,唯 (LY 4.15); 宋玉對曰,唯,然,有之 (XX 1.16).

emp, ‘only’, a mostly empha­sizing anteposed O: 父母唯其疾之 憂 (LY 2.6); cf. 14.2; b emphasizing the S: 唯我與爾有是夫 (LY 7.11); c emphasizing the predicate: 王唯信子,故處子於蔡 (Zuo 10.15); 吾唯不知務而輕用吾 身,吾是以亡足 (Zhuang 5).

tr [caus < 亡], ‘lose; destroy’, a tr: 非吳喪越,越必喪吳 (Lü 14.5); 其喪師也,不亦宜乎 (Zuo 1.11); b pass: 弱喪而不知歸 (Zhuang 2); 師喪分 焉 (GYang 3.4).

善  shàn, *danʔ, v , 1 ‘be good, excellent’, a st

itr: 善哉問也 (Meng 1B4); b caus, ‘approve of’: 武王善之 (Xun 25); ‘make sth. good, remedy’: 雖有智者,不能善其後矣 (Sun 2); c comp: 治地莫善於助,莫不善於貢 (Meng 3A3); d mod: 善人,吾不得而見 之矣 (LY 7.26); 2 adv, ‘do well, be skilled in’: 巧者善度,知者善豫 (HNan 16); 3 nc, ‘goodness’: 教人以善,謂之忠 (Meng 3A4); 善者,王教之化也 (CQFL 36). Syn. 良 [138.1]; ant. 惡 [61.8]; 拙 [64.5].

啼  tí, *dê, v , ‘cry, howl’, only itr: 穆嬴日 n

抱太子以啼于朝 (Zuo 6.7); 老人兒啼 (SJ 119).

喜  xı̌ , *həʔ, vi, ‘be happy, glad; enjoy, rejoice (in)’, a itr: 問於若國之士,孰喜

428

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

孰懼 (Mo 10); 天子獨喜 (SJ 12); 鄭伯喜 於王命 (Zuo 5.5); b pseudo-­tr: 褒姒大 說,喜之 (Lü 22.3); c caus: 王不如設 戎 . . . 以喜其民 (GY 19.1); 喜之勿怒 (SY 7). Ant. 怒 [61.5].

13

器  qì, *khrəs/khrəts, nc, 1 ‘tool, implement; weapon; vessel’: 使有什伯之器而不 用 (Lao 80); 金石絲竹,樂之器也 (Li 19); 2 vi, ‘serve as a tool’, a itr: 君子不器 (LY 2.12); b caus: 及其使人也,器之 (LY 13.25).

喻  yú, *lo, var. of 愉 [61.9]. 喻  yù, *loh, v , ‘understand, be compe- 噭噭然  jiàojiàorán, *kiâuk-­kiâuk-nan, n

tent in’; explain (by ana­logy), illus­trate, a itr: 君子喻於義,小人喻於利 (LY 4.16); 先生就坐,寡人喻矣 (SY 12); 王好 戰,請以戰喻 (Meng 1A3); b tr: 師也 者,教之以事而喻諸德者也 (Li 8); c tr with PrP: 臣不能以喻臣之子 (Zhuang 13). Note: 喻explicitly means explaining by analogy, in con­trast to more general words for ‘explain’ like 說 [149.7] or 釋 [165.13].

adv, ‘sobbingly’, hapax, adv: 我噭噭然隨 而哭之 (Zhuang 18).

17

嚴  yán, *ŋam, nc, ‘strictness’: 將者智信 仁勇嚴也 (Sun 1); 臨制不犯謂之嚴,反 嚴為軟 (XS 8.3). 18

囂囂  xiāoxiāo, *hâu-­hâu, v , 1 ‘be casi

10

嗣  sì, *s-­ləh, vn, 1 ‘follow, succeed (as ruler)’, a itr: 魯莊公以賊嗣 (SY 8); b ­tr: 子產而死,誰其嗣之 (Zuo 9.30); 臣不 足以嗣之 (Zuo 10.3); c mod [!]: 內事曰 孝王某,外事曰嗣王某 (Li 2). 2 nc, ‘successor’: 誡其嗣懿子 (SJ 47). 嗚呼  wūhū, *ʔâ-­hâ, interj, ‘alas!’: 宣 子曰,嗚呼 (Zuo 7.2).

嗜  shì, *gih, v , 1 ‘enjoy, like, be fond n

of ’, a tr: 曾皙嗜羊棗 (Meng 7B36); 公儀 休相魯而嗜魚 (HSWZ 3); b itr: 無廉恥 而嗜乎飲食 (Xun 2); c aux: 不嗜殺人者 能一之 (Meng 1A6); 2 nc, ‘liking, predilection’: 嗜欲不同 (Li 5). Syn. 好 [38.3], 欲 [76.7].

ual, at ease; indifferent’, itr: 尊德樂義,則 可以囂囂矣 (Meng 7A9); 天下囂囂,新 主之資也 (XS 1.2); 2 adv: 囂囂然曰 (Meng 5A7).

Classifier 31  囗 2

四  sì, *slis, v , ‘be four’: a itr: 故三代不 i

四 (SJS 7); b caus: 四其本,則鄉里給 (Guan 78); c mod: 四海之內皆兄弟也 (LY 12.5). Cf. Box 3.

3

回/囘  Huí, *wə̂i, n , personal name: p

吾與囘言終日 (LY 2.9).

因  yīn, *ʔin, v , 1 ‘rely, depend on; foln

11

嘗  cháng, *daŋ, v , 1 ‘taste; try, attempt’, n

only tr: 少嘗苦曰苦,多嘗苦曰甘 (Mo 17); 請先嘗沮之 (SJ 47); 2 adv, ‘once, on one occasion’: 孟子嘗與我言於宋 (Meng 3A2); cf. 10.5.

嗾  sǒu,

*sôʔ/h, vn, ‘set (a dog) on’, hapax, tr: 公嗾夫獒焉 (Zuo 7.2).

嘆  tàn, *nhâns, var. of 歎 [76.11].

low, accord with’, a tr: 因事而制禮 (SJS 1); 取用於國,因糧於敵 (Sun 2); b itr: 殷 因於夏禮 (LY 2.23); 2 conj, ‘thereupon, consequently, accord­ingly’: 因釋其耒而 守株 (HFei 49); cf. 16.4.

4

困  kùn, *khûns, v , ‘be in distress, diffii

culty’, a itr: 困而不學,民斯為下矣 (LY 16.9); 困於陳蔡之閒(SJ 47); b caus: 君子

429

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

不困人於阨 (SJ 38); c mod: 困獸猶 鬬,況國相乎 (Zuo 7.12).

5

固  gù, *kâ(k)h, v , 1 ‘be firm, solid’, a itr: st

斲輪,徐則甘而不固 (Zhuang 13); 使睦 而疾楚,以固於晉 (Zuo 9.13); b caus: 天 下陰燕陽魏,連荊固齊 (HFei 1); c comp: 重而閉之,莫固於口 (HNan 10); syn. 堅 [32.9]; 2 adv, ‘cer­ tainly, surely, ab­solutely’, a in vb clause: 仁人固如是乎 (Meng 5A3); b in nom clause: 子固非魚也 (Zhuang 17); cf. 3.3.2. Cf. note at 古 [30.2].

6

囿  yòu, *wəkh, nc, ‘park, reserve’: 棄田 以為園囿,使民不得衣食 (Meng 3B9). Syn. 圃 [31.7], 園 [31.10]. 7

圃  pǔ, *pâʔ/pâh, n , 1 ‘(vegetable) garc

den’: 今有一人,入人園圃,竊其桃李 (Mo 17); 2 ‘garden­er’: 請學為圃 (LY 13.4). Syn. 囿 [31.6], 園 [31.10].

8

國  guó, *kwə̂k, nc, 1 ‘capital city’: 都城過 百雉,國之害也 (Zuo 1.1); 遍國中無與 立談者 (Meng 4B33); ‘country, polity, state’: 國之所以為國者,民體以為國 (Guan 31); 諸侯各愛其國,不愛異國 (Mo 14); 2 vi, ‘be a polity/state’, itr: 微二子 者,楚不國矣 (Zuo 12.16). Syn. 邦 [163.4]. Note: With the emer­gence of territorial polities in the Chun­qiu period, 國, which original­ly referred to a ‘city’, took on the meaning ‘regional polity/state’. It means an enfeoffed polity in con­trast to the realm, 天 下 [37.1], on the one hand and the ‘noble hous­es’, 家 [40.7], on the other hand. - 國家  guójiā, *kwə̂k-­krâ, n , ‘polc

ity and noble houses’: 國家之敗,由官 邪也 (Zuo 2.2); cf. 家 [40.7].

莅中國,而撫四夷也 (Meng 1A7); 中國 者,聰明叡知之所居也 (ZGC 19.4); 夫 戎辟而遼遠,未聞中國之聲也 (SY 20).

9

圍  wéi, *wəi, v , ‘encircle, lay siege to’, a tr

tr: 王怒,遂圍蕭 (Zuo 7.12); b pass: 寡人 一城圍,食不甘味,臥不便席 (ZGC 5.10); c mod: 曷為久居此圍城之中而不 去也 (ZGC 20.13).

10

園  yuán, *wan, nc, ‘garden, orchard’: 趙 穿殺靈公於桃園 (Zuo 7.2). Syn. 囿 [31.6], 圃 [31.7].

圓/員  yuán, *wen, nc, 1 ‘circle’: 一手 畫方,一手畫圓,莫能成 (CQFL 44); as measure: 雖有詩書,鄉一束,家一員, 猶無益於治也 (SJS 3); 2 vi, ‘be round’, a itr: 槃圓而水圓 (Xun 12); 圓者中規,方者 中矩 (Xun 26); b mod: 執而圓機,獨成 而意 (Zhuang 29). Ant. 方 [70.0]. 11

圖  tú, *dâ, vn, ‘plan, make provisions (for)’, a tr: 君其圖之 (SJS 1); 今王非越是 圖 (GY 19.3); 王受九鼎而圖犯 (SJ 4); b itr: 棄安而圖,遠於孝矣 (GY 7.4). Classifier 32  土 0

土  tǔ, *thâʔ, n , ‘ground, soil, land’: 民 c

之有口,猶土之有山川也 (GY 1.3); 何 非君土 (Zuo 10.7).

3

地  dì, *draih, nc, ‘earth’: 天能生物,不 能辨物也,地能載人,不能治人也 (Xun 19). Ant. 天 [37.1].

圩  yú, *wa (?), vi, ‘be indented’, caus: 生 而首上圩頂,故因名曰丘云 (SJ 47).

- 中國  Zhōng­guó, *truŋ-­kwə̂k, n , 在  zài, *dzə̂ʔ, v , 1 ‘be p

‘Mid­dle Poli­ties’, of the North China plain:

present (in/at/ n on); be alive’, a itr: 祭如在,祭神如神在

430

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(LY 3.12); 父在,觀其志,父沒,觀其 行 (LY 1.11); 衛雖小,其君在焉 (Zuo 11.9); esp. ‘be due to, depend on’: 在於凶德 (Zuo 6.18); b tr: 樂亦在其中矣 (LY 7.16); 故知之難,不在見人,在自見 (HFei 21); 2 prep, ‘in, at, on’: 禹往見之,則耕在 野 (Zhuang 12).

4

均  jūn, *kwin, vi, ‘be equal, alike; well balanced’, a itr: 四民均則王道興而百姓 寧 (SY 7); b caus: 節以山川,均其寒暑 (HSWZ 3).

坐  zuò, *dzôiʔ, v , ‘sit, kneel’: a itr: 王坐 i

於堂上 (Meng 1A7); b caus: 坐之起之 (Li 11); c pseudo-­tr: 寡人自坐地,二三子皆 坐地 (SY 17). Note: 坐means kneeling (on a mat) with the feet under­neath the buttocks, which was the proper way to ‘sit’ in Ancient China (there were no chairs). A more casual way of sitting—cross-­ legged or with legs pending—was called 踞 [157.8]. Sitting with outstretched legs, 箕踞, was considered highly improper.

5

坤  kūn, *khwə̂n, np, name of a trigram and a hexagram, associated with earth, female, passivity, etc.: 天尊地卑,乾坤定 矣(Yi 7); 乾道成男,坤道成女 (Yi 7); 坤,地也,故稱乎母 (Yi 9). Ant. 乾 [5.10]. 6

城  chéng,

*deŋ, nc, 1 ‘city-­wall, wall; city’: 長城之界 (SJ 69); 城非不高也,池 非不深也 (Meng 2B1); 賂秦伯以河外列 城五 (Zuo 5.15); 2 vn, ‘wall, fortify’, a tr: 費 伯帥師城郎 (Zuo 1.1); 靖郭君將城薛 (HFei 23); b itr: 今吳是懼,而城於郢 (Zuo 10.23). Cf. note at 郭 [163.8].

8

堂  táng,

*dâŋ, nc, ‘hall’: 天子之堂九 尺,諸侯七尺,大夫五尺,士三尺 (Li 10); 登堂而跪 (SY 3.6). Cf. note at 室 [40.6].

埸  yì, *lek, n , ‘border’: 疆埸無主,則 c

啟戎心 (GY 7.6).

執  zhí, *təp, vtr, ‘seize, apprehend; hold’, a tr: 楚人執而殺之 (Zuo 10.8); ‘take charge of ’: 居處恭,執事敬,與人 忠 (LY 13.19); ‘maintain, believe; insist’: 今執無鬼者言曰 (Mo 31); b pass: 然則是 大臣執於朝,而列陳之士執於賞也 (Guan 80). 執政  zhízhèng, *təp-­teŋ, n , ‘adminisc

trator, governor’: 鄭人游于鄉校,以論 執政 (Zuo 9.31).

9

報  bào, *pûh, vtr, ‘repay, requite, avenge; decide (on a punishment); re­ward’, a tr: 以 德報怨 . . . 何以報德 (LY 14.34); 為善者 天報之以福,為不善者天報之以禍 (Xun 28); b pass: 則王之怨報矣 (ZGC 25.26); 然則將軍之仇報 (SJ 86); 夫造禍 而求其福報 (SJ 70). 報  bào, *pûh, v , ‘report

(to a supen rior)’, a tr: 荀息牽馬操璧而報獻公 (HFei 10); 使者以報楚王 (SY 12); b itr: 可使千里外一日反報 (Guan 39); 子以 是報矣 (Lü 15.7).

堅  jiān,

*kîn, vst, 1 ‘be tough, solid, firm’, a itr: 仰之彌高,鑽之彌堅 (LY 9.11); b caus: 堅其行陣,連其什伍,以 禁淫非 (SY 15); c comp: 仰之彌高,鑽 之彌堅 (LY 9.11); d mod: 可使制梃以撻 秦,楚之堅甲利兵矣 (Meng 1A5); 2 nc, ‘hard­ness, firm­ness’: 石可破也,而不可 奪堅 (Lü 12.4). Syn. 固 [31.5], 剛 [18.8].

堪  kān, *khə̂m, v , ‘endure, bear’, tr: 人 n

不堪其憂,回也不改其樂 (LY 6.11); 民 不堪命矣 (GY 1.3).

堯  Yáo, *ŋiâu, n , a mythical sage ruler: c

堯舜之道,孝弟而已矣 (Meng 6B2); 堯 欲傳天下於舜 (HFei 13). Cf. Focus 7.

10

填/塡  tián, *dîn, vn, 1 ‘fill, clog’, tr: 毋 壞室,毋填井,毋伐樹木 (SY 15); 民填

431

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

溝壑 (XS 3) 2 adv, onomatopoetic for the sound of drums: 填然鼓之 (Meng 1A3); cf. note at 鼓 [207.0].

Classifier 33  士 0

塗  tú, *lâ, n , ‘mud, clay’: 嬰兒 . . .以塗 士  shì, *s-­rəʔ, n , 1 a minor nobleman, c

c

為羹 (HFei 32); ‘street, way’: 往拜之,遇 諸塗 (LY 17.1); 塗有餓莩而不知發 (Meng 1A3); syn. 道 [162.9].

11

塵  chén, *drən, n , ‘dust’: 嬰兒 . . . 以塵 為飯 (HFei 32).

c

境/竟  jìng, *kraŋʔ, n , ‘border’: 臣始 c

至於境,問國之大禁,然後敢入 (Meng 1B2); 燕王欲殺之,走且出境 (ZGC 31.2); 邊境不寧,土地分裂 (Li 6). Note: 境 means state borders that define the scope of a ruler’s power, whereas 邊 [162.15] means borders of cities or other smaller units within a state; 四海 refers to the loosely demar­ cated borderlands on the edges of the Chinese realm.

12

墜  zhuì, *drus, vi, a itr, ‘fall’: 其劍自舟 中墜於水 (Lü 15.8); b caus: 自先王莫墜 其國 (GY 18.4); c mod: 有墜星下東郡 (SJ 6). 墦  fán, *ban, nc, ‘grave’: 卒之東郭墦間 之祭者 (Meng 4B33).

‘knight, officer; retainer; scholar’: 王臣 公,公臣大夫,大夫臣士 (Zuo 10.7); 大夫食邑,士食田,庶人食力 (GY 10.4); 行己有恥,使于四方,不辱君 命,可謂士矣 (LY 13.20); 好士者 強,不好士者弱 (Xun 15); 古者有四 民, 有士民,有商民,有農民,有工 民 (GLiang 8.1). Note: The shi belonged to the lowest ranks of the West Zhou and Chunqiu aris­ tocracy, beneath the 卿 [26.9] and the 大夫 [37.0]. Origi­ nally serving (quite literally, the word 士 is related to 事) mainly as warriors, their roles di­ versified in Zhanguo times: as scholars—people like Confucius, Mengzi belonged to the shi—they rose to become an intellectual elite that played a crucial role in the service of regional rulers. Vir­ tually all classical literature derives from the shi: ultimately, Classical Chinese was the language of the shi.

1

壬  rén, *nəm, ntemp, ninth of the ‘Heavenly stems’: 以壬癸殺黑龍於北方 (Mo 47). Cf. Focus 8. 4

墨  mò, *hmə̂k, n , 1 ‘ink’: 宋元君將畫 壯  zhuàng, *tsraŋh, v , ‘be full-­grown, c

圖,眾史皆至,受揖而立,舐筆和墨 (Zhuang 21); 卜人定龜,史定墨 (Li 13); 2 vst, ‘be black’, itr: 面深墨 (Meng 3A2); 3 np, family name: 仲尼墨翟,窮為匹夫 (Zhuang 29); short for Mo Di and his follow­ers: 楊墨之道不息,孔子之道不 著 (Meng 3B9); 儒者偽辭,墨者兼愛 (Zhuang 29).

13

壅  yōng, *ʔoŋ, vtr, ‘block, obstruct, dam up’, a tr: 人主者,固壅其言談 (HFei 9); b pass: 川壅而潰,傷人必多 (GY 1.3).

st

able­bodied, strong’: 人生 . . . 三十曰 壯,有室 (Li 1); a itr: 夫人幼而學 之,壯而欲行之 (Meng 1B9); b caus: 滕 公奇其言,壯其貌 (SJ 92); c mod: 帝毋 壯子,太后畏君等 (SJ 9). Cf. note at幼 [52.2].

5

垂  chuí, *doi, v , ‘­hang

down, droop, i dangle’, a itr: 凡侍於君,紳垂 (Li 13); b caus, ‘let hang, drop’ 黃帝堯舜垂衣裳而 天下治 (Yi 8); 則己垂涕泣而道之 (Meng 6A3); 垂其股肱之力而不相勞來

432

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

也 (Mo 10); c pseudo-­tr: 翼若垂天之雲 (Zhuang 1).

多謀伐寡人者 (Meng 1B11); 多見黑曰白 (Mo 17). Ant. 少 [42.1].

9

5

夜  yè, *jah, n

壹  yī, *ʔit, var. of 一 [1.0].

temp, 1 ‘night’: 是月也,日 夜分 (Li 6); 2 adv, ‘at night’: 夜行晝居 (Zhuang 14); 吾嘗終日不食, 終夜不寢 (LY 15.31); 文伯之喪,晝夜哭 (Li 4). Ant. 日 [72.0], 晝 [72.7].

Classifier 35  夊 7

夏  xià, *grâʔ, n

, ‘summer’: 春秋教 以禮樂,冬夏教以詩書 (Li 5); 天有四 時,春秋冬夏 (Li 29); 夏者,天之德也 (CQFL 79). temp

夏  Xià, *grâʔ, n , the first of the ancient c

Three Dynasties: 夏禮,吾能言之 (LY 3.9); 今有搆木鑽燧於夏后氏之世者 (HFei 49). Cf. Focus 7.

Classifier 36  夕

夢  mèng, *məŋ, vn, ‘dream (of)’, 夢,臥 而以為然也 (Mo 40); a tr: 夜夢之 (Zuo 7.15); with infinitive O: 晉侯夢與楚子 搏 (Zuo 5.28); 不知周之夢為胡蝶與,胡 蝶之夢為周與 (Zhuang 2); b itr: 方其夢 也,不知其夢也 (Zhuang 2). Classifier 37  大 0

2

外  wài, *ŋwâts, nloc, 1 ‘(the) outside’, a head: 大門之外 (Meng 5B6); 大隧之 外,其樂也洩洩 (Zuo 1.1); 外無敵國之 患,內無亂臣之憂 (HFei 14); b mod: 故 外戶而不閉,是謂大同 (Li 9); 2 adv: 外 舉不避讎,內舉不避子 (Lü 1.5); 3 vi, ‘be external, outside’, a itr: 其為形也亦外矣 (Zhuang 18); b caus: 天非私齊魯之民而 外秦人也 (Xun 23); 故內其國而外諸 夏,內諸夏而外夷狄 (CQFL 6). Ant. 內 [11.2]. 3

多  duō, *tâi, v , 1 ‘be many, numerous’, a st

11

itr: 吾樂甚多 (SY 17); b caus, ‘increase’: 晉 平公好樂,多其賦斂 (SY 9); 吾多其義 耳 (SY 12); ‘appreciate, esteem’: 君安能少 趙人而令趙人多君(ZGC 20.14); with locative S, ‘have/be many’: 國多財則遠者 來 (Guan 1); 醫門多疾 (Zhuang 4); c comp: 寡人之民不加多 (Meng 1A3); 無 望民之多於鄰國也 (Meng 1A3); 2 adv, ‘often, mostly; much, many, a lot of ’: 諸侯

大  dà, *dâs, vst, 1 ‘be big, large’, a itr: 畢 萬之後必大 (Zuo 4.1); b caus: 君子不自 大其事 (Li 32); 民不得大其美而小其惡 (Li 33); c comp: 故操彌約,而事彌大 (Xun 3); 齊大於魯 (HFei 39); d mod:大國 之樹必巨 (SY 12); 2 adv, ‘greatly, abundant­ly; plenty’: 大事于大廟 (Zuo 6.2); 齊王方大飲左右 (Lü 20.6); cf. 10.2– ③. Syn. 巨[48.2]; ant. 小[42.0]; dist. 太 [37.1]; 犬 [94.0]. 大夫  dàfū, *dâs-­pa, n , ‘dignitary’: 禮 c

不下庶人,刑不上大夫 (Li 1). Cf. note at 士 [33.0].

大子  tàizı̌,

[37.1].

1

*thâs-­tsəʔ, var. of 太子

夫  fú, *ba, prdem, 1 ‘this, that’, mod: 非夫 人之為慟而誰為 (LY 11.10); 夫人不言, 言必有中 (LY 11.14); 公嗾夫獒焉 (Zuo 7.2); cf. 彼 [60.6], and 20.3–㊲; 2 ptemp, ‘well, now’: 夫上之所為,民之歸也 (Zuo 10.21); 王往而征之,夫誰與王敵 (Meng 1A5); cf. 今 [9.2], and 18.3–⑱.

433

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

夫  fú, *ba, pt [< 不乎], ‘isn't it?’, final: 天下  tiānxià, *thîn-­grâʔ, n , ‘all under c

亦可以弗畔矣夫 (LY 6.27). Cf. Box 16.

夫  fū,

*pa, nc, ‘husband’: 夫妻不 和,家室大凶 (SY 10.7). Cf. 丈夫 [1.2]; ant. 妻 [38.5], 婦 [38.8].

夫人  fūrén, *pa-­nin, n , ‘lady, wife (of c

a ruler)’: 邦君之妻, 君稱之曰夫 人 . . . 邦人稱之曰君夫人 (LY 16.14); 夫 人將啟之 (Zuo 1.1).

夫子  fūzı̌, *pa-­tsəʔ, n , ‘master’: 吾聞 c

諸夫子 (LY 19.17).

太/大  tài, *thâs, vst, ‘be big, great, grand’, mod: 觀於太廟右陛之前 (SY 10). Dist. 犬 [94.0]. 太子  /大子 tàizı̌, *thâs-­tsəʔ, n , a title, c

‘crown prince’:君既許我殺太子而立奚 齊矣 (GY 8.1); 初楚子將以商臣為大子 (Zuo 6.1). Cf. note at 公子 [12.2].

heaven’, used in several senses: ‘oikoumenē, civilized world’: 凡天下戰國七 (SJ 69); 國治而後天下平 (Li 42); ‘kingdom, realm’: 堯欲傳天下於舜 (HFei 34); ‘all the people’ as in tout le monde: 天下皆知 美之為美,斯惡已 (Lao 2); 恐天下之議 我也 (SJS 1).

天子  tiānzı̌,

*thîn-­tsəʔ, nc, ‘Son of Heaven’, i.e. the Zhou king: 故天子聽 政,使公卿至於列士獻詩 (GY 1.3). Cf. note at 天 [37.1].

2

失  shī, *lhit, vtr, ‘let slip, lose (control of); neglect, be remiss’, a tr: 為善者得道,為 惡者失道 (SY 16); b pass: 臣聞酒入舌 出,舌出者言失,言失者身棄 (SY 10); c mod: 是故民無廢事,而國無失利也 (Guan 73). Ant. 得 [60.8]; dist. 矢 [111.0].

天  tiān, *thîn, n , 1 ‘Heaven’, as a deity: 失  yì, *lit, var. of 佚 [9.5]. c

天將以夫子為木鐸 (LY 3.24); 獲罪於 天,無所禱也 (LY 3.13); ‘sky’: 有鳥 焉 . . . 翼若垂天之雲 (Zhuang 1); 天油然 作雲,沛然下雨 (Meng 1A6); ant. 地 [32.3]; ‘nature’: 天行有常,不為堯 存,不為桀亡 (Xun 17); 天有四時,春 秋冬夏 (Li 29); 2 vi, ‘act like Heaven’, itr: 故明主之行制也天,其用人也鬼 (HFei 48). Note: 天 was the highest god of the Zhou. Perhaps origi­nally an anthropomorphic clan deity, it was successively transformed to a transcendent god, a guard­ian of virtuous conduct. Heaven’s repre­ sen­ tative on earth, the ‘Son of Heaven’, 天子, ruled by virtue of the ‘Heavenly Mandate’, 天命, which Heaven could revoke in case of immoral behav­ iour. In an increasingly rational society, 天was described in terms of a mechanistic principle, ‘nature’, or it simply designated the ‘sky’.

天地  tiāndì, *thîn-­draih, n , ‘heaven and earth, (Zhuang 2).

world’:

c

天地與我並生

3

夷  Yí, *ləi, n , a non-­Chinese people: 雖 p

之夷狄,不可棄也 (LY 13.19). Note: The foreign people whose lands bordered on the North China plain were con­sidered by the Chinese as un­ civilized ‘barbar­ ians’; the term 四夷 subsumed the ‘bar­ barians’ in the four cardinal directions: the Yi 夷 to the east, the Rong 戎 to west, the Di 狄 to the north, and the Man 蠻 to the south.

5

奉  fèng, *phoŋʔ, n , ‘service’: 車甲之奉 c

(Sun 2). Dist. 舉 [134.10], 秦 [115.5].

奔  bēn,

*pə̂n, vi, ‘run, flee; hasten, hurry’, a itr: 虢公其奔 (Zuo 5.5); 昭公師 敗,奔於齊 (SJ 47); b caus: 捐殯而奔其 父之使者,是以奔父也 (GLiang 7.18); c pseudo-­tr: 大叔出奔共 (Zuo 1.1); d mod: 奔車之上無仲尼,覆舟之下無伯夷 (HFei 25). Syn. 亡 [8.1]; 走 [156.0]; 逃 [162.6].

434

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

奈/柰  nài, *nâs, v , ‘approach, deal n

with’, al­ most exclusively with 何, ‘what about it? how to deal with it?’: 市義柰何 (ZGC 11.1); 今為之奈何 (SY 18); 禍成 矣,無可奈何 (SJ 4).

6

好 (SY 3.21); 於是選齊國中女子好者八十 人 (SJ 47); b mod: 雖好色, 必惡心, 不 可謂好 (GY 7.3).

好  hào, *hûh, vn, ‘love to (do sth.)’: 不 好犯上,而好作亂者,未之有也 (LY 1.2). Syn. 嗜 [30.10], 欲 [76.7]; ant. 惡 [61.8]; cf. 4.4.1.

契  qì, *khêts, n , ‘contract, agreement, 如  rú, c

tally’: 上古結繩而治,後世聖人易之以 書契 (Yi 8); 載券契而行 (ZGC 11.1). Syn. 券 [18.6].

7

奚  xī, *gê, pr

int, 1 ‘what?’, a as O: 子將奚先 (LY 13.3); b with prep: 雖多,亦奚以為 (LY 13.5); 奚由相得 (Lü 16.3); 君長有齊,奚以 薛為 (HFei 23); 2 adv, ‘why?, how?’: 子奚不 為政 (LY 2.21); 夫奚說書其不義以遺後 世哉 (Mo 17). Syn. 何 [9.5], 胡 [130.5].

11

奪  duó, *lôt, 1 vn ‘steal, rob’, a tr: 盜思奪 之矣 (Yi 7); b tr2: 王奪之人, 霸奪之 與, 彊奪之地 (Xun 9); 勿奪其時 (Meng 1A3); cf. 5.2.2; c itr: 奪然後義,殺 然後仁 (Xun 13); 怒不過奪,喜不過予 (Xun 2); ant. 與/予 [134.7]; 2 nc, ‘robbery’: 故争奪生而辭讓亡焉 (Xun 23). Classifier 38  女 0

女  nǚ, *nraʔ, n , 1 ‘woman; girl’: 男女居 c

室,人之大倫也 (Meng 5A2); esp. ‘daughter’: 昔秦伯嫁其女於晉公子 (HFei 32). Ant. 男 [102.2].

女  rǔ, *naʔ, var. of 汝 [85.3]. 2

奴  nú, *nâ, n , ‘slave’: 箕子為之奴 (LY c

18.1); 田事聽此奴, 必與同食 (SJ 100).

3

好  hǎo, *hûʔ, vst, ‘be good, pleasant, fine; beauti­ful’, a itr: 蒼梧之弟娶妻而美

*na, vn, ‘go to’, a tr: 鄭伯如 周,始朝桓王也 (Zuo 1.6); 孔子遂 行,復如陳 (SJ 47); b tr with C, ‘approach, come to’: 匡人其如予何 (LY 9.5); 君使臣,臣事君,如之何 (LY 3.19). Syn. 若 [140.5].

如  rú, *na(h), vn, 1 perhaps < preceding (‘ap­proach’ > ‘come close to’ >), ‘be like, equal’, a cop with subject C: 仁人固如是 乎 (Meng 5A3); b in object C: 文王視民 如傷 (Meng 4B20); cf. 11.4; 2 conj, ‘if ’: 子 如不言,則小子何述焉 (LY 17.16); cf. 15.4.1. Syn. 若 [140.5]. 妄  wàng, *maŋh, vi, ‘be unreasonable, reckless; useless’, a itr: 彼好專利而妄 (Zuo 12.25); 上無量則民乃妄 (Guan 1); b mod: 此亦妄人也已矣 (Meng 4B28). 4

妙  miào, *miauh, vi, ‘be marvellous, profound, subtle’, a itr: 故常無,欲以觀 其妙 (Lao 1); 至精而妙乎天地之間 者,德也 (HSWZ 5); b mod: 我以為妙 道之行也 (Zhuang 2); 殆物妙言,治之 害也 (HFei 51). Syn. 微 [60.10]. 5

妻  qī, *tshə̂i, nc, ‘(main) wife’: 齊人有 一妻一妾而處室者 (Meng 4B33); 殺孔 父而取其妻 (Zuo 2.2); 父之兄妻為世 母,父之弟妻為叔母 (Erya 4). Ant. 夫 [37.1]. Note: Ancient Chinese society was polygamous. While a man was expected to have only one pri­mary wife, 妻, he could have several ‘companions’, 媵, which were female relatives that accompanied the primary wife into marriage, in addition to lower-­ranking concu­bines, 妾

435

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

[38.5], which came from other families. Cf. note at 婦 [38.8].

妻  qì, *tshə̂ih [< preceding], v , ‘give a n

wife (to)’, a tr: 齊又欲妻之 (SY 13); b tr with PrP: 以其子妻之 (LY 5.1); 翟以其少 女妻重耳,長女妻趙衰而生盾 (SJ 43). Cf. 娶 [38.8]; 嫁 [38.10].

妾  qiè, *tshap, nc, ‘concubine’: 聘則為 妻,奔則為妾 (Li 12); 無以妾為妻 (Meng 6B7); 齊人有一妻一妾而處室者 (Meng 4B33). Cf. note at 妻 [38.5].

始  shı̌, *lhəʔ, ntemp, 1 ‘begin­ning’, a ini: 始吾於人也,聽其言而信其行 (LY 5.11); b head: 無名天地之始 (Lao 1); 2 vi ‘begin, start’, a itr: 征商,自此賤丈夫始 矣 (Meng 2B10); 吾欲為難,安始而可 (GY 7.6); b caus: 無始亂,無怙富 (Zuo 11.4); 昔者其父始之,我終之 (GY 14.9); 3 adv, ‘initially, first’: 吾始聞之懼 (Zhuang 14); 于是始作八卦 (Yi 8). Ant. 卒 [24.6], 終 [120.5]. 委  wěi,

*ʔoiʔ, vn, ‘surrender, deliver, hand over’, a tr: 魏委國於王 (ZGC 25.12); b itr: 王使委於三吏 (Zuo 8.2); 以膠東委 於燕,以濟西委於趙 (SJ 69). Cf. 8.2.4.

委吏  wěilì,

*ʔoiʔ-­rəh, nc, ‘commissioner’: 孔子嘗為委吏矣 (Meng 5B5).

姓  xìng, *seŋh, n , ‘clan; clan name’: 周 c

姬姓也,而彼姜姓也 (Xun 12); thence ‘family name’: 字仲尼,姓孔氏 (SJ 47). Cf. 百姓 [106.1]. Note: This term, originally refer­ ring to aristocratic clans, changed its mean­ ing with the social changes in Zhanguo times. When com­ moners, too, began using 姓, they became family names. Cf. Focus 5 and note at 宗 [40.5].

姑  gū, *kâ, adv, ‘now, for a while/time

being/a moment’, expresses ‘temporary’ aspect: 子姑待之 (Zuo 1.1); 若子之為人 吏,宜受上戮,子姑默矣 (YTL 24). Cf. 10.5.

姑/家  gū, *kâ, n , ‘father’s sister; husc

band’s mother’: 婦人不飾,不敢見舅姑

(Li 4); 父母舅姑之命,勿逆勿怠 (Li 12); 長鋏歸來乎,無以為家(ZGC 11.1).

6

威/畏  wēi, *ʔui, n , 1 ‘power, authorc

ity’: 所謂威者, 擅權勢而輕重者也 (HFei 52); 誠動可畏謂之威,反威為圂 some; (XS 8.3); 2 vst ‘be majestic, awe­ intimi­dating’: a itr: 君子不重則不威 (LY 1.8); 子溫而厲,威而不猛,恭而安 (LY 7.38); b caus, ‘consider awesome, fear’: 宗邑無主,則民不威 (GY 7.6); but c tr [!], ‘intimidate, over­awe’: 嚴斷刑罰,以 威其淫,懼其未也 (Zuo 10.6); 立刑罰 以威其下 (Guan 67); 立民所惡,以禁 其邪,故為刑罰以畏之 (Guan 67); d comp: 德莫厚焉,刑莫威焉 (Zuo 5.15); e mod: 威令不聞 (Guan 1). Cf. 畏[102.4].

姦/菅  jiān, *krân, v , ‘be treacherous, i

deceitful, in­sidious’, a itr: 八風從律而不 姦 (SY19); 野蕪曠則民乃菅 (Guan 1); b mod: 如是則姦言姦說姦事姦謀姦 譽姦愬,莫之試也 (Xun 14). Syn. 邪 [163.4], 怪 [61.5]; ant. 忠 [61.4].

7

娣  dì, *dîʔ, n , ‘younger

sister’: 娣者 c 何,弟也 (GYang 3.19); 其娣生卓子 (GY 7.3).

姬  Jī, *kə (*kjə), n , a clan: 君姬氏之愛 p

子也 (Zuo 7.2).

8

婢  bèi/bì,

*beʔ, nc, ‘female servant, maid’: 而母婢也 (ZGC 20.13).

婦  fù, *bəʔ, n , ‘married woman’: 父子 c

有親,君臣有義,夫婦有別,長幼有 序,朋友有信 (Meng 3A4). Ant. 夫 [37.1]. Note: While 婦 is the most common word for a married woman, CC has a number of specific words for wives of different ranks: 天子之妃曰后,諸侯曰 夫人,大夫曰孺人, 士曰婦人,庶人 曰妻 (Li 2). Cf. note at 妻 [38.5].

436

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

婦人  fùrén, *bəʔ-nin, n , ‘woman; wife c

(of an officer)’: 古者,丈夫不耕 . . . 婦人 不織 (HFei 49). Ant. 丈夫 [1.2].

娶  qǔ, *tshoh, v , ‘marry, take as wife’, a n

tr: 蒼梧之弟娶妻而美好 (SY 3.21); b itr: 舜不告而娶 (Meng 4A27); 椒舉娶于申 公子牟 (Zuo 9.26). Cf. 嫁 [38.10]; 妻 [38.5].

10

嫁  jià, *krâh, v , ‘give [a daughter] as a tr

wife, marry off ’, a tr: 凡諸侯嫁女,同姓 媵之,異姓則否 (Zuo 8.8); 晉獻公筮嫁 伯姬於秦 (Zuo 5.15); b pass, ‘get married’, said of women: 公女,嫁于敵國 (Zuo 2.3); 女子許嫁 (Li 1); c pseudo-­tr2: 太后嫁 女諸侯 (ZGC 30.6); d mod: 共粉飾 之,如嫁女床席 (SJ 126). Cf. 娶 [38.8], 妻 [38.5].

嫉/疾  jí, *dzit(s), vi, ‘be jealous (of); hate’ tr: 不肖嫉賢,愚者嫉智 (XX 2); 荀 卿嫉濁世之政,亡國亂君相屬 (SJ 74). Syn. 惡 [61.8]. 12

itr: 君君,臣臣,父父,子子 (LY 12.11); b caus, ‘treat like a child’: 故人不獨親其 親,不獨子其子 (Li 9); 不拊愛子其民 (ZGC 11.1). Ant. 父 [88.0]; cf. 兒 [10.6], 嬰 兒 [38.14], 童 [117.7].

子  zı̌, *tsəʔ, nc, ‘viscount’, second-­lowest of the five aris­tocratic ranks of the Zhou: 王者之制祿爵,公侯伯子男,凡五等 (Li 5); 箕子為之奴 (LY 18.1); cf. Box 15; perhaps through reduction in status, ‘mas­ ter’: 孔子為魯司寇 (SY 14); ‘sir, you’, re­ spectful form of address: 子非魚 (Zhuang 17); 二三子,何患於喪乎 (LY 3.24); anteposed as O in negated clauses: 不子欺 (Lü 15.7); 功不子逮 (Yan 2.24); honorific ele­ment in courtesy names: 子路 使門人為臣 (LY 9.12); cf. Focus 5. 子  zı̌, *tsəʔ, n

temp, first of the ‘Earthly branches’: 甲子日,紂兵敗 (SJ 3). Cf. Focus 8.

1

孔  Kǒng, *khôŋʔ, np, family name: 孔 子. . . 名曰丘云,字仲尼,姓孔氏 (SJ 47). 3

嬉  xī, *hə, v , ‘amuse oneself’: 江沅之間 存  cún, *dzə̂n, v , a itr, ‘exist, be there, 謂戲為媱,或謂之愓,或謂之嬉 (FY n

i

10); only itr: 孔子為兒嬉戲 (SJ 47); 孟子 之少也,嬉遊為墓間之事 (LNZ 1.11).

survive’: 父母俱存 (Meng 7A20); b caus, ‘keep, retain’: 存其心,養其性,所以事 天也 (Meng 7A1). Ant. 亡 [8.1], 去 [28.3].

字  zì, *dzəh, n , 1 ‘(courtesy) name’: 幼

14

c

嬰兒  yīngér,

*ʔeŋ-­ŋe, nc, ‘toddler, 名,冠字 (Li 3); 名曰丘云,字仲 child’: 嬰兒 . . . 以塵為飯,以塗為羹,以 尼,姓孔氏 (SJ 47); 2 vtr, ‘give a courtesy name’, a tr: 吾不知其名,字之曰道 (Lao 木為胾 (HFei 32). Cf. 子 [39.0]. 25); b pass: 男子二十冠而字(Li 1). Cf. note at 名 [30.3] and Focus 5.

Classifier 39  子

4

0

子  zı̌, *tsəʔ, n , 1 ‘child’: 干越夷貉之 c

子,生而同聲 (Xun 1); esp. ‘son’: 子女不 雜處 (XS 8.1); ant. 女 [38.0]; more generally, ‘youngers’, i.e. male rela­ tives of younger gener­ations: 自王以下,各有父 兄子弟 (Zuo 9.14); 2 vst, ‘be childlike’, a

孝  xiào, *hrâuh, nc, 1 ‘filial piety, obedience’: 致孝乎鬼神 (LY 8.21); 孝子之 至,莫大乎尊親 (Meng 5A4); 君義,臣 行,父慈,子孝,兄愛,弟敬,所謂 六順也 (Zuo 1.3); 子愛利親謂之孝,反 孝為孽 (XS 8.3); 2 vst, ‘be filial’, a itr: 堯舜 之道,孝弟而已矣 (Meng 6B2); 孝於父

437

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

母,信於交友 (SY 16); 入孝於親,出忠 於君 (HNZ 9); b pseudo-­tr: 曾參孝其 親,天下願以為子 (ZGC 7.8); c mod: 孝 子不失時以養 (HSWZ 3).

5

孤  gū, *kwâ, v , ‘be fatherless, or­phaned, i

soli­tary’: 幼而無父曰孤 (Meng 1B5); a itr: 德不孤,必有鄰 (Meng 1B17); 矜寡 孤獨廢疾者,皆有所養 (Li 9); b caus: 殺妾父,孤妾身 (Yan 2.27); fig. ‘isolate’: 莫如從親以孤秦 (SJ 69); c mod: 孤子當 室,冠衣不純采 (Li 1).

孟  mèng,

*mrâŋh, nc, 1 ‘senior’, first-­ born of a secondary wife: 謂之吳孟子 (LY 7.31); cf. Focus 5; 2 np, family name: 孟 軻,騶人也 (SJ 74).

季  jì, *kwis, n , 1 ‘junior’, youngest in a c

row of brothers, used in courtesy names: 髙 辛氏有才子八人,以伯仲叔季爲 序(Zuo 6.18); cf. Focus 5; 2 a family name: 孔子…嘗為季氏史 (SJ 47).

6

孩/咳  hái, *gə̂, vi, ‘smile (like a toddler)’, a itr: 撫其首咳而名之 (Li 12); 孩 提之童 (Meng 7A15); b mod: 今王播棄 黎老,而孩童焉比謀 (GY 19.5). 7

孫  sūn, *sûn, nc, ‘grandchild’: 子之子為 何,曰為孫 (SJ 75); 子又有五子,大父 未死而有二十五孫 (HFei 49). 8

於文 (LY 6.27); syn. 習 [124.5]; 2 nc, ‘learning’: 吾必謂之學矣 (LY 1.7).

Classifier 40  宀 2

它  tā/tuō, *lhâi, var. of 他 [9.3]. 3

安  ān, *ʔân, vst, 1 ‘be quiet, satisfied; peaceful, secure’, a itr: 政治而國安也 (Mo 12); 吾生於陵而安於陵 (Zhuang 19); b caus: ‘pacify’: 明主,救天下之 禍,安天下之危者也 (Guan 64); ‘acquiesce, find peace in’: 美其服,安其居 (Lao 80); 法有,民安其次 (SJS 20); c comp: 易於反掌,安於太山 (SY 9); 2 nc, ‘peace, security; well-­being’: 見國之利而 忘君安,非子道也 (XX 7.10). Ant. 危 [26.4]. 安  ān, *ʔân, adv, ‘how?’: 吾欲為難,安

始而可 (GY 7.6); 子非魚,安知魚之樂 (Zhuang 17); ‘where?’: 沛公安在 (SJ 7); 安 見君子 (SY 3.6). Cf. 焉 [86.7].

守  shǒu, *hjuʔ, vn, 1 ‘guard, protect’, a tr: 陽虎務取之,我務守之 (HFei 33); 因釋 其耒而守株 (HFei 49); in military context, ‘defend’: 守城之道,盛力也 (SJS 12); b itr: 建國諸侯之君分土而守 (Xun 11); 不令則不可以守,不可以戰 (Lü 26.3); c mod (!): 四戰之國,貴守戰 (SJS 12); ant. 攻 [66.3], 戰 [62.12]; 2 nc, ‘protection’: 務在四時,守在倉廩 (Guan 1).

孰  shú, *duk, adv [< 誰, 149.8], 1 ‘which 宅  zhái, one?’: 禮與食孰重 (Meng 6B1); 2 print, ‘who, who­ever’: 孰謂鄹人之子知禮乎 (LY 3.15); 孰殺子產,吾其與之 (Lü 16.5); Cf. 7.3.

孰  shú/shóu, *duk, var. of 熟 [86.11].

*drâk, nc, 1 ‘dwelling, residence’: 五畝之宅, 樹之以桑, 五十者 可以衣帛矣 (Meng 1A3); 夫陳善田利宅 所以戰士卒也 (HFei 45); 2 vn, ‘live, dwell (at)’, only itr: 身今得見王,而家宅乎齊 (Lü 23.1).

宇  yǔ, *waʔ, n , ‘eaves’: 上棟下宇,以 c

13

學  xué, *grûk, v , 1 ‘learn, study’, a tr: 小 n

子何莫學夫詩 (LY 17.8); b itr: 君子博學

待風雨 (Yi 8); by extension: ‘space’: 宇,東西家南北 (Mo 40); 有實而無乎 處者,宇也 (Zhuang 23).

438

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

4

宋  Sòng, *sûŋh, n , a polity: 宋,小國 p

也 (Meng 3B5); 孔丘,聖人之後,滅於 宋 (SJ 47). Note: Song, a small polity in the east of modern Henan, was given as a fief to the descendants of the Shang dynasty (Zi 子 clan) in the eleventh century bc. Confucius family is said to have come from Song. Although Duke Xiang of Song (r. 650–637) dis­played ambitions to become hegemon, Song’s power declined rapidly thereafter. Following sev­eral coups d’état in the fourth century, it was finally annihilated by Qi in 286 bc.

完  wán, *gôn, vi, ‘be complete, whole, intact, ready’, a itr: 城郭不完,兵甲不多 (Meng 4A1); 彼其事王必完矣 (ZGC 23.15); b caus, ‘keep intact, pre­serve’: 邑以 賈怠, 不如完舊 (Zuo 10.15); 不能完父 兄子弟 (SJ 8); ‘repair; finish; strengthen, fortify’: 父母使舜完廩 (Meng 5A2); 完堤 防,謹壅塞 (Li 6). 5

定  dìng,

*dêŋh, vtr, 1 ‘settle, steady, estab­lish; determine, decide’, a tr: 中天下 而立,定四海之民 (Meng 7A21); b pass: 名定而實辨,道行而志通 (Xun 22); 天 下惡乎定 . . . 定于一 (Meng 1A6); c mod: 言無定術,行無常議 (HFei 19.50); 2 nc, ‘steadiness, stability’: 知止而後有定 (Li 42); 3 np, ‘Orderly’, a posth. name: 定公十 四年,孔子年五十六 (SJ 47).

官  guān,

*kwân, nc, 1 ‘office; official, officer’: 凡官者,以治為任,以亂為罪 (Lü 17.3); fig. ‘organ’: 官知止而神欲行 (Zhuang 3); 2 vi ‘be an officer, hold office’, a itr: 大德不官,大道不器 (Li 18); 官於 大夫者之為之服也,自管仲始也 (Li 21); b caus: 今君分之土而官之 (GY 7.7). Note: 官 is the common word for offices and, by extension, their occupants; 吏 and 有事 mean generic ‘offi­cers’, whereas 司 徒, 司馬, 司空, 司寇, and others de­signate specific officers.

宜  yí, *ŋai, vst, 1 ‘be appropriate’, a itr: 宜 乎百姓之謂我愛也 (Meng 1A1); b caus, ‘find appropriate, approve of ’: 子甚宜其 妻 (Li 12); c mod: 中國夷蠻戎狄,皆有 安居和味宜服利用備器 (Li 5); b aux, ‘should (do sth.)’: 是宜為君,有恤民之 心 (Zuo 3.11); cf. 4.4.1; 2 nc, ‘proper place, position, method’, etc.: 羣道當則萬物皆 得其宜 (Xun 9); 禮從宜,使從俗 (Li 1).

宗  zōng, *tsûŋ, nc, ‘lineage, common descent group; ancestor; ancestral home’: 敬宗廟,恭祖舊 (Guan1); 父母宗族 . . . 為戮沒 (ZGC 31.5); 晉,吾宗也 (Zuo 5.5). Note: The 宗, members of which claimed descent from a common ancestor, was part of a larger clan, 姓, which claimed descent from a mythical ancestor; it was larger than the 族, which were limited to common ancestry over five generations. 6

宦  huàn, *grôns, vi, ‘serve, be on duty’, a itr: 宦三年矣,未知母之存否 (Zuo 7.2); b caus: 宦公子於四鄰,四鄰必重之 (HFei 31).

客  kè, *khrâk, nc, 1 ‘guest, client’: 七 教,父子兄弟夫婦君臣長幼朋友賓客 (Li 5); 2 vst, ‘be a guest, client’, a itr: 子 貧,客於此 (SJ 85); b caus: 夫子見 妾,勃然不悅,是客妾也 (LNZ 1.11); 孟嘗君客我 (ZGC 11.1).

室  shì, *lhit, nc, ‘chamber, room; house (build­ing)’: 仲子所居之室,伯夷之所 築與 (Meng 3B10); ‘house, family’: 故視 人之室若其室,誰竊 (Mo 14); 夫妻不 和,家室大凶 (SY 10.7); esp. the royal house: 為文王卿士,勳在王室 (Zuo 5.5); 孔子西藏書於周室 (Zhuang 13); cf. note at 家 [40.7]. Note: 室 is the most general word for a room in a house or the house itself; 寢 [40.11] specifically means a bedroom; 堂 [32.8] were ceremonial halls for rulers, built on an ele­vated platform; 庭 [53.7] is a more gener­al word for a hall and the courtyard in its front.

439

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

宣  xuān, *swan, vi, 1 ‘be perspicuous, lucid’, a itr: 君人者,宣則直言至矣 (Xun 21); 繼文之業,而信宣於諸侯 (Zuo 5.25); b caus ‘[make clear >] spread, proclaim, propagate’: 文王欲伐崇,先 宣言曰 (SY 15); 為民者宣之使言 (GY 1.3); 2 np, ‘Per­spicuous’, a posth. name: 周 宣王殺其臣杜伯而不辜 (Mo 31).  

宰  zǎi, *tsə̂ʔ, n , ‘superintendent’: 宰者 c

何,官也 (GYang 1.1); 原思為之宰 (LY 6.5).

宰夫  zǎifū, *tsə̂ʔ-­pa, nc, ‘assistant minister; chef, cook’: 宰夫胹熊蹯不熟 (Zuo 7.2). 8

寇  kòu, *khôh, n , ‘brigand, marauder’: c

7

宮  gōng, *kuŋ, n , ‘residence, abode’: 故 c

宮室可得而居也 (Xun 9); esp. ‘palace’: 古 之王者,擇天下之中而立國,擇國之 中而立宮,擇宮之中而立廟 (Lü 17.6).

害  hài, *gâts, nc, 1 ‘damage, harm’: 害 者, 利之反也 (HFei 46); 此天下之害 與,天下之利與 (Mo 16); 2 vn, ‘damage, harm, hurt’, a itr: 生有益於人,死不害 於人 (Li 3); 利則行之,害則捨之 (SJ 79); b tr: 不以私害公 (LNZ 5.6); 豈害我 哉 (Zuo 5.5). Syn. 禍 [113.9]; ant. 利 [18.6]. 家  jiā, *krâ, n , ‘family, house, estate’: 若 c

大國之攻小國也,大家之亂小家也 (Mo 16); ‘school (of thought)’: 法家 . . . 名 家 . . . 道家 (SJ 130). Cf. 國[31.8]. Note: The term means lower aristo­cratic families and their estates in contrast to the re­gional polities (國); the house of the Zhou rulers was called 周室. In Zhanguo times, 家 came to mean simply ‘household’.

- 大家  dàjiā, *dâs-­krâ, n , ‘great c

(aristocratic) family’: 大家之亂小家 也 . . . 此天下之害也 (Mo 16).

-百家  bǎijiā,

*prâk-­ krâ, nc, ‘Hundred Schools (of thought)’: 今諸侯 異政,百家異說 (Xun 21).

家  gū, *kâ, var. of 姑 [38.5]. 容  róng, *loŋ, n , ‘face, look; expression, c

counte­nance’: 孔子為兒嬉戲,常陳俎 豆,設禮容 (SJ 47); 舜見瞽瞍,其容有 蹙 (Meng 5A4).

宴  yàn, *ʔêns, n , ‘feast, banquet’: 臣侍 c

君宴,過三爵,非禮也 (Zuo 7.2).

寇不可翫 (Zuo 5.5); 即戎寇至,傳鼓相 告 (Lü 22.3).

寅  yín, *jən/jəi, ntemp, third of the ‘Earthly branch­es’: 四年正月庚寅,孝公 生 (SJ 5). Cf. Focus 8.

寄  jì, *kaih, v , ‘stay, sojourn, lodge’, a itr: i

寄於上最親,意傷之 (SJ 59); 為丘墓之 寄於荊也 (SJ 79); b caus, ‘send, entrust to’: 可以託六尺之孤,可以寄百里之命 (LY 8.6); 孟嘗君寄客於齊王 (SY 11); c pseudo-­tr: 寡人得寄僻陋蠻夷之鄉 (SY 12); d pseudo-­tr2: 因為之書,寄之秦王 (SY 11). Syn. 託 [149.3].

9

富  fù, *pəkh, nc, 1 ‘wealth, affluence’: 視 生如死,視富如貧,視人如豕,視吾 如人 (Lie 4); 2 vst, ‘be rich’, a itr: 今君富於 季孫,而齊大於魯 (HFei 39); b caus: 其 摶力也,以富國強兵也 (SJS 8).

寒  hán, *gân, vst, 1 ‘be cold’, a itr: 脣亡 齒寒 (Zuo 5.5); b caus: 一日暴之,十日 寒之 (Meng 6A9); c comp: 冰,水為 之,而寒於水 (Xun 1); d mod: 寒水之所 積也 (HNan 4); 2 nc, ‘cold’: 馬,蹄可以 踐霜雪,毛可以禦風寒 (Zhuang 9); 日 月運行,一寒一暑 (Yi 7). Ant. 暑 [72.8]. 寐  mèi, *mis, v , ‘sleep’, itr: 吾聞之,喜 n

而不寐 (Meng 6B13); 獻公為之不寐 (CQFL 14).

10

寘  zhì, *tes, v , ‘put, place; set aside’, a tr: n

寘諸橐以與之 (Zuo 7.2); b itr: 疾使侍人 誘其初妻之娣,寘於犁,而為之一宮 (Zuo 12.11)

440

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

11

察  chá, *tshrêt, vn, ‘clarify, investigate, analyse; be clear (about)’, a tr: 願孰察之 (SJS 1); b itr: 不察乎人之性偽之分者也 (Xun 23); 百官以治,萬民以察 (Yi 8). 寡  guǎ, *kwrâʔ, v , 1 ‘be solitary, few’, a itr: i

地澙鹵,人民寡 (SJ 129); 儒者華於言而 寡於實 (YTL 24); b caus: 其所以寡人之道 多 (Mo 20); 約其所守,寡其所求 (HNZ 1); ‘have few’: 多事而寡功 (Xun 6); 2 nc, ‘widow’: 老而無夫曰寡 (Meng 1B5); 恤鰥 寡,問疾病 (Guan 10); cf. 鰥 [195.10].

寡人  guǎrén, *kwrâʔ-­nin, nc, ‘my humble self, I’, used by the ruler: 寡人託國 於子 (HFei 34).

寧  níng, *nêŋ, v , ‘be peaceful, tranquil, st

at ease’, a itr: 四民均則王道興而百姓寧 (SY 7); 國多盜賊,邊境不寧 (Li 6); b caus: 無畏,寧爾也 (Meng 7B4); 龍反其 淵,安寧其處 (SY 6).

寢  qı̌n, *tshəmʔ, vi, 1 ‘sleep, rest’, a itr: 吾 嘗終日不食, 終夜不寢 (LY 15.31); b caus, ‘put to rest’: 今一日飲酒,而三日 寢之 (Yan 1.3); c mod: 必有寢衣,長一 身有半 (LY 10.5); 見寢石以為伏虎也 (Xun 21); 2 nc, ‘private room, bed­room’: 寢 門闢矣 (Zuo 7.2); cf. note at 室 [40.6].

實  shí, *dit, vi, 1 ‘be full, filled up; solid, substan­tial, genuine, true’, a itr: 田野開 闢,府倉實,民眾殷 (GY 21.1); b caus: 虛其心,實其腹,弱其志,強其骨 (Lao 3); syn. 滿 [85.11]; ant. 虛 [141.6], 空 [116.3]; 2 nc, ‘reality’: 名定而實辨,道行 而志通 (Xun 22); ant. 名 [30.3]. 實  shí, *m-­lit, n , ‘fruit’: 草木之實足食 也 (HFei 49).

c

實  shí, *m-­dit, prdem, ‘he, she, it, they’, used as re­sumptive ‘dummy’, a for the S: 人實不 知,非龍實知 (Zuo 10.29); 雖楚有 材,晉實用之 (Zuo 9.26); b for the O: 鬼 神非人實親,惟德是依 (Zuo 5.5); cf. 14.2. 寤  wù, *ŋâh, 1 var. of 悟 [61.7]; 2 var. of 忤 [61.4].

12

寬  kuān,

*khwân, vn, ‘be generous, indulgent, lenient’, a itr: 寬則得眾 (LY 17.5); b tr: 致彊而能以寬弱 (Xun 7).

審  shěn, *lhəmʔ, v , ‘be clear, obvious, i

manifest’, a itr: 是白馬之非馬,審矣 (GLong 1.2); 名之審於是非也,猶繩之 審於曲直也 (CQFL 35); b caus, ‘make clear, examine, study’: 欲知億萬,則審 一二,欲知上世,則審周道,欲審周 道,則審其人所貴君子 (Xun 5); 審法 制,明賞罰 (WLiao 4).

16

寵  chǒng, *rhoŋʔ, n , 1 ‘favour, honour’, c

a head: 臣有寵矣,然而臣卑 (HFei 36); 無恃寵 (Zuo 11.4); b mod: 蔓草猶不可 除,況君之寵弟乎 (Zuo 1.1); 2 vn, ‘favour, honour’, only tr: 不自退而寵其政 (GY 14.17); 寵神其祖,以取威於民 (GY 18.1).

17

寶  bǎo, *pûʔ, n , 1 ‘jewellery, treasure, c

valu­ables’: 玉人以為寶也 (Zuo 9.15); 2 vst, ‘be valu­able, precious’, only caus: 寶 之,珍之,貴之,神之 (Xun 5); 吾有 三寶,持而寶之 (HFei 20); syn. 珍 [96.5].

Classifier 41  寸 0

寸  cùn, *tshûns, n , 1 measure of length, c

‘inch’: 孔子長九尺有六寸 (SJ 47); 2 adv: 子腸亦且寸絕 (ZGC 31.2).

6

封  fēng, *poŋ, vtr, ‘cover with a mound; en­feoff ’: a tr: 制其畿方千里,而封樹之 (ZL 18); 今不封蔡 (Zuo 10.13); 遂封唐叔 虞於晉 (SY 1); b pass: 蔡不封矣 (Zuo 10.13); 大公封於營丘 (Li 3).

441

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

7

射  shè, *m-­lakh, vn, 1 ‘shoot (an arrow)’, a tr: 公如囿射鴻 (Lü 25.6); 斬而射 之,通於犀革 (SY 3); b itr: 由射於百步 之外也 (Meng 5B1); 2 nc, ‘arch­ery’: 是故 古者天子,以射選諸侯卿大夫士,射 者,男子之事也 (Li 46). 8

且格 (LY 2.3); 導其妻子,使養其老 (Meng 7A22); b itr: 孔子以三代之道,教 導於後世 (SY 5); 是故為川者决之使導 (GY 1.3); 2 nc, ‘instruction, guid­ance’: 故必 將有師法之化,禮義之道 (Xun 23).

Classifier 42  小 0

將  jiāng, *tsaŋ, adv, ‘be about to; intend 小  xiǎo, *siauʔ, v , a itr, ‘be small’: 聖人 to’: 天將以夫子為木鐸 (LY 3.24); 孟子 st

將朝王 (Meng 2B2); 其所善者,吾將行 之;其所惡者,吾將改之 (XX 4). Cf. 10.5; syn. 且 [1.4].

將  jiàng, *tsaŋh, n , ‘leader, general’: 將 c

者智信仁勇嚴也 (Sun 1); 夫將者,國之 輔也 (Sun 3). Syn. 帥 [50.6].

尉  wèi, *ʔuts, nc, a title, ‘commandant’: 分天下為三十六郡,郡置守尉監 (SJ 6). 專  zhuān, *ton, v , 1 ‘monopolize, conn

centrate (on)’, a tr: 臣得盡其力,而主得 專其柄 (SJS 6); 不敢專其智而委之屈春 (SY 2); b itr: 人其以朕為忘賢有德者而專 於子 (SJ 10); 2 adv, ‘alone, only, especially’: 今王之地方五千里,帶甲百萬,而專 屬之昭奚恤 (ZGC 14.3).

9

之心小矣 (HNan 9); b caus: 孔子登東山 而小魯,登泰山而小天下 (Meng 7A24); 匠人斵而小之 (Meng 1B9); c mod: 若大國之攻小國也,大家之亂小 家也 (Mo 16). Ant. 大 [37.0], 巨 [48.2].

小人  xiǎorén,

*siauʔ-­nin, nc, ‘petty man’: 君子和而不同,小人同而不和 (LY 13.23). Ant. 君子 [30.4].

1

少  shǎo, *hjau, v , 1 ‘be little, few’, a itr: st

其為天下多,其私樂少矣 (Xun 22); b caus, ‘belittle, de­ni­grate’: 君安能少趙人 而令趙人多君(ZGC 20.14); 顯王左右素 習知蘇秦 ,皆少之 (SJ 69); c comp: 鄰 國之民不加少 (Meng 1A3); 一少於二而 多於五 (Mo 42); 2 adv: 少嘗苦曰苦,多 嘗苦曰甘 (Mo 17). Ant. 多 [36.3].

尊  zūn, *tsûn, v , ‘be honourable, ven- 少  shào, *hjauh, v , ‘be young’, a itr: 人 erable, re­spected’, a itr: 德成而教尊,教 st

st

尊而官正 (Li 8); b caus, ‘honour, respect’: 莫不尊親 (Li 31); c comp: 三加彌尊,喻 其志也 (Li 11); 故主上愈卑,私門益尊 (HFei 11). Ant. 卑 [24.6].

11

對  duì, *tûps [< 答, 118.6], v , 1 ‘confront, n

face; answer, re­ply’, a itr: 子路不對 (LY 7.19); b tr: 魯君之使者至,顏闔自對之 (Lü 2.2); 三公問夫子政一也,夫子對之不同 (HFei 38); 2 adv: 孟子對曰 (Meng 1A1).

13

導/道  dǎo, *lûh, v , 1 ‘instruct, lead, n

steer’, a tr: 道之以德,齊之以禮,有恥

少, 則慕父母 (Meng 5A1); 人之少也愚 (Lü 1.4); b caus: 公其毋少少,毋賤賤 (Guan 51); c comp: 韓起少於欒黶 (Zuo 9.9); d mod: 少子亡而不得. Cf. note at 幼 [52.2].

5

尚  shàng, *daŋh, v , a itr, ‘be supreme, i

superior’: 夫至仁尚矣 (Zhuang 14); b  caus, ‘elevate, honour, esteem’: 不尚 賢,使民不爭 (Lao 3); syn. 上 [1.2].

尚  shàng, *daŋh, adv, ‘still’: 尚早,坐

而假寐 (Zuo 7.2); 王孫滿尚幼 (Zuo 5.33); ‘even’: 其子之肉尚食之,其誰不 食 (ZGC 22.3).

442

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 43  尢

home’: 問其名居 (Zuo 7.2); 山林者,鳥 獸之居也 (Xun 14). Syn. 處 [141.5].

9

就  jiù, *dzuh, v , 1 ‘approach, advance; n

return to; attain, achieve’, a tr: 木受繩則 直,金就礪則利 (Xun 1); 處工,就官 府, 處商,就市井, 處農,就田野 (GY 6.1); 諸侯罷戲下,各就國 (SJ 7); b itr: 寡人如就見者也 (Meng 2B2); ant. 去 [28.3]; 2 nc, ‘achieve­ment’: 凡性者,天之 就也 (Xun 23).

屈  jué, *gut, v , ‘wear out’, a tr: 堯且屈 tr

力,而況眾人乎 (Lü 17.6); b pass: 攻城 則力屈 (Sun 2).

6

屏  bı̌ng, *peŋʔ/h, v , ‘shield’, tr: 屏王之 n

耳目,使不聰明 (Zuo 10.27).

11

屢  lǚ/lǜ, *roh, adv, ‘repeatedly, several

Classifier 44  尸

times’: 孔子屢嘆之 (Li 11).

0

尸  shī, *lhi, n , 1 ‘corpse, (dead) body’: c

12

隨梧之死,楊朱撫其尸而哭 (Lie 4); 2 vi, ‘be like fa corpse’, a itr: 寢不尸,居不 容 (LY 10.24); b caus, ‘expose as a corpse’: 祭仲殺雍糾,尸諸周氏之汪 (Zuo 2.15); 3 adv, ‘like a corpse’: 君子 . . . 尸居而龍見 (Zhuang 11).

履  lǚ/lı̌, *riʔ, v , ‘step, tread (on)’, a tr:嘗

1

to; as­semble’ a tr:宋穆公疾,召大司馬 孔父而屬殤公焉 (Zuo 1.3); 乃屬其耆老 而告之曰 (Meng 1B15); b itr: 黥布蒲將 軍亦以兵屬焉 (SJ 7); c pseudo-­tr: 使人屬 孟嘗君 (ZGC 11.1); d pseudo-­tr2: 專屬之 昭奚恤 (ZGC 14.3).

尺  chı̌, *thak, nc, measure of length, ‘foot’, a head: 孔子長九尺有六寸 (SJ 47); b mod: 尺地莫非其有也 (Meng 2A1). 5

n

與汝登高山,履危石 (Zhuang 21); b itr: 行不能正履 (Li 8). Syn. 踐 [157.8].

18

屬  zhǔ, *tok, v , ‘attach, entrust, assign n

居  jū, *kaʔ/kah, v , ‘sit, rest’, a itr: 夜行晝 屬  shǔ, i

居 (Zhuang 20); 不在所與居,在所與謀 也 (HFei 33); b caus, ‘settle, place’: 萬民弗 忍,居王于彘 (Zuo 10.26); 真可以居吾 子矣 (LNZ 1.11); c mod: 皆有遠志,無有 居心 (Lü 26.3). Ant. 起 [156.3], 興 [134.9].

居  jū, *kaʔ/kah, vn, 1 ‘reside, dwell (in/ on)’, a itr: 邑于岐山之下居焉 (Meng 1B15); 居於陋巷 (Meng 4B29); b tr: 男女 居室,人之大倫也 (Meng 5A2); 君子所 居而安者,易之序也 (Yi 7); ‘at­tend’: 居 君之母與妻之喪 (Li 3); ‘settle for, choose’: 兩者大王何居焉 (ZGC 14.17); with temp expressions, ‘take up (some time); after (some time)’: 居有間,平公 又問祁黃羊曰 (Lü 1.5); 居數月,其馬將 胡駿馬而歸 (HNan 18); 2 nc, ‘residence,

*dok, vn, ‘be attached, connected to’, itr: 眇乎小哉, 所以屬於人 也 (Zhuang 5); 不能屬於禮義,則歸之 庶人 (Xun 9); 其夫屬乎父道者,妻皆 母道也 (Li 16).

Classifier 46  山 0

山  shān, *srân, n , ‘mountain’: 積土成 c

山, 風雨興焉 (Xun 1); 知者樂水,仁 者樂山 (LY 6.23). Syn. 丘 [1.4], 阜 [170.0].

4

岐  Qí, *ge, n , a mountain: 邑于岐山之 p

下居焉 (Meng 1B15).

443

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

8

2

崩  bēng, *pə̂ŋ, v , ‘collapse; pass away, 左  zuǒ, *tsâiʔ, n i

die’, esp. a ruler: 天子死曰崩 (Li 2); a itr: 且以文王之德,百年而後崩 (Meng 2A1); 平王崩 (Zuo 1.3); b mod: 初為霖雨 之操,更造崩山之音 (Lie 5). Cf. note at 死 [78.2].

18

巍巍  wéiwéi, *ŋui-­ŋui, v , 1 ‘be towerst

ing, lofty; majestic, magnif­icent’, itr: 巍巍 乎,其有成功也 (LY 8.19); 高皇帝受命 平暴亂,功德巍巍 (YTL 52); 2 adv: 泰 山之容,巍巍然高 (HNZ 16).

19

巖  yán, *ŋrâm, v , ‘be high, lofty; prei

cipitous, steep, dangerous’, a itr: 山陵岑 巖,淵泉閎流 (Guan 11); b mod: 是故知 命者,不立乎巖牆之下 (Meng 7A2); 制,巖邑也,虢叔死焉 (Zuo 1.1).

loc, 1 ‘left (side)’: 君子居 則貴左,用兵則貴右 . . . 吉事尚左,兇 事尚右,偏將軍居左,上將軍居右 (Lao 31); 男子行乎塗右,女子行乎塗 左 (Lü 16.5); 2 vi: ‘be/turn to the left’, a itr: 客入門而左 (Li 1); b caus: 左之不左,右 之不右 (HFei 34). Ant. 右 [30.2].

左右  zuǒyòu,

*tsâiʔ- wəʔ/h, nc, ‘entourage, ret­inue’: 齊王方大飲左右 (Lü 20.6).

巨/鉅  jù, *gaʔ, vst, 1 ‘be large, great; mighty, powerful’, a itr: 大國之樹必巨 (SY 12); 宜鉅者鉅,宜小者小 (SJ 23); b comp: 齊臨菑十萬戶 . . . 巨於長安 (SJ 52); c mod: 於是使力士舉巨囊 (GYang 12.6); 2 adv: 國者,巨用之則大,小用之則小 (Xun 11). Syn. 大 [37.0]; ant. 小 [42.0]; dist. 臣 [131.0]. 巧  qiǎo, *khrûʔ/h, v , 1 ‘be skilled, artn

至,百川灌河 (Zhuang 17).

ful, crafty, cunning’, a itr: 昔舜巧於使 民,而造父巧於使馬 (Xun 31); 古之民 樸以厚,今之民巧以偽 (SJS 7); b tr: 多 詐則巧法令 (Lü 26.3); c mod (!): 巧言令 色,鮮矣仁 (LY 1.3); 2 nc, ‘skill; skilfulness, clev­er­ness’: 大巧若拙 (Zhuang 10); 若夫規矩勾繩,巧之具也 (Wen 6). Syn. 智 [72.8]; ant. 拙 [64.5].

8

4

巢  cháo,

*dzrâu, nc, ‘nest’: 有聖人 作,搆木為巢 (HFei 49).

*ma, nc, ‘shaman’: 若國大 旱,則帥巫而舞雩,國有大災,則帥 巫而造巫恒 (ZL 49).

Classifier 48  工

Classifier 49  已

0

0

Classifier 47  川 0

川  chuān, *k-­hlun, n , ‘river’: 秋水時 c

巫  wū,

工  gōng, *kôŋ, n , 1 ‘artisan, craftsman’: 己  jı̌, *kəʔ, pr c

所謂四民者,士農工商也 (SY 7); 2 vi, ‘be skilled, skil­ful’, a itr: 羿工乎中微而拙 於使人無己譽,聖人工乎天而拙乎人 (Zhuang 23); 昔者舜工於使人,造父工 於使馬 (HSWZ 2); b pseudo-­tr: 工文學 者非所用 (HFei 49).

refl, ‘my-/your-/him-/her-/ itself ’, a head: 克己復禮為仁 (LY 12.1); as O, with neg: 不患人之不己知 (LY 1.16); 羿工乎中微而拙於使人無己譽 (Zhuang 23); b mod: 己所不欲,勿施於 人 (LY 12.2); 大叔又收貳以為己邑 (Zuo 1.1); c in O clause: 不患人之不己知 (LY

444

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

1.16); cf. 12.3–㉜. Syn. 自 [132.0]; dist. 巳 [49.0], 已 [49.0].

己  jı̌, *kəʔ, n

, sixth of the ‘Heavenly stems’: 夏四月己巳,晉人使陽處父盟 公以耻之 (Zuo 6.2). Cf. Focus 8; dist. 巳 [49.0], 已 [49.0]. temp

巳  sì, *s-­ləʔ, n

temp, sixth of the ‘Earthly branches’: 夏,四月丁巳,王入于王城 (XX 9.2). Cf. Focus 8; dist. 己 [49.0], 已 [49.0].

已  yı̌, *ləʔ, v , 1 ‘finish, stop, end’, a itr: 非 i

擇而取之,不得已也 (Meng 1B21); cf. 而已 [126.0]; b caus: 三已之, 無慍色 (LY 5.19); 已此三者, 然後刑可即也 (Xun 28); 2 adv, ‘already’: 門已閉矣 (Zuo 12.15); cf. 10.4; dist. 巳 [49.0]; 己 [49.0].

已  yı̌, *ləʔ, var. of 矣 [111.2]. 9

巷  xiàng, *grôŋh, n , ‘alley, lane’: 顏子 c

當亂世,居於陋巷 (Meng 4B29); 居期 而生子, 以為不祥, 棄之隘巷 (SJ 4).

6

帝  dì, *têh, nc, ‘god, ancestor; thearch’: 天帝令我長百獸 (XX 2.4). Cf. 上帝 [1.2], 皇帝 [106.4], 黃帝 [201.0].

帥/率  shuài, *srut(s), v , ‘command, n

lead’, tr: 命子封帥車二百乘以伐京 (Zuo 1.1); 天子作師,公帥之,以征不德 (GY 5.2); 越王句踐率其眾以朝吳 (SJ 31).

帥/率  shuài, *sruts, nc, ‘commander, leader, officer’: 三軍可奪帥也,匹夫不 可奪志也 (LY 9.26); 將率不能則兵弱 (Xun10). Syn. 將 [41.8]; dist. 師 [50.7]. 7

師  shī, *sri, n , 1 ‘troops’: 師行而糧食 c

(Meng 1B4); ‘teacher’: 禮者,所以正身 也,師者,所以正禮也 (Xun 2); 溫故 而知新,可以為師矣 (LY 2.11); a title, ‘music master’: 師摯之始,關雎之 亂,洋洋乎盈耳哉 (LY 8.15); 2 vn, ‘take as teach­er’, tr: 吾即沒,若必師之 (SJ 47). Dist. 帥 [50.6].

8

常  cháng, *daŋ, v , 1 ‘be constant, per-

Classifier 50  巾

i

2

布  bù, *pâh, n , ‘cloth, linen’: 賜母金布 (LNZ 6.2).

c

市  shì, *dəʔ, nc, 1 ‘market’: 日中為市, 致天下之民, 聚天下之貨, 交易而 退, 各得其所 (Yi 8); 爭名者於朝,爭 利者於市 (ZGC 3.7); 五十里有市 (ZL 25); 2 vn, ‘trade, sell’, a itr: 鄭商人弦高將 市於周 (Zuo 5.33); 良賈不為折閱不 市(Xun 2); b tr: 以其所有,易其所 無,市賤鬻貴 (GY 6.1); 天下必以王為 能市馬 (ZGC 29.11). 5

帛  bó, *brâk, n , ‘silk’: 子女玉帛,則君 c

有之 (Zuo 5.23); 五十者可以衣帛矣 (Meng 1A3); 書之竹帛,琢之槃盂,傳 以遺後世子孫 (Mo 10).

petual, un­changing; regular, ordinary’, a itr: 作服不常,以笑諸侯 (Yan 2.13); b caus: 五常得之以常其位 (HFei 20); c mod: 道可道,非常道 (Lao 1); 夫常人 安於故習 (SJS 1); 2 nc, ‘constancy; rule, regula­tion’: 言未始有常(Zhuang 2); 冠 帶有常 (Li 6).

帶  dài, *tâs, 1 n , ‘sash, belt’: 束帶立於 c

朝 (LY 5.8); 2 vn, ‘carry (at belt), take along’, only tr: 固不帶劍 (Guan 18).

一帶甲  dàijiǎ, *tâs-­krâp, nc, ‘man-­ at-­arms, soldier’: 帶甲百萬 (ZGC 14.3). 12

幣  bì, *bets, n , ‘silk; offering, gift; curc

rency’: 狄人侵之,事之以皮幣 (Meng 1B15); 以珠玉為上幣,以黃金為中幣, 以刀布為下幣 (Guan 73). Dist. 弊 [55.12].

445

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 51  干 2

平  píng, *breŋ, v , 1 ‘be level, balanced; i

peaceful’: 據當不傾謂之平,反平為險 (XS 8.3); a itr: 當堯之時,天下猶未平 (Meng 3A4); b caus, ‘pacify’: 武王已克殷 紂,平天下,封功臣昆弟 (SJ 35); ant. 險 [170.13]; 2 np, ‘Serene’, a posth. name: 晉平公問於祁黃羊曰 (Lü 1.5).

幼有序,朋友有信 (Meng 3A4). Ant. 老 [125.0], 長 [168.0]. Note: Accord­ing to the idealized descrip­tion in Li 1, a ten-­ year old is called 幼, a twenty-­year old弱, a thirty-­year old壯, a forty-­year old強, a fifty-­year old艾, a sixty-­year old耆, and a seventy-­year old老, an eighty- to ninety-­ year old 耄, and a centenarian 期.

9

幾  jī, *kəi, v , 1 ‘be/come close to, near; n

3

并/幷  bìng,

*peŋh, vtr, ‘combine; annex, occupy (territory)’, a tr: 果以弱燕 并彊齊 (SY 1); 招致賓客游士, 欲以并 天下 (SJ 6); b pass: 殷并於周,陳亡於 楚 (HSWZ 7).

年  nián, *nîn, n

, ‘harvest; year; age’: 五穀皆熟,為有年也 (GLiang 2.3); 樂 歲終身飽,凶年免於死亡 (Meng 1A7); 三年之喪何也 (Li 38); 父母之年,不可 不知也 (LY 4.21). Syn. 歲 [77.9]. temp

5

幸  xìng, *grêŋʔ, v , ‘favour, grace’, a tr: 襄 tr

公有賤妾,幸之 (SJ 37); 楚王召入幸 之,遂生子男 (SJ 78); b pass, ‘be favoured; lucky, fortunate’: 丘也幸,茍有過人必知 之 (LY 7.31); 不幸而有疾,不能造朝 (Meng 2B2); c mod, ‘favoured, favour­ite’: 善人在上,則國無幸民 (Zuo 7.16).

be im­minent’, a itr: 時日及矣,公子幾 矣 (GY 10.1); 求致社稷之福,必不幾矣 (HFei 49); 吾國幾於亡乎 (Xun 32); b tr: 漢之為漢幾四十歲矣 (XS 4.4); 2 adv, ‘almost’: 幾不免虎口哉 (Zhuang 29).

幾  jı̌, *kəiʔ, print, ‘how much/many?’: a head: 畏首畏尾,身其餘幾 (Zuo 6.17); 其人能靖者與,有幾 (Zuo 5.23); b mod: 子來幾日矣 (Meng 4A24).

幾何  jı̌hé, *kəiʔ-­gâi, print, ‘how much/ many/big/long?’: 人生幾何 (Zuo 9.31); 秦所亡民者幾何,民客之兵不得事本 者幾何,臣竊以為不可數矣(SJS 15); cf. 10.3–⑮. Note: Questions with 幾何 ask for abso­ lute numbers or measures, whereas 如何 asks for comparative ­measures. Classifier 53  广 4

10

xù, *s-­laʔ, n , 1 ‘order, sequence’: 父 幹  gàn, *kâns, n , ‘stem, framework’: 禮 序  子有親,君臣有義,夫婦有別,長幼 c

c

國之幹也,敬禮之輿也 (Zuo 5.11).

Classifier 52  幺 2

有序,朋友有信 (Meng 3A4); 春夏 先,秋冬後,四時之序也 (Zhuang 13); 2 vtr, ‘order, arrange’, a tr: 故序四時,裁 萬物 (Xun 9); 以歲時序其祭祀 (ZL 37); b pass: 論賢才之義,別所長之能,則 百官序矣 (CQFL 12); 離至不可得而序 (GYang 9.5). Syn. 次 [76.2].

幼  yòu, *ʔiuh, vst, 1 ‘be young; immature’, a itr: 幼而無父曰孤 (Meng 1B5); 王 孫滿尚幼 (Zuo 5.33); b caus, ‘treat as 5 young’: 幼吾幼,以及人之幼 (Meng 1A7); ster’: 府  fǔ, *poʔ, nc, ‘trove, archive’: 詩書義 2 nc, ‘younger one, junior, young­ 父子有親,君臣有義,夫婦有別,長 之府也 (Zuo 5.27). 446

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

庚  gēng,

*krâŋ, ntemp, seventh of the ‘Heavenly stems’: 庚戌天王崩 (SY 19). Cf. Focus 8.

庖  páo, *brû, nc, ‘kitchen; cook’: 庖有 肥肉 (Meng 1A4); 庖人調和而弗敢 食,故可以為庖 (Lü 1.5); 良庖歲更刀 (Zhuang 3).

康  kāng,

*khâŋ, vst, 1 ‘be peaceful, prosperous, content’, a itr: 繁文簡節之音 作,而民康樂 (Li 19); b caus: 康周 公,故以賜魯也 (Li 25); c comp: 治莫康 於立君 (SJS 7); 2 nc, ‘tranquility, peace’: 是謂小康 (Li 9); 3 np, ‘Tranquil’, posth. name: 恭王聞之,立康王為太子 (SY 3).

10

6

lián, *rem, n , ‘integrity, modesty’: 度  dù, *dâkh, n , 1 ‘measure, norm, rule; 廉  辭利刻謙謂之廉,反廉為貪 (XS 8.3); c

c

disci­pline’: 以人自觀謂之度,反度為妄 (XS 8.3); 守國之度,在飾四維 (Guan 1); 2 vi, ‘conform/comply with (a measure)’, itr: 今京不度,非制也 (Zuo 1.1).

度  duó,

*dâk, vn, ‘measure, calculate, plan’, a itr: 度,然後知長短 (Meng 1A6); 度於閎夭而謀於南宮 (GY 10.13); b tr: 度其功勞,論其慶賞 (Xun 9).

廉不蔽惡,恥不從枉 (Guan 1); 凡牧民 者,欲民之有廉也 (Guan 3); 今兄弟被 侵必攻者,廉也 (HFei 49).

廋  sōu, *sru, vn,‘hide’, only itr: 人焉廋哉 (LY 2.10); 日月廋而無溉於志 (HNan 14). 11

廄  jiù, *kuh, n , ‘stable’: 取戈劍者,其 c

不義又甚入人欄廄,取人馬牛 (Mo 17).

7

庭  tíng, *lêŋ n , ‘hall; courtyard’: 鯉趨 c

而過庭 (LY 16.13); 公庭不言婦女 (Li 2). Syn. 朝 [74.8]; cf. note at室 [40.6].

8

庶人  shùrén, *lha(k)h-­nin, n , ‘comc

mon people’: 自天子以至於庶人 (Li 42).

庶幾  shùjī, *lha(k)h-­kəi, adv, ‘may; perhaps’: 吾王庶幾無疾病與 (Meng 1B1); 王庶幾改之 (Meng 2B12). Cf. 7.2.

庸  yōng/yóng, *loŋ, vtr, ‘need; use, employ’, a tr: 王其無庸戰 (GY 19.1); 昭侯 顯列,申子庸之 (SJ 130); b aux, ‘need to’: 子婦未孝未敬,勿庸疾怨 (Li 12); 不如聽之以卒秦,勿庸稱也以為天下 (ZGC 5.10); c mod, ‘usual, regular; ordinary, mediocre’: 臣,君之庸臣也 (GY 6.1); 庸主安能以御一國之民(SJS 25); 人 有五儀,有庸人,有士,有君子,有 賢人,有大聖 (Xun 31). Cf. 用 [101.0]. 庸  yōng/yóng, *loŋ, adv, ‘how?’, in rhetorical ques­tions: 而謀召君者,庸非二 乎 (Zuo 3.14); 沮之而不可則致地,庸 遲乎 (SJ 47). Cf. 其 [12.6]; 豈 [151.3].

12

廢  fèi, *pats, v , ‘abandon, discard, cast tr

aside, dismiss’, a tr: 君子不以言舉人,不 以人廢言 (LY 15.23); b pass: 聲則非雅聲 者舉廢 (Xun 9); 鄉飲酒之禮廢,則長幼 之序失 (Li 26); c mod: 是故民無廢事,而 國無失利也 (Guan 73). Cf. 乏 [4.4].

廣  guǎng, *kwâŋʔ, vst, ‘be extensive, broad, wide’: a itr: 臺謝甚高,園囿甚廣 (Xun 11); b caus: 是以聖人不高山,不 廣河 (HNan 19). Ant. 狹 [94.7]. 廟  miào, *mrauh, n , ‘ancestral temple’: c

秋八月丁卯,大事于大廟 (Zuo 6.2).

13

廩  lı̌n, *rəmʔ, nc, ‘granary’: 守在倉廩 (Guan 1). Classifier 55  廾 12

弊  bì, *bets, nc, ‘weakness’: 屈力殫 貨,則諸侯乘其弊而起 (Sun 2). Dist. 幣 [50.12].

447

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 56  弋

弟  tì, *dîh, var. of 悌 [61.7]. 弟子  dìzı̌, *dîʔ-­tsəʔ, n , ‘disciple’: 弟

10

c

弒  shì, *lhəkh, v , ‘assassinate (a supetr

rior)’, a tr: 萬乘之國弒其君者,必千乘 之家;千乘之國弒其君者,必百乘之 家 (Meng 1A1); b pass: 田常徒用德而簡 公弒 (HFei 7). Cf. note at 殺 [79.6].

Classifier 57  弓 0

弓  gōng,

*kwəŋ, nc, ‘bow’: 今日我疾 作,不可以執弓 (Meng 4B24); 軍士引 弓將射之 (LNZ 5.6).

1

子事夫子有年矣 (HSWZ 8); 孔子以詩 書禮樂教,弟子蓋三千焉 (SJ 47). Note: 弟子, literally ‘younger broth­ers and sons’, is specifically used for the disciples of Confucius. Indeed, it appears that quite a few of the master’s followers were his relatives. Confu­cius less prominent disciples were called 門人 [169.0]: literally ‘those [waiting] at the gate’.

5

弢  tāo, *lhû ?, n ,‘bow case’: 弢無弓,服 c

無矢 (Guan 20); 伏弢而死 (Mo 31).

6

引  yı̌n, *linʔ, v , ‘pull; stretch, draw (a 弭/彌  mı̌, *meʔ, v , ‘stop, cease, end’, a tr

bow); attract, incur; lead on, encourage; withdraw’, a tr: 使烏獲疾引牛尾 (Lü 1.3); 天下之民皆引領而望之矣 (Meng 1A6); 軍士引弓將射之 (LNZ 5.6); 江東已 定,急引兵西擊秦 (SJ 7); b pass: 引於 外物,亂於玩好也 (HFei 20).

2

弗  fú, *pət, fus [< 不+之], 1 ‘not . . . him/

her/it/them’: 女弗能救與 (LY 3.6); 飢者 弗食,勞者弗息 (Meng 1B11); 愛之而弗 仁 (Meng 7A45); 2 adv,in late CC texts occasional­ly used like 不, without inherent O: 粥羔豚者弗飾賈 (SJ 47); 弗失之矣 (Li 31). Cf. 5.2.3.

3

弛  shı̌,

*lhaiʔ, vtr, ‘loosen, slacken, release; ease, lessen’, a tr: 復二文之業,弛 周室之憂 (Zuo 10.32); 弛罪廢法,而伏 其辜 (HSWZ 2); b pass: 弓張而不弛 (Mo 7); 及彌子色衰愛弛 (HFei 12); c mod: 張 弓尚筋,弛弓尚角 (Li 1). Ant. 張 [57.8].

4

弟  dì, *dîʔ, n , ‘younger brother’: 四海 c

之內皆兄弟也 (LY 12.5). Ant. 兄 [10.3], 昆 [72.4].

i

itr: 暴亂之姦彌矣 (XS 1.2); b caus: 吾能 弭謗矣 (GY 1.3).

7

弱  ruò, *niauk, vst, ‘be weak; young’, a itr: 果以弱燕并彊齊 (SY 1); b caus: 虛其 心,實其腹,弱其志,強其骨 (Lao 3); 弱天下之民 (XS 1.1)); c comp: 兼人而兵 俞弱 (Xun 15); 兵莫弱是矣 (Xun 15). Syn. 柔 [75.5], 懦 [61.14]; ant. 強 [57.8]; cf. note at 幼 [52.2]. 8

張  zhāng,

*traŋ, vtr, ‘stretch, extend; prevail, un­fold’, a tr: 我張吾三軍而被吾 甲兵 (Zuo 2.6); b pass: 四維張則君令行 (Guan 1); c mod: 張弓尚筋,弛弓尚角 (Li 1). Cf. 長 [168.0]; ant. 弛 [57.3].

強/彊/强  qiáng, *gaŋ, vst, ‘be strong’, a itr: 輔周則國必強 (Sun 3); b caus: 故惟明君知好言之不可以強兵闢 土也 (SJS 3); 虛其心,實其腹,弱其 志,強其骨 (Lao 3); c comp: 晉國,天 下莫強焉 (Meng 1A5); 兼人而兵俞強 (Xun 15); d mod: 果以弱燕并彊齊 (SY 1). Ant. 弱 [57.7], 懦 [61.14], 柔 [75.5]; cf. note at 幼 [52.2].

448

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

1); c comp: 伯夷叔齊餓于首陽,而志益 彰 (HSWZ 2); 彰人之善而美人之功 (Li 32). Ant. 隱 [170.14].

9

强  qiáng, *gaŋ, var. of 強 [58.8]. 12

彈  tán, *dân, v , ‘catapult, fling, hurl, n

shoot’, tr: 從臺上彈人 (Zuo 7.2).

Classifier 60  彳 4

役  yì, *wek, nc, ‘conscript’: 善用兵 者,役不再籍 (Sun 2).

13

彊  qiáng, *gaŋ, var. of 強 [57.8].

5

14

彌  mí, *me, adv, ‘ increasingly, more’, expres­sing the comparative of vst: 仰之彌 高,鑽之彌堅 (LY 9.11); 其虧彌甚者 也,其尊彌薄 (Lü 2.2). Cf. Box 24.

Classifier 59  彡 4

形/刑  xíng, *gêŋ, nc, 1 ‘form, shape’: 物成生理謂之形 (Zhuang 12); 形色,天 性也 (Meng 7A38); 在天成象,在地成 形 (Li 19); 形者,生之具也 (SJ 130); 2 vi, ‘form, take shape, ap­pear’, a itr: 冬與夏不 能兩刑 (Lü 24.5); 勝敗存亡之機固已形 於胸中矣 (SJ 69); b caus: 長短相形 (Lao 2); 道生之,德畜之,物形之,勢成之 (Lao 51); 著有過,刑仁講讓 (Li 9). 8

彫  diāo, *tiû, vtr, ‘engrave, carve, decorate’, a tr: 厚斂以彫牆 (Zuo 7.2); b pass: 丹漆不文,白玉不彫,寶珠不飾 (SY 20); c mod: 天子彫弓,諸侯彤弓,大 夫黑弓,禮也 (Xun 27). 11

彼  bı̌, *paiʔ, prdem [< 夫], ‘that, those’, a head: 此與彼同類 (Mo 45); referring to people: 彼君之讎也,天或者將棄彼矣 (Zuo 9.27); 知彼知己,百戰不殆 (Sun 3); b mod: 彼國安而天下安 (Xun 20); subordinated by 其: 彼其道遠而險 (Zhuang 20); 彼其髪短而心長 (Zuo 10.3). Cf. note at 此 [77.2].

往  wǎng, *waŋʔ, vn, ‘go (to)’, a itr: 寒往 則暑來, 暑往則寒來 (Yi 8); 不與陰陽 俱往來 (CQFL 16); b tr: 宋使人往之 (SY 12); c mod (!), ‘gone, past’: 往言不可及也 (GY 8.1); 享往古之勛 (ZGC 19.4); 往日 先生至燕 (SJ 69); ant. 來 [9.6]. 征  zhēng,

*teŋ, vn, ‘take, seize; levy (taxes); at­tack’: a tr: 上下交征利而國危 矣 (Meng 1A1); 今燕虐其民,王往而征 之 (Meng 1B11); b itr: 關市譏而不征 (Meng 1B5); 民務勝而力征 (SJS 7). Note: 征, related to 正 [77.1] and 政 [66.4], means a punitive, ‘corrective’ military campagin by the legitimate ruler, where­as 伐 [9.4] is a more gen­eral term for an attack, 侵 [9.7] designates a ma­levolent raid, and 襲 [145.16] a surprise attack.

6

彰/章/璋 zhāng, *taŋ, vst, ‘be clear, obvious, dis­tinct; illus­trious’, a itr: 順風而呼,聲非加疾 也,而聞者彰 (Xun 1); 黔首已親矣,名 號已章矣 (Lü 17.4); b caus: 尊賢育 才,以彰有德 (Meng 6B7); 文巧不禁 則民乃淫,不璋兩原則刑乃繁 (Guan

待  dài, *də̂ʔ, vn, 1 ‘wait (for)’, a tr: 衛君 待子而為政,子將奚先 (LY 13.3); ‘need, depend on’: 伯成子高不待問而知之 (Lü 20.2); 國待農戰而安,主待農戰而 尊 (SJS 3); ‘wait on, treat as a guest’: 相待 如賓 (Zuo 5.33); ‘treat, face, deal with’: 諸 侯多謀伐寡人者,何以待之 (Meng

449

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

1B11); 上棟下宇,以待風雨 (Yi 8); b itr: 左右屏而待 (Li 1); 魯復善待,由是反 魯 (SJ 47).

後/后  hòu, *ɦôʔ, n

, 1 ‘(the) back, rear, back-­ side; later; consequence’, a head: 子隨我後 (ZGC 14.3); 雖有智 者,不能善其後矣 (Sun 2); b mod: 百 王之道,後王是也 (Xun 3); 2 nc, ‘successor’: 所愛美女欲以其子為後 (HFei er; behind, at the 31); 3 vi, ‘be last, lat­ back’, a itr: 三子者出,曾皙後 (LY 11.24); 法古則後於時 (SJS 7); b caus, ‘defer, give last prior­ity’: 苟為後義而先 利,不奪不饜 (Meng 1A1). Ant. 先 [10.4], 前 [18.7]. loc

7

徒  tú, *dâ, nc, ‘footsoldier’: 休卒徒 (Zhuang 29); ‘follower, disciple’: 仲尼之 徒,無道桓文之事者 (Meng 1A7); ‘people like . . .’: 鄧析惠施之徒也 (HSWZ 4); ‘corvée labourer’: 被刑之徒 (SY 5). Dist. 徙 [60.8]. 徒  tú,

*dâ, adv, ‘only, just’: 王如用 予,則豈徒齊民安,天下之民舉安 (Meng 2B12); 強秦之所以不敢加兵于趙 者,徒以吾兩人在也 (SJ 81); often negated: 非徒無益,而又害之 (Meng 2A2); 非徒無生也,而本無形 (Zhuang 18); emphatic, ‘sim­ply’: 夫子徒無所不知 (Xun 29). Dist. 徙 [60.8].

徐  xú, *s-­la, v , ‘be slow, leisurely’, a itr: i

斲輪,徐則甘而不固 (Zhuang 13); b caus: 臣請受邊城,徐其公而留其日 (ZGC 32.4). Ant. 疾 [104.5]; dist. 除 [170.7].

8

從  cóng,

*dzoŋ, vn, 1 ‘follow, comply with, agree to’, a itr: 祝先,子從,宰宗 人從 (Li 7); 年少則從乎父母,出嫁則 從乎夫,夫死則從乎子,禮也 (LNZ 1.11); b tr: 不從其所令,從其所行 (Li 33); 聖人復起,必從吾言矣 (Meng 2A2); 遂從諸死鳥 (Zuo 10.20); c pseudo-

t­ r2: 魯人從君戰 (HFei 49); 2 prep, ‘from’: 從臺上彈人 (Zuo 7.2).

一  從政,

‘conduct the government, govern’: 仲由可使從政也與 (LY 6.8).

從  zòng, *dzoŋh, nc [< preceding], ‘follower’: 田車數百乘,從數千,人滿野 (Mo 31).

從容  cóngróng/cōngróng, *dzoŋ-­loŋ, vi, 1 ‘be carefree, untroubled, lighthearted’, itr: 儵魚出遊從容 (Zhuang 17); 2 adv: 從 容中道,聖人也 (Li 31); 坐談話,從容 言至於士 (SY 11). 得  dé, *tə̂k, v , 1 ‘find; get, obtain’, a tr: n

宋人或得玉 (Zuo 9.15); 為善者得 道,為惡者失道 (SY 16); b itr: 是故得乎 丘民而為天子,得乎天子為諸侯,得 乎諸侯為大夫 (Meng 7B14); 單于 得,欲殺之 (SJ 110); c pseudo-­tr2: 得之 分義也 (Xun 9); c aux, ‘get to’: 王之所大 欲可得聞與 (Meng 1A7); cf. 4.4.1. Ant. 失 [37.2].

徙  xı̌, *seʔ, v , ‘move away/on, migrate’, i

a itr: 使民重死而不遠徙 (Lao 80); 南徙 於楚 (SY 4); b caus: 徙其兵以臨晉,晉 必聽命矣 (Lü 23.3). Syn. 遷 [162.12]; dist. 徒 [60.7].

御  yù, *ŋah, vn, ‘steer (a vehicle); direct, control’: 國之善射御之士 (Mo 8); 將能 而君不御者勝 (Sun 3). 9

復  fù, *buk, vi, ‘return’, a itr: 昭王之不 復,君其問諸水濱 (Zuo 5.4); 有生則復 於不生,有形則復於无形 (Lie 1); b caus: 公父定叔出奔衛,三年而復之 (Zuo 3.16); c pseudo-­tr: 使之行商容而復其位 (Li 19); 吳師既退,昭王復國 (XX 7.11).

復  fù, *bukh [< preceding], adv, ‘again, once more, anew’: 使民復結繩而用之 (Lao 80); 久矣吾不復夢見周公 (LY 7.5).

循  xún, *s-­lun, vn, ‘follow; abide by, stick to, accord with’: a tr: 循海而歸,其 可也 (Zuo 5.4); 請循其本 (Zhuang 17); b

450

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

itr: 上不循於亂世之君 (Xun 23); 循於道 之謂備 (Zhuang 12).

徧  biàn, *pêns, var. of 遍 [162.9]. 10

微  wēi, *məi, conj, ‘if not, unless; if it were not for, without’: 微君言,臣故將 謁之 (HFei 36); 微禹,吾其魚乎 (Zuo 10.1); 微二子者,楚不國矣 (Zuo 12.16). Cf. 15.4.1. 微  wéi, *məi, vst, ‘be imperceptible, hidden, in­conspicuous, subtle’, a itr: 世衰道 微,邪說暴行有作 (Meng 3B9); b caus: 白公奔山而縊,其徒微之 (Zuo 12.16); c comp: 精微乎毫毛 (Xun 26); d mod: 季桓 子微服往觀再三 (SJ 47). Syn. 妙 [38.4]. 12

德  dé, *tə̂k, n , ‘virtue, charisma; power; c

capabil­ity, quality’: 施行得理謂之 德,反德為怨 (XS 8.3); 德蕩乎名,知 出乎爭 (Zhuang 4).

徵  zhēng, *trəŋ, vtr, 1 ‘demand; levy, summon; seek; inquire; verify’, a tr: 齊徵 諸侯而戍周 (Zuo 5.16); 文獻不足故 也,足則吾能徵之矣 (LY 3.9); b pass: 徵於色,發於聲,而後喻 (Meng 6B15); 久則徵,徵則悠遠 (Li 31); 叔孫通 . . . 以 文學徵 (SJ 99); 2 nc, ‘evidence, manifestation, symptom’: 歌者樂之徵也, 哭者 悲之效也 (HNZ 19); 亡徵者,非曰必 亡,言其可亡也 (HFei 15). 13

徼  jiào, *kiâuh, nc, ‘boundary, margin’: 常有,欲以觀其徼 (Lao 1). Classifier 61  心

1

必  bì, *pit, vi, 1 ‘be inevitable, certain, necessary’, a itr: 其措兵於寡人必矣 (HFei 10); b caus, ‘in­sist on, force’: 無參 驗而必之者,愚也 (HFei 50); 成不可必 也 (Xun 15); 2 adv, a in vb clause: 何必曰 利 (Meng 1A1); 人與人不相愛則必相賊 (Mo 15); b in nom clause: 亡鄧國者,必 此人也 (Zuo 3.6); 必非誠士 (Xun 5); cf. 3.3.2. 3

忍  rěn, *nənʔ, vn, ‘[be indifferent (to) >] endure; be hard-­hearted’, a tr: 天子不忍 小忿 (Zuo 5.24); 不忍人,必自忍也 (GY 7.6); b itr: 臣固知王之不忍也 (Meng 1A7); c aux: 余不女忍殺 (Zuo 10.1).

忘  wàng, *maŋ, v , ‘forget’, a tr: 余聞而 n

弗忘 (Zuo 10.3); 回忘禮樂矣 (Zhuang 6); b itr: 回坐忘矣 (Zhuang 6); 不可以須 臾忘於法 (SJS 20).

志  zhì, *təh, n , 1 ‘will, purpose, aim, c

in­tention, ambition’: 願聞子之志 (LY 5.27); 志士仁人,無求生以害仁 (LY 15.9); syn. 心 [61.0]; 2 vn, ‘be deter­mined, dedicat­ed, intent on’, a itr: 吾十有五而 志於學 (LY 2.4); 苟不志於仁,終身憂 辱 (Meng 4A9); b tr: 有法而無志其義 (Xun 2).

志  zhì, *təh, var. of 識 [149.12]. 4

0

心  xīn, *səm, n , ‘heart;

遠志,無有居心 (Lü 26.3). Syn. 志 [61.3]. Note: In an­cient China, the heart was considered the organ of thought, feeling, desire, and under­standing; hence ‘mind’ is often an appropriate translation.

mind, sense, c rationality; attitude’: 心主身之血脈 (HDNJ 44); 是宜為君,有恤民之心 (Zuo 3.11); 人皆有不忍人之心 (Meng 2A6); 心知道,然後可道 (Xun 21); 皆有

忽  hū, *hmə̂t, adv, ‘suddenly’: 龍忽不 見 (LNZ 7.3); 忽然出,勃然動 (Zhuang 12); 人生天地之間,若白駒之過郤,忽 然而已 (Zhuang 22).

忠  zhōng, *truŋ, n , 1 ‘loyalty, integrity, c

faithful­ness’: 中能應外,忠也 (GY 1.14);

451

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

愛利出中謂之忠,反忠為倍 (XS 8.3); 2 vst, ‘be loyal, faithful’, a itr: 為人謀而不忠 乎 (LY 1.4); b pseudo-­tr: 子胥忠其君,天 下願以為臣 (ZGC 7.8); 為人臣而忠其 君者 (SY 9); c comp: 比干且死,而諫愈 忠 (HSWZ 2); d mod: 是先王之道,忠 臣孝子之極也 (Xun 19). Ant. 姦 [38.6].

忤/牾/寤  wǔ/wù, *ŋâh, v , 1 ‘go/be n

against, oppose, resist’, a itr: 灌將軍得罪 丞相,與太后家忤 (SJ 107); b tr: 居下位 而上忤其君 (SY 17); 2 adv, ‘adversely’: 莊 公寤生,驚姜氏 (Zuo 1.1).

5

怙  hù, *gâʔ, v , ‘rely/presume on’, only n

tr: 無怙富 (Zuo 11.4). Syn. 恃 [61.6].

怠  dài, *lə̂ʔ, vi, ‘be slack, negligent, indifferent; impolite’, a itr: 得時無怠,時 不再來 (GY 21.6); 往觀終日,怠於政事 (SJ 47); b caus, ‘neglect, dis­regard’: 父母 舅姑之命,勿逆勿怠 (Li 12). 怫然  fúrán,

*bət-­nan, adv, ‘furiously, fuming (with anger)’: 公怫然怒 (HFei 34).

急  jí, *kəp, vst, 1 ‘be urgent; hurried, quick’, a itr: 江東已定,急引兵西擊秦 (SJ 7); b caus: 且人所急無如其身 (HFei 32); 夫珠玉, 人主之所急也 (HFei 13); c comp: 謂之衣食孰急於人 (HFei 43); 凡治人之道,莫急於禮 (Li 25); 2 nc, ‘hurry, urgency’: 魏伐趙,趙告急齊 (SJ 44). 怒  nù, *nâh, vi, 1 ‘be angry, furious’, a itr: 公怒,乃止 (Zuo 1.5); b caus: 樂之勿 苦,喜之勿怒 (SY 7); 伐無罪之國,以 怒明神 (Yan 1.22); syn. 慍 [61.10]; 2 nc, ‘anger’, 喜怒哀懼愛惡欲七者,弗學而 能 (Li 9). Ant. 喜 [30.9]; dist. 恕 [61.6].

怒  nù, *nâh, vtr, ‘arouse, stimulate’, a tr: 父母怒之弗為改 (HFei 49); 弗許而後 戰,所以怒我而怠寇也 (Zuo 2.8); b pass: 怒而飛,其翼若垂天之雲 (Zhuang 1). Dist. 恕 [61.6].

思  sī, *sə, vn, ‘think (of), contemplate; ponder’, a itr: 學而不思則罔,思而不學

則殆 (LY 2.15); 終夜不竊,以思 (LY 15.31); b tr: 君子有九思,視思明,聽思 聰,色思溫,貌思恭,言思忠,事思 敬,疑思問,忿思難,見得思義 (LY 16.10); 家貧則思良妻, 國亂則思良相 (SJ 44). Syn. 慮 [61.11].

性  xìng, *seŋh, nc, ‘(human) nature, natural con­stitution; character, essence’: 生之謂性 (Meng 6A3); 性者,生之質也 (Zhuang 23); 性者,本始材樸也 (Xun 19); 性者,天質之樸也 (CQFL 36); 人性 之善也,猶水之就下也 (Meng 6A2). Cf. 18.4–⑲.

怨  yuàn, *ʔons, nc, 1 ‘anger; hatred, resentment’: 以德報怨 . . . 何以報德 (LY 14.34); 2 vn, ‘hate; ex­press resentment’, a tr: 鄭伯怨王 (Zuo 1.3); 民怨其上,不遂 亡者,未之有也 (SY 3); b itr: 詩 . . . 可以 怨 (LY 17.9); 樂不樂者, 其民必怨 (Lü 5.3); c mod (!): 國無怨民曰強國 (SJS 4).

怪  guài, *kwrə̂h, vst, ‘be strange, odd, anomalous’, a mod: 見怪物,皆曰神 (Li 23); 凡有怪徵者,必有怪行 (Zhuang 24); b caus: 怪之,可也,而畏之,非 也 (Xun 17); 吾固怪吾主之數逆天 道,至今無禍 (SY 15). Syn. 姦 [38.6]. 6

恥/耻  chı̌, *nhrəʔ, nc, 1 ‘sense of shame’: 道之以德,齊之以禮,有恥且 格 (LY 2.3); 禮不踰節,義不自進,廉 不蔽惡,恥不從枉 (Guan 1); 2 vn, ‘be ashamed (of)’, tr: 寡人恥之 (Meng 1A5).

恭  gōng, *kroŋ, n , 1 ‘humility; reverc

ence, re­spect’: 不忘恭敬,民之主也 (Zuo 7.2); 接遇慎容謂之恭,反恭為媟 (XS 8.3); 2 vn, ‘be humble, reverent, respectful’, a tr: 敬宗廟,恭祖舊 (Guan 1); b itr: 君之揖朝也恭 (Lü 18.3).

恆  héng, *gə̂ŋ, v , 1 ‘be constant, regui

lar’: a itr: 其不言有,不使麟不恆於中 國也 (GLiang 12.14); b caus: 而欲恆其 生,盡其終,惑於數也 (Lie 1); c mod: 有恆產者有恆心,無恆產者無恆心 (Meng 3A3); 夏,恆星不見,夜明也

452

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Zuo 3.7); 2 adv: 愛人者人恆愛之,敬 人者人恆敬之 (Meng 4B28).

恐  kǒng, *khoŋʔ, vn, ‘fear, be afraid of ’, a itr: 齊王聞之,君臣恐懼 (ZGC 11.1); b tr, typically with O clause: 吾恐其為天下 笑 (Zhuang 24); 恐天下之議我也 (SJS 1). Syn. 懼 [61.18]. 恃  shì, *dəʔ, v , ‘trust in, rely on’, a tr: 何 n

恃而不恐 (Zuo 5.26); 忠則人與之,信則 人恃之 (SY 10); b itr: 恃於民而不輕,因 於物而不去 (Zhuang 11). Syn. 怙 [61.6].

恕  shù, *nhah, n , 1 ‘empathy, tolerance’: c

以己量人謂之恕,反恕為荒 (XS 8.3); 2 vn, ‘feel empathy, be tolerant’, a itr: 所藏乎 身不恕,而能喻諸人者,未之有也 (Li 42); 且恕於人,順於天 (CQFL 80); b tr: 下為非則矜而恕之 (XS 9.1). Dist. 怒 [61.5].

息  xí, *sək, vi, 1 ‘rest, take a break’, a itr: 楊墨之道不息,孔子之道不著 (Meng 3B9); b caus, ‘let rest’: 飢者弗食,勞者弗 息 (Meng 1B11); ant. 勞 [19.10]; 2 ‘ex­hale’, itr: 喟然太息 (Lü 21.3). 恤  xù,

*swit, vn, ‘worry about; look after, take care of ’, a tr: 是宜為君,有恤 民之心 (Zuo 3.11); 賞死事,恤孤寡 (Li 6); b itr: 愚者反此,不能法天而恤於人 (Zhuang 31).

7

悖  bèi, *bûs, vi, ‘be adverse, contradictory, ab­surd; contrary, rebellious’, a itr: 是 以事行而不悖 (GY 1.3); 毋悖于時,毋 或作為淫巧以蕩上心 (Li 6); b caus: 萬 物并育而不相害,道并行而不相悖 (Li 31). Syn. 逆 [162.6]. 患  huàn, *grôns, v , 1 ‘fear, worry, be tr

anxious about’, a tr: 寡人患之,吾將奈 何 (HFei 10); in rhe­torical questions: 子何 患焉 (HSWZ 9); 二三子何患乎無君 (Meng 1B15); typi­cally with O clause: 不患 人之不已知,患不知人也 (LY 1.16); 先 王患禮之不達於下也 (Li 9); ‘loathe, take offense’: 宣子驟諫,公患之 (Zuo

7.2); b pass: 故君之所以患於軍者三 (Sun 3); 2 nc, ‘anger, trouble, an­noyance, disap­pointment’: 此人不死,必為吳患 (SJ 31); syn. 憂 [61.11].

悔  huı̌, *hmə̂h, v , 1 ‘regret, repent’, a tr: n

君必悔之 (Zuo 5.5); 太甲悔過 (Meng 5A6); 寡人雖死,亦無悔焉 (Zuo 1.3); b itr: 雖悔無及矣 (SY 3); 2 nc, ‘remorse’: 悔 吝者,懮虞之象也 (Yi 7).

悌/弟  tì, *dîh, n , 1 ‘fraternal love’: 孝 c

弟也者,其為仁之本與 (LY 1.2); 不恭 祖舊則孝悌不備 (Guan 1); 弟敬愛兄謂 之悌,反悌為敖 (XS 8.3); 2 vst, ‘show fraternal love’, a itr: 入則孝,出則悌 (Meng  3B4); 為人弟必悌 (Mo 16); b caus: 父不得子其子,兄不得弟其弟 (Guan 80).

悟/寤  wù, *ŋâh, v , ‘awake;

realize, n become aware of; understand’, a tr: 孝文 皇帝悟兵之不可宿也 (XX 10.13); b itr: 秦王懼然而悟 (XX 3.3); 不明鬼神則陋 民不悟 (Guan 1).

悅/說  yuè, *lot, v , ‘be pleased, con­ i

tent; pleas­ing’, a itr: 若時雨降,民大悅 (Meng 3B5); 孔子不悅於魯衛 (Meng 5A8); 學而時習之,不亦說乎 (LY 1.1); b caus: 民致行己以說其上矣 (Li 33); 夫欲 得之君,不可說以利 (ZGC 31.2); c pseudo-­tr, ‘take pleasure in’: 不知說 生,不知惡死 (Zhuang 6); 王后悅其 狀,高其知 (ZGC 7.5).

悉  xī, *sit, vtr, 1 ‘exhaust, employ fully’, a tr: 悉其聰明,致其忠愛以盡之 (Li 5); 悉國中武力以伐齊 (SY 9); b pass: 以其 用民力為已悉矣 (GLiang 3.29); 2 adv, ‘all, completely’, a referring to the S: 悉來 合券 (ZGC 11.1); 魯下令丁男悉發,五 尺童子皆至 (SY 13); b referring to the O: 臣願悉言所聞 (HFei 1); 不穀國雖 小,已悉發之矣 (SJ 45); cf. Box 30. 8

惡  è, *ʔâk, n , 1 ‘evil, vice’: 為善者得 c

道,為惡者失道 (SY 16); 2 vst, ‘be bad,

453

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

evil; ugly’: a itr: 色惡,不食,臭惡,不 食 (LY 10.6); b comp: 甲惡於此 . . . 衣無 惡於甲矣 (XX 5.10); c mod: 士志於 道,而恥惡衣惡食者,未足與議也 (LY 4.9). Ant. 善 [30.9], 美 [123.3].

sorrow for’: 臣等不惜臣之身,惜陛下 國之亡耳 (SY 20).

*ʔâkh, vn, [< preceding] ‘hate, despise, loathe’, a tr: 惟仁者能好人,能 惡人 (LY 4.3); 其所惡者,吾將改之 (XX 4.16); b itr: 衛侯始惡於公叔戍 (Zuo 11.13); 愛親者,不敢惡於人 (XJ 2); c aux: 彼唯人言之惡聞 (Zhuang 18). Syn. 嫉 [38.10], 憎 [61.12]; ant. 好 [38.3]; 欲 [76.7].

with; treat with care; support; love’: 不能 分人,又不能自養,謂之愛 (Yan 4.23); a tr: 愛人者必見愛也 (Mo 16); 我何愛一 牛 (Meng 1A7); b pass: 今夫人事太 子,甚愛而無子 (SJ 85); 彌子瑕愛於衛 君 (SY 17); c comp (!): 令貴於寶 . . . 法愛 於人 (Guan 6); d mod: 故善為國者遇 民,如父母之愛子,兄之愛弟 (SY 7); syn. 吝 [30.4]; 2 nc, ‘love, care; solidarity’: 故父母之愛不足以教子 (HFei 49). Ant. 惡 [61.8], 憎 [61.12].

惡  wù,

惡  wū, *ʔâ, print, ‘where? how?’, a adv: 彼 惡知之 (Meng 1A7); 棄父之命,惡用子 矣 (Zuo 2.16); 視父兄與君若其身,惡施 不孝 (Mo 14); b in PrP: 學惡乎始,惡乎 終 (Xun 1); 天下惡乎定 (Meng 1A6).

惠  huì, *wîs, n , 1 ‘kindness, benevolence, c

fa­vor’: 分人以財謂之惠 (Meng 3A4); 君 子懷刑,小人懷惠 (LY 4.11); 心省恤人 謂之惠,反惠為困 (XS 8.3); 2 vi, ‘be thoughtful, kind, be­nevolent’, a itr: 惠而不 知為政 (Meng 4B4); 不惠於其父乎 (GY 7.8); b caus: 天下之人,皆歸其德而惠其 義 (Guan 57); c pseudo-­tr: 庶人駭政,則 莫若惠之 (Xun 9); 夫豈惠其民 (GY 7.8); d mod: 惠人也 (LY 14.9); 3 np, 莊子與惠子 遊於濠梁之上 (Zhuang 17); as posth. name, ‘Benevolent’: 秦惠王曰 (Lü 1.5).

惑  huò, *wə̂k, vi, ‘be in doubt, unsure, confused’, a itr: 四十而不惑 (LY 2.4); 余甚惑焉, 儻所謂天道, 是邪非邪 (SJ 61); b caus: 惑主敗法,以亂士民 (HFei 11).

情  qíng, *dzeŋ, nc, 1 ‘nature (of a thing); emo­tions, feelings’: 性之好惡喜怒哀樂 謂之情 (Xun 22); 何謂人情,喜怒哀懼 愛惡欲七者,弗學而能 (Li 9); 內外上 下節者,義之情也 (Xun 16); 2 adv, ‘really, actually’: 情不知其不義也 (Mo 17); cf. 誠 [149.6].

9

愛  ài, *ʔə̂s, v , 1 ‘begrudge; be stingy tr

意  yì, *ʔəkh, nc, ‘thought, intention, will’: 道者,令民與上同意也 (Sun 1); 先生不羞,乃有意欲為收責於薛乎 (ZGC 11.1).

愚  yú, *ŋo, vst, ‘be stupid, ignorant’: a itr: 人之少也愚 (Lü 1.4); b caus: 愚其民 (Yan 8.1); 夫差智太宰嚭而愚子胥 (HFei 38); c comp: 君今自閉鉗天下之口而日 益愚 (SJ 101). Ant. 知 [111.3], 智 [72.8].

愉/喻  yú, *lo, vi, ‘be pleased, satisfied, happy’, a itr: 桀之治天下也,使天下瘁 瘁焉人苦其性, 是不愉也 (Zhuang 11); 自喻適志與 (Zhuang 2); b caus: 上稱三 皇五帝之業以愉其意 (Lü 7.4).

愈  yù, *loʔ, vst, 1 ‘be better, superior’, a itr: 女與囘也孰愈 (LY 5.9); esp. of diseases, ‘get better’: 昔者疾,今日愈 (Meng 2B11); b caus, ‘increase’: 愈疏,不 孝也 (Meng 6A3); c comp: 丹之治水也 愈於禹 (Meng 6B11); 法雖不善,猶愈 於無法 (Shen 1); d pseudo-­tr: 思天下惟 羿為愈己 (Meng 4B24); 2 adv, ‘increasingly, more’, expressing the comp of vst: 身 貴而愈恭,家富而愈儉,勝敵而愈戒 (Xun 8). Cf. Box 24.

10 惟  wéi, *wi, var. of 唯 [30.8]. 惜  xī, *sak, v ,‘be a pity’, a itr: 惜也,越竟 慈  cí, i

乃免 (Zuo 7.2); b caus, ‘pity, regret, feel

*dzə, nc, 1 ‘compassion, kindheartedness’: 親愛利子謂之慈,反慈為

454

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

嚚. . . 惻隱憐人謂之慈,反慈為忍 (XS 8.3); 2 vn, ‘be affectionate, kind (to); care for’, a itr: 慈於子者不敢絕衣食,慈於 身者不敢離法度,慈於方圓者不敢舍 規矩 (HFei 20); 父慈,子孝,兄良,弟弟 (Li 9); b tr: 皆慈其子而嚴其上 (HNan 11); 老其老,慈其幼 (GY 19.9).

慍  yùn, *ʔuns, vi, ‘be angry, annoyed’, a itr: 人不知而不慍,不亦君子乎 (LY 1.1); b pseudo-­tr: 然亦慍知伯 (SJ 43); c mod: 三已之,無慍色 (LY 5.19). Syn. 怒 [61.5].

慎  shèn, *dins, vn, ‘be careful, attentive, mindful; heedful, cautious’, a itr: 敏於事 而慎於言 (LY 1.14); 民平則慎,慎則難 變 (SJS 2); b tr: 戒之慎之 (Mo 31); 君子 必慎其獨也 (Li 42); with ne­gated infinitive O, ‘take care (not to . . .)’: 若漢挑戰,慎 勿與戰 (SJ 8). Syn. 戒 [62.3], 謹 [149.11]. 11

慢  màn, *mrâns, vn, ‘be neglectful, careless (a­bout); disrespect’, a itr: 君使民 慢,亂將作矣(Zuo 3.8); b tr: 民慢其 政,國之患也 (Zuo 3.28); syn. 傲 [9.11], 輕 [159.7]; c mod (!): 夫人知極,鮮有慢 心 (GY 7.6). 慮  lǜ,

*rah, vn, 1 ‘ponder, consider, think (about)’, a itr: 安而後能慮,慮而 後能得 (Li 42); b tr: 使知者慮之,則與 愚者論之 (Xun 12); 2 nc, ‘thoughts, deliberation’: 君亟定變法之慮 (SJS 1). Syn. 思 [61.5].

慶  qìng,

*khraŋ(h), nc, ‘blessing; recompensa­tion’: 勉之以慶賞 (Xun 9).

憂  yōu, *ʔu, vn, 1 ‘worry (about), be worried, sad’, a tr: 父母唯其疾之憂 (LY 2.6); b itr: 君子不憂不懼 (LY 12.4); syn. 懼 [61.18]; 2 adv, ‘anxious­ly’: 司馬牛憂曰(LY 12.5); 3 nc, ‘sadness, sorrow, worry’: 外無敵國之 患,內無亂臣之憂 (HFei 14); 悔吝 者,憂虞之象也 (Yi 7); syn. 患 [61.7].

慟  tòng, *dôŋh, vi, ‘be moved (to tears), lament, grieve’, only itr: 子慟矣 . . . 非夫人

之為慟而誰為 (LY 11.10); 悲心感慟 (LNZ 1.7). Cf. 動 [19.9].

12

憯  cǎn, *tshə̂mʔ, vst, 1 ‘be painful, hurt’, a itr: 苦痛雜於腸胃之間,則傷人也憯 (HFei 20); 怨之憯於骨髓 (HNan 18); b comp: 咎莫憯於欲得 (HFei 21); 其知憯 於蠣蠆之尾 (Zhuang 14); 2 adv: 重裘而 立,猶憯然有寒氣 (XS 7.3). 憎  zēng, *tsə̂ŋ, v , ‘hate, despise’, a tr: tr

子產憎其為人也 (Zuo 10.19); 好憎人 者,亦為人所憎 (SY 16); b pass: 禦人以 口給,屢憎於人 (LY 5.5). Syn. 惡 [61.8]; ant. 愛 [61.9]; 欲 [76.7].

憒  kuì, *kûs/kwə̂s, vi, ‘be distracted, troubled’, only itr: 文倦於事,憒於憂 (ZGC 11.1); 肅而不悖,剛而不憒 (HNan 5).

憫/閔  mı̌n, *mrən, vn, ‘be deplorable, depres­sed; worried (about), troubled, concerned’, a tr: 宋人有閔其苗之不長而 揠之者 (Meng 2A2); 文公於是憫中國之 微 (SY 10); b itr: 由此觀之,則人之生 憫矣 (HNan 9). 13

應  yìng, *ʔəŋh, v , ‘react; answer, reply’, a n

tr: 而可以應有馬,而不可以應有白馬 (GLong 1.2); b itr: 能定然後能應 (Xun 1).

14

懦/懧  nuò,

*no, vi, ‘be soft, weak, timid, cow­ ardly, effete’, a itr: 夫火形 嚴,故人鮮灼,水形懦,人多溺 (HFei 30); 宮之奇之爲人也,懦而不能 強諫 (Zuo 5.2); 憒於憂,而性懧愚 (ZGC 11.1); b mod: 懦夫有立志 (Meng 5B1). Syn. 弱 [57.7], 柔 [75.5]; ant. 強 [57.8]; cf. note at 儒 [9.14].

懧  nuò, *no, var. of 懦 [61.14]. 15

懲  chéng, *drəŋ, v , ‘correct, reprove, tr

repri­mand’, a tr: 勉之以慶賞,懲之以

455

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

刑罰 (Xun 9); b pass: 教而不誅,則姦民 不懲 (Xun 10).

16

懷  huái, *grûi, vn, ‘embrace, carry in one’s bos­om; hold dear, cherish’, a tr: 君子 懷德,小人懷土,君子懷刑,小人懷 惠 (LY 4.11); b itr: 民者固服於勢,寡能 懷於義 (HFei 49). 懸/縣  xuán,

*gwên, vi, ‘hang; be attached to’, a itr: 怨懸於天下 (HFei 2); 海 內之心懸於天子 (CQFL 34); b caus, ‘hang up, suspend’ 懸琴於城門以為寡人符 (SY 1); c pseudo-­tr2: 乃懸書宮門曰 (SY 6).

18

懼  jù, *gwakh, v , 1 ‘be scared, afraid n

(of)’, a itr: 一怒而諸侯懼, 安居而天下 熄 (Meng 3B2); 公懼於晉,殺子叢以說 焉 (Zuo 5.28); 齊王聞之,君臣恐懼 (ZGC 11.1); b tr: 吾不免是懼 (Zuo 9.22); 及昭公即位,懼其殺己也 (HFei 39); c pseudo-­tr: 吏乃皆悚懼其所 (HFei 30); syn. 恐 [61.6]; 憂 [61.11]; 2 nc, ‘fear’: 多男 子而授之職,則何懼之有 (Zhuang 12).

懼  jù,

*gwakh, vtr, ‘scare, frighten, intimidate’, a tr: 民不畏死,奈何以死懼 之 (Lao 74); 吾懼君以兵,罪莫大焉 (Zuo 3.19); 懼之以患害 (HFei 9); b mod: 彼見吾貌,必右懼心 (Zuo 12.2).

戌  xū, *sut/swit, n

temp, eleventh of the ‘Earthly branches’: 庚戌天王崩 (SY 19). Cf. Focus 8.

3

我  wǒ, *ŋâiʔ, prper, ‘I, we’, a as S: 我肉也 (Lü 11.4); 人之所教,我亦教之 (Lao 42); b as O: 既已知吾知之而問我 (Zhuang 17); 我之求也 (Zuo 2.16); 莫我 知也夫 (LY 14.35); c mod: 凡我國之忠信 之士,我將賞貴之 (Mo 13). Cf. Box 5. 戒/誡  jiè, *krə̂h, v , ‘guard against, be i

on guard’, a itr: 天地人皆以告矣,而王 不知戒焉 (ZGC 13.1); 聰者聽於無 聲,慮者戒於未成 (SY 16); b caus, ‘warn, admonish’: 父以戒子,兄以戒弟 (Mo 26); 誡其嗣懿子 (SJ 47); c pseudo– tr: 戒之慎之 (Mo 31); 簡士卒以戒不虞 (Wen 12). Syn. 警 [149.13]; 慎 [61.10].

4

或  huò, *wə̂k, adv [< 有, 74.2], ‘some, someone’: 宋人或得玉 (Zuo 9.15); 或問 子產 (LY 14.9). Cf. 7.3.

或者  huozhě, *wə̂k-­taʔ, adv, ‘maybe,

Classifier 62  戈

perhaps’: 或者不可乎 (Meng 2B2).

0

戈  gē, *kwâi/kôi, n , ‘battle ax’: 取戈劍 c

者,其不義又甚入人欄廄,取人馬牛 (Mo 17).

1

戊  wù, *muh, ntemp, fifth of the ‘Heavenly stems’: 以戊己日為大黃龍一 (CQFL 74). Cf. Focus 8. 2

成  chéng, *deŋ, v , 1 ‘finish, complete; tr

(Zuo 5.22); b pass: 德成而教尊 (Li 8); c mod: 愚者闇於成事 (SJS 1); 2 nc, ‘comple­ tion’: 疑行無成,疑事無功 (SJS 1); 3 np, ‘Accomplished’, a posth. name: 十八 年,宣公卒,子成公黑肱立,是為成 公 (SJ 33). Syn. 考 [125.1]; ant. 敗 [66.7].

achieve, accomplish’, a tr: 宋人既成列

8

戟  jı̌, *krak, n , ‘halberd’: 既而與為公 c

介,倒戟以禦公徒而免之 (Zuo 7.2).

11

戮  lù, *ruk, vtr, 1 ‘kill, execute; humiliate, dis­grace’, a tr: 無罪而戮民,則士可以徙 (Meng 4B4); b pass: 俄則束乎有司,而 戮乎大市 (Xun 5); 2 nc, ‘disgrace’: 妻子為 戮 (Zuo 6.13); 為天下戮 (Lü 9.2). Note: 戮, related to 僇, ‘disgrace’, means a particularly shameful ex­ ecution, including

456

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

public exposure of the muti­lated corpse; cf. note at 殺 [79.6].

12

戰  zhàn, *tans, vi, ‘fight, battle, wage war’, a itr: 鄒人與楚人戰 (Meng 1A7); b caus: 貧窮不足者與之,其足以戰民乎 (HFei 34); 夫陳善田利宅所以戰士卒也 (HFei 45); c mod: 凡天下戰國七 (SJ 69); ant. 守 [40.3]; 2 nc, ‘battle, war’: 國之所以 興者,農戰也 (SJS 3); 邲之戰,楚大勝 晉 (SY 4). Note: 戰 is the most general term for ‘war’, but it can specifically mean offensive war­fare, as opposed to defensive war­fare, 守 [40.3]; cf. also 攻 [66.3] and the note at 征 [60.5]. 13

戲  xì, *haih, v , ‘play (with), joke, divert’, n

a itr: 孔子為兒嬉戲,常陳俎豆,設禮 容 (SJ 47); b tr: 前言戲之耳 (LY 17.4); 天 子無戲言,言則史書之 (SY 1).

Classifier 63  戶 0

戶  hù, *gâʔ, n , ‘door, gate’: 不出戶而知

才  cái, *dzə̂, vi, ‘be talented, gifted, able’, a itr: 才不才,亦各言其子也 (LY 11.8); 公脩而才,可謂小儒矣 (Xun 8); b mod: 昔高陽氏有才子八人 (SJ 1). 3

扡  tuō, *lhâi(h), var. of 拖 [64.5]. 4

承  chéng,

*dəŋ, vtr, ‘receive, accept; uphold’, a tr: 用能協於上下,以承天休 (Zuo 7.3); ‘hold up; of­fer, present’: 出門如 見大賓,使民如承大祭 (LY 12.2); b pass: 故天下承而不絕 (SJ 79).

扶  fú, *ba, vtr, ‘support, assist’, a tr: 邠 人負幼扶老從之 (SY 14); b pass: 蓬生麻 中,不扶而直 (Xun 1); 壯者趨而 進,老者扶而至 (HSWZ 7). Syn. 攜[64.18]. 抑  yì, *ʔək, conj, ‘or; rather’, in alt. ques-

tions: 求之與,抑與之與 (LY 1.10); 南方 之強與,北方之強與,抑而強與 (Li 31).

折  zhé, *tet, v , ‘bend; break’, a tr: 折頸 tr

天下 (HSWZ 3).

而死 (HFei 49); b pass: 折而不撓,勇也 (Xun 30); 族庖月更刀,折也 (Zhuang 3).

4

5

c

所  suǒ, *sraʔ, 1 pr

rel, ‘where, what’: 所居 (HSWZ 8); by ex­tension ‘that which, what’, a sub­stituting for the O of a verb: 不從其所 令,從其所行 (Li 33); implying obligation: 人不忘其所忘,而忘其所不忘 (Zhuang 5); 其所厚者薄,而其所薄者厚(Li 42); b for the O of a prep: 舜之所以事堯 (Meng 4A2); c sub­stituting for a PrP: 聖人有所遊 (Zhuang 5); cf. 12.2 and Review 3; 2 nc,‘place’: 鄭公子忽在王所 (Zuo 1.7); syn. 處 [141.5].

Classifier 64  手

tr

拔人之城而非攻也 (Sun 3); b pass: 兵至 釐而六城拔矣 (HFei 19).

拜  bài, *prêts, v , 1 ‘bow (before)’, a itr: n

武王又再拜稽首,乃出 (SJ 4); 敵者不 在,拜於其室 (Li 13); b tr: 不得受於其 家,則往拜其門 (Meng 3B7); 2 adv: 大 夫親賜士,士拜受 (Li 13).

拂  fú, *phət, vn, ‘stroke gently; brush, wipe’, a tr: 周公拂其首 (SY 3.6); b itr: 忠 言拂於耳 (HFei 32). 拘  jū,

0

手  shǒu, *hjuʔ/nhuʔ, n , ‘hand’: 足,啟予手 (LY 8.3).

拔  bá, *brât, v , ‘capture, conquer’, a tr:

c

啟予

*ko, vtr, ‘seize, grasp; cling to, adhere to’, a tr: 拘禮之人,不足與言事 (SJS 1); b pass/pseudo-­tr: 比干見刳,孔 子拘匡 (Xun 26); 周西伯拘羑里 (SJ 32).

457

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

招  zhāo,

*tau, vn, ‘wave (to), signal (s.o.)’, a itr: 登高而招,臂非加長也,而 見者遠 (Xun 1); b tr: 旃以招大夫,弓以 招士,皮冠以招虞人 (Zuo 10.20).

拙  zhuō,

*tot, vi, 1 ‘be clumsy, inept, incompe­tent’, a itr: 夫楚劍利,倡優拙 (SY 15); b caus: 徑省其說則以為不智而拙之 (HFei 12); c mod: 大匠不為拙工改廢繩墨 (Meng 7A41); 2 nc, ‘clumsi­ness’: 大巧若拙 (Zhuang 10). Ant. 巧 [48.2], 善 [30.9].

拖/扡/拕  tuō, *lhâi(h), vn, ‘pull, draw; tear off; take away’, only tr: 至殺不 辜人也,扡其衣裘 (Mo 17); 奪之車 馬,解其橐笥,拖其衣被 (HNan 18).

挾貴,不挾兄弟而友 (Meng 5B3); b itr: 制度以陳,政令以挾 (Xun 11).

8

授  shòu, *duh, vn, only tr, ‘give, hand over, deliv­er’, a tr with Oi: 堯授舜,舜授 禹(XX 7.1); b tr2: 多男子而授之職,則 何懼之有 (Zhuang 12); c tr with Oi and PrP: 授之以政 (LY 13.5).

推  tuī, *thûi, vn, ‘push; promote, encourage, ad­ vance; remove, replace’, a tr: 夫二子者,或輓之,或推之 (Zuo 9.14); 剛柔相推而生變化 (Yi 7); b itr: 天 子三推,三公五推,卿諸侯九推 (Li 6); 言以虛靜推於天地 (Zhuang 13).

拕  tuō, *lhâi(h), var. of preceding. 掌  zhǎng, *taŋʔ, n , ‘palm of the 拊  fǔ, *phoʔ, v , ‘lay hands on; com- hand’: 及生,有文在掌曰友,遂以名 c

n

fort, console; strike, clap’, a tr: 不拊愛子 其民 (ZGC 11.1); 王大喜,拊手而笑 (SY 13); b itr: 拊不得其所白 (GLong 5).

6

拾  shí, *gip, v , ‘pick up’, a tr: 塗不拾遺 n

(SJ 47); b itr: 是君子之所棄, 而愚者拾 以為己寶 (Xun 22).

掘/闕  jué, *got, vn, ‘dig (out/up), excavate’, a tr: 若闕地及泉,隧而相見 (Zuo 1.1); 掘平王冢,笞其墳 (SY 12); 掘 地為臼 (Yi 8); b itr: 闕為深溝,通於商 魯之閒 (GY 20.6). 9

提  tí, *dê, v , ‘hold in hands, grasp; lift tr

7

挫  cuò, *tsôih, v , ‘break, wear away’, a tr

tr: 其用戰也勝,久則鈍兵挫銳 (Sun 2); b pass: 今有五錐,此其銛,銛者必先 挫 (Mo 1); 思以一豪挫於人,若撻之於 市朝 (Meng 2A2).

挺  tı̌ng, *lhêŋʔ, vi, ‘be straight’, a itr: 雖 有槁曓,不復挺者,輮使之然也 (Xun 1); 猿得木而挺 (SY 16); b mod: 將劍與挺 劍異 (Mo 44).

振/賑  zhèn, *təns, vtr, ‘support, assist’, a tr: 不資者振之 (Guan 80); 持節發河南 倉粟以振貧民 (SJ 120); b pass: 禍必不振 矣 (ZGC 31.5); 戰士以奉,飢民以賑 (YTL 2). 挾  xié,

之 (SJ 33).

*gêp, vn, ‘hold, grasp; tuck under the arm’, a tr: 執朱弓,挾朱矢 (Mo 31); fig. ‘claim; rely on’: 不挾長,不

up’, a tr: 列子提屨,跣而走 (Zhuang 32); 提刀而立(Zhuang 3); b mod: 孩提之 童,無不知愛其親者 (Meng 7A15).

揭  jiē, *g/k/k rat, v , ‘raise, lift; carry’, a h

tr

tr: 於是乘其車,揭其劍 (ZGC 11.1); 昭 昭乎若揭日月而行也 (Zhuang 19); 斬木 為兵,揭竿為旗 (XS 1.1); b pass: 唇揭者 其齒寒 (ZGC 27.1).

揠  yà,

*ʔrât, vn, ‘pull up/out’, only tr: 宋人有閔其苗之不長而揠之者 (Meng 2A2); 助之長者,揠苗者也 (Meng 2A2).

揉  róu, *nuh, var. of 輮 [159.9]. 10

搏  bó, *pâk, v , ‘strike, hit, beat’, a tr: 公 n

嗾夫獒焉,明搏而殺之 (Zuo 7.2); ‘grab, seize’: 有故而去,則君搏執之 (Meng 4B3); b itr: 慎罪邪,務搏執 (Li 6).

458

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

搆  gòu, *kôh, v , ‘build, put together’, a tr

tr: 有聖人作,搆木為巢 (HFei 49); b pass: 夫一戰而不勝,則禍搆矣 (HFei 2); 事已搆,可遂殺楚使者 (SJ 91).

11

抱一兒,攜一兒而行 (LNZ 5.6); b itr: 長者與之提攜 (Li 1). Syn. 扶[64.4].

Classifier 66  攴 2

摽  biāo, *piau, v , ‘wave, signal’, tr: 摽使 收  shōu, *hju, v , ‘take, receive, collect; 者出諸大門之外 (Meng 5B6). harvest’, a tr: 必死是間,余收爾骨焉 摩  mó, *mâi, v , ‘rub, polish’, a tr: 摩頂 (Zuo 5.32); 秦國之所以收君者,豈其微 n

tr

n

放踵利天下 (Meng 7A26); b itr: 是故剛 柔相摩,八卦相盪 (Yi 7); 兩木相摩而 然,金火相守而流 (HNan 1); 男女切 踦,肩摩於道 (HNan 11).

哉 (SJ 68); b pass: 五穀不收謂之饑 (Mo 5); 責畢收乎 (ZGC 11.1); 春耕,夏 耘,秋收,冬藏 (Xun 9).

13

3

gǎi, *kə̂ʔ, v , ‘change; improve’, a itr: 擊  jī, *kêk, v , ‘strike, hit’ a tr:以肱擊之 改  過而能改,善莫大焉 (Zuo 7.2); 吾能改 i

n

(Zuo 8.2); b itr: 以擊則不及 (Lü 8.3).

據  jù, *kah, v , ‘hold onto, cling to; rely n

on’, a tr: 據法而治者,吏習而民安 (SJS 1); b itr: 志於道,據於德,依於仁,游 於藝 (LY 7.6).

擒  qín,

*gəm, nc, ‘prey’: 故小敵之 堅,大敵之擒也 (Sun 3).

撻  tà, *thât, vn, ‘beat, flog, wallop’, a tr: 可使制梃以撻秦,楚之堅甲利兵矣 (Meng 1A5); b itr: 雖日撻而求其齊 也,不可得矣 (Meng 3B6).

擇  zé, *drak, vn, ‘choose, select’, a tr: 鳥 則擇木 (Zuo 12.11); b itr: 君請擇於斯二 者 (Meng 1B15). 17

攘  ráng, *naŋ, vn, ‘take away, steal, rob’, a tr: 其父攘羊,而子證之 (LY 13.18); 西 征,攘白狄之地,至於西河 (GY 6.5); b itr: 早晏不禁,則攘奪竊盜 (Guan 13). 18

攝相  shèxiàng,

*nhep-­ saŋh, nc, ‘deputy chan­cellor’: 由大司寇行攝相事 (SJ 47).

攜  xié, *wê, vn, ‘take by the hand, lead’, a tr: 民扶老攜幼,迎君道中 (ZGC 11.1);

矣 (Zuo 7.9); b caus: 王庶幾改之 (Meng 2B12); 三年無改於父之道,可謂孝矣 (LY 1.11).

攻  gōng,

*kôŋ, vn, 1 ‘attack, charge, strike’, a tr: 里克率國人以攻殺之 (Lü 23.6); b itr: 以難攻 (SJS 3); 2 nc, ‘attack; offensive war’: 此謀攻之法也 (Sun 3). Ant. 守 [40.3]; cf. note at 戰 [62.12].

4

放  fàng, *paŋh, v , ‘abandon, get rid of, tr

discard, exile’, a tr: 湯放桀,武王伐 紂,有諸 (Meng 1B15); b pass: ‘be abandoned/exiled; get lost’: 屈原遂放於 外,乃作離騷 (XX 7.19); 人有雞犬放 (Meng 6A11); fig. ‘abandon oneself, be heedless’: 葛伯放而不祀 (Meng 3B5); 聖 王不作,諸侯放恣 (Meng 3B9).

5

故  gù,

*kâh, nc, 1 ‘reason for, cause; tradi­tion’, a head: 司馬以吾故亡其良子 (Zuo 10.21); 不循秦國之故 (SJS 1); 何故 之法 (SJS 1); b mod, ‘old; well known, familiar’: 故人至暮不來 (HFei 32); cf. note at 古 [30.2]; 2 conj, a resultive, ‘there­ fore’: 故父母之愛不足以教子 (HFei 49); 不勝其哀,故哭也 (SY 6); cf. 16.4; b

459

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

consecutive, ‘then’: 順是,故争奪生而辭 讓亡焉 (Xun 23); 因糧於敵,故軍食可 足也 (Sun 2); 3 adv, ‘surely, certainly’: 微 君言,臣故將謁之 (HFei 36).

政  zhèng,

*teŋh, nc, ‘government, administra­tion’: 夫國之所以強者,政 也 (HFei 54); 天下有道, 則政不在大 夫 (LY 16.2); 政者,正也,正也者,所 以正定萬物之命也 (Guan 16).

6

效/校  xiào, *grâuh, vn, ‘imitate’, a tr: 不敬宗廟則民乃上校 (Guan 1); 崇效 天,卑法地 (Yi 7); b itr: 故效於古者,先 德而治,效於今者,前刑而法 (SJS 7). Syn. 法 [85.5]. 7

敖  áo, *ŋâu, vn, ‘saunter, ramble; play about’, a itr: 官無邪則民不敖,民不敖 則業不敗 (SJS 2); b tr: 無敖禮 (Zuo 11.4). 敗  bài, *prâts, v , ‘defeat, destroy’, only tr: n

是夫也將不唯衛國之敗 (Zuo 8.14); 利之 而勿害,成之勿敗 (SY 7). Ant. 成 [62.2].

敗  bài, *brâts, 1 vi [< preceding], ‘be defeated, destroyed’, itr: 師敗矣 (Zuo 8.2); 東敗於齊 (Meng 1A5); 2 nc, ‘destruction, defeat; decay’: 口之宣言也,善敗 於是乎興 (Guo 1.3). Ant. 勝 [19.10]. 教/敎  jiào,

*krâuh, vn, 1 ‘teach, instruct, in­form’, a tr: 不教民而用之 (Meng 6B8); b tr2: 父教之忠 (Zuo 5.23); cf. 5.2.2; c itr: 春秋教以禮樂,冬夏教以 詩書 (Li 5); 聖人不易民而教 (SJS 1); 堯 教於隸屬而民不聽 (HFei 40); 2 nc, ‘teach­ings, doctrine’: 七教,父子兄弟夫 婦君臣長幼朋友賓客 (Li 5).

救  jiù, *kuh, vn, ‘save, rescue, relieve (from)’, a tr: 諸侯之兵皆至救天子 (Lü 22.3); 湯之時,民樂其救之於患害也 (CQFL 1); b itr: 山澤救於火,草木植 成,國之富也 (Guan 4); 今人不務使患 無生,而務施救於患 (Wen 7); c pseu­ do-­tr: 明主救天下之禍,安天下之危者

也 (Guan 20.64); 是救病而飲之以堇也 (Lü 4.2).

敏  mı̌n, *mə̂ʔ, vn, ‘be clever, quick-­ witted, adroit’, a itr: 雍雖不敏,請事斯 語矣 (LY 12.2); 訥於言,敏於行 (SJ 130); b tr: 人道敏政, 地道敏樹 (Li 31). 啟  qı̌, *khîʔ, v , ‘open (up); set up, begin; tr

arouse, incite’, a tr: 乃啟其口 (XS 3.8); 齊桓 公并國三十,啟地三千里 (HFei 6); 今臣作亂,而君不禁,以啟寇讎 (Zuo 6.7); b pass: 雷之動也萬物啟 (Wen 6).

8

敢  gǎn, *kâmʔ, vn, ‘be courageous, audacious, impudent’, a itr: 惡果敢而窒 者 (LY 17.24); b tr: 遵汶之上,不敢左右 (GY 19.4); c typically aux, ‘dare, venture, presume to’: 楚不敢伐 (Zuo 2.6); 子無敢 食我也 (ZGC 14.3); cf. 4.4.1. 散  sàn, *sâns, v , ‘scatter, disperse’, a tr: 夫 tr

散其本教,而施之刑辟 (HSWZ 3); b pass: 兄弟妻子離散 (Meng 1A5); c mod: 四 時之散精為萬物 (HNan 3). Ant. 聚 [128.8].

9

敬  jìng, *kreŋh, nc, 1 ‘respect, deference’: 禮國之幹也,敬禮之輿也 (Zuo 5.11); 接遇肅正謂之敬,反敬為嫚 (XS 8.3); 2 vn, ‘be respectful (of), revere; hold in high regards’, a tr: 愛人者人恆愛 之,敬人者人恆敬之 (Meng 4B28); b itr: 太廟之內敬矣 (Li 10); 仁而有智,敬 於事上 (LNZ 3.7). Syn. 祗 [113.5]. 11

敵  dí, *dêk, n , 1 ‘enemy, foe, rival’: 外無 c

敵國之患,內無亂臣之憂 (HFei 14); 仁 者無敵 (Meng 1A5); syn. 讎 [149.16]; 2 vn, ‘resist, oppose, be a match for, equal’, a tr: 然則小固不可以敵大,寡固不可以敵 眾,弱固不可以敵強 (Meng 1A7); b itr: 夫誰與王敵 (Meng 1A5); 倍則分之,敵 則能戰之,少則能逃之 (Sun 3).

數  shù, *sroh, nc, ‘number, figure, amount’, a head: 強國知十三數,境內

460

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

倉口之數,壯男壯女之數 . . . (SJS 4); b mod, ‘a number of, some, sever­al’: 加我數 年,五十以學易 (LY 7.17); 孟嘗君為相 數十年 (ZGC 11.1); cf. Box 28.

數  shuò,

*srôk, adv, ‘repeatedly, frequently’: 數奪民時,大饑乃來 (Lü 26.3); 多飲數小便 (HDNJ 43).

數  shù, *sroh, n , ‘method, skill, techc

nique’: 口不能言,有數存焉於其間 (Zhuang 13); 故為國之數,務在墾草 (SJS 6); cf. 術 [144.5].

13

斂  liǎn, *ramʔ, rams, vn, 1 ‘collect, accumulate; levy (imposts)’, a itr: 厚斂以 彫牆 (Zuo 7.2); b tr: 斂幣玉,藏諸兩階 之間 (Li 7); 2 nc, ‘levy, impost’: 薄稅斂 (Meng 1A5). 14

斃  bì,

*be(t)s, vi, ‘fall, topple; drop dead, die’, a itr: 鄭人擊簡子中肩,斃于 車中 (Zuo 12.2); 驪姬與犬肉,犬 斃,飲小臣酒,亦斃 (GY 8.1); 齒堅於 舌而先斃 (Wen 1); b caus, ‘kill’: 射 之 , 斃 一 人  . . . 又 斃 二 人 , 每 斃 一 人,掩其目 (Li 4).

Classifier 67  文 0

文  wén, *mən, n , 1 ‘pattern, ornament; c

refine­ment’: 君子博學於文,約之以 禮,亦可以弗畔矣夫 (LY 6.27); esp. ‘writing’: 及生,有文在掌曰友,遂以 名之 (SJ 33); ant. 樸 [75.12]; 質 [154.8]; 2 np, ‘Civil’, a posth. name: 我文王之為子 (Xun 32); ant. 武 [77.5].

Classifier 68  斗 9

斟酌  zhēnzhuó, *təm-­tiauk, vn, ‘ponder, con­sider’, a itr: 而後王斟酌焉 (GY 1.3); b tr: 斟酌之者眾也 (HNan 9). Classifier 69  斤 1

斥  chì, *k-­lhak, v , ‘reject, turn down; tr

blame’, a tr: 大國之求, 無禮以斥之, 何饜之有 (Zuo 10.16); b pass: 斥乎 齊,逐乎宋衛 (SJ 47); 秦女必貴而夫人 斥矣 (SJ 70).

8

斯  sī, *se, pr

dem, 1 ‘this’, a head: 困而不 學,民斯為下矣 (LY 16.9); b mod: 請事 斯語矣 (LY 12.2); cf. note at 此 [77.2]; 2 conj, ‘then’: 我欲仁,斯仁至矣 (LY 7.30); 得其民,斯得天下矣 . . . 得其 心,斯得民矣 (Meng 4A10); 天下皆知 美之為美,斯惡已 (Lao 2); cf. 15.4.2.

9

新  xīn, *sin, vst, 1 ‘be new, recent’, a itr: 所謂古之言服者,皆嘗新矣 (Mo 39); b caus: 正靜不失,日新其德 (Guan 37); 遠者近之,而舊者新之 (Mo 46); c mod: 今所謂新王必改制者,. . . 非繼前王而 王也 (CQFL 1); 2 nc, ‘sth. new’: 舍其舊而 新是謀 (Zuo 5.28); ant. 舊 [134.12]. 11

斲  zhuó, *trôk, vn, ‘carve; chop, hack’, a tr: 輪扁斲輪於堂下 (Zhuang 13); b itr: 隨 繩而斲,因攢而縫 (HFei 27). 14

文  wèn, *mǝns, v , ‘decorate, ornament’, 斷  duàn, *tônʔ/s, v , ‘cut off, sever; septr

a tr: 飾邪說,文姦言 (Xun 4); 東方曰 夷,被髮文身 (Li 5); b pass: 丹漆不 文,白玉不彫,寶珠不飾 (SY 20); c mod: 文馬三十駟 (SJ 47). Syn. 飾 [184.5].

tr

arate; de­cide’, a tr: 斷木為杵,掘地為臼 (Yi 8); 寡人嘗學斷斯獄矣 (Li 4); 誅 紂,斷其首 (Xun 18); b pass: 動靜有 常,剛柔斷矣 (Yi 7); 曹觸龍斷於軍

461

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Xun 15); 故王者刑賞斷於民心,器用 斷於家 (SJS 5).

Classifier 70  方 0

方  fāng, *paŋ, nc, 1 ‘quarter, region; direction’: 有朋自遠方來 (LY 1.1); 遊必有方 (LY 4.19); 使於四方,不能專對 (LY 13.5); 2 vi, ‘be square’, a itr: 圓者中規,方者中矩 (Xun 26); 繼之以規矩準繩,以為方員 平直 (Meng 4A1); b caus: 以直其政,方 其義也 (Li 39); 3 adv: 地方百里而可以王 (Meng 1A5); cf. 18.3–⑯. Ant. 員 [30.7]. 方  fāng, *paŋ, adv, ‘at that moment, just

then’: 國家方危,諸侯方貳,將以襲 敵,不亦難乎 (Zuo 11.4); 齊王方大飲 (Lü 20.6). Cf. 10.5.

方  fāng, *paŋ, n , ‘method, technique’: c

多方以誤之 . . . 必大克之 (Zuo 10.30); 能 近取譬,可謂仁之方也已 (LY 6.30); 惠 施多方,其書五車 (Zhuang 33).

4

於  yú/yū, *ʔa, prep, 1 ‘in, at, on; from’, a local: 牛鳴於門外 (HFei 20); b temp: 主 相晉國,於今八年 (Zuo 10.1); ini: 於是 齊人侵魯疆 (Zuo 2.17); c with 乎: 財用 於是乎出 . . .衣食於是乎生 (GY 1.3); cf. Box 23; d with anteposed O: 君於何有 (Zuo 9.23); e with C: 今君富於季孫,而 齊大於魯 (HFei 39); f with pass, ‘by (s.o.)’: 必見欺於張儀 (SJ 40); g with what is logic­al­ly the Oi: 寡人託國於子 (HFei 34); cf. 8.2; 2 vn, ‘be/act at; relate to’, only tr: 人之過也,各於其黨 (LY 4.7); 善影者不於影於形,為天下者不於天 下於身 (Lü 3.3); 爭名者於朝,爭利者 於市 (ZGC 3.7); mostly in nom­inalized clauses: 君子之於禮,敬而安之 (Xun 12); cf. 13.4. Syn. 乎 [4.4], 于 [7.1]. 5

施  shī, *lhai, v , 1 ‘spread out, extend’, a i

itr: 令聞廣譽施於身 (Meng 6A17); 德之

理盡施於人 (XS 8.5); b caus, ‘bestow, apply’: 己所不欲,勿施於人 (LY 12.2); 今王發政施仁 (Meng 1A7); 施民所 善,去民所惡 (GY 19.3); 2 nc, ‘kindness, generosity’: 如有博施於民而能濟眾, 何如 (LY 6.30); 報施救患,取威定霸, 於是乎在矣 (Zuo 5.27).

施  yì/yí *lai, v , ‘spread, continue, transi

fer’, itr: 愛其母,施及莊公 (Zuo 1.1); 盎 於背,施於四體 (Meng 7A21); 夫施及三王而天下大駭矣 (Zhuang 11).

7

旌  jīng, *tseŋ, nc, 1 ‘banner, standard, flag’: 全羽為旞, 析羽為旌 (ZL 53); 2 vn, ‘signal, honour’, tr: 乃可以威民而懼戎, 且旌君伐 (GY 7.6). 旅  lǚ, *raʔ, n , ‘battalion’, supposedly conc

sisting of 500 men: 五人為伍,五伍為 兩,四兩為卒,五卒為旅 (ZL 20); 凡用 兵之法 . . . 全旅為上,破旅次之 (Sun 3).

族  zú,

*dzôk, nc, ‘lineage, extended family’: 父母宗族 . . . 為戮沒 (ZGC 31.5). Cf. note at 宗 [40.5].

Classifier 71  无 0

无  wú, *ma, var. of無 [86.8]. 无  wú, *mə, var. of 毋 [80.0]. 5

既/旣  jì, *kəs, vi, 1 ‘finish; be finished, done’, a itr: 萬物之情既矣 (HNan11); b caus: 吾與汝既其文, 未既其實 (Zhuang 7); 2 adv, ‘already’: 今周德既衰 (Zuo 5.24); cf. 10.4; 3 conj ‘when, after, since’: 君既定矣,又何求 (Zuo 11.4); cf. 16.3. 既而  jì'ér, *kəs-­nə, conj, ‘afterwards, later on’: 既而與為公介 (Zuo 7.2); 楚子 賜之金,既而悔之 (Zuo 5.18); cf. 16.3.

462

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 72  日

日適越而昔至也 (Zhuang 2). Ant. 今 [9.2]; cf. 古 [30.2].

0

日  rì, *nit, n , 1 ‘sun’: 日月所照 . . . 莫不 c

尊親 (Li 31); 2 ntemp, ‘day’: 吾與囘言終日 (LY 2.9); 日以暴之,夜以息之 (HNan 20); 3 adv, ‘daily’: 鵠不日浴而白,烏不 日黔而黑 (Zhuang 14). Syn. 晝 [72.7]; ant. 夜 [36.5]; cf. 月 [74.0].

2

早  zǎo, *tsûʔ, v , 1 ‘be early; early in the i

morning; timely’, itr: 尚早,坐而假寐 (Zuo 7.2); 其遠害也早 (Xun 2); 2 adv: 不 如早為之所 (Zuo 1.1).

3

旱  hàn, *gânʔ, nc, ‘aridity, drought’: 夫 水旱俱天下陰陽所為也 (SY 19). Ant. 水 [85.0]. 4

昆  kūn, *kûn, n , ‘elder brother’: 吾不 c

與子為昆弟矣 (HFei 23); a later (dialectal?) syn. of 兄 [10.3]. Ant. 弟 [57.4].

明  míng, *mraŋ, v , ‘be

clear, bright; st evident; clear-­sighted’, 所謂明者,無所 不見 (SJS 18); a itr: 用此觀之,人之性 惡明矣 (Xun 23); 得下之情,則是明於 民之善非也 (Mo 13); b caus: 搢紳先生多 能明之 (Zhuang 33); 大學之道,在明明 德 (Li 42); c comp: 在天者莫明於日月 (Xun 17); 視而形之,莫明於目 (HNan 10); d mod: 明主救天下之禍,安天下 之危者也 (Guan 64). Ant. 闇 [169.9].

昔  xī/xí, *sak, n

, 1 ‘former/ancient times’, typically as A, a ini, ‘formerly’: 昔文 公與秦伐鄭 (Zuo 9.14); ranging from antediluvian times: 昔高陽氏有才子八 人 (SJ 1) to the day before: 昔者疾,今日 愈 (Meng 2B2); often em­phasized with 者: 昔者由也聞諸夫子曰 (LY 17.7); cf. 10.2; b mod: 昔歲入陳,今茲入鄭 (Zuo 7.20); 昔日臣夢見君 (ZGC 20.17); 2 adv: 是今 temp

易  yì, *lek, v , 1 ‘change, exchange’, a tr: tr

以羊易之 (Meng 1A7); 夏殷之滅也,不 易禮而亡 (SJS 1); b pass: 政成而民 聽,易則生亂 (Zuo 2.2); 2 np, ‘(Canon of ) Changes’: 周史有以周易見陳侯 者,陳侯使筮之 (Zuo 3.22); 易者,察 人之精德之理與弗循,而占其吉凶 (XS 8.5).

易  yì, *lekh, vst, 1 ‘be easy (to do)’, a itr: 言之易,行之難 (Lü 24.1); 天下莫易於 為善 (Wen 9); b caus, ‘find easy’: 民 愚,則易力而難巧 (SJS 6); 仲尼賞而魯 民易降北 (HFei 49); c aux for pass: 上好 禮,則民易使也 (LY 14.41); 民農則 樸,樸則易用 (Lü 26.3); cf. 4.4.2; d comp: 易於反掌,安於太山 (SY 9); 2 nc, ‘sth. easy’: 難易相成 (Lao 2); 以易攻者 必危 (SJS 3). Syn. 簡 [118.12]; ant. 難 [172.11]. 5

春  chūn,

*thun, ntemp, ‘spring’: 二年 春,公會戎于潛 (Zuo 1.2); 天有四 時,春秋冬夏 (Li 29); 春者,天之和也 (CQFL 79).

春秋  chūnqiū, *thun-­ tshiu, nc, 1 ‘spring and autumn’, synecdoche for ‘year’: 先生老矣,春秋高矣 (HSWZ 10); in book titles, ‘annals’, esp. the ‘Spring and Autumn Annals’: 孔子成春秋而亂臣賊 子懼 (Meng 3B9); 春秋者,守往事之合 德之理與不合,而紀其成敗,以為來 事師法 (XS 8.5).

是  shì, *deʔ, vst, 1 ‘be right, correct, hold true’, a itr: 偃之言是也 (LY 17.3); 自 天子以至於庶人,壹是 (Li 42); b caus, ‘regard as right, approve’: 欲是其所非而 非其所是,則莫若以明 (Zhuang 2); 不 是師法,而好自用 (Xun 2); 2 nc, ‘right’: 以是為非 (Lü 18.4). Ant. 非 [175.0]. 是  shì, *deʔ, pr

dem, ‘this, these’, referring to a top­ic in question, a S: 是宜為君 (Zuo

463

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

3.11); b re­sum­ing an anteposed O: 唯 敵是 求 (Zuo 7.12); cf. 14.2; c mod: 是夫也將不 唯衛國之敗 (Zuo 8.14); also ‘the same’: 官以是豕來也 (Lü 17.3). Cf. note at 此 [77.2].

昭  zhāo, *tiao, nc, ‘ancestor (2nd, 4th, 6th, etc. in order)’: 大王之昭也 (Zuo 5.5). 6

晉  Jìn, *tsins, n , name of a polity: 晉唐 p

叔虞者,周武王子而成王弟 (SJ 39). Note: Jin, located at the intersection of present-­day Shanxi, Hebei, Shaanxi, and Henan. Given as a fief to a son of king Wu, it became one of the most powerful polities of the Chunqiu period, rising to hegem­ony under Duke Wen (r. 636–628). After military defeats by Chu, Jin’s power waned, and it was finally divided in 453 bc by mighty families into the states of Han韓, Wei魏, and Zhao趙.

時  shí, *dəʔ, n

, 1 ‘season; (point in) time, prop­ er time’: 天有四時,春秋冬 夏(Li 29); ini: 是時孔子當阨 (Meng 5A8); 2 vn, ‘be timely’, itr: 故養長時則六 畜育 (Xun 9); 制度時,則國俗可化 (SJS 8); 3 adv, ‘timely, in time; at times’: 學 而時習之 (LY 1.1); 秋水時至,百川灌 河 (Zhuang 17). temp

時  shí, *də, pr

dem [cognate with 之, 是], ‘this’, Opr: 少事長, 賤事貴, 共帥時 (Li 12).

7

晨  chén, *(m-)dən, ntemp, ‘morning’: 晨 往,寢門闢矣,盛服將朝 (Zuo 7.2). 晝  zhòu, *trukh/truh, n

, 1 ‘day’ (vs. night): 中山之俗,以晝為夜 (SY 13); 故 晝夜均而寒暑平 (CQFL 50); 2 adv, ‘at daytime’: 宰予晝寢 (LY 5.10); 夜行晝居 (Zhuang 20). Syn. 日 [72.0]; ant. 夜 [36.5]. temp

8

景  Jı ̌ng, *kraŋʔ, n , ‘Glorious’, a posth. p

name: 齊景公問政於孔子 (LY 12.11).

智/知  zhì, *treh, nc, 1 ‘wisdom, knowledge’: 深知禍福謂之知,反知為愚 (XS 8.3); 聞而知之,聖也,見而知之,智 也 (Wen 5); 2 vst, ‘be wise, knowledgeable’ a itr: 仁且智,夫子既聖矣 (Meng 2A2); b caus: 其家甚智其子,而疑鄰人之父 (HFei 12); c comp: 蟲莫知於龍 (Zuo 10.29). Syn. 巧 [48.2]; ant. 愚 [61.9]. 暑  shǔ, *lhaʔ, vst, ‘be hot’, a itr: 冬日則 寒,夏日則暑 (Lü 15.5); b mod: 喜則為 暑氣而有養長也 (CQFL 44); 2 nc, ‘heat, hot weather’: 日月運行,一寒一暑 (Yi 7); 寒往則暑來,暑往則寒來 (Yi 8); 日 南,則景短多暑 (ZL 2). Ant. 寒 [40.9]. 9

暇  xiá,

*grâh, nc, ‘leisure’: 壯者以暇 日,修其孝弟忠信 (Meng 1A5).

11

暴  bào, *bâukh, v , 1 ‘be violent, cruel, i

forceful’, a itr: 近古之世,桀紂暴亂,而 湯武征伐 (HFei 49); b mod: 邪說暴行又 作 (Meng 3B9); 伐無道,誅暴秦 (SJ 48); 2 nc, ‘violence’: 合於犯分亂理而歸於暴 (Xun 23). Syn. 苛 [140.5], 虐 [141.3].

暴/曓  pù, *bôk, v , ‘expose

(to the tr sun), dry in the sun’, a tr: 秋陽以暴之 (Meng 3A4); 一日暴之,十日寒之 (Meng 6A9); 久暴師則國用不足 (Sun 2); b pass: 雖有槁曓,不復挺者,輮使 之然也 (Xun 1).

暮  mù, *mâkh, n

temp, 1 ‘evening’: 故人 至暮不來 (HFei 32); 2 adv: 朝食於戚, 暮宿於百泉 (Xun 8). Ant. 朝 [74.8].

暱  nì, *nrək, vst, ‘be near, close, intimate’: a itr: 諸夏親暱,不可棄也 (Zuo 4.1); 夫正國者不可以暱於權 (GY 14.1); b caus: 宮之奇 . . . 少長於君,君暱之 (Zuo 5.2); 其誰暱我 (Zuo 9.2). Syn. 親 [147.9]; 私 [115.2]. 13

曓  pù, *bôk, var. of 暴 [72.11].

464

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

15

曠  kuàng,

*khwâŋh, vi, ‘be desolate, empty; unin­habited; neglected’, a itr: 野 蕪曠則民乃菅 (Guan 1); b caus: 君不失 其威,下不曠其產而莫相德也 (Guan 30); c mod: 內無怨女,外無曠夫 (Meng 1B5); 無曠土,無游民 (Li 5).

其言以遺後世 (Mo 17); b itr: 受命於君 前,則書於笏 (Li 13); 書於竹帛 (Mo 27); c pseudo-­tr: 乃丹書帛曰陳勝王 (SJ 48).

一尚書  Shàngshū,

*daŋh-­ lha, np, since Han times ‘(Canon of) Docu­ments’: 孝文帝時,欲求能治尚書者,天下無 有 (SJ 121).

Classifier 73  曰

8

0

*tsôs, adv, ‘extremely, most’, expressing the superlative of vst: 畫孰最難者 (HFei 32); 智氏最強 (SY 10); 最為天下貴 也 (Xun 9); with vn: 趙襄子最怨智伯 (SJ 86). Cf. Box 24.

最  zuì,

曰  yuē, *wat, v , ‘say, call’, a tr: 天子 tr

曰,天降禍于周 (Zuo 10.32); b pass, with C, ‘be called’: 北方有獸,其名曰蟨 (SY 6.1); c in object C: 生穆公,名之曰 蘭 (Zuo 7.3). Syn. 云 [7.2].

2

曲  qū, *khok, v , 1 ‘be crooked, dishoni

est’, a itr: 其小枝卷曲而不中規矩 (Zhuang 1); 以為直於君而曲於父 (HFei 49); b caus: 曲其頸,欲攫而食之也 (HSWZ); 2 nc, ‘curvature; curve’: 其曲中 規 (Xun 1). Ant. 直 [109.3].

曲  qǔ/qū * khok, n , 1 ‘tune, air, piece c

9

會  huì,

*gôts, nc, 1 ‘meeting, confer­ ence’: 齊桓為幽之會,衛不到 (CQFL 8); 2 vn, ‘meet, come to­gether (with)’, a tr: 秦會諸侯於周 (SJ 4); b itr: 十六年,會 于曹 (SJ 33); 陰陽會於中冬 (CQFL 46); c pseudo-­tr: 與魯定公好會夾谷 (SJ 32).

會  kuài, *gwât/gôts, n , ‘account, financ

cial rec­ord’: 會計當而已矣 (Meng 5B5); 誰習計會 (ZGC 11.1).

(of music)’: 使公卿至於列士獻詩,瞽 獻曲 (GY 1.3).

Classifier 74  月

3

0

更  gēng, *krâŋ, v , ‘change, alter’ a tr: 今 月  yuè, tr

吾欲變法以治,更禮以教百姓 (SJS 1); b pass: 則谿陵山谷之神之祭更,應國 之稱號亦更矣 (Guan 35).

6

書  shū, *lha, nc, 1 ‘(piece of) writing’: 書 足以記名姓而已 (SJ 7); 何必讀書然後 為學 (SJ 47); spe­cifically ‘letter’: 為書與其 傅鞠武曰 (YDan 1); or ‘document’ esp. such in the ‘(Canon of) Documents’: 書 者,政事之紀也 (Xun 1); 書者,著德 之理於竹帛而陳之 (XS 8.5); 順先王詩 書禮樂以造士 (Li 5); 2 vn, ‘write (down), record’, a tr: 子張書諸紳 (LY 15.6); 故書

*ŋwat, nc, 1 ‘moon’: 日月所 照 . . . 莫不尊親 (Li 31); 2 ntemp, ‘month’: 自十月不雨至于五月 (Zuo 5.3).

2

有  yǒu, *wəʔ, vn, 1 ‘have, own, possess’, a tr: 父子有親,君臣有義,夫婦有別,長 幼有序,朋友有信 (Meng 3A4); ‘be there, occur’ 有齊無秦,有秦無齊 (SJ 69); typi­ cally with locative S: 田中有株 (HFei 49); b tr with C: 爾有母遺 (Zuo 1.1); 2 nc, ‘pos­ session’: 汝身非汝有也 (Zhuang 22). Ant. 無 [86.8]; cf. Box 25 and 有司 [30.2]. 有  yǒu, *wәʔ, pr

ind, ‘something, somebody’, a as O, preceding the P: 將有待也

465

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Zuo 6.12); 子駟與尉止有爭 (Zuo 9.10); preceding a prep: 亦將有以利吾國乎 (Meng 1A1); cf. 5.2.1; b mod, ‘some, a’: 有朋 自遠方來,不亦樂乎 (LY 1.1); 有聖人 作,搆木為巢 (HFei 49); but cf. Box 25. Ant. 無 [86.8].

有  yǒu, *wəʔ, nc, 1 ‘exist­ence, being’: 故 有無相生 (Lao 2); ant. 無 [86.8]; 2 vi, ‘exist, be there’, on­ly itr: 君子而不仁者有 矣夫 (LY 14.6); 法有,民安其次 (SJS 20); 若其未有,君亦圖之 (Zuo 10.5); 法 令滋章,盜賊多有 (SJ 122); cf. Box 25. 有  yòu, *wəh, var. of 又 [29.0]. 4

服  fú, *bək, nc, 1 ‘clothes; mourning clothes’: 晨往,寢門闢矣,盛服將朝 (Zuo 7.2); 2 vn, ‘wear’, a tr: 齊桓公好服 紫,一國盡服紫 (HFei 32); b itr: 蘭芷不 為莫服而不芳 (Wen 6).

服  fú, *bək, vi, ‘comply, submit, serve’, a itr: 東西南北,莫敢不服 (ZGC 11.5); 人 之情心服於德,不服於力 (Wen 4); b caus, ‘subdue’: 人君不能服強敵,破大 國也 (SJS 3); c pseudo-­tr: 上服度則六親 固 (Guan 1). 朋  péng, *bə̂ŋ, n , ‘comrade, mate’: 有 c

朋自遠方來,不亦樂乎 (LY 1.1); 朋友 死 (LY 10.19). Note: 朋 seems to denote a fellow member (of a school or another institution), a ‘colleague’, while 友 re­fers to a friend by mutual affection.

7

望  wàng, *maŋ, v , 1 ‘look, gaze at; hope n

for’, a tr: 從山下望木者,十仞之木若箸 (Xun 21); 予日望之 (Meng 2B12); b itr: 吾 嘗跂而望矣 (Xun 1); 且民之望於上也 甚矣 (HFei 36); 2 nc, ‘hope’: 吾猶有望 (Zuo 5.24); 吾子楚國之望也 (Zuo 10.12).

8

朝  cháo, *drau, nc, 1 ‘court(yard); audience’: 使天下仕者皆欲立於王之朝

(Meng 1A7); 朝將視朝,不識可使寡人 得見乎 (Meng 2B2); syn. 庭 [53.7]; 2 vi, ‘attend court/audience’, said both of ruler and sub­jects, a itr: 固將朝也,聞王命而 遂不果(Meng 2B2); 諸侯朝於天子曰述 職 (Meng 1B4); 王朝而有憂色 (Xun 32); b caus, ‘cause to attend court, give audi­ence (to)’: 武丁朝諸侯有天下 (Meng 2A1); 文 王以為令尹 . . . 朝陳蔡 (Zuo 12.17); c pseudo-­tr: 孟子將朝王 (Meng 2B2).

朝  zhāo, *trau, ntemp, 1 ‘morning’: 病非 一朝一夕之故 (Lie 6); 2 adv: 朝將視 朝,不識可使寡人得見乎 (Meng 2B2). Ant. 暮 [72.11]. 期  qī/qí, *gə, v , ‘set (a time); set (one’s n

mind) on, aspire to, aim at’, a tr: 晉文公 伐原,與士期七日 (Lü 19.6); 是以聖人 不期脩古 (HFei 49); b itr: 至於聲,天 下期於師曠 (Meng 6A7); 與老人期, 後, 何也 (SJ 55); 2 nc, ‘appointed time’: 公聞其期,曰可矣 (Zuo 1.1); 會天大 雨,道不通,度已失期 (SJ 48).

期  jī, *kə, var. of 朞 [74.8]. 朞/期  jī, *kə, v , ‘be full, entire’, mod: i

後期年,齊王謂孟嘗君曰 (ZGC 11.1); 吾田朞歲不收 (HSWZ 9); 苟有用我 者,朞月而已 (SJ 47).

Classifier 75  木 0

木  mù ,*môk, nc, ‘tree, wood’: 車木 也 . . . 船木也 (Mo 45). 1

本  běn, *pə̂nʔ, n , 1 ‘root, origin; fundac

mentals’: 孝弟也者,其為仁之本與 (LY 1.2); 吾所謂利者,義之本也(SJS 7); esp. referring to farming: 徠三晉之民,而使 之事本 (SJS 15); ant. 末 [75.1]; 2 vi, ‘be root­ed in’, itr: 善不善本於義 (Lü 19.4); 凡 戰法必本於政勝 (SJS 10).

末  mò, *mât, n , ‘treetop; branches; tip, c

end’: 末大必折 (Zuo 10.11); 明足以察秋

466

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

毫之末 (Meng 1A7); fig. ‘non­essentials’, esp. referring to trade as opposed to farming: 舍本而事末則不令 (Lü 26.3); 其本 亂而末治者否矣 (Li 42). Ant. 本 [75.1]; dist. 未 [75.1].

未  wèi, *məs, adv, 1 ‘not (yet)’: 未知 生,焉知死 (LY 11.12); 吾未之聞也 (Meng 4A19); with vst, ‘not quite; not really’: 未仁乎 (LY 14.16); 未知,焉得仁 (LY 5.19); 吾未 之樂也,亦未之不樂也 (Zhuang 18); cf. 10.4; 2 conj, ‘before’: 公未言,榮成子曰 (GY 5.5); cf. 16.3. Dist. 末 [75.1]. 未  wèi, *məs, n

temp, eighth of the ‘Earthly branches’: 八月辛未,治兵,建而不旆 (Zuo 10.13). Cf. Focus 8; dist. 末 [75.1].

2

朱  zhū, *to, vi, ‘be bright red, vermilion’, a itr: 鞸,君朱,大夫素 (Li 13); b caus, ‘colour bright red’: 公取而朱其尾鬣以與 之 (Zuo 11.10); c mod: 乘朱路,駕赤 騮,載赤旂,衣朱衣 (Li 6). 3

材  cái,

*dzə̂, nc, ‘material, (natural) resources’: 雖楚有材,晉實用之 (Zuo 9.26).

果  guǒ, *kôiʔ, n , 1 ‘fruit’: 民食果蓏蜯 c

蛤 (HFei 49); 2 vi, ‘come to fruition, be realized’, a itr: 大人者,言不必信,行 不必果,惟義所在 (Meng 4B11); b caus, ‘realize, put into effect’: 梁惠王不果所言 (SJ 74); 2 adv, ‘in effect, really, actually’: 君 不果行 (Zuo 10.7); 果能此道矣,雖愚 必明(Li 31); cf. 4.3.

林  lín, *rəm, n , ‘forest’: 山林者,鳥獸 c

之居也 (Xun 14).

松  sōng, *s-­loŋ, n , ‘pine tree’: 松耶柏 c

耶 (SJ 46).

枉  wǎng, *ʔwaŋʔ, n , ‘crookery; corrupc

tion’: 禮不踰節,義不自進,廉不蔽 惡,恥不從枉 (Guan 1).

5

柏  bǎi/bó, *prâk, n , ‘cypress’: 松耶柏 c

耶 (SJ 46).

某  mǒu, *môʔ, pr

ind, ‘a certain; someone; so and so’: 某某也 (SY 3.6).

柱  zhù, *droʔ, nc, ‘pillar, post, column’: 散木也 . . .以為柱則蠹,是不材之木也 (Zhuang 4); 倚柱彈其劍 (ZGC 11.1).

李  lı̌, *rəʔ, n , ‘plum’: 其不義又甚入人 柰  nài, *nâs, var. of 奈 [37.5]. 園圃竊桃李 (Mo 17). 柔  róu, *nu, v , ‘be soft, weak, gentle, yielding’ , a itr: 其聲能短能長,能柔能 束  shù, *lhok, n , ‘bundle’, as measure: c

st

c

4

剛 (Zhuang 14); 其聲和以柔 (Li 19); 雖 愚必明,雖柔必強 (Li 42); b caus, ‘soften, placate, mollify’: 去其餌,柔其肉 (Li 12); 嘉善而矜不能,所以柔遠人也 (Li 31); c comp: 天下莫柔弱於水 (Lao 78); d mod: 柔聲以諫 (Li 12). Syn. 弱 [57.7]; 懦 [61.14]; ant. 剛 [18.8]; 強 [57.8].

東  dōng, *tôŋ, n

6

雖有詩書,鄉一束,家一員,猶無益 於治也 (SJS 3).

杯  bēi, *pə̂, nc, ‘cup’, a head: 母歿而杯 圈不能飲焉 (Li 13), b as measure: 一杯水 (Meng 6A18). Cf. Box 13. loc, 1 ‘east’: 孔子登東山 而小魯,登泰山而小天下 (Meng 7A24); 2 vi, a itr: ‘go/turn east’, a itr: 齊宋 衛燕舉師而東 (CQFL 33); 日東則景夕 多風 (ZL 18); b caus, ‘orient, direct towards the east’: 使齊之封內,盡東其 畝 (Zuo 8.2). Ant. 西 [146.0].

格  gé, *krâk, v , ‘investigate, examine’, a tr

tr: 致知在格物 (Li 42); b pass: 物格而後 知至 (Li 42).

桀  Jié, *gat, np, the last Xia king: 湯放 桀,武王伐紂 (Meng 1B15).

467

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

桃  táo, *lâu, n , ‘peach’: 竊其桃李 (Mo 棲  qī/xī, *sî, v , ‘roost, dwell’: a itr: 越王 c

17).

i

校  xiào, *grâuh, nc, ‘barracks’: 鄭人游 于鄉校,以論執政 (Zuo 9.31).

句踐乃以兵五千人棲於會稽山上 (SY 9); b caus: 吳王東伐越,棲諸會稽 (Mo  49); 於是舉兵而攻齊,棲閔王於 莒 (SY 8).

校  xiào, *grâuh, var. of 效xiào [66.6]. 棟  dòng, *tôŋh, n , ‘ridgepole’: 任重莫 栩栩  xǔxǔ, *hwaʔ-­hwaʔ, v , 1 ‘be 若棟,任國莫若德 (Wen 9); 上棟下 c

i

blithe, cheer­ful’: itr: 今視子之鼻間栩栩 然 (Zhuang 21); 2 adv (?): 栩栩然胡蝶也 (Zhuang 2).

株  zhū, *tro, n , ‘tree trunk, tree stump’: c

田中有株 (HFei 49).

宇,以待風雨 (Yi 8); fig. 太子,國之棟 也 (GY 7.7).

9

楚  Chǔ, *tshraʔ, n , 1 a polity: 吾楚人 p

7

梁  liáng, *raŋ, n , ‘bridge, beam’: 莊子 c

與惠子遊於濠梁之上 (Zhuang 17).

梃  tı̌ng, *lêŋʔ, n , ‘stick, club’: 可使制 c

梃以撻秦楚之堅甲利兵矣 (Meng 1A5).

梓  zı̌, *tsəʔ, n , ‘catalpa’, a tree: 二子者 c

往觀乎南山之陰,見梓勃焉實而俯 (SY 3.6).

8

楮  chǔ, *thraʔ/tâʔ, n , ‘mulberry tree’: c

宋人有為其君以象為楮葉者 (HFei 21).

也 (ZGC 7.5); 2 vn, ‘behave/speak in the way of Chu’: 雖日撻而求其楚,亦不可 得矣 (Meng 3B6). Note: Chu, founded in the eleventh century bc in the re­gion of modern Hubei and ruled by the Mi 羋 clan, was long considered a ‘barbarian’ polity by the northerners. Yet, in Chunqiu times, it expanded from the middle Yangzi region into the Yangzi delta and the North China plain, even reaching hegemony under king Zhuang莊 (r. 613– 591 bc). A major rival of Qin in the Zhanguo period, Chu was finally annihilated by the latter in 223 bc.

椎  chuí, *drui, n , ‘hammer’: 釋椎鑿而 業  yè, *ŋap, n , ‘work, business, trade; 上 (Zhuang 13).

c

c

棺  guān, *kôn, n , ‘(inner) coffin’: 人不 c

死則棺不買 (HFei 17).

椁  guǒ, *kwâk, nc, ‘(outer) coffin’: 古之 喪禮 . . . 天子棺槨七重 (Zhuang 33); 顏 淵死,顏路請子之車以為之椁 (LY 11.8).

task, enterprise; achievement’: 願留而受 業於門 (Meng 6B2); 顧爭於戎狄,去王 業遠矣 (ZGC 3.7); 繼文之業,而信宣 於諸侯 (Zuo 5.25).

概然  gàirán,

*kə̂s-nan, vi, ‘be sad’, hapax, itr: 我獨何能无概然 (Zhuang 18).

10

極  jí, *gək, n , ‘highest point, top, peak, 槁  gǎo, zenith; limit’: 是先王之道,忠臣孝子之 c

極也 (Xun 19); 早處之,使知其極 (GY 7.6).

棄  qì, *khits, vtr, ‘abandon, discard, reject, throw away’, a tr: 棄君之命,不信 (Zuo 7.2); b pass: 水官棄矣, 故龍不生 得 (Zuo 10.29); c mod: 人無棄人,物無 棄物 (HNan 12); 問國之棄人, 何族之 子弟也 (Guan 24).

*khâuʔ, vi, ‘dry up, wither, shrivel’, a itr: 七八月之間旱,則苗槁矣 (Meng 1A6); 百仞之松,本傷於下,而 末槁於上 (Lü 3.3); b mod: 形固可使如 槁木 (Zhuang 2).

槐  huái, *grûi, nc, ‘pagoda tree’: 觸槐 而死 (Zuo 7.2).

榮  róng, *wreŋ, nc, ‘blossom; glory, honour, fame’: 衣食足則知榮辱 (Guan 1).

468

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

11

樂  lè, *râuk, nc, 1 ‘happi­ness, joy, delight’: 吾樂甚多 (SY 17); 2 vst, ‘be happy, joyful’, a itr: 有朋自遠方來,不亦樂乎 (LY 1.1); b caus, ‘make happy, please’: 樂之勿苦 (SY 7); c comp: 聽人以言,樂於鐘鼓琴瑟 (Xun 5). Ant. 苦 [140.5]. 樂  yuè,

*ŋrâuk [probably cognate to pre­ceding], nc, 1 ‘music’: 臣請為王言樂 (Meng 1B1); 2 np, the ‘Canon of Music’: 順 先王詩書禮樂以造士 (Li 5).

樂  yào, *ŋrâuks, v , ‘rejoice, delight in’, n

tr: 知者樂水,仁者樂山 (LY 6.23); 安其 居,樂其俗 (Lao 80); 君明則樂官不明 則樂音 (ZGC 22.6).

若馭樸馬,若養赤子 (Xun 13). Syn. 素 [120.4]; 質 [154.8]; ant. 文 [67.0].

13

檝  jí, *tsap, n , ‘oar’: 假舟檝者,非能 c

水也,而絶江河 (Xun 1).

17

欄  lán, *rân, nc, ‘corral, pen, enclosure’: 取戈劍者,其不義又甚入人欄廄,取 人馬牛 (Mo 17). 18

權  quán, *gon, n , ‘scale, balance’: 謹權 c

槪  gài, *kə̂s, v , ‘be moved; sad’, itr: 我獨

量,審法度 (LY 20); ‘power, authority’: 國之所以治者三:一曰法,二曰 信,三曰權 (SJS 14); 不知三軍之 權,而同三軍之任 (Sun 3).

12

Classifier 76  欠

i

何能無槪然 (Zhuang 18).

橋  qiáo, *gau, n , a tree: 橋者父道也 c

2

樹  shù, *doh, nc, 1 ‘tree’: 大國之樹必巨 (SY 12); 2 vn, ‘plant (with trees), establish’, a tr: 五畝之宅,樹之以桑 (Meng 1A3); 葬之中野,不封不樹 (Yi 8); b tr2: 天生 民而樹之君,以利之也 (Zuo 6.13).

次  cì, *tshih, nc, [perhaps < 二], 1 ‘the second, next, following’: 學而知之者次 也,困而學之,又其次也 (LY 16.9); ‘sequence, order’: 抑晉國之舉也,不失 其次 (GY 2.11); ‘position’: 法有,民安其 次 (SJS 20); 2 vn, ‘follow, be next’, tr: 全卒 為上,破卒次之 (Sun 3). Syn. 序 [53.4]; cf. note at 二 [7.0].

橫  héng,

7

(SY 3.6).

橐  tuó, *thâk, nc, ‘bag, sack’: 寘諸橐以 與之 (Zuo 7.2).

*gwrâŋ, vi, ‘be horizontal, oblique’, a itr: 天下未嘗無事,不從則橫 (SY 11); b caus: 蓺麻如之何,橫從其畝 (Li 30).

橫  hèng, *gwrâŋs, vi, 1 ‘be perverse, unruly, harsh, arbitrary, offensive’, a itr: 文帝寬,不忍罰,以此吳日益橫 (SJ 106); b mod: 橫政之所出,橫民之所 止,不忍居也 (Meng 5B1); 2 adv: 又非 吾敢橫失而能盡之難也 (HFei 12). 樸  pǔ,

*phrôk, vi, ‘be natural, raw; plain, simple, unso­phisticated’, a itr: 其民 愚而樸,少私而寡欲 (Zhuang 20); 民農則樸,樸則易用 (Lü 26.3); b mod:

欲  yù, *lok, vn, 1 ‘want, desire, crave’, a tr: 我欲仁,斯仁至矣 (LY 7.30); 故常無 欲,以觀其妙;常有欲,以觀其徼 (Lao 1); with O clause: 治國者欲民之農 也 (SJS 3); b aux, ‘want to’: 堯欲傳天下 於舜 (HFei 34); cf. 4.4.1; syn. 嗜 [30.10], 好 [38.3]; ant. 惡 [61.8], 憎 [61.12]; 2 nc, 生 而有耳目之欲,有好聲色焉 (Xun 23). 8

欺  qī, *khə, vn, ‘cheat, lie to, deceive’, a tr: 吾誰欺,欺天乎 (LY 9.12); b itr: 臣下 百吏誕詐之人乘是而後欺 (Xun 12); 侮 蔽惡以欺於君 (HFei 34); but c pass [!]:

469

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

故主必欺於上,而臣必重於下矣 (HFei 14); 王欺於張儀 (SJ 40).

10

歌  gē, *kâi, v , 1 ‘sing’, a itr: 莊子則方箕 n

踞鼓盆而歌 (Zhuang 18); 今王子頹歌舞 不倦 (Zuo 3.20); 夜聞漢軍四面皆楚歌 (SJ 7); b tr: 歌鹿鳴之三,三拜 (Zuo 9.4); 公飲之酒,使大師歌巧言之卒章 (Zuo 9.14); 2 nc, ‘song’: 賜聞聲歌各有宜也 (Li 19).

11

歎/嘆  tàn, *nhân/nhâns, vn, ‘sigh’, a itr: 顏淵喟然歎 (LY 9.11); b tr: 蓋嘆魯也 (Li 9); 平城之危,七日不食,天下歎 之 (XX 10.13). 14

歟/與  yú, *la [< 也乎], pt

int, a with verbal P: 女弗能救與 (LY 3.6); in alt questions: 求之與,抑與之與 (LY 1.10); cf. 5.3.1; b with nom P: 是魯孔丘與 (LY 18.6); 今孔丘年少好禮,其達者歟 (SJ 47); cf. 3.4.1; in alt questions: 天與 . . . 人與 (Zhuang 3); cf. 3.4.2. Cf. Box 16.

3

步  bù, *bâh, nc, ‘step, pace’, unit of length, approx. 1.38 m: 能見百步之 外,而不能自見其睫 (HFei 21); 廣百 步,長百步,為百畝 (HSWZ 4). 4

武  Wǔ,

*maʔ, np, ‘Martial’, a posth. name; ant. 文 [67.0]; esp. that of the Zhou founder: 武王伐紂 (Meng 1B15); cf. Focus 7.

9

歲  suì, *swats, n , ‘harvest; year’: 非我 c

也歲也 (Meng 1A3); 武留十餘歲 (XX 7.29). Syn. 年 [51.3].

14

歸  guī, *kwəi, v , 1 ‘return’, a itr: 行而不

Classifier 77  止

i

0

止  zhı̌, *təʔ, vi, ‘stop, stand still’, a itr: 終 日在地上行止 (Lie 1); b caus: 其誰能止 之 (SY 9); c mod: 人莫鑑於流水,而鑑 於止水 (Zhuang 5). Ant. 行 [144.0]. 1

正  zhèng, *teŋh, v , ‘be straight, correct, st

right’: 方直不曲謂之正,反正為邪 (XS 8.3); a itr: 名不正,則言不順 (LY 13.3); b caus, ‘correct, reg­ulate’: 吾誰使正之 (Zhuang 2); c comp: 道莫正於中 (CQFL 77); d mod: 子為正卿 (Zuo 7.2).

2

此  cı̌, *tsheʔ, pr

人君之操也 (SY 1); 誰可順此事 (SJ 1). Note: 此, which refers to something close at hand, is often opposed to 彼 [60.5] which refers to something more distant; finally, 是 [72.5] refers to something that has just been men­tioned; 斯 [69.8] was a Lu dialect syno­nym of 此.

dem, ‘this’, a head: 此與彼 同類 (Mo 45); 此天也 (XS 3.9); b mod: 此

知歸,失家者也 (Lie 1); 取國子以歸於 吳 (Mo 49); b caus: 不如歸尸以內攻之 (Lü 15.6); 齊人來歸鄆 讙 龜陰田 (Zuo 11.10); c mod: 夫言死人為歸人,則生人 為行人矣 (Lie 1).

歸  guī, *kwəi, vn, 1 ‘turn to, go to; give alle­giance to’, a itr: 逃墨必歸於楊,逃楊 必歸於儒 (Meng 7B26); b tr: 天下歸湯 若流水 (Guan 80); 從之者如歸巿 (Meng 1B15); 是無歸夫 (SY 5); 2 nc, ‘refuge; home’: 夫上之所為,民之歸也 (Zuo 10.21); 男有分,女有歸 (Li 9). 歸  guī,

*kwəi, vn, 1 ‘give, consign, assign (to)’, a tr: 天下之民歸心焉 (LY 20.1); 天下歸仁焉 (LY 12.1); 然而天下獨 歸咎於齊者,何也 (ZGC 12.1); b tr2: 百 姓歸之名 (Li 27); c pseudo-­tr2: 少取以成 秋,其余以歸之冬 (CQFL 49).

470

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 78  歹

de­struction’: 故殘賊生而忠信亡焉 (Xun 23). Cf. note at 殺 [79.6].

2

死  sı,̌ *siʔ, n , 1 ‘death’: 未知生,焉知 c

死 (LY 11.12); 2 vn, ‘die, be dead’, a itr: 人不 死則棺不買 (HFei 17); 予死於道路乎 (LY 9.12); b tr‚ ‘die for’, 提彌明死之 (Zuo 7.2); 民親其上,死其長矣 (Meng 1B12); c pseudo-­tr: 與其死夫人所者,不若賜 死君前 (HFei 14); 重徙則死處而無二慮 (Lü 26.3); d mod [!]: 生人立焉,死人入 焉 (SY 2). Ant. 生 [100.0]. Note: Differ­ent words for ‘to die’ were used in CC, depending on the rank of the de­ceased: 天子死 曰崩, 諸侯曰薨, 大夫曰卒, 士曰不 祿, 庶人曰死 (Li 2). More­over, 沒 [85.4] and 終 [120.5] are euphemistic expressions for a natural death; 殪 [78.12] denotes a violent death, typically after being shot by an arrow.

5

殆  dài, *lə̂ʔ, vst, 1 ‘be dangerous, hazardous; be endangered, at risk’, itr: 學而 不思則罔,思而不學則殆 (LY 2.15); b caus: 戰必能殆之而羞與之鬥 (Xun 7); 與小人處者,吾殆之也 (Xun 28); syn. 危 [26.4]; 2 adv, ‘(there is a risk of >) perhaps, prob­ably, should’, a in verbal clause: 君亟定變法之慮,殆無顧天下之議之 也 (SJS 1); 殆必不可 (Zuo 10.25); b in nom. clause: 是殆非周公之行 (Xun 8); cf. 7.2. 殃  yāng, *ʔaŋ, 1 n , ‘calamity, disaster; c

misfor­tune’: 無德而祿,殃也,殃將至 矣 (Zuo 4.2); 眾以為殃 (Li 9); 2 vtr, ‘harm, damage’, a tr: 不教民而用之,謂 之殃民 (Meng 6B8); 今以梗陽之賄殃 之,不可 (GY 15.7); b pass: 果實早 成,民殃於疫 (Li 6).

8

殘  cán, *dzân, v , 1 ‘kill, exterminate, tr

destroy’, a tr: 陽人不服,公圍之,將殘 其民 (GY 10.16); b pass: 夫士死於外,民 殘於內 (ZGC 12.1); 2 nc, ‘carnage,

殖  zhí, *dək, v , ‘plant, cultivate; raise, tr

establish’, a tr: 其子曰農,能殖百穀 (Li 23); b pass: 樹之而五穀蕃焉,草木殖 焉,禽獸育焉 (Xun 32); 六畜不育於 家,五穀不殖於野 (YTL 15); c mod: 人 民之眾,山川之觀,殖物之阜 (Lie 5). Syn. 育 [130.4]; 蕃 [140.11].

12

殫  dān, *tân, vtr, ‘exhaust, deplete, drain’, a tr: 夫鈍兵挫銳,屈力殫貨(Sun 2); 殫 天下之財,而澹一人之欲,禍莫深焉 (HNan 15); b pass: 玩好充, 故家貨殫 (Yan 3.11). 殪  yì, *ʔits, v , ‘be killed, drop dead’, a i

itr: 前後擊之,盡殪 (Zuo 1.9); 偃,且 射子鉏,中頰,殪 (Zuo 11.8); b pseudo­tr: 殪車中,伏弢而死 (Mo 31); c caus, ‘kill’: 莊子儀荷朱杖而擊之,殪之車上 (Mo 31). Cf. note at死 [78.2].

Classifier 79  殳 6

殺  shā, *srât, vtr, ‘kill, slay’, a tr: 楚人執 而殺之 (Zuo 10.8); b pass: 先令者殺 (HFei 19); 隱公慈而殺於弟,叔武賢而 殺於兄 (HSWZ 7); 公子不可以入,入 則殺矣 (GYang 5.1); 姦人者殺 (SY 15). Note: Violence being ram­pant in ancient China, there were several distinct words for taking some­one’s life. Whereas 殺 is the most general term, 弒 means assassinating a superior, 賊 de­notes malicious murder; 戮 means a dishonourable execution; 殘 means cruel (mass) slaughter; 滅 denotes the extinction of a whole group. Cf. also the notes at 死 [78.2] and 刑 [18.4]. 9

毀  huı̌, *hmaiʔ, v , ‘destroy, demolish; tr

ruin’, a tr: 家必自毀,而後人毀之 (Meng 4A8); b pass: 龜玉毀於櫝中

471

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(LY 16.1); 械用,則凡非舊器者舉毀 (Xun 9). Ant. 譽 [149.14].

Classifier 82  毛 0

毛  máo, *mâu, n , ‘hair, fur’: 馬,蹄可

Classifier 80  毋

c

以踐霜雪,毛可以禦風寒 (Zhuang 9); 人之所以為人者,非特以二足而無毛 也 (Xun 5).

0

毋/無/无 wú, *mə, adv, ‘(should) not, do not’: 王如知 此,則無望民之多於鄰國也 (Meng 1A3); 所惡於上,毋以使下,所惡於下,毋 以事上 (Li 42); 无傷吾行 . . . 无傷吾足 (Zhuang 4); 爾無我叛 (Zuo 10.16). Cf. 4.3.

1

母  mǔ,

*mə, nc, ‘mother’: 女子之嫁 也,母命之 (Meng 3B2); 子之於母,其 親愛也 (LNZ 5.6); 父母者,子婦之所 受教也 (Guan 64). Ant. 父 [88.0].

3

每  měi, *mə̂ʔ, pr

ind, 1 ‘each, every’, mod: 子入太廟,每事問 (LY 3.15); 故為政 者,每人而悅之,日亦不足矣 (Meng 4B2); 2 adv, ‘every time, whenever’: 每至 於族,吾見其難為,怵然為戒 (Zhuang 3); 每見王,常掩鼻 (HFei 31).

Classifier 81 

Classifier 83  氏 0

氏  shì, *geʔ, n

c 1, ‘lineage; family’: 鄭人 取貨於印氏 (Zuo 9.26); 2 ‘Mr.’, after family name; ‘Mrs.’, after clan name: 何必公 山氏之之也 (LY 17.4). Cf. Focus 5.

1

民  mín, *min, n , ‘man; people; populac

tion’: 民為貴,社稷次之,君為輕 (Meng 7B14). Note: The precise meaning of 民 has been subject to much scholarly debate. It has been said to mean ‘slaves’, ‘peasants’, ‘man’ as a biological creature as op­posed to a social being (人), common people as opposed to the aristocracy (百 姓), subjects as opposed to the ruler (上), etc. Like English ‘peo­ ple’, its meaning seems to be highly variable: 民 are those ‘people’ that matter in a given context.

- 黎民  límín, *rî-­min, n , ‘black

0

c

比  bì, *bih, v , ‘follow, go along, associn

ate, com­ply with; be cliquish’, a itr: 君子 之於天下也 . . . 義之與比 (LY 4.10); 君子 周而不比,小人比而不周 (LY 2.14); b tr: 先王之道 . . . 比中而行之 (Xun 8).

比  bì, *bih, 1 conj, ‘when, by the time’: 比 至,已誅之矣 (HFei 34); 比已食,三歎 (GY 15.7); com­bined with 及: 比及五 世,皆反葬於周 (Li 3); 2 prep, ‘for’: 願比 死者壹洒之 (Meng 1A5); ‘when, until, by the time’: 比其反也,非彼死則臣必死矣 (Lü 16.4); com­bined with 及: 公索氏比及 三年必亡矣 (SY 13); syn. 及 [29.2], 至 [133.0].

(-haired) people’, i.e. the com­mon people: 黎 民不飢不寒 (Meng 1A3).

- 四民  sìmín,

*slis-­ min, nc, ‘four estates of peo­ple’: 士農工商四民者,國 之石民也,不可使雜處 (Guan 20) 古者 有四民,有士民,有商民,有農 民,有工民 (GLiang 8.1).

Classifier 84  气 6

氣  qì, *khəs, nc, ‘breath; pneuma, vital spirit; energy; vapour’: 水火有氣而無生 (Xun 9); 氣也者,虛而待物者也

472

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Zhuang 4); 氣也者,神之盛也 (Li 24); 氣衰則生物不遂 (Li 19).

Classifier 85  水 0

水  shuı̌, *lhuiʔ, n , 1 ‘water; river’: 黃帝 c

遊乎赤水之北 (Zhuang 12); ant. 火 [86.0]; esp. ‘flood’: 夫水旱俱天下陰陽所 為也 (SY 19); ant. 旱 [72.3]; 2 vtr, ‘flood, inundate’, a tr: 圍晉陽而水之 (ZGC 18.1); 冬日至,令剝陰木而水之 (ZL 70); b pass: 中古之世,天下大水 (HFei 49); 3 adv, ‘on water’: 主人疏食水飲 (Li 22).

1

永  yǒng, *wraŋʔ, vi, 1 ‘be long-­lasting, enduring, eternal’, a itr: 行客之人,嗟然 永久 (LNZ 6); b caus: 永其世而豐其年 (YTL 29); 2 adv: 以遺後之人,使永監 焉 (GY 3.9). Note: This word is very rare in CC, it mostly occurs in quotations of pre-­Classical texts like the Odes or the Documents. The CC equivalent is 久 [4.2]. 2

求  qiú, *gu, vn, ‘search, seek, strive for, demand’, a tr: 唯敵是求 (Zuo 7.12); 虎求 百獸而食之 (ZGC 14.3); with infinitive O: 求致社稷之福 (HFei 49); b itr: 先生 非求於人,人則獻之 (SY 4); c pseudo­tr: 國之所以求民者少 (SJS 6). 3

江  Jiāng, *krôŋ, n , 1 the Yangzi river: p

江東已定,急引兵西擊秦 (SJ 7); 2 nc, ‘river’: 彼其道遠而險,又有江山 (Zhuang 20).

汝/女  rǔ, *naʔ, pr

per, ‘you’: a S: 女弗能 救與 (LY 3.6); b O: 堯何以資汝 (Zhuang 18); 余不女忍殺 (Zuo 10.1); c mod: 汝身非 汝有也 (Zhuang 22). Cf. Box 5.

池  chí, *d-­lai, n , ‘pool, pond; moat’: 然 c

後踐華爲城,因河爲池 (XS 1.1); 城郭 溝池以為固 (Li 9). Syn. 溝 [85.10].

4

決/决  jué, *kwêt, vtr, 1 ‘open up, breach; sever, cut off; decide’, a tr: 決諸東 方則東流 (Meng 6A2); 決晉水而灌之 (ZGC 18.2); 決獄訟,必端平(Li 6); b pass: 濡肉齒決,乾肉不齒決 (Li 1); 治 國是非,不以術斷而決於寵人(HFei 47); c mod: 天下一歲決獄幾何 (SJ 56).

沒  mò, *mə̂t, vi, ‘sink; disappear; die (naturally), pass away’, a itr: 始浮,行數 十里乃沒 (SJ 126); 夢為魚而沒於淵 (Zhuang 6); 吾即沒,若必師之 (SJ 47); b caus: 行并植於晉國,不沒其身(Li 4); 竭心力以沒其身 (Yan 4.18); c pseudo-­tr: 君子疾沒世而名不稱焉 (LY 15.20). Cf. note at 死 [78.2]. 沃  wò, *ʔâuk, nc, ‘grasslands’: 猶其有 原隰衍沃也 (GY 1.3).

沈/沉  chén, *d-­ləm, v , ‘sink, drown; i

be sub­ merged, immersed’, a itr: 散木 也,以為舟則沈(Zhuang 4); 沉於國家 之事 (ZGC 11.1); 常民溺於習俗,學者 沉於所聞 (ZGC 19.4); b caus: 沈其卒二 萬人於河中 (SJ 73); c pseudo-­tr2: 殺出子 及其母,沈之淵旁 (SJ 5).

沉  chén, *d-­ləm, var. of 沈 [85.4]. 5

泉  quán, *dzwan, nc, ‘spring, source; subterrane­an water’: 短綆不可以汲深井 之泉 (Xun 4).

-黃泉  huángquán, * gwâŋ-­dzwan, low springs’, i.e. ‘the nethernc, ‘the yel­ world’: 不及黃泉,無相見也 (Zuo 1.1).

治  chí, *drə, vn, ‘regulate, order, govern, admin­ister, control’, a tr: 天 . . . 不能治人 也 (Xun 19); b itr: 無為而治者,其舜也 與 (LY 15.5); 聖人不易民而教,知者不 變法而治 (SJS 1); 百官以治,萬民以察 (Yi 8); but c pass (!): 勞力者治於人 (Meng 3A4). 治  zhì, *drəh, n

c [< preceding], 1 ‘order’: 禮義者,治之始也 (Xun 9); 治亂天邪

473

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Xun 17); 2 vst, ‘be well-­ordered/orderly’, a itr: 舜有臣五人而天下治 (LY 8.20); 家 齊而後國治 (Li 42); b comp: 管夷吾治 於高傒 (Zuo 3.9); c mod: 治國刑多而賞 少 (SJS 7). Ant. 亂 [5.12].

法  fǎ, *pap, n , 1 ‘model, norm, rule, c

law’: 予之法制,告之訓典 (Zuo 5.23); 人主釋法用私 (HFei 6); ant. 私 [115.2]; 2 vn, ‘take as model/example, follow, model sth. on’, tr: 是以聖人苟可以彊國,不法 其故 (SJS 1); syn. 效 [66.6].

河  Hé, *gâi, np, the Yellow River: 河內 凶,則移其民於河東 (Meng 1A3); 河水 出昆侖東北陬,貫渤海 (HNan 4).

況  kuàng, *hwaŋh, v , ‘compare with, n

be equal to’, a itr: 成名況乎諸侯 (Xun 6); typically in rhe­tor­ical questions: 臣猶知 之,況於君乎 (SJ 39); b tr: 臣猶知 之,而況君乎 (GY 10.1).

沮  jǔ, *dzaʔ, vtr, ‘obstruct; stop, hinder; dissuade’, a tr: 請先嘗沮之 (SJ 47); b pass: 是以為善者勸,為不善者沮 (Xun 16). 泰山  Tàishān,

*thâs-­ srân, np, a mountain: 孔子登東山而小魯,登泰山 而小天下 (Meng 7A24).

6

洗/洒  xı̌, *sə̂iʔ, vn, ‘cleanse, wash (clean)’, a tr: 願比死者壹洒之 (Meng 1A5); b itr: 足垢,燂湯請洗 (Li 12).

洒  xı̌, *sə̂iʔ, var. of 洗 [85.6]. 洩洩  yìyì, *lats-­lats, v (?) ‘overflow, i

spread’, hapax, itr: 大隧之外,其樂也洩 洩 (Zuo 1.1).

7

流  liú, *r(i)u, vi, a itr, ‘flow; wander’: 上 下與天地同流 (Meng 7A13); b caus, ‘banish, exile’: 三年乃流王於彘 (GY 1.3); c mod: 人莫鑑於流水,而鑑於止水 (Zhuang 5).

海  hǎi, *hmə̂ʔ, nc, ‘sea, ocean’: 江海之 魚吞舟,大國之樹必巨 (SY 12); ‘bor-

derland’: 四海之內皆兄弟也 (LY 12.5); cf. note at 境 [32.11].

浴  yù, *lok, vn, ‘bathe’, itr: 鵠不日浴而 白,烏不日黔而黑 (Zhuang 14). 8

淺  qiǎn, *tshenʔ, v , ‘be shallow, superst

ficial (also fig.)’, 多聞曰博,少聞曰淺 (Xun 2); a itr: 知行淺薄 (Xun 3.5); 其所 以聞見甚淺 (Lü 17.3); b caus: 若夫斷之 繼之,博之淺之,益之損之,類之盡 之,盛之美之 (Xun 19); 子深其深,淺 其淺,益其益,尊其尊 (Mo 44); c mod: 夫淺知之所爭者末矣 (Lie 8). Ant. 深 [85.8], 博 [24.10].

深  shēn, *nhəm, v , ‘be deep, profound’, st

a itr: 水益深,. . . 火益熱 (Meng 1B17); b caus: 子深其深,淺其淺,益其益,尊 其尊 (Mo 44); c comp: 傷人之言,深於 矛戟 (Xun 4); 天下之患,莫深於獄 (SY 5). Ant. 淺[85.8].

淫  yín, *ləm, n , 1 ‘depravity, licentiousc

ness, excess’: 故淫亂生而禮義文理亡焉 (Xun 23); 2 vi, ‘be licentious, immoderate, immoral, depraved’, a int: 關雎,樂而不 淫 (LY 3.20); 文巧不禁則民乃淫,不璋 兩原則刑乃繁 (Guan 1); b caus: 天下不 淫其性,不遷其德 (Zhuang 11).

淵  yuān,

*ʔwîn, nc, 1 ‘deep water; whirlpool’: 鳥飛於空,魚游於淵 (Shen 9); 2 np, personal name: 顏淵喟然歎 (LY 9.11).

9

渡  dù, *dâkh, v , ‘cross, ford’, a tr: 乃西 n

南渡淮水,之衡山南郡 (SJ 6); b itr: 河 水間耳,君不能自渡 (SY 17); 武王伐 紂,渡于孟津 (HNan 6).

湯  Tāng,

*lhâŋ, np, founder of the Shang dynasty: 湯與桀戰於鳴條之野 (SJS 17); 湯放桀,武王伐紂, (Meng 1B15). Cf. Focus 7.

湍  tuān, *thôn, v , ‘whirl, flow rapidly’, i

a itr: 水湍悍,難以行平地 (SJ 29);

474

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

b  caus: 道果便近,而水湍石,不可漕 (SJ 29); c mod: 性猶湍水也 (Meng 6A2).

游/遊  yóu, *ju, vi, ‘wander, stroll, drift’, a itr: 乃語魯君為周道游 (SJ 47); b mod: 游士危於戰陳 (HFei 13); ‘swim’: 鳥 飛於空,魚游於淵 (Shen 9); 越人善游 矣 (HFei 40). 10

滅  miè, *met, v , ‘defeat, destroy, extertr

minate’, a tr: 獻公伐驪戎,克之,滅驪 子 (GY 7.3); b pass: 人有此三數行 者,以為上則必危,為下則必滅 (Xun 3.5); 是周滅於從,衛亡於衡也 (HFei 49); c mod: 興滅國,繼絕世 (LY 20.1). Syn. 亡 [8.1.]; cf. note at 殺 [79.6].

溺  nì, *niâuk, vi, ‘be immersed, submerged; drown’, a itr: 夫常人安於故 習,學者溺於所聞 (SJS 1); b caus: 即不 為河伯娶婦,水來漂沒,溺其人民 (SJ 126); c mod: 假人於越而救溺子,越人 雖善遊,子必不生矣 (HFei 22).

滕  Téng, *lə̂ŋ, np, a polity: 滕小國 也,間於齊楚(Meng 1B13). Note: Teng was a small polity in southeastern Shandong ruled by the Ji 姬 clan; it was annihi­lated by Song in 316 bc.

源/原  yuán, *ŋwan, nc, ‘source, origin’: 不璋兩原則刑乃繁 (Guan 1).

滋/兹  zī, *tsə, v , 1 ‘increase, grow, i

proliferate’, itr: 不如早為之所,無使滋 蔓 (Zuo 1.1); 物生而後有象,象而後有 滋,滋而後有數 (Zuo 5.15); 2 adv, ‘more, increasingly’: 民之多違,事滋無 成 (Zuo 9.8); esp. expressing the comp of vst:苟虧人愈多,其不仁兹甚,罪益厚 (Mo 17); cf. Box 24.

漢興三十年矣 (XS 3.9). Note: When Liu Bang proclaimed himself emper­or in 202 bc, setting up his capital in Chang’an, he named his new dynasty after his fief at the Han river. The Han dynasty, which lasted (with a short interruption) for over four centuries, was a model for later rulers and eventually became epony­mous for the Chinese as an ethnic group.

漆  qī, (Sun 2).

*tshit, nc, ‘varnish’: 膠漆之材

漠/莫  mò, *mâk, nc, ‘expanse, region’: 狄之廣莫,於晉為都 (GY 7.6).

滿  mǎn, *mânʔ, vi, 1 ‘be full; swell’, a itr: 百姓虛而府庫滿 (Xun 10); 虛則 欹,中則正,滿則覆 (Xun 28); b caus, ‘fill (up), cover, permeate’: 在谷滿谷,在 阬滿阬 (Zhuang 14); 田車數百乘,從數 千,人滿野 (Mo 31); ‘satisfy’: 彼利太子 之戰攻,而欲滿其意者眾 (ZGC 32.7); 2 adv, ‘fully’: 故妾雖老,年未滿五十 (Li 12). Syn. 實 [40.11]; ant. 虛 [141.6]. 12

潔/絜  jié, *kêt, v , ‘be clean, spotless’, a i

itr: 其為人也,小心精潔 (GY 7.6); 君子 謂齊姜潔而不瀆 (LNZ 2.3); b caus: 欲潔 其身 (LY 18.7); 絜其居,美其服 (GY 20.1).

潰  kuì, *(g)wə̂s, v , ‘leak out, overflow; i

break out/up; scatter, desert’: 凡民逃其 上曰潰 (Zuo 6.3); a itr: 川壅而潰,傷人 必多 (GY 1.3); b caus: 秦王有病召 醫,破癰潰痤者得車一乘 (Zhuang 32).

夫為井,井間廣四尺,深四尺,謂之 溝 (ZL 85); 城郭溝池以為固(Li 9). Syn. 池 [85.3].

潤  rùn, *nuns, vi, ‘be moist, smooth, glossy’, a itr: 溫潤而澤,仁也 (Li 48); 玉 在山而草木潤 (Xun 1); b caus, ‘moisten, water’: 鼓之以雷霆,潤之以風雨 (Yi 7); fig. ‘add lustre (to), en­rich’: 富潤屋,德 潤身 (Li 42).

11

13

a river 漢東之 p 國,隨為大 (Zuo 2.6); 2 a dynasty: 今

sure, pump’, a tr: 激而行之,可使在

溝  gōu, *kô, n , ‘ditch, trench, moat’: 九 c

漢  Hàn, *hâns, n , 1

激  jī, *kiâuk, v , ‘dam up; apply pres-

475

tr

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

山(Meng 6A2); b pass: 夫弩弱而矢高 者,激於風也 (HFei 40); c mod: 激水之 疾,至于漂石者,勢也 (Sun 5).

14

濟  jì, *tsə̂i, v , 1 ‘ford, cross (a river); i

arrive at a goal, succeed’: a itr: 宋人既成 列,楚人未既濟 (Zuo 5.22); 昔我先君 駒王西討濟於河 (Li 4); 事雖大,必濟 (Zuo 8.9); b caus: 子產聽鄭國之政,以 其乘輿濟人於溱洧 (Meng 4B2); 開其 兌,濟其事,終身不救 (Lao 52); c pseudo-­tr: 濟河而西 (Li 19).

15

瀆  dú, *lôk, n , ‘moat, drainage ditch’: c

中古之世,天下大水,而鯀禹決瀆 (HFei 49).

18

灌  guàn, *kwâns, vn,‘pour; irrigate, water’, a tr: 秋水時至,百川灌河 (Zhuang 17); b itr: 既灌,然後迎牲 (Li 11). Classifier 86  火 0

火  huǒ, *hmə̂iʔ, nc, 1 ‘fire’: 水益 深,. . . 火益熱 (Meng 1B17); 火居南方 而主夏氣 (CQFL 42); ant. 水 [85.0]; 2 vi, ‘burn’, a itr: 新宮火,亦三日哭 (Li 4); b caus: 夏,成周宣榭火,人火之也 (Zuo 7.16). 5

為/爲  wéi, *wai, v , 1 ‘do’, a tr: 夫上之 n

所為,民之歸也 (Zuo 10.21); ‘make, create, contrive, produce’: 嬰兒 . . . 以塵為 飯,以塗為羹,以木為胾 (HFei 32); 初 楚子將以商臣為大子 (Zuo 6.1); ‘attend to’: 孔子為政必霸 (SJ 47); ‘regard as, consider’: 吾以夫子為天地 (Zhuang 5); 百 姓皆以我為愛 (Meng 1A7); b tr2: 顏淵 死,顏路請子之車以為之椁 (LY 11.8); 為之簞食與肉 (Zuo 7.2); cf. 5.2.2; c aux:

唯天為大 (LY 8.19); 苟為後義而先利 (Meng 1A1); 2 cop, a with C: ‘act as, be’: 孔 子為魯司寇 (Meng 6B6); 子路使門人為 臣 (LY 9.12); b with Oi and C: 原思為之 宰 (LY 6.5); 身為宋國笑 (HFei 49). Cf. Box 14 and 11.2.

為/爲  wèi, *waih, prep [< preceding],

1 ‘for’: 吾將為女問之 (Xun 29); ‘be­cause of, due to’: 獻公為之不寐 (CQFL 14); cf. 9.3; 2 vn, ‘act on behalf of/for the sake of; aim for’, tr: 我楚國之為 (Zuo 9.28); 天不 為秦國 (Lü 16.4).

6

烏  wū, *ʔâ, nc, ‘crow, raven’: 鵠不日浴 而白,烏不日黔而黑 (Zhuang 14). 7

焉  yān, *ʔan, adv [< 於 + pr

int], ‘how?’: 鶴實有祿位,余焉能戰 (Zuo 4.2); 未知 生,焉知死 (LY 11.12); ‘where?’: 人焉廋 哉 (LY 2.10); 君焉在 (Zuo 9.30). Cf. 8.2.5 and Box 16; syn. 安 [40.3].

焉  yán, *ʔan, fus [< 於 + pr

dem], ‘in/on/ from it; there’, final: 其君在焉 (Zuo 11.9); with pass, ‘by (s.o.)’: 君必欺焉 (HFei 18.47); in bracket con­structions: 於此有人 焉 (Meng 3B4). Cf. 8.2.5, Box 16, and Box 23.

8

然  rán, *nan, vi, ‘burn’, itr: 若火之始 然,泉之始達 (Meng 2A6); 兩木相摩而 然,金火相守而流 (HNan 1).

然  rán, *nan, vst [< 如 + prdem], ‘be so, like (sth.)’, a itr: 何以知其然也 (HFei 1); cf. 雖然 [172.9]; b caus: 沛公然其計,從 之 (SJ 8); c suffix for adv, ‘-ly’: 囂囂然曰 (Meng 5A7); d in bracket constructions: 聖人若天然 . . . 若地然 (Guan 37). Cf. 7.5, Box 16, and Box 23.

然而  ránér, *nan-­nə, conj, ‘however, neverthe­less’: 七十者衣帛食肉,黎民 不飢不寒,然而不王者,未之有也 (Meng 1A3).

476

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

然後  ránhòu, *nan-­ɦôʔ, conj, ‘(only)

then’: 事必成然後舉,身必安而後行 (YDan 1); cf. 16.3.

然則  ránzé, *nan-­tsə̂k, conj, ‘that being so, therefore, consequently’: 然則 君之所讀者,古人之糟魄已夫 (Zhuang 13); 然則是大臣執於朝,而列 陳之士執於賞也 (Guan 80).

無/无  wú, *ma, v , 1 ‘have not, lack’, a n

tr: 自今子其無事矣 (Zuo 10.3); 有說則 可,无說則死 (Zhuang 13); with locative S: ‘there is no’: 外無敵國之患,內無亂 臣之憂 (HFei 14); b itr: 人皆有兄弟,我 獨無 (LY 12.5); 爾有母遺,繄我獨無 (Zuo 1.1); 梁則無矣,麤則有之 (Zuo 12.13); 2 adv, ‘not’, negates VP: 殆無顧天下 之議之也 (SJS 1) . Syn. 亡 [8.1]; ant. 有 [74.2].

無  wú, *ma, pr

ind, ‘noth­ing, nobody’, as O, pre­ceding the P: 今老矣,無能為也已 (Zuo 5.30). Ant. 有 [74.2]; cf. 5.2.1.

無  wú,

*ma, nc, ‘non-­exist­ence, non-­ being’: 有無相生 (Lao 2). Ant. 有 [74.2].

無  wú, *mə, var. of 毋 [80.0].

漁  yú, *ŋa, vn [< 魚, 195.0], ‘fish, catch fish’, a itr: 作結繩而為罔罟,以佃以漁 (Yi 8); 故君子仕則不稼,田則不漁 (Li 30); b tr, fig. ‘squeeze, exploit’: 牧漁其民 以富其家 (Guan 67); 貪主暴君,涸漁 其下 (Wen 8); c mod (!): 命漁師伐蛟取 鼉 (Li 6); 此漁夫之勇悍也 (SY 11). 12

燕  Yān, *ʔên, n , a polity: 武王 . . .封召 p

公奭於燕 (SJ 4). Note: Yan, located in the north of modern Hebei (its capital, Ji 薊, was close to modern Bei­jing), was supposedly founded in the eleventh centu­ry bc as a fief for the Duke of Shao, a scion of the royal Ji 姬 clan. A powerful rival of Qi 齊, Yan became one of the seven big states of the Zhan­guo period; it was finally destroyed by Qin in 222 bc.

燒  shāo, *ŋhiau, v , ‘burn’, a itr: 不時焚 i

9

照/炤  zhào, *tiau, vn, ‘shine (on), illuminate’, a itr: 日月有明,容光必照焉 (Meng 7A24); b tr: 左光照右,右光照左 (SY 20). 10

熊  xióng, *wəm, n , ‘bear’: 宰夫胹熊蹯

不熟 (Zuo 7.2).

7.2); fig. 行端直則思慮熟,思慮熟則得 事理 (HFei 20) ; 秦王明而熟於計 (ZGC 4.15); b caus: 夫仁亦在乎熟之而已矣 (Meng 6A19); 中歲而熟之 (CQFL 77); 2 adv, ‘maturely, thoroughly, carefully’: 願 孰察之 (SJS 1); 願君熟思之 (HSWZ 10).

c

燒,無所臧之 (Xun 10); 藏寶之臺燒 (HSWZ 10); b caus: 因燒其券 (ZGC 11.1); 燒其軍,火三日不滅 (SY 1); c mod: 燒 荅覆之,沙石雨之 (Mo 63).

13

燧  suì, *s-­jus, n , ‘kindler, fire drill’: 有 c

聖人作,鑽燧取火 (HFei 49).

燭  zhú, *tok, n , 1 ‘torch, candle’: 夜行以 c

11

熱  rè, *ŋet, vst, ‘be hot’, a itr: 天久不 雨,髮將焦,身將熱 (SY 18); 夏寒冬 溫,春熱秋榮 (HSWZ 2); b comp: 水益 深,. . . 火益熱 (Meng 1B17); c mod: 朝寒 請進熱食 (SY 2).

熟/孰  shú/shóu, *duk, v , 1 ‘be ripe, i

mature; well cooked, done’, a itr: 實熟則 剝 (Zhuang 4); 宰夫胹熊蹯不熟 (Zuo

燭,無燭則止 (Li 12); 以人魚膏為燭 (SJ 6); 2 vn,‘shine on, illuminate’, tr: 燭之以日月 之明 (Zhuang 14); 日月無私燭也 (Lü 1.5).

Classifier 87  爪 4

爭/争  zhēng, *tsrêŋ, v

n, 1 ‘fight, struggle, com­pete (for)’, a itr: 不尚賢,使民不

477

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

爭 (Lao 3); 大臣爭於私而不顧其民 (SJS 14); b tr: 民爭利且不服 (Lü 20.2); 民猶爭利而忘義 (Li 30); 2 nc, ‘fight, struggle, quarrel’: 德蕩乎名,知出乎爭 (Zhuang 4).

8

爲  wéi/wèi, *wai/*waih, var. of 為 [86.5]. 13

爵  jué,

*tsiauk, nc, ‘goblet’: 臣侍君 宴,過三爵,非禮也 (Zuo 7.2); ‘nobility, dignity, rank’: 官爵貴重,朋黨又 眾,而一國為之訟 (HFei 11).

Classifier 88  父 0

父  fù, *baʔ, nc, ‘father’: 故父母之愛不 足以教子 (HFei 49); more generally, ‘elders’, meaning male relatives of elder generations: 視父兄與君若其身 (Mo 14). Ant. 子 [39.0], 母 [80.1]. - 大父  dàfù, *dâs-baʔ, n , ‘grandfac

ther’: 大父未死而有二十五孫 (HFei 49).

Classifier 89  爻 0

爻  yáo, *grâu, n , ‘line

(of Yijing tric gram or hexa­gram)’: 爻者,言乎變者也 (Yi 7); 六爻之動,三極之道也 (Yi 7); 爻 也者,效天下之動者也 (Yi 8). Cf. 32.1.

10

爾  ěr, *neʔ, pr

, ‘you’, a as S: 爾非吾君 也 (SY 9); 唯我與爾有是夫 (LY 7.11); b as O: 吾將射爾 (LNZ 5.6); 我無爾詐,爾 無我虞 (Zuo 7.15); c mod: 走者爾母耶 (LNZ 5.6). Cf. Box 5. per

Classifier 90  爿 13

(Zuo 7.2).

11

牖  yǒu, *luʔ/juʔ, n , ‘window’: 不出於 c

戶而知天下,不窺於牖而知天道 17.2).

(Lü

Classifier 93  牛 0

牛  niú, *ŋwə, n , ‘ox’: 有牽牛而過堂下 c

者 (Meng 1A7). Cf. note at 畜 [102.5].

4

牧  mù, *mək, vtr, ‘shepherd, tend (cattle)’, a tr: 胡人不敢南下而牧馬 (XS 1.1); metaphorical: 此聖王之所以牧臣下也 (HFei 44); b pass: 民之不牧者,非吾民 也 (Guan 3). 物  wù, *mət, nc, ‘thing, creature’: 天地 與我並生,而萬物與我為一 (Zhuang 2); ‘sign, mark, symbol’: 叔孫氏之甲有 物 (Zuo 11.10); 夫天下之為聞見鬼神之 物者,不可勝計也 (Mo 31).

牽  qiān, *khîn, vn,‘pull/lead by a rope’, a tr: 有牽牛而過堂下者 (Meng 1A7); b itr: 故 從山上望牛者若羊,而求羊者不下牽 也 (Xun 21); 王乃牽而上殿 (Zhuang 30). 7

牾  wǔ/wù, *ŋâh, var. of 忤 [61.4]. Classifier 94  犬 0

犬  quǎn, *khwînʔ, n , ‘dog’: 狗,犬也 c

(Mo 10); 與犬肉 (GY 8.1). Syn. 狗 [94.5]; dist. 大 [37.0], 太 [37.1]; cf. note at 畜 [102.5].

2

牆  qiáng, *dzaŋ, n , ‘wall’: 厚斂以彫牆 c

Classifier 91  片

犯  fàn, *bamʔ, vn, ‘oppose, resist; offend against’, a tr: 不好犯上,而好作亂 者,未之有也 (LY 1.2); 君子犯義,小人

478

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

犯刑 (Meng 4A1); b itr: 上不酌民言,則 犯也 (Li 30).

13

4

one’s own’, esp. ‘old and childless’: 老而無子 曰獨 (Meng 1B5); a itr: 故君子必慎其獨也 (Li 42); b caus, ‘treat as one’s own, have for oneself’: 姓所同也,名所獨也 (Meng 7B36); 2 adv, ‘alone, only’: 人皆有兄弟,我 獨無 (LY 12.5); cf. 7.3; ‘possibly’, em­pha­tic, with Print: 我獨何能無槪然 (Zhuang 18); 宗不余辟,余獨焉辟之 (Zuo 9.28).

狄/翟  Dí,

*lêk, np, a non-­Chinese people: 雖之夷狄,不可棄也 (LY 13.19); 夷翟遇之 (SJ 5). Cf. note at 夷 [37.3].

5

狐  hú, *gwâ, n , ‘fox’: 虎不知獸畏己而 c

走也,以為畏狐也 (ZGC 14.3).

狗  gǒu, *kôʔ, n , ‘dog’: 狗,犬也 (Mo c

10); 歸公乘馬, 祭服五稱, 牛羊豕雞 狗皆三百 (Zuo 4.2); 狗馬實外廄 (ZGC 11.1). Syn. 犬 [94.0]; cf. note at 畜 [102.5].

7

狼  láng, *râŋ, n , ‘wolf ’: 吾國 . . . 虎狼之 c

所棲也 (Guan 80).

狹  xiá, *grêp, v , ‘be narrow, confined’, a st

itr: 君子之路,行止之道固狹耳 (YTL 31); b caus: 若欲狹之,乃是離之,若欲 飾之,乃是賊之 (Wen 8); c comp: 使其妻 織組而幅狹於度 (HFei 34). Ant. 廣 [53.12].

獨  dú, *dôk, v , 1 ‘be solitary, alone, on i

14

獲  huò,

*gwrâk, vn, ‘capture, catch; obtain’, a tr: 獲驪姬以歸 (GY 7.3); 獲罪 於天,無所禱也 (LY 3.13); b itr: 獲於上 有道 (Meng 4A12).

15

獵  liè, *rap, vn, ‘hunt’, a itr: 今王田獵於 此 (Meng 1B1); b tr: 田獵禽獸者,野虞 教道之 (Li 6).

獸  shòu, *hjuh, n , 1 ‘animal, beast’: 山林 c

者,鳥獸之居也 (Xun 14); 2 adv, ‘like beasts’: 獸處群居(Guan 31). Cf. 禽獸 [114.8].

16

8

猛  měng, *mrâŋʔ, v , ‘be wild’, a itr: 子 獻  xiàn, *hŋans, v , 1 ‘give, present (to), n

st

溫而厲,威而不猛,恭而安 (LY 7.38); 棄人用犬,雖猛何為 (Zuo 7.2); b comp: 小子識之,苛政猛於虎也(Li 4); c mod: 周公兼夷狄,驅猛獸,而百姓寧 (Meng 3B9).

offer’, a tr: 列士獻詩,瞽獻曲,史獻書 (GY 1.3); 士有獻於國君 (Li 2); b itr: 奉犧 牲玉鬯往獻焉 (GY 1.12); 持鹿獻於二世 (SJ 6); 2 np, ‘Exemplary’, posth. name: 獻公 為之不寐 (CQFL 14).

9

Classifier 95  玄

猶  yóu, *ju, v , 1 cop, ‘be like’: 性猶湍水 n

也 (Meng 6A2); 2 adv, ‘still, yet’: 當堯之 時,天下猶未平 (Meng 3A4); cf. 10.5; 2 conj. ‘but, but still, yet’: 雖亡子,猶不亡 族 (Zuo 6.16); 民猶以為小也 (Meng 2B12); ‘if even’: 臣猶知之,而況君乎 (GY 10.1); cf. 16.2.

11

獒  áo, *ŋâu, n , ‘hound, hunting dog’: c

公嗾夫獒焉 (Zuo 7.2).

0

玄  xuán,

*gwîn, vst, ‘be dark, black; obscure, mysterious’, a itr: 玄之又玄 (Lao 1); b caus: 故先王明之,豈特玄之耳哉 (Xun 18).

6

率  shuài,

*srut, vn, ‘go along, follow’, tr: 周仁之謂信,率義之謂勇 (Zuo 12.16); 率性而行謂之道 (HNan 11).

479

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

率  shuài, *sruts, v , 1 ‘lead (s.o. to do

sth.)’, tr: 此率獸而食人也 (Meng 1A4); 越 王句踐率其眾以朝吳 (SJ 31); 里克率國 人以攻殺之 (Lü 23.6); 2 var. of 帥 [50.6].

亡焉 (Xun 23); 2 vtr, ‘bring order to, manage, arrange’, a tr: 堯理天下 (Lü 20.2); 禮 也者 . . . 理萬物者也 (Li 10); b pass: 聖人 之所在,則天下理焉 (Lü 4.2); 聖人寡 為而天下理矣 (SY 1).

Classifier 96  玉

11

n

0

王  wáng, *waŋ, n , ‘king’: 其曰王者,民 c

之所歸往也 (GLiang 3.3); 文王視民如傷 (Meng 4B20); 田贊衣補衣而見荊王 (Lü 15.5). Cf. Box 15 and note at 主 [3.4].

王  wàng, *waŋh, v

i [< preceding], ‘be king(ly), rule (as/like a king)’, a itr: 保民 而王 (Meng 1A7); 伊尹相湯以王於天下 (Meng 5A6); b caus, ‘make/treat as king’: 今可以王齊王而壽黔首之命 (Lü 21.5); 尉佗平南越,因王之 (SY 12); c pseudo­tr: 使王天下 (HFei 49).

玉  yù, *ŋok, n , ‘jade, nephrite’: 宋人或 得玉 (Zuo 9.15).

c

璋  zhāng, *taŋ, var. of 彰 [59.11]. 13

環  huán, *wrên, nc, 1 ‘bracelet; ring’: 宣 子有環,其一在鄭商 (Zuo 10.16); 2 vn, ‘circle, whirl, re­volve (around); encircle, sur­round’, a itr: 水之性…倚則環,環則 中 (Guan 57); b tr: 慶氏以其甲環公宮 (Zuo 9.28); 三江環之,民無所移 (GY 20.1). Classifier 99  甘 0

甘  gān, *kâm, v , ‘be sweet, delicious’, st

4

玩  wán, *ŋwâns, vtr, ‘toy, trifle, amuse oneself (with); appreciate, enjoy; ponder’, a tr: 所樂而玩者,爻之辭也 (Yi 7); 其 知甚少,而子玩之 (SY 2); b pass: 夫兵 不可玩,玩則無威 (SY 15); c mod: 君之 玩物,衣以文繡 (SY 14).

a mod: 直木先伐,甘井先竭 (Zhuang 20); b caus: 甘其食,美其服 (Lao 80). Ant. 苦 [140.5].

4

珍  zhēn, *trən, v , ‘be precious, exqui-

甚  shèn, *dəmʔ/s, vst, 1 ‘be serious, terrible, bad; extreme, great’, a itr: 甚矣汝之 不惠 (Lie 5); b caus: 天王殺其弟佞 夫,甚之也 (GLiang 9.30); c comp: 防民 之口,甚於防川 (GY 1.3); 事之彌 煩,其侵人愈甚 (Xun 10); d pseudo-­tr: 其助甚此矣 (HFei 30); 其不義又甚入人 園圃竊桃李 (Mo 17); 2 adv, ‘very, extremely’: 吾樂甚多 (SY 17).

6

Classifier 100  生

5

st

site’, a itr: 其言神而珍 (SY 11); b caus: 書曰,公子遂,珍之也 (Zuo 6.8); 寶之,珍之,貴之,神之 (Xun 5); c mod: 不愛珍器重寶肥饒之地 (XS 1.1). Syn. 寶 [40.17].

珠  zhū,

*to, nc, ‘pearl’: 含而不見於 口,而辟千金者,珠也 (Guan 80); 珠 足以御火災,則寶之 (GY 18.7).

7

理  lı̌, *rəʔ, n , 1 ‘line; pattern, structure; c

princi­ple; order’: 故淫亂生而禮義文理

0

生  shēng, *sreŋ, nc, 1 ‘life, be­ings’: 天地 者,生之始也 (Xun 9); 死生有命,富 貴在天 (LY 12.5); 羣生皆得其命 (Xun 9); 2 vi, ‘be born; live, exist; grow’, a itr: 文 王生於岐周,卒於畢郢 (Meng 4B1); 寡

480

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

人生於深宮之中 (Xun 31); 有桑穀生於 庭 (SY 1); b caus, ‘give birth (to), beget, produce’: 道生一,一生二,二生三, 三生萬物 (Lao 42); c pseudo-­tr: 孔子生 魯昌平鄉陬邑 (SJ 47); d mod: 氣衰則生 物不遂(Li 19); 生人立焉,死人入焉 (SY 2); but e tr2 (!): 驪姬既遠大子,乃生 之言 (GY 7.6); 我有圃,生之杞乎 (Zuo 10.12). Ant. 死 [78.2].

6

産  chǎn, *srânʔ/srênʔ, vtr, 1 ‘breed, produce; bear’, a tr: 父母之於子也,產男則 相賀,產女則殺之 (HFei 46); b pass: 陳 良,楚產也 (Meng 3A4); 有名產於無 名,無名者有名之母也 (Wen 1); 2 nc, ‘(agricultural) produce, products’: 若 民,則無恆產,因無恆心 (Meng 1A7); 民農則其產復,其產復則重徙 (Lü 26.3). Classifier 101  用

甲  jiǎ, *krâp, n , ‘armour’: 可使制梃以 c

撻秦楚之堅甲利兵矣 (Meng 1A5); by extension ‘ar­moured soldier’: 秋九 月,晉侯飲趙盾酒,伏甲,將攻之 (Zuo 7.2).

甲  jiǎ, *krâp, ntemp, first of the ‘Heavenly stems’: 是故春三月,以甲乙之日發五 政 (Guan 40). Cf. Focus 8. 申  shēn,

*lhin, ntemp, ninth of the ‘Earthly branches’: 壬申,公薨於高寢 (SY 19). Cf. Focus 8.

由  yóu, *ju, v , 1 ‘proceed from, go along, n

follow’, tr: 誰能出不由戶 (LY 6.17); 塗有 所不由, 軍有所不擊 (Sun 8); 2 prep, ‘from’, a local: 水由地中行 (Meng 3B9); 由 君子觀之 (Meng 4B33); b temp: 由湯至於 武丁 (Meng 2A1); 由九月以下何也 (Xun19); c with print: 何由知吾可也 (Meng 1A7); 奚由相得 (Lü 16.3); d with prind: 無由 相得 (Lü 16.3); 有由然也 (Zuo 12.7).

2

0

用  yòng, *loŋh [< 庸, 53.8], v , 1 ‘use, tr

utilize, employ, de­ploy’, a tr: 不教民而用 之 (Meng 6B8); 靈公老,怠於政,不用 孔子 (SJ 47); b pass: 上下相親,兵刃不 用矣 (Guan 53); 今道不用於齊 (LNZ 1.11); c pseudo-­tr: 其用戰也勝,久則鈍 兵挫銳 (Sun 2); 2 prep, ‘with, by means of; according to’: 用此觀之,然則人之 性惡明矣 (Xun 23); 故謀用是作,而兵 由此起 (Li 9); syn. 以 [9.3]; 3 nc, ‘asset, means’: 財用於是乎出 (GY 1.3); 久暴師 則國用不足 (Sun 2).

Classifier 102  田 0

田  tián,

*lîn, nc, ‘field (planted with crops)’: 回有郭外之田五十畝 (Zhuang 28); ant. 野 [166.4].

田  tián/diàn, *lîns, var. of 佃 [9.5].

男  nán, *nə̂m, n , 1 ‘man; boy’: 男女居 c

室,人之大倫也 (Meng 5A2); ant. 女 [38.0]; 2 lowest of the five aristo­ cratic ranks of the Zhou, ‘baron’: 王者之制祿 爵,公侯伯子男,凡五等 (Li 5). Cf. Box 15.

4

界  jiè, *krê(t)s, n , ‘border’: 長城之界 c

(SJ 69).

畏  wèi, *ʔuih, v

n [< 威, 38.6], ‘(be in) fear, dread; be in awe (of), re­spect’, a itr: 未賞而民勸,未罰而民畏 (Lü 20.2); 子 畏於匡(LY 9.5); b tr: 君子有三畏,畏天 命,畏大人,畏聖人之言 (LY 16.8); 吾 聞忠不畔上,勇不畏死 (HSWZ 9).

畏  wēi, *ʔui, var. of 威 [38.6]. 5

畚  běn, *pə̂nʔ, n , ‘basket’: 殺之,寘諸 畚 (Zuo 7.2).

481

c

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

留  liú, *ru, v , ‘stay, remain’, a itr: 願留而 i

受業於門 (Meng 6B2); 辛請留於趙以觀 之 (XX 2.15); b caus, ‘retain, hold back, stop, hinder’: 寧其死為留骨而貴乎 (Zhuang 17); 孟子謝,遂留其婦 (LNZ 1.11); 君行矣,無留吾事 (XX 7.1); c pseudo-­tr: 地辟舉則民留處 (Guan 1).

物皆得其宜 (Xun 9); 2 prep, ‘in, on, at; when’, a temp.: 當堯之時,天下猶未平 (Meng 3A4); 顏子當亂世,居於陋巷 (Meng 4B29); b local: 當其馬前曰 (Lü 19.1); 既歌而入,當戶而坐 (Li 3).

畔  pàn, *bâns, var. of 叛 [29.7]. 畜  xù, *huk, v , ‘breed, rear’, a tr: 仰足

當  dàng, *tâŋh, vst, ‘be right, proper, adequate’: a itr: 會計當而已矣 (Meng 5B5); 其言多當矣,而未諭也,其行多 當矣,而未安也,其知慮多當矣,而 未周密也 (Xun 8); b aux, ‘should, ought to’: 今臣使楚,不當從此門入 (Yan 6.9); 吾聞二世少子也,不當立,當立者乃 公子扶蘇 (SJ 48).

畜  chù,

14

tr

以事父母,俯足以畜妻子 (Meng 1A7); 子產使校人畜之池 (Meng 5A1); b pass: 安職則畜,不安職則棄 (Xun 9); c mod: 仲尼之畜狗死 (Li 4).

*rhuk, nc, [< the preceding] ‘livestock, domestic animal’: 六畜皆得其 長 (Xun 9); 嘗為司職吏而畜蕃息 (SJ 47). Note: The ‘six kinds of do­mestic animals’, 六畜, were oxen, horses, sheep, chicken, dogs, and pigs.

6

疆  jiāng, *kaŋ, nc, ‘boundary, border’: 入其疆,土地辟,田野治 (Meng 6B7); 蒲與二屈,君之疆也 (GY 7.6). Classifier 103  疋 7

畢  bì, *pit, v , 1 ‘finish, be complete’, a itr: 疏  shū, *sra, var. of 蔬 [140.12]. 王者之事畢矣 (Xun 9); 三年之喪,二 i

十五月而畢 (Li 49); 天數畢於十 (CQFL 24); b caus: 乃畢山川之祀 (Li 6); c pseudo-­tr: 天下之道畢是矣 (Xun 8); 2 adv, ‘all, complete­ly’: 國人畢作 (Li 11); 天 瑞並至,徵祥畢見 (SY 11); 責畢收乎 (ZGC 11.1).

異  yì, *ləkh, v , 1 ‘be different, strange’, st

a itr: 吾黨之直者異於是 (LY 13.18); 殺人 以梃與刃,有以異乎 (Meng 1A4); b caus: 人皆謂之長人而異之 (SJ 47); 王勿 異也 (Meng 5B9); c mod: 諸侯各愛其 國,不愛異國 (Mo 14); 2 nc, ‘difference; peculi­arity, sth. unusual’: 吾以子為異之 問 (LY 11.22). Ant. 同 [30.3].

8

當  dāng, *tâŋ, v , 1 ‘be

apposite (to), n match; face, attend to, confront; oppose’, a tr: 辨而不當論,信而不當理,勇而不 當義,法而不當務 (Lü 11.4); 周公旦代 成王治,當國 (SJ 37); b itr: 羣道當則萬

9

疑  yí, *ŋə, v ,‘doubt, distrust; be unsure, n

hesi­tate; impugn, throw doubt upon’, a tr: 仁者無敵,王請勿疑 (Meng 1A5); 盛容 服而飾辯說,以疑當世之法 (HFei 49); b itr: 周人作會而民始疑 (Li 4); 眾 疑,赦之 (Li 5). Ant. 信 [9.7].

Classifier 104  疒 5

病  bìng, *braŋh, n , 1 ‘disease, illness’: 吾 c

王庶幾無疾病與 (Meng 1B1); 2 vst, ‘be ill; trouble­some’, a itr: 今日病矣, 予助苗長 矣 (Meng 2A2); 以逆來者,於我未病 (Zuo 9.9); b caus, ‘think ill of, be disturbed by’: 君子病無能焉,不病人之不己知也 (LY 15.19); 視不明, 聽不聰, 行不正, 不知哀, 君子病之 (Li 21); c mod: 且夫 良醫之門多病人 (Xun 30). Syn. 疾 [104.5].

482

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

疾  jí, *dzit, v , 1 ‘be ill, sick; suffer’, a itr: st

昔者疾,今日愈 (Meng 2B11); b caus, ‘think ill of, wish ill, hate’: 百姓疾之 (Xun 10); 夫子疾由之為仁義乎 (HFei 34); c mod: 上醫醫國,其次疾人 (GY 14.17); 2 nc, ‘ailment, pain, disease’: 吾王庶幾無疾 病與 (Meng 1B1). Syn. 病 [104.5].

疾  jí, *dzit, v , ‘be quick, fast; industrist

ous; in­tense’, a itr: 疾則苦而不入 (Zhuang 13); 然則往速徠疾 (SJS 2); 我疾 農,先實公倉 (SJS 3); b comp: 順風而 呼,聲非加疾也,而聞者彰 (Xun 1); 2 adv: 疾諷誦,謹司聞 (Lü 4.3). Syn. 速 [162.7]; ant. 徐 [60.7]; 遲 [162.12].

疾  jí, *dzit, var. of嫉 [38.10]. 疲/罷  pí, *bai, v , 1 ‘be exhausted, i

weary; feeble(-minded), incapable’, a itr: 民人日駭,勞罷死轉 (Zuo 10.19); 數戰 則民疲,數勝則主驕 (HSWZ 10); b caus: 彼躁我靜,則是疲其力也 (HNan 15); 2 nc, ‘incompetent (person)’: 罷不能 不待須而廢 (Xun 9).

1

百  bǎi/bó, *prâk, vi, 1 ‘be a hundred, hundred­fold; numerous’, a itr: 利不 百,不變法 (SJS 1); b caus: 人一能之己 百之 (Li 31); c mod: 秋水時至,百川灌 河 (Zhuang 17); 2 nc, ‘a hundred’: 千取百 焉,不為不多矣 (Meng 1A1). Note: 百, used as mod in expressions like 百家, 百 姓, 百官, 百工, etc., simp­ly denotes plurality: ‘all (the many)’; cf. also Box 3. 一百姓  bǎixìng, *prâk-­seŋh, nc, ‘the hundred clans, aristocracy’: 今吾欲變法 以治,更禮以教百姓 (SJS 1); thence ‘the people’: 四民均則王道興而百姓寧 (SY 7). 4

皆  jiē, *krî, adv, ‘every; all’, a referring to

the S: 人皆有兄弟,我獨無(LY 12.5); 眾 人皆醉,我獨醒 (ChuC 7); 四海之內皆 兄弟也 (LY 12.5); b re­ferring to the O: 宣 子皆獻馬焉 (Zuo 10.16). Cf. 7.3.

皇帝  huángdì, *gwâŋ-­têh, n , ‘emper­or’, c

Classifier 105  癶 4

癸  guı̌, *kwiʔ, ntemp, tenth of the ‘Heavenly stems’: 以壬癸殺黑龍於北方 (Mo). Cf. Focus 8. 7

title created upon the founding of the empire in 221 bc: 秦王政立二十六 年,初并天下為三十六郡,號為始皇 帝 (SJ 5). Cf. Box 15.

Classifier 107  皮 0

登  dēng, *tə̂ŋ, v , ‘climb, mount’, a tr: 孔 皮  pí, *bai, n , ‘hide, fur, leather’: 婦人 子登東山而小魯,登泰山而小天下 n

c

(Meng 7A24); b itr: 黃帝遊乎赤水之 北,登乎崑崙之丘 (Zhuang 12).

Classifier 106  白 white, clear’, a itr: 鵠不日浴而白,烏不日黔而黑 (Zhuang 14); b caus: 有知其…白白而黑 黑也 (CQFL 1); c comp: 身死而名彌白 (Xun 4). Syn. 素 [120.4]; ant. 黑 [203.0]. st

Classifier 108  皿 4

0

白  bái/bó, *brâk, v , ‘be

不織,禽獸之皮足衣也 (HFei 49).

盈  yíng, *leŋ, vi, a itr, ‘be full, abundant’: 不務地利則倉廩不盈 (Guan 1); 爵盈而不飲(Zuo 10.5); b caus, ‘fill up’: 舉先王言仁義者盈廷,而政不免於亂 (HFei 49). Ant. 竭 [117.9], 盡 [108.9], 虛 [141.6].

483

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

盆  pén, *bə̂n, n , ‘pot, bowl’: 莊子則方 c

箕踞鼓盆而歌 (Zhuang 18).

5

盍  hé, *gâp, adv [< 胡不], ‘why not?

how not?’, as encouragement: 二子盍相 與觀乎南山之陽有木焉 (SY 3.6); 盍致 地焉 (SJ 47). Cf. 胡 [130.5] and Box 16.

益  yì, *ʔek, v , 1 ‘be beneficial, advantai

geous’, a itr: 友直,友諒,友多聞,益 矣 (LY 16.4); b caus, ‘add, increase, benefit’: 家務相益,不務厚國 (HFei 6); 2 adv, ‘increasingly, more’, for comp of vst: 主上愈 卑,私門益尊 (HFei 11); 水益深,. . . 火 益熱 (Meng 1B17); 3 nc, ‘profit, gain, advantage’: 吾嘗終日不食,終夜不 寢,以思,無益,不如學也 (LY 15.31).

6

盛  shèng, *deŋh, v , ‘be full; abundant, st

plentiful, flourishing’, a itr: 樹木盛,則 飛鳥歸之 (Lü 2.5); 爝火甚盛 (Lü 19.8); b caus: 王者盛其德而遠人歸 (SY 5); c comp: 自生民以來,未有盛於孔子也 (Meng 2A2); 德彌盛者文彌縟 (SY 19); b mod: 晨往,寢門闢矣,盛服將朝 (Zuo 7.2). Ant. 衰 [145.4].

7

盜  dào, *dâuh, n , ‘robber, bandit’: 竊人 c

之財, 猶謂之盜, 況貪天之功以為己 力乎 (Zuo 5.24); 不貴難得之貨,使民 不為盜 (Lao 3). Syn. 竊 [116.18].

8

盟  méng, *mraŋ, nc, ‘covenant, oath (of alliance)’: 藏於盟府 (Zuo 5.5). 9

監  jiān, *krâm, v , ‘observe, oversee’, a n

tr: 得衛巫,使監謗者 (GY 1.3); b itr: 以 遺後之人,使永監焉 (GY 3.9).

盡  jı̌n,

*tsinʔ, vn, ‘fully use, entirely devote; treat fully, ex­haust’, a tr: 無盡百姓 之勞 (ZGC 19.4); 書不盡言,言不盡意

(Yi 7); b itr: 去惡莫如盡 (Zuo 12.1); 殷周 之樂,盡矣 (Li 19). Syn. 竭 [117.9], 虛 [141.6]; ant. 盈 [108.4].

盡  jìn,

*dzinʔ, vi [< preceding] 1 ‘be exhausted’, itr: 五戰而三勝,陣卒盡矣 (SJ 70); 2 adv, ‘all, en­tirely, fully, utterly’, a referring to the S: 人其盡死,而我獨存 乎 (Zhuang 11); b referring to the O: 不盡 知用兵之害者,則不能盡知用兵之利 也 (Sun 2); 盡食其肉 (HSWZ 7); cf. Box 30. Syn. 竭 [117.9], 虛 [141.6]; ant. 盈 [108.4].

12

盪  dàng, *lâŋh, var. of 蕩 [140.12]. Classifier 109  目 0

目  mù,

*muk, nc, ‘eye’: 屏王之耳 目,使不聰明 (Zuo 10.27); 目 . . . 不能自 見其睫 (HFei 21); ‘de­tail’: 請問其目 (LY 12.1).

3

直  zhí,

*drək, vi, 1 ‘be straight; upright, honest’, a itr: 以為直於君而曲 於父 (HFei 49); b caus: 以直其政,方 其義也 (Li 39); c mod: 直木先伐,甘 井先竭 (Zhuang 20); 2 nc, ‘straight­ ness’: 於曲之中,必見其直;於直之 中,必見其曲 (CQFL 20). Ant. 輮 [159.9]; 曲 [73.2].

4

省  shěng, *sreŋʔ, vtr, ‘be sparing, reduce, dimin­ish’, a tr: 故省刑之要,在禁文巧 (Guan 1); 夫楚王好小腰,而美人省食 (Guan 52); b pass: 是故刑罰省而威流 (Xun 15). 省  xı̌ng,

*seŋʔ, vn, ‘examine, inspect, scrutinize’, a tr: 吾日三省吾身 (LY 1.4); 退而省其私,亦足以發 (LY 2.9); b itr: 君子博學而日參省乎己 (Xun 1).

484

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

相  xiāng,

*saŋ, adv, ‘each other, one another’, for the O: 性相近也,習相遠也 (LY 17.2); 人與人不相愛則必相賊 (Mo 15); 父子相食 (SJ 120); 剛柔相摩,八卦 相盪 (Yi 7); cf. 7.5–㉕.

相  xiàng, *saŋh, nc, 1 ‘chancellor’: 相 者,論列百官之長,要百事之聽 (Xun 11); 孟嘗君為相數十年 (ZGC 11.1); 2 vn ‘be chancellor’, a tr: 管仲相桓公,霸諸 侯 (LY 14.17); 孔子相魯 (SJ 39); b itr: 曰 頰谷之會,孔子相焉 (GLiang 11.10). 5

真/眞  zhēn, *tin, v , ‘be true, real’, a i

itr: 其為人也真 (Zhuang 21); 其知情 信,其德甚真 (Zhuang 7); b caus: 真其 義也,真真情也,乃以為名 (CQFL 36); c mod: 且有真人,而後有真知 (Zhuang 6).

6

眾/衆  zhòng, *tuŋh, v , 1 ‘be numerst

ous, many’, a itr: 上古之世,人民少而 禽獸眾 (HFei 49); b caus: 人民寡則從事 乎眾之 (Mo 25); c comp: 魯之群室,眾 於齊之兵車 (Zuo 12.11); 其人彌眾,其 虧彌大矣 (HFei 20); d mod: 眾妙之門 (Lao 1); cf. Box 28; 2 nc, ‘mass, multitude’: 成大功者,不謀於眾 (ZGC 19.4).

8

睹  dǔ, *tâʔ, v , ‘see’, tr: 臣不睹臣之母 n

三年矣 (Lü 9.5); 未睹巧之久也 (Sun 2). Syn. 見 [147.0].

睫  jié, *tsap, n , ‘eyelash’: 目…不能自 c

見其睫 (HFei 21).

睦  mù, *mruk, v , 1 ‘get along well, be in i

con­cord’, a itr: 吾兄弟之不協,焉能怨 諸侯之不睦 (Zuo 5.22); 諸侯方睦於晉 (Zuo 10.18); b caus: 以正君臣,以篤父 子,以睦兄弟,以和夫婦 (Li 9); 2 nc, ‘agreement, concord, understanding’: 選 賢與能,講信修睦 (Li 9). Syn. 和 [30.5], 友 [29.2].

12

瞷  xián, *grên, vn, ‘peek, spy out’, tr: 吾 將瞷良人之所之也 (Meng 4B33). 13

瞽  gǔ, *kâʔ, nc, ‘blind man’: 兩瞽相 扶,不觸墻木,不陷井阱,則其幸也 (HSWZ 5); ‘musi­cian’: 史為書,瞽為詩 (Zuo 9.14). Note: In ancient times, blind men were supposedly employed as official musicians because of their acute hearing, hence the double meaning of the word. Classifier 110  矛 4

矜  jīn, *grin, vn, ‘be arrogant, self-­ important; flaunt, brag, boast; praise’, a itr: 不矜而莊 (Li 32); b tr: 夫人主年少而 矜材 (ZGC 17.9); 今矜狄之善,其志益 廣 (GY 8.1). 矜  guān, *kwrə̂n, var. of 鰥 [195.10]. Classifier 111  矢 0

矢  shı̌, *lhiʔ, nc, ‘arrow’: 弦木為弧,剡 木為矢 (Yi 8); 執朱弓,挾朱矢,追周 宣王,射之車上 (Mo 31). 2

矣/已  yı̌, *ləʔ, pt, marks perfective aspect, a fi­nal: 臣不睹臣之母三年矣 (Lü 9.45); 吾以女為死矣 (LY 11.21); 天下皆 知美之為美,斯惡已 (Lao 2); 我非 子,固不知子矣 (Zhuang 17); cf. 10.4; b with inversion: 甚矣,吾衰也 (LY 7.5); 久矣哉,由之行詐也 (LY 9.12); cf. 14.3. 3

知  zhī, *tre, vn,‘be knowledgeable, know (a­bout), recognize, under­stand; a tr: 不出 戶而知天下 (HSWZ 3); with O clause: 知

485

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

國之將亡也 (SY 13); ‘know how to (do sth.)’; with inf. O: 惠而不知為政 (Meng 4B4); 古之真人,不知說生,不知惡死 (Zhuang 6); cf. 4.4.1; b itr: 豈不愚而知矣 (Xun 6); 乾以易知,坤以簡能 (Yi 7); c mod (!): 知農不離其故事,則草必墾矣 (SJS 2). Ant. 愚 [61.9]; syn. 叡 [29.14].

知  zhì, *treh, var. of 智 [72.8]. 5

矩  jǔ, *kwaʔ, nc, ‘right measure’:七十而 從心所欲,不踰矩 (LY 2.4).

0

示  shì, *gih, v , [caus < 視, 147.5], ‘show, n

demon­strate’: a tr2: 於是乎大蒐以示之 禮 (Zuo 5.27); 示民有常 (Li 9); b only Oi: 邦之利器不可以示人 (HFei 21); c only Od: 陰相善而陽相惡,以示無私 (HFei 17); d itr: 大史書曰趙盾弒其君,以示 於朝 (Zuo 7.2).

3

社  shè, *daʔ, n , ‘altar of (the god of) c

7

短  duǎn, *tônʔ, vst, ‘be short, small, low’, a itr: 晏子短 (SY 12); 陰陽長短 (Li 24); b  caus: 天短之而人養之者,其短益 (CQFL 77); c comp: 莫長於是,莫短於 是 (Mo 41). 12

矯  jiǎo, *kauʔ, vn, ‘bend; rectify; distort, falsify’, a tr: 矯重人之姦行 (HFei 11); 起 矯命以責賜諸民 (ZGC 11.1); b itr: 為上 而矯,悖也 (Guan 30). Classifier 112  石 0

石  shí, *dak, nc, ‘stone, rock’: 嘗與汝登 高山,履危石 (Zhuang 21); a weight measure, ‘picul’, c. 30 kg: 楚國之法,得 伍胥者賜粟五萬石 (SJ 66). 5

破  pò, *phâih, v , ‘destroy, break’, a tr: 燒 tr

山林, 破增藪, 焚沛澤, 逐禽獸, 實 以益人 (Guan 78); 大敗知伯,破其首 以為飲器 (HNan 12); b pass: 卵破子死 (Xun 1); c mod: 全國為上,破國次之 (Sun 3).

15

Classifier 113  示

礪  lì, *rats, nc, ‘whetstone’: 故木受繩則 直,金就礪則利 (Xun 1).

the soil’: 諸侯自立社 (Li 23).

-社稷  shèjí, *daʔ-­tsək, nc, ‘altars of (the gods of) soil and millet’, symbolizing the state: 諸侯不仁,不保社稷 (Meng 4A3); 民為貴,社稷次之,君為輕 (Meng 7B14); 代位不忘 社亡社稷,君 之道也 (SJS 1). 5

神  shén, *m-­lin, nc, 1 ‘ghost, spirit, deity’: 山林川谷丘陵,能出雲為風雨,見怪 物,皆曰神 (Li 23); 神者生之本也 (SJ 130); 致孝乎鬼神 (LY 8.21); 2 vst, ‘be divine, delightful’, a itr: 其言神而珍 (SY 11); b caus, ‘deify, worship’: 寵神其祖,以 取威於民 (GY 18.1); 寶之,珍之,貴 之,神之 (Xun 5). Cf. note at 鬼 [194.0]. 神農  Shénnóng, *m-­lin-­nûŋ, np,‘Divine Farm­er’, a mythical ruler: 伏羲神農教而不 誅 (SJS 1); 神農教耕而王天下 (SJS 6); 包 犧氏沒,神農氏作 (Yi 8). Cf. Fo­cus 7.

祗  zhī, *ke, vn, ‘venerate, respect, honour’, a tr: 案飾其辭而祗敬之 (Xun 6); 不 祗山川則威令不聞 (Guan 1); b itr: 父不 慈,子不祗,兄不友,弟不共,不相 及也 (Zuo 5.33). Syn. 敬 [66.9].

祖  zǔ, *tsâʔ, n , ‘ancestor’: 先祖者,類 c

之本也 (Xun 19); 敬宗廟,恭祖舊 (Guan 1); 天地者,萬物之本,先祖之 所出也 (CQFL 33).

祐  yòu, *wəh, var. of 佑 [9.5].

486

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

6

*s-­ jaŋ, vst, ‘be auspicious’, mostly ne­gated, a itr: 其人而不教,不祥 (Xun 27); 背天不祥 (Zuo 5.15); b comp: 離則不祥莫大焉 (Meng 4A18); c mod: 是不祥人也 (Zuo 8.2). Cf. 吉 [30.3].

者,所以正禮也 (Xun 2); 禮節也,故 成 (Xun 27); 夫禮者所以定親疏,決嫌 疑,別同異,明是非也 (Li 1); 禮也 者,小事大,大字小之謂 (Zuo 10.30); 禮者,因人之情而為之節文 (Li 30); 動 有文體謂之禮,反禮為濫 (XS 8.3); 夫 禮者,民之紀, 紀亂則民失 (Yan 2.12); 2 np, the ‘(Can­on of) Rites’: 順先王 詩書禮樂以造士 (Li 5); 禮者,體德理 而為之節文,成人事 (XS 8.5); 3 vn, ‘be courteous, pay reverence to’, a tr: 人主兼 禮之 (HFei 49); 於是嬰乃禮文 (SJ 75); b itr: 何以禮焉 (Zuo 5.12); 夫子禮於賈季 (Zuo 6.6).

8

14

祭  jì, *tsets, v , 1 ‘sacrifice’, 祭者, 所以 n

追養繼孝也 (Li 25); a itr: 祭如在 (LY 3.12); 祭於公,不宿肉 (LY 10.8); b tr with Oi: 祭神如神在 (LY 3.12); 死葬之以禮,祭 之以禮 (LY 2.5); 罷兵休卒,收養昆 弟,共祭先祖 (Zhuang 29); 2 nc, ‘sacrifice, of­fering’: 使民如承大祭 (LY 12.2).

祥  xiáng,

禁  jìn, *krəms, v , 1 ‘forbid, prohibit; 禱  dǎo, *tûʔ, v , ‘pray (for)’, a itr: 禱於尼 tr

prevent’, a tr: 禁之不可,桓公立,乃老 (Zuo 1.3); b pass: 文巧不禁則民乃淫 (Guan 1); 2 nc, ‘prohibition’: 俠以武犯禁 (HFei 49); 士則學習法令辟禁 (SJ 6).

9

福  fú, *pək, nc, 1 ‘good fortune, blessing’: 全壽富貴之謂福 (HFei 20); 夫利為害 本,而福為禍先 (HSWZ 1); 聞君子禍至 不懼,福至不喜 (SJ 47); 2 vi, ‘be fortunate’, a itr: 不禱祠而福 (SY 20); b caus: 小 信未孚,神弗福也 (Zuo 3.10); 聖王明知 天鬼之所福 (Mo 27). Ant. 禍 [113.9].

禍  huò, *gôiʔ, n , 1 ‘misfortune, disaster’: c

天降禍于周 (Zuo 10.32); 福不可請,而 禍不可諱 (Mo 35); 2 vi, ‘suffer bad fortune’, a itr: 子必禍矣 (Zuo 11.13); b caus: 神福仁 而禍淫 (Zuo 8.5); 賊其身而禍其家 (CQFL 31). Syn. 害 [40.7]; ant. 福 [113.9].

11

禦  yù, *ŋaʔ, v , ‘fend off, resist, withn

stand; ward off, repel’, tr: 馬,蹄可以踐 霜雪,毛可以禦風寒 (Zhuang 9).

13

禮  lı̌, *rîʔ, n , 1 ‘ritual, rites; manners, c

etiquette’: 禮者,法之大分,類之綱紀 也 (Xun 1); 禮者,所以正身也,師

n

丘得孔子 (SJ 47); b tr: 獲罪於天,無所 禱也 (LY 3.13); 禱之於山川 (CQFL 30).

Classifier 114  禸 4

禹  Yǔ, *waʔ, np, mythical flood fighter and found­er of the Xia dynasty: 禹之治天 下,使民心變 (Zhuang 14); 微禹,吾其 魚乎 (Zuo 10.1). Cf. Focus 7. 8

禽  qín, *gəm, nc, ‘bird’: 終日而不獲一 禽 (Meng 3B1). - 禽獸  qínshòu,

*gəm-­hjuh, nc, ‘birds and beasts, wild animals’: 禽獸有 知而無義 (Xun 9); 人民不勝禽獸蟲蛇 (HFei 49).

Classifier 115  禾 2

私  sī, *siʔ, vi, 1 ‘be personal, private, familial; self­i sh; partial’: 自環者謂之 私,背私謂之公 (HFei 49); 兼覆無私謂 之公,反公為私 (XS 8.3); a itr: 送糧者 不私 (SJS 2); 凡將 . . . 不私於一人 (WLiao 9); b caus, ‘treat as one’s own/partially’:

487

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

吾妻之美我者,私我也 (ZGC 8.12); 子,人之所私也 (Lü 1.5); c mod: 少私義 則公法立 (Lü 26.3); 私視使目盲,私聽 使耳聾,私慮使心狂 (Lü 12.6); 2 nc, ‘self-­ interest, private profit, personal benefit’: 君臣釋法任私,必亂 (SJS 14); 不以私 害公 (LNZ 5.6). Syn. 暱 [72.11]; ant. 公 [12.2], 法 [85.5].

15.20); d pass with C: 稱鄭伯,譏失教也 (Zuo 1.1); 夫舉事者或為善而不稱善, 或不善而稱善者,何 (XY 5).

10

稽  qı̌, *khîʔ, v , ‘bow (the head to the n

ground), kowtow’, only tr: 武王又再拜稽 首,乃出 (SJ 4).

稼  jià, *krâh, n , ‘crops’:

4

秋  qiū, *tshiu, ntemp, ‘autumn’: 秋八月 丁卯,大事于大廟 (Zuo 6.2); 天有四 時,春秋冬夏 (Li 29); 秋者,天之平也 (CQFL 79). 5

秦  Qín, *dzin, n , a polity: 昔文公與秦 p

伐鄭 (Zuo 9.14). Dist. 奉 [37.5]. Note: 秦 was founded in the eighth century bc in Guanzhong (modern Shaan­ xi), the old home of the Zhou. It was ruled by the Ying 嬴 clan. Located on the Western periphery of the realm, Qin expanded through several conquests in the fourth and third cen­ turies bc, finally subduing all other states and founding the first empire in 221 bc.

6

移  yí, *lai, v , ‘move, transfer’, a itr: 相地 i

而衰征, 則民不移 (GY 6.2); 下修今而 不時移 (SJS 8); b caus: 河內凶,則移其 民於河東 (Meng 1A3); c pseudo-­tr2: 不 變,移之郊 (Li 5).

冬與夏不能 c 兩刑,草與稼不能兩成 (Lü 24.5).

11

積  jī, *tsek, v , ‘accumulate, collect, pile n

up’, a tr: 積土成山,風雨興焉 (Xun 1); 君宮中積珍寶 (ZGC 11.1); b itr: 積於不 涸之倉, 藏於不竭之府 (Guan 1); 且夫 水之積也不厚,則負大舟也無力 (Zhuang 1); c mod (!): 我有積怨深怒於 齊 (XX 3.6); 積羽沈舟,群輕折軸 (HNan 10).

穆  mù,

*muk, nc, 1 ‘ancestor’, first, third, fifth, etc. in sacrificial order: 王季之 穆也 (Zuo 5.5); 2 np, ‘Splendid’, a posth. name: 九月衛穆公卒 (Zuo 8.2).

Classifier 116  穴 0

穴  xué, *wît, n , 1 ‘pit, hole; cave’: 堀穴深 c

不通於泉 (Mo 21); 蛇之穴也,如虎之室 (SY 1); 2 adv, ‘in caves’: 上古穴居而野處 (Yi 8); 北方曰狄,衣羽毛穴居 (Li 5).

3

7

稅  shuì, *lhots, n , ‘taxes’: 民之飢,以其 空  kōng, *khôŋ, v , 1 ‘be empty, hollow; c

上食稅之多 (Lao 75); 省刑罰,薄稅斂 (Meng 1A5); 縣鄙則將輕田野之稅 (Xun 11); 為田開阡陌封疆,而賦稅平 (SJ 68).

9

稱  chēng, *thəŋ, vtr, ‘call, name (s.o.); praise, acclaim’, a tr: 讓國餓死,天下稱 之 (SJ 130); b tr with C: 邦君之妻,君稱 之曰夫人,夫人自稱曰小童 (LY 16.14); c pass: 君子疾沒世而名不稱焉 (LY

st

vain’, a itr: 不信仁賢,則國空虛 (Meng 7B12); b mod: 空言虛語,非所守也 (SJ 8); 2 nc, ‘emptiness; air, sky’: 鳥飛於 空,魚游於淵 (Shen 9). Syn. 虛 [141.6]; ant. 實 [40.11].

10

窮  qióng, *guŋ, vi, ‘be ex­hausted, destitute, de­pleted, wanting (for)’, a itr: 讒人 困窮 (Lü 3.4); 孔子窮於陳蔡之間,七

488

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

日不火食 (Zhuang 20); 故君不窮於智 (HFei 5); b caus, ‘exhaust’: 舜不窮其 民,造父不窮其馬 (Xun 31); ‘reach the end (of)’: 故聖人之事,廣之則極宇 宙,窮日月 (Lü 17.8); c mod: 昔上世之 窮民,貪於飲食,惰於從事 (Mo 35). Syn. 貧 [154.4], 乏 [4.4]; ant. 達 [162.9].

11

窺  kuī,

*khwe, vn ‘spy, peek (at/ through)’, a itr: 不窺於牖而知天道 (Lü 17.2); b tr: 窺敵觀變欲潛以深 (Xun 15).

18

竊  qiè, *tshêt, v , 1 ‘steal, rob’, a tr: 竊人 n

之財, 猶謂之盜, 況貪天之功以為己 力乎 (Zuo 5.24); b itr: 君位危,則大臣 倍,小臣竊 (Li 9); 羿請不死之藥於西 王母,姮娥竊以奔月 (HNan 6); c aux, ‘take the liberty, dare’: 臣竊計 (ZGC 11.1); 衛國之法,竊駕君車罪刖 (SY 17); 2 nc, ‘thief; theft’: 盜竊亂賊而不作 (Li 9). Syn. 盜 [108.7].

Classifier 117  立

9

竭  jié, *gat, vtr, ‘dry up, exhaust’: a tr: 事 父母能竭其力 (LY 1.7); b pass: 直木先 伐,甘井先竭 (Zhuang 20); 地侵於 齊,貨竭於晉 (SY 13); 耳目竭於聲 色,精神竭於外貌 (HFei 21). Syn. 虛 [141.6]; 盡 [108.9]; ant. 盈 [108.4]. Classifier 118  竹 0

竹  zhú, *truk, nc, ‘bamboo’: 南山有 竹,弗揉自直,斬而射之,通於犀革 (SY 3); 著於竹帛,鏤於金石 (Wen 2); by metonymy, instruments made of bamboo: 金石絲竹,樂之器也 (Li 19). 4

笑  xiào, *siauh, vn, 1 ‘laugh (at)’, a itr: 王 笑而不言 (Meng 1A7); b tr: 大夫多笑之 (Zuo 10.2); 2 nc, ‘laughing stock, butt’:身 為宋國笑 (HFei 49). 5

0

立  lì, *rəp, vi, ‘stand’, a itr: 立于門外 (Zuo 7.2); 本立而道生 (LY 1.2); ant. 臥[131.2]; esp. ‘assume pow­er/the throne’: 伯成子高立為諸侯焉 (XX 7.1); b caus, ‘put up, establish; install (as ruler)’: 諸侯 自立社 (Li 23); 君既許我殺太子而立奚 齊矣 (GY 8.1); ‘bring to a stand, stop’: 孔 子立輿而問曰 (XX 5); c mod: 楚人日徵 敝邑,以不朝立王之故 (Zuo 10.3). 6

笞  chī, *rhə, vtr, ‘beat, flog’, a tr: 他日笞 子未嘗見泣 (SY 3); 掘平王冢,笞其墳 (SY 12); b pass: 三見而三笞 (SY 3). 6

答  dá, *tûp, vn, 1 ‘answer, respond (to)’, a itr: 夫子不答 (LY 14.5); b tr: 於答是也 何有 (Meng 6B1); 答之曰 (CQFL 1); 2 nc, ‘answer’: 今者吾忘吾答,因失吾問 (Zhuang 23). Ant. 問 [30.8]; cf. 對 [41.11]. 7

節  jié, *tsît, n , a ‘joints

章  zhāng, *taŋ, var. of 彰 [59.11]. 7

竦  sǒng, *soŋʔ, vi, ‘rise up; stand upright’, itr: 見橋竦焉實而仰 (SY 3.6).

on bamboo; c measure; division of time; moderation’: 內外上下節者,義之情也 (Xun 16); 費 弗過適謂之節,反節為靡 (XS 8.3).

8

童  tóng, *dôŋ, n , ‘boy, child’: 孩提之 箸  zhù, *drah, n , ‘chopsticks’: 十仞之 c

童,無不知愛其親者 (Meng 7A15). Cf. 子 [39.0].

c

木若箸 (Xun 21); 紂為象箸而箕子怖 (HFei 21).

489

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

箸  zhù, *trakh, var. of 著 [140.8]. 箕  jī, *kə, n , ‘winnowing fan, basket’: c

良弓之子,必學為箕 (Li 18). Cf. Focus 4.

10

築  zhú, *truk, v , ‘build, construct’, tr: n

仲子所居之室,伯夷之所築與 (Meng 3B10).

篤  dǔ, *tûk, v , 1 ‘be firm, stable; loyal, i

devoted; honest’, a itr: 君子篤於親,則 民興於仁 (LY 8.2); 言忠信,行篤 敬,雖蠻貊之邦行矣 (LY 15.6); b caus: 以正君臣,以篤父子 (Li 9); 2 adv, ‘faithful­ly, devotedly’: 明辨之,篤行之 (Li 31).

12

簞  dān, *tân, n , ‘basket’: 使盡之,而 c

為之簞食與肉 (Zuo 7.2).

簡  jiǎn, *krênʔ, nc, ‘bamboo slip (for writing)’: 昭王讀法十餘簡而睡臥矣 (HFei 32); 執簡以往,聞既書矣 (Zuo 9.25). Cf. Focus 19.

簡  jiǎn, *krênʔ, vst, ‘be short, simple, brief; insig­nificant’, a itr: 易則易知,簡則 易從 (Yi 7); 其作始也簡,其將畢也必 巨 (Zhuang 4); 大樂必易,大禮必簡 (Li 19); b caus, ‘slight, neglect, ignore’: 孟子獨 不與驩言,是簡驩也 (Meng 4B27); 簡 禮賤義者其國亂 (Xun 15); 士輕祿,民 簡賞 (Guan 68); c mod: 國之簡士待臣而 後舉火者數百家 (SY 2). Syn. 易 [72.4]. 簡  Jiǎn,

*krênʔ, np, ‘Perspicacious’, a posth. name: 昔者,燕簡公殺其臣莊子 儀而不辜 (Mo 31).

14

籍  jí, *dzak, n , 1 ‘register, record’: 夫學 c

者載籍極博, 猶考信於六蓺 (SJ 60); 2 vn, ‘register, re­cord; enlist’, tr: 非禮也,勿 籍 (Zuo 8.2); 義再拜受命而籍之 (SJ 43).

Classifier 119  米 4

料  liào,

*riâuh, vtr, ‘measure, count, estimate’, a tr: 夫古者不料民而知其少 多 (GY 1.9); 臣竊料匈奴之眾,不過漢 一千石大縣 (XS 4.2); b mod: 嘗為季氏 史,料量平 (SJ 47).

5

粕/魄  pò, *phrâk, n , ‘dregs’: 然則君 c

之所讀者,古人之糟魄已夫 (Zhuang 13); 凶年饑歲,糟粕不厭 (XX 2). Syn. 糟 [119.11].

6

粟  sù, *sok, nc, ‘millet; grain’: 雖有 粟,吾得而食諸 (LY 12.11). Note: Specifically denoting foxtail millet, 粟 was also used as a general term for all kinds of millet and sorghum. 粥  zhōu, *tuk, n , ‘congee’: 饘於是,粥 c

於是,以餬余口 (SJ 47).

粥  yù, *luk, var. of 鬻 [193.12]. 8

精  jīng, *tseŋ, v , ‘be subtle, pure, fine; st

refined, skilful, accomplished’, a itr: 其為 人也,小心精潔 (GY 7.6); 農精於 田 . . . 賈精於市 . . . 工精於器 (Xun 21); b caus: 慎其行,精其思慮 (Mo 9); c comp: 莫精於氣 (CQFL 56).

11

糟  zāo, *tsû, n , ‘sediment, dregs’: 然則 c

君之所讀者,古人之糟魄已夫 (Zhuang 13). Syn. 粕 [119.5].

12

糧  liáng, *raŋ, n , ‘provisions, supplies’: c

師行而糧食 (Meng 1B4); 善用兵者,役 不再籍,糧不三載 (Sun 2).

490

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 120  糸

紳  shēn, *lhin, n , ‘sash’: 子張書諸紳

3

終  zhōng, *tuŋ, vtr, 1 ‘end, finish, complete’, a tr: 雖終其業,其去之必速 (Li 18); 終身也者,非終父母之身,終其 身也 (Li 12), b pass, ‘be finished; die’: 今君 雖終,言猶在耳 (Zuo 6.7); ant. 始 [38.5]; cf. note at 死 [78.2]; c mod, typi­cally with ntemp, ‘the whole’: 吾嘗終日不食, 終夜 不寢 (LY 15.31); with nc: 子路終身誦之 (LY 9.27); 2 adv, ‘always’: 昔者孟子嘗與 我言於宋,於心終不忘 (Meng 3A2).

約  yuē, *ʔɨɑk, v , ‘be restrained, const

stricted; condensed; few’ a itr: 事約而易 從, 求寡而易足 (Guan 56); 能以眾擊 寡者,則吾之所與戰者,約矣 (Sun 6); b caus, ‘tie, fasten; rein in, restrain’: 君子 博學於文,約之以禮,亦可以弗畔矣 夫 (LY 6.27); 於是約車治裝 (ZGC 11.1); c comp: 故操彌約,而事彌大 (Xun 3); d mod: 然張耳陳餘始居約時 (SJ 89).

紂  Zhòu, *druʔ, np, the last Shang king: 湯放桀,武王伐紂 (Meng 1B15); 桀紂暴 亂,而湯武征伐 (HFei 49). Cf. Focus 7.

紀  jì/jı̌, *kəʔ, n , ‘(main) thread, strand; c

leading principle, norm, rule’: 無變天之 道,無絕地之理,無亂人之紀 (Lü 1.1); 夫孝,三皇五帝之本務,而萬事之紀 也 (Lü 14.1); 夫禮者,民之紀, 紀亂則 民失 (Yan 2.12).

4

純  chún,

*dun, vi, ‘be pure; upright; silken’, a itr: 潁考叔純孝也 (Zuo 1.1); 非德 不純,形勢弱也 (SJ 17); b caus: 詩書具 其志,禮樂純其養,易春秋明其知 (CQFL 2); c mod: 石碏,純臣也 (Zuo 1.4).

素  sù, *sâh, v , ‘be plain, undyed, coli

ourless; white; pristine’: 素也者,五色 之質也 (Guan 39); 夫素之質白,染之 以涅則黑 (HNan 11); a itr: 其心愉而不 偽,其事素而不飾 (HNan 8); b mod: 杜 伯乘白馬素車 (Mo 31). Syn. 樸 [75.12]; 白 [106.0]; ant. 飾 [184.5].

6

結  jié, *kît, v , ‘knot, tie; connect’, a tr: tr

使民復結繩而用之 (Lao 80); b pass: 如 此,則上下之恩結矣 (HFei 27); 故君子 結於一也 (Xun 1).

絶/絕  jué,

*dzot, vtr, a ‘cut, tear, sever, break’: 一維絶則傾,二維絶則 危,三維絶則覆,四維絶則滅 (Guan 1); ‘cut across, cross, traverse’: 非能水 也,而絶江河 (Xun 1); b pass: 一旦命已 絕矣 (Zhuang 30); c mod: 興滅國,繼絕 世 (LY 20.1). Ant. 繼 [120.14].

絲  sī, *sə, nc, ‘silk; thread’: 春秋冬夏皆 有麻枲絲繭之功 (Lü 26.3); 繭之性為絲 (HSWZ 5); ‘stringed instruments’: 金石 絲竹,樂之器也 (Li 19). 給  gěi/jı̌, *kəp, v , ‘be well-­supplied, sufi

fice’, a itr: 乘民之不給,百倍其本 (Guan 73); 家給人足 (HSWZ 9); b caus, ‘supply, provide’: 寡人之國貧,恐不能給也 (ZGC 6.8); 孟嘗君使人給其食用 (ZGC 11.1).

絜  jié, *kêt, var. of 潔 [85.12].

5

累  lěi, *roiʔ, v , ‘pile

c

(LY 15.6).

up, accumulate, tr collect, assemble’, a tr: 其少時,家累千 金 (SJ 65); 雖倍賞累罰而不免於亂 (HFei 49); b pass: 累而成文,名之麗也 (Xun 22); c mod: 九層之臺,起於累土 (Lao 64).

7

經  jīng, *kêŋ, nc, ‘warp (in a loom)’, fig. ‘guide­ line, guiding principle’: 順民之經 (Guan 1); 民不罷勞,君無怨讟,政有經 矣 (Zuo 7.12); fig. ‘guide, canonical text’: 其

491

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

數則始乎誦經, 終乎讀禮 (Xun 1); 通一 經之士不能獨知其辭,皆集會五經 家,相與共講習讀之 (SJ 24); 是以孔子 論六經,紀異而說不書 (SJ 27); cf. Focus 1.

8

維  wéi, *wi, nc, ‘rope; tie, bond’, fig.: 四 維張則君令行 (Guan 1). 10

縣  xiàn, *gwêns, n , 1 ‘district, county’: c

雖與之縣,不亦可乎 (Zuo 10.28); 2 vi, ‘be a district’, caus: 文王以為令尹,實縣 申息 (Zuo 12.17).

縣  xuán, var. of 懸 [61.16]. 11

繁  fán, *ban, v , a ‘be abundant, prost

fuse; nu­merous; complex’, a itr: 不璋兩原 則刑乃繁 (Guan 1); b caus: 亂其教,繁 其刑 (Xun 28); c comp: 是以刑彌繁,而 邪不勝 (Xun 28); 昔秦法繁於秋荼,而 網密於凝脂 (YTL 55).

縻  mí, *mai, v , ‘fetter, entangle’, tr: 是

謂縻軍 (Sun 3).

n

縱  zòng,

*tsoŋh, conj, ‘even if, even though’: 縱子忘之,山川鬼神其忘諸 乎 (Zuo 11.1). Cf. 16.2.

繫/係  xì, *kêh, v , ‘bind, attach, hang’, n

a tr: 聖人設卦觀象,繫辭焉而明吉凶 (Yi 7); 繫馬千駟,弗視也 (Meng 5A7); 百越之君俛首係頸 (SJ 6); b pseudo-­tr2: 折紂而繫之赤環 (Mo 31); 編之以 髮,繫之葦苕 (Xun 1).

繫  xì, *gêh, v [< preceding], ‘be bound, i

attached (to)’, a itr: 飽食而敖遊,汎若 不繫之舟 (Zhuang 32); 少繫於父母,長 繫於夫,老繫於子 (LNZ 1.12); 不固溺 於流俗,不拘繫於左右 (SY 1); b mod: 似繫馬而止也 (Zhuang 13).

14

繼  jì, *kêh, vtr, ‘continue; follow up’, a tr: 興滅國, 繼絕世 (LY 20); 工匠之 子,莫不繼事 (Xun 8); 非繼前王而王 也 (CQFL 1); b pass: 苟信不繼,盟無益 也 (Zuo 2.12); 王道廢而不起,禮義絕 而不繼 (HSWZ 5). Ant. 絶 [120.6]. Classifier 122  网 3

罕  hǎn, *hŋânʔ, adv, ‘seldom, rarely’: 子 罕言利與命與仁 (LY 9.1); 役,無奪農時 (Xun 11).

罕舉力

繄  yī, *ʔî, interj, ‘alas’: 爾有母遺,繄 罔  wǎng, *maŋʔ, n , 1 ‘net’: 作結繩而為 c

我獨無 (Zuo 1.1).

12

織  zhī, *tək, v , ‘weave’, a itr: 婦人不 n

織,禽獸之皮足衣也 (HFei 49); b tr: 織 席以為食 (Meng 3A4).

繕  shàn, *dans, vn, ‘repair, restore, put in order’, a tr: 繕甲兵,具卒乘 (Zuo 1.1); 修城郭,繕牆垣 (CQFL 61); b itr: 大夫 家長不建繕,則農事不傷 (SJS 2). 13

繩  shéng, *m-­ləŋ, n , ‘rope, cord; (carc

penter’s) line’: 使民復結繩而用之 (Lao 80); 聖人既竭目力焉,繼之以規矩準 繩 (Meng 4A1).

罔罟,以佃以漁 (Yi 8); 罔罟毒藥不入 澤 (Xun 9); 2 vn, ‘net, catch, entangle’, tr: 以左右望而罔市利 (Meng 2B10); 不能 以非功罔上利 (SJS 17).

罔  wǎng,

*maŋʔ, vi, 1 ‘be confused, unclear, dis­oriented’, a itr: 學而不思則 罔,思而不學則殆 (LY 2.15); 離離 然,偷儒而罔 (Xun 6); b caus, ‘confuse, deceive, cheat’: 及陷於罪,然後從而刑 之,是罔民也 (Meng 1A7); 君子…可欺 也,不可罔也 (LY 6.26).

5

罟  gǔ, *kâʔ, nc, ‘(fishing) net’: 數罟不 入洿池,魚鼈不可勝食也 (Meng 1A3); 作結繩而為罔罟,以佃以漁 (Yi 8).

492

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

8

罪  zuì, *dzûiʔ, nc, 1 ‘crime, offence, guilt, fault’: 無功不賞, 無罪不罰 (Xun 9); 君若以臣為有罪 (Zuo 10.31); 臣寧得 罪於王,無負於先王 (SY 9); 2 vtr, ‘blame, accuse, condemn’, a tr: 我將罪賤 之 (Mo 13); 命吏罪之 (Lü 17.3); b pass: 夫 離法者罪 (HFei 49); c pass with C: 竊駕 君車罪刖 (SY 17).

4

羔  gāo, *kâu, n , ‘lamb’: 粥羔豚者弗飾 賈 (SJ 47).

c

5

羞  xiū,

*snu, vn, ‘be ashamed (of)’ tr: 柳下惠不羞汙君,不卑小官 (Meng 2A9); 是仁人之所羞而不為也 (Xun 16).

7

置  zhì, *trəkh, v , ‘put, place, set up’, a tr: 羣/群  qún, *gwən, n , 1 a ‘herd, flock; 置鼓其上,遠近相聞 (Lü 22.3); fig. n

c

‘appoint, install’: 置君而後去之 (Meng 1B11); 舉賢而置之 (Li 10); b itr: 人主不 察則敵廢置矣 (HFei 31).

署  shǔ, *da(k)h, vn, ‘sign up, register; write’, a itr: 馮諼署曰能 (ZGC 11.1); b tr: 使主節必疏書,署其情 (Mo 71). 9

罰  fá, *bat, v , 1 ‘punish, fine’, tr: 上為政 n

者得則罰之 (Mo 17); 2 nc ‘penalty, fine’: 掌受士之金罰貨罰,入于司兵 (ZL 69); 重罰輕賞,則上愛民,民死上 (SJS 4). Cf. note at 刑 [18.4]; ant. 賞 [154.8].

10

罷  pí, var. of 疲 [104.5]. Classifier 123  羊

group, community’, 禽獸固有群矣 (Zhuang 13); 小人成群,何足禮哉 (HSWZ 1); syn. 聚 [128.8]; b mod, ‘multitude, host of ’: 以避羣害 (HFei 49); 羣生 皆得其命 (Xun 9); cf. Box 28; 2 vi, ‘form a group/com­munity’, itr: 人能群,彼不能 群也 (Xun 9); 3 adv, ‘in groups’: 獸處群 居(Guan 31).

義  yì, *ŋaih, vst, 1 ‘be righteous, upright, moral’: 義理也,故行 (Xun 27); 義者,宜也 (Li 31); 義,人之正路 也 (Meng 4A11); a itr:其不義又甚入人 園圃竊桃李 (Mo 17); b caus: 督以 正,義其名 (Mo 19); 於是襄子大義之 (SJ 86); 2 nc, ‘right conduct, righteous­ ness, (sense of) justice, moral­ity’: 義之 實,從兄是也 (Meng 4A27); 內外上下 節者,義之情也 (Xun 16); 堅剛而不 屈,義也 (Xun 30); 行充其宜謂之 義,反義為懵 (XS 8.3). 13

0

gēng, *krâŋ, n , ‘soup, broth’: 嬰 羊  yáng, *jaŋ/laŋ, n , ‘sheep, goat’: 其父 羹  兒 . . . 以塵為飯,以塗為羹,以木為胾 c

c

攘羊,而子證之 (LY 13.18); 故從山上望 牛者若羊 (Xun 21). Cf. note at 畜 [102.5].

(HFei 32).

Classifier 124  羽

3

美  měi, *muiʔ, v , 1 ‘be beautiful, fine, st

excellent’, a itr: 牛山之木嘗美矣 (Meng 6A8); b caus: 君子之學也,以美其身 (Xun 1); 美其服,安其居 (Lao 80); c mod: 所愛美女欲以其子為後 (HFei 31); 2 nc, ‘beauty’: 天下皆知美之為美 (Lao 2). Ant. 惡 [61.8].

0

羽  yǔ, *waʔ, nc, ‘feather’: 福輕乎 羽 . . . 禍重乎地 (Zhuang 4). 5

習  xí, *s-­ləp, vn, 1 ‘repeat, practice, review’, a tr: 學而時習之,不亦說乎

493

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(LY 1.1); 士則學習法令辟禁 (SJ 6); b itr: 臣得之矣,而未習也 (HFei 10); thence ‘be experienced, acquainted/familiar (with)’: 夫夫也,為習於禮者 (Li 3); 吏 習而民安 (SJS 1); c pseudo-­tr: 誰習計會 (ZGC 11.1); syn. 學 [39.13]; 2 nc, ‘custom, habit’: 常人安於故習 (SJS 1).

8

翟  dí, *liâuk, var. of 狄 [94.4].

必見惡也 (Mo 16); b ad rem: 知我者其 惟春秋乎,罪我者其惟春秋乎 (Meng 3B9); cf. 6.3, 12.1; Review 3; c used to emphasize the S: 道者,令民與上同意 也 (Sun 1); cf. 13.2; d used to conclude a 所 phrase: 臣以所學者觀之 (ZGC 30.9); 2 conj, ‘if ’: 因民而教者,不勞而功成 (SJS 1); 地廣者粟多,國大者人眾,兵 彊則士勇 (SJ 87); 水濁者魚噞,令苛者 民亂 (HNan 10); cf. 則 [18.7].

耆  qí, *gri, n , ‘sexagenarian, old man’: c

六十曰耆,指使 (Li 1); 養耆老以致孝 (Li 5); 瞽史教誨,耆艾修之 (GY 1.3). Cf. note at 幼 [52.2].

Classifier 125  老 0

老  lǎo, *rûʔ, vst, 1 ‘be old; worn out’, a itr: 吾老矣,不能用也 (LY 18.3); 且楚師 老矣,必敗 (GY 10.7); b caus, ‘treat as old’ (i.e. with re­spect): 老吾老,以及人 之老 (Meng 1A7); ‘wear out’: 老師費 財,亦無益也 (Zuo 5.33); ‘grow old’: 與 人居,長子老身 (Zhuang 18); c mod: 老 人兒啼 (SJ 119); 2 adv, ‘at old age’: 是以行 年七十而老斲輪 (Zhuang 13); 3 nc, ‘old person, elder, senior’: 老吾老,以及人 之老 (Meng 1A7). Cf. note at 幼 [52.2].

老子  Lǎozi, *rûʔ-­tsəʔ, np, a philosopher: 老子者,楚苦縣厲鄉曲仁里人 也,姓李氏,名耳,字耼,周守藏室 之史也 (SJ 63). 1

考  kǎo, *khûʔ, vn, ‘examine, analyse’, a tr: 以其時考之則可矣 (Meng 2B13); 所 以考其善不善者,豈有他哉 (Meng 6A14); b itr: 天子歲試天下,三試而一 考,前後三考而絀陟 (CQFL 21). 考  kǎo, *khûʔ, v , ‘complete’, a tr: 言以 n

考典,典以志經 (Zuo 10.15); 以著其 義,以考其信 (Li 9); 婦人不得剡麻考 縷 (HNan 18); b itr: 上帝不考,時反是 守 (GY 21.2). Syn. 成 [62.2].

4

者  zhě, *taʔ, pt, 1 nominalizes VP, a ad

personam: 愛人者必見愛也,而惡人者

Classifier 126  而 0

而  ér, *nə, prper, [< 爾, 89.10], ‘your’, only mod: 而母婢也 (ZGC 20.13). Cf. Box 5.

而  ér, *nə, conj [perhaps < 如], 1 ‘and, but’, a subordinating two VP, expressing manner: 哭而過市(Zuo 6.18); 保民而王 (Meng 1A7); conjunction: 三見而三笞 (SY 3.6); or other kinds of logical relations, such as sequential: 殺孔父而取其妻 (Zuo 2.2); 卒聞之而惑 (Zhuang 14); causal: 見國之利而忘君安 (XX 7.10); 鳥 聞之而飛 (Zhuang 12); conditional: 政治 而國安也 (Mo 12); often adver­sative: 君 子和而不同,小人同而不和 (LY 13.23); 不教民而用之 (Meng 6B8); final: 虎求 百獸而食之 (ZGC 14.3); with O=: 擇能而 使之 (Zuo 9.31); often with deletion of subordinated O: 余聞而弗忘 (Zuo 10.3); 勞而食之 (SY 3.6); cf. 7.6. b subordi­nating clauses: 今君富於季孫,而齊大於魯 (HFei 39); cf. 且 [1.4], and 15.3; 2 ‘if ’, following the S: 子產而死,誰其嗣之 (Zuo 9.30); 人而無信,不知其可也 (LY 2.22); cf. 15.4.1–⑳. 而後  érhòu, *nə-­ɦôʔ, conj, ‘(only) then’: 事必成然後舉,身必安而後行 (YDan 1); cf. 16.3.

494

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

而已  éryı̌, *nə-­ləʔ, ptemp, ‘only, simply; that’s all’: 堯舜之道,孝弟而已矣 (Meng 6B2). Syn. 耳 [128.0].

耳  ěr, *nəʔ, ptemp [< 而已], ‘simply; that’s all, that’s it’, final: 山澤之婦人耳 (LNZ 5.6); 君家所寡有者以義耳 (ZGC 11.1). Cf. Box 16; syn. 而已 [126.0].

Classifier 127  耒

3

耶/邪  yé, *la/ja [< 也乎], pt

0

耒  lěi, *ruiʔ, n , ‘plough’: 因釋其耒而 c

守株 (HFei 49); 斲木為耜,揉木為耒 (Yi 8).

int, 子之 師誰邪 (Zhuang 21); in alt. questions: 松 耶柏耶 (SJ 46). Cf. 3.4 and Box 16.

4

耻  chı̌, *nhrəʔ, var. of 恥 [61.6].

4

耕  gēng, *krêŋ, v , ‘plough’, a tr: 宋人有 n

耕田者 (HFei 49); 故人情者,聖王之田 也,脩禮以耕之,陳義以種之,講學 以耨之 (Li 9); b itr: 春耕,夏耘,秋 收,冬藏 (Xun 9); 禹往見之,則耕在 野 (Zhuang 12).

耘  yún, * v , ‘weed (out)’, a itr: 農夫春 n

耕,夏耘,秋斂,冬藏 (Mo 7); 有地君 國,而不務耕耘 (Guan 13); fig. ‘eliminate’: 不戰而耘, 利莫大焉 (SJ 114); b tr: 不耘苗者也 (Meng 2A2).

4

耜  sì, *s-­ləʔ, n , ‘spade’: 斲木為耜,揉 c

木為耒 (Yi 8); 匠人為溝洫,耜廣五 寸,二耜為耦 (ZL 85).

10

耨  nòu, *nûk/nôkh, nc, 1 ‘hoe’: 耨柄 尺,此其度也,其耨六寸,所以間稼 也 (Lü 26.4); 耒耨之利,以教天下 (Yi 8); 2 vn, ‘weed’, a itr: 彼奪其民時,使不 得耕耨,以養其父母 (Meng 1A5); b tr: 耕之耰之,鉏之耨之 (SJ 128); 故人情 者,聖王之田也,脩禮以耕之,陳義 以種之,講學以耨之 (Li 9). Classifier 128  耳 0

耳  ěr, *nəʔ, n , ‘ear’: 耳目之官 (Meng c

6A15); 今君雖終,言猶在耳 (Zuo 6.7).

7

聖  shèng, *lheŋh, [< 聲?] vst, ‘be perspicacious, sage, wise’: 聖也者,盡倫者也 (Xun 21); 聖也者,達於情而遂於命也 (Zhuang 14); 聞其末而達其本者聖也 (HSWZ 5); 聞而知之,聖也,見而知 之,智也 (Wen 5); a itr: 仁且智,夫子 既聖矣 (Meng 2A2); b mod: 聖王者,不 貴義而貴法 (SJS 18).

聖人  shèngrén, *lheŋh-­nin, n , ‘sage, c

wise man’: 聖人,百世之師也 (Meng 7B15); 聖人者,原天地之美而達萬物 之理 (Zhuang 22); 以天為宗,以德為 本,以道為門,兆於變化,謂之聖 人(Zhuang 33); esp. an ideal ruler: 是以聖 人之治,虛其心,實其腹 (Lao 3); 聖人 治國也,審壹而已矣 (SJS 17). Cf. note at 主 [3.4].

8

聞  wén, *mən, v , ‘hear’, a tr: 齊王聞 n

之,君臣恐懼 (ZGC 11.1); ‘understand, accept’: 敬聞命矣 (SY 11); 軍中聞將軍 令,不聞天子之詔 (SJ 57); b itr: 齊人聞 而懼 (SJ 47). Note: 聞, as op­posed to 聽, does not imply active listening; by the same token, one can listen with­out hearing: 聽之弗聞 (Shi 1); 聽而不聞 (Li 42).

聞  wèn, *məns, vst [< preceding], 1 ‘be audi­ble, loud; re­nowned, famous’, a itr: 不 祗山川則威令不聞 (Guan 1); 主以不 賄,聞於諸侯 (Zuo 10.28); b caus: ‘make

495

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

known’: 國人道之,聞之於宋君 (Lü 22.6); c comp: 不言之言,聞於雷鼓 (Guan 37); 2 nc, ‘fame, repute’: 子有令 聞,而美其室 (Zuo 9.15).

聚  jù, *dzoh, vtr, 1 ‘collect, gather, assemble, store’, a tr: 朝之室也, 聚酒千 鍾 (Lie 7); 聚天下之貨, 交易而退, 各得其所 (Yi 8); b pass: 夫諸侯之 賄,聚於公室 (Zuo 9.24); 人之生,氣 之聚也,聚則為生,散則為死 (Zhuang 22); ant. 散 [66.8]; 2 nc, ‘assemblage, group, troops’: 禹無百人之聚,以 王諸侯 (SJ 69); syn. 羣 [123.7]. 11

聰  cōng, *tshôŋ, vst, ‘have acute hearing’, a itr: 目明而耳聰 (Xun 23); 視不明, 聽 不聰, 行不正, 不知哀, 君子病之 (Li 21); b comp: 聽而精之,莫聰於耳 (HNan 10); c mod: 是以聰耳明目相與視 聽乎 (Mo 3). 聲  shēng, *hjeŋ, n , ‘sound, noise; voice’: c

生而同聲,長而異俗 (Xun 1); 音聲相 和 (Lao 2).

12

肯  kěn/kěng,

*khêŋʔ, vn, ‘be willing, ready to’, aux: 幼而不肯事長,賤而不 肯事貴,不肖而不肯事賢,是人之三 不祥也 (Xun 5). Cf. 4.4.1.

育  yù, *luk, v , ‘give birth (to); nurture, tr

rear, breed; cultivate’, a tr: 養桑麻,育六 畜也 (Guan 1); 尊賢育才,以彰有德 (Meng 6B7); b pass: 五穀熟,而民人育 (Meng 3A4); 故養長時則六畜育 (Xun 9). Syn. 殖 [78.8]; 養 [184.6].

5

胡  hú,

*gâ, adv, ‘how, why’: 成而行 之,胡可壅也 (GY 1.3). Syn. 何 [9.5], 奚 [37.7]; cf. also 盍 [108.5].

胡蝶  húdié,

*gâ-­lêp, var. of 蝴蝶

[142.9].

胃  wèi, *wəs, n , ‘stomach’: 腥臊惡臭 c

而傷害腹胃 (HFei 49). Cf. 腹 [130.9].

背  bèi, *pə̂kh, n , 1 ‘back (of the body)’: c

其生色也,睟然見於面,盎於背 (Meng 7A21).

背  bèi, *bə̂kh, var. of 偝 [9.9]. 6

職  zhí, *tək, n , ‘office, function; duty’: 胹  ér, *nə, v , ‘boil in water’, a tr: 宰夫 多男子而授之職,則何懼之有 c

tr

胹熊蹯不熟 (Zuo 7.2); b pass: 及食熊 蹯,胹不熟 (SJ 43).

(Zhuang 12).

16

聽  tīng, *lhêŋ, v , ‘hear; listen to’, a tr: 君 n

不吾聽 (SY 1); b itr: 政成而民聽 (Zuo 2.2); 視乎不可見,聽乎不可聞,為乎 不可成 (Xun 12). Cf. note at 聞 [128.8].

Classifier 130  肉 0

肉  ròu, *nuk, nc, ‘meat’: 子肉也,我肉 也 (Lü 11.4). 4

能  néng, *nə̂ŋ/nə̂ʔ, v , 1 ‘be able (to do n

sth.), ca­pable (of doing sth.)’, a tr: 唯賢者 能之 (Li 25); 客何能,曰客無能也 (ZGC 11.1); b itr: 夫子欲寡其過而未能 也 (LY 14.25); 將能而君不御者勝 (Sun 3); 乾以易知,坤以簡能 (Yi 7); c typically as aux: 冬與夏不能兩刑,草與稼 不能兩成 (Lü 24.5); cf. 4.4.1; 2 nc, ‘ability, capa­ bility; capable person’: 我且賢之 用,能之使 (HFei 33).

胾  zì, *tsrəh, nc, ‘minced meat’: 嬰 兒 . . . 以塵為飯,以塗為羹,以木為胾 (HFei 32).

肱  gōng, *kwə̂ŋ, n , ‘(upper) arm’: 以肱 脊  jı̌, *tsek, n , ‘spine, backbone’: 射之

擊之 (Zuo 8.2).

c

c

車上,中心折脊 (Mo 31); fig: 夫秦攻樑

496

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

者,是示天下要斷山東之脊也 25.1).

(ZGC

7

脣  chún, *m-­dun, n , ‘lips’: 脣亡齒寒 c

(Zuo 5.5).

脩  xiū, *siu, var. of 修 [9.8].

ject’, a itr: 君君,臣 vi, ‘be/act as sub­ 臣,父父,子子 (LY 12.11); 夫韓雖臣 於秦,未嘗不為秦病 (HFei 2); b caus: 見且由不得亟,而況得而臣之乎 (Meng 7A8); c pseudo-­tr: 左儒友於杜 伯,皆臣周宣王 (SY 4). Ant. 君 [30.4]; dist. 巨 [48.2].

2

8

腸  cháng, *d-­laŋ, n , ‘intestines’: 大腸 臥/卧  wò, c

者,傳道之官,變化出焉;小腸者, 受盛之官, 化物出焉 (HDNJ 8).

9

腹  fù, *puk, n , ‘belly’: 實其腹 (Lao 3); c

燕王必當殺子,刳子腹及子之腸矣 (ZGC 31.2). Cf. 胃 [130.5].

腥  xīng, *sêŋ(h), n , 1 ‘flesh’: 君賜腥,必 c

熟而薦之 (LY 18); 2 ‘stench (of foul flesh), fetor’: 香臭芬鬱腥臊漏庮奇臭以鼻異 (Xun 22); 腥臊惡臭而傷害腹胃(HFei 49); 3 vi, ‘be rancid, foul, smelly’, itr: 其味辛,其 臭腥 (Li 6). Syn. 臊 [130.13].

11

膠  jiāo, *kriû, n ,‘glue’: 膠漆之材 (Sun 2). c

12

膰  fán, *ban, n , ‘sacrificial meat’: 魯今 c

且郊,如致膰乎大夫,則吾猶可以止 (SJ 47).

13

臂  bì, *pekh, n , ‘arm’: 登高而招,臂 c

非加長也,而見者遠 (Xun 1).

臊  sāo, *sâu, 1 n , ‘stench (of meat)’: 腥 c

臊惡臭而傷害腹胃 (HFei 49); 2 vi, ‘be putrid, rancid, fetid’, itr: 其政腥臊,馨香 不登 (SY 18). Syn. 腥 [130.9].

Classifier 131  臣 0

臣  chén, *gin, nc, 1 ‘servant, subject; minister’: 臣,王之臣也 (ZGC 19.5); 2

*ŋôih, vi, ‘lie down; fall asleep; sleep’, a itr: 醉而臥 (HFei 19); 心 臥則夢 (Xun 21); b caus: 臥名利者寫生 危 (Guan 38). Ant. 立 [117.0].

11

臨  lín, *rəm, vn, ‘approach; overlook, look down at; oversee, supervise’, tr: 不臨 深谿,不知地之厚也 (Xun 1); 臨長晉 國者,非女其誰 (GY 11.3). Classifier 132  自 0

自  zì, *dzih, prep, ‘from; since’, a local:

其劍自舟中墜於水 (Lü 15.8); 季孟自南 門入,出自東門 (Zuo 11.6); b temp: 自 十月不雨至于五月 (Zuo 5.3); c with anteposed O: 康公我之自出 (Zuo 8.13). Cf. 8.3.

自  zì, *dzih, adv, ‘him-/her-/itself ’, a with itr P: 以虧人自利也 (Mo 17); b with tr P: 目. . . 不能自見其睫 (HFei 21); c in O clause: 故明主不窮烏獲,以其 不能自舉 (HFei 24); 12.3–㉜. Syn. 己 [49.0]. 4

臭  chòu, *k-­hjuh, nc, 1 ‘smell, stench’: 色 惡,不食,臭惡,不食 (LY 10.8); 2 vi, ‘be malodorous, smelly, stink’, a itr: 腥臊 惡臭而傷害腹胃 (HFei 49); 酤酒而 酸,買肉而臭 (HNan 17); b mod: 廚中 有臭肉,則門下無死士矣 (SY 8). Ant. 香 [186.0].

497

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 133  至 0

至/致  zhì, *tits, v , 1 ‘arrive; reach’, a itr: i

道雖邇, 不行不至 (Xun 2); 夫子至於是 邦也,必聞其政 (LY 1.10); thematic,‘when it comes to; even’: 至於犬馬皆能有養 (LY 2.7); b pseudo-­tr: 非利足也,而致千里 (Xun 1); 地東至海暨朝鮮 (SJ 6); c mod, ‘the ultimate/highest’: 周之德,其可謂至 德也已矣 (LY 8.20); 大學之道 . . . 在止於 至善 (Li 42); syn. 到 [18.6]; 2 adv, ‘most, ex­tremely, perfectly’, expressing the superlative of vst: 舜其至孝矣 (Meng 6A3); 其為 氣也,至大至剛 (Meng 2A2); cf. Box 24 3 prep, ‘until’: a local: 驅車至茆門 (HFei 34); 又伐其東南,至于陽丘 (Zuo 6.16); b temp: 故人至暮不來 (HFei 32); 至今為患 (Zuo 9.26); 自十月不雨至于五月 (Zuo 5.3); cf. 8.3; syn. 及 [29.2], 比 [81.0]; 4 nc,‘ultimate, extreme’: 此謂知之至也 (Li 42).

叔 (Zuo 1.1); c tr with Oi and PrP: 堯以天 下與舜 (Meng 5A5); d tr2: 與犬肉 (GY 8.1); 予之法制,告之訓典 (Zuo 5.23); cf. 5.2.2; e itr: 怒不過奪,喜不過予 (Xun 2); 上以功勞與,則民戰;上以詩書 與,則民學問 (SJS 23). Ant. 奪 [37.11].

與  yǔ, *laʔ, conj, 1 ‘and, or’: 天地 與我 (Zhuang 2); 利與命與仁 (LY 9.1); cf. 2.3; 2 prep, ‘with’: 與朋友交而不信乎 (LY 1.4); cf. 9.4 and note at 及 [29.2].

與  yù, *lah, vn, 1 ‘join, associate with; participate, be involved in’, a tr: 誰子之與也 (Xun 11); b itr: 後死者不得與於斯文也 (LY 9.5); 小功可以與於祭乎 (Li 7); c mod (!):齊韓,周之與國也 (ZGC 3.7); d pass with C: 民不可與慮始,而可與樂成 (SJS 1); 拘禮之人,不足與言事 (SJS 1); 2 adv: 既而與為公介 (Zuo 7.2); 與聞國政三月 (SJ 47). 與  yú, *la, var. of 歟 [76.14]. 9

4

致  zhì, *trits, v

[caus < preceding], ‘cause to come, bring (about), con­vey, present’, a tr: 致孝乎鬼神 (LY 8.21); 莫之致而至者,命 也 (Meng 5A6); 欲誠其意者,先致其知 (Li 42); ‘give up (an office), resign’: 孟子致為 臣而歸 (Meng 2B10); 七十致政 (Li 5); b itr: 作羹致於樂羊 (ZGC 33.9); 輕取於其民而 重致於其君 (Guan 14); c pseudo-­tr2: 寡人 終何塗之從而致之齊 (ZGC 1.1). Cf. 來 [9.6]; ant. 去 [28.3], 取 [29.6]. n

致  zhì, *tits, var. of 至 [133.0]. 8

臺  tái, *də̂, nc, ‘terrace, platform’: 九層 之臺, 起於累土 (Lao 64); 從臺上彈人 (Zuo 7.2).

興  xīng, *həŋ, v , 1 ‘rise, come up, evolve’, i

a itr: 四民均則王道興而百姓寧 (SY 7); ant. 居 [44.5], 衰 [145.4]; b caus, ‘raise’: 難 將由我,我不為難,誰敢興之 (SY 5); ant. 廢 [53.12]; 2 nc, ‘prosperity; rise’: 漢興 已六十餘歲矣 (SJ 12).

10

舉  jǔ, *klaʔ, v , ‘lift, raise, take up’, a tr: tr

吾力足以舉百鈞,而不足以舉一羽 (Meng 1A7); ‘elect, promote’: 君子不以言 舉人 (LY 15.23); b pass: 然後十萬之師舉 矣 (Sun 2);五日而國舉 (SJ 69); 事必成 然後舉 (YDan 1). Dist. 奉 [37.5].

舉  jǔ,

Classifier 134  臼

*klaʔ, adv, ‘all, completely’, a referring to the S: 天下之民舉安 (Meng 2B12); 聲則非雅聲者舉廢 (Xun 9); b referring to the O: 君舉不信群臣乎 (Zuo 12.6). Dist. 奉 [37.5].

7

12

與/予  yǔ, *laʔ, v , ‘give’, a tr with O : 舊  jiù, *gwəʔ, v , 1 ‘be old’, a itr: 衣新而不 n

d

秦又多予金 (SJ 46); b tr with Oi: 欲與大

i

舊,則是脩也 (SY 11); b mod: 孟軻守舊

498

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

術,不通世務 (YTL 11); 2 nc, ‘sth. old/ ancient’: 舍其舊而新是謀 (Zuo 5.28); 不恭 祖舊則孝悌不備 (Guan 1). Ant. 新 [69.9].

Classifier 135  舌 2

舍  shě, *lhaʔ, v , ‘put to rest, let go, dis­ n

card, re­ject, leave behind, put aside’, a tr: 食舍肉 (Zuo 1.1); 舍其舊而新是謀 (Zuo 5.28); 以為無益而舍之者,不耘苗者也 (Meng 2A2); b itr: 操則存,舍則亡 (Meng 6A8). Ant. 取 [29.6].

舍  shè, *lhah, v , 1 ‘rest’, a itr: 原泉混 n

混,不舍晝夜 (Meng 4B18); esp. ‘lodge, spend the night’: 初,宣子田於首 山,舍于翳桑 (Zuo 7.2); b tr: 商君亡至 關下,欲舍客舍 (SJ 68); 2 nc, ‘lodging place’: 欲舍客舍 (SJ 68).

Classifier 136  舛

(HSWZ 8); b comp: 存乎人者,莫良於 眸子 (Meng 4A15); c mod: 司馬以吾故亡 其良子 (Zuo 10.21); 2 nc ‘good person’: 鄭 有叔詹堵叔師叔三良為政 (Zuo 5.7); 政 令時則百姓一,賢良服 (Xun 9). Syn. 善 [30.9].

良人  liángrén,

*raŋ-­ nin, nc, ‘goodman, hus­band’, used by a wife referring to her spouse: 其妻告其妾曰,良人出 (Meng 4B33).

Classifier 139  色 0

色  sè, *srək, n , ‘colour; countenance, c

facial expression; beauty’: 吾未見好德如 好色者也 (LY 9.18).

Classifier 140  艸 3

6

舜  Shùn, *hwins, n , a mythical sage ruler: p

堯欲傳天下於舜 (HFei 13). Cf. Focus 7.

芒芒  mángmáng, *mâŋ- mâŋ, adv, ‘wearily, exhaustedly’: 芒芒然歸 (Meng 2A2). 芒芴  mángwù,

8

舞  wǔ, *maʔ, v , ‘dance, mime’, a itr: 八 n

*mâŋ-­hmə̂t, nc, ‘tohubohu’: 雜乎芒芴之間 (Zhuang 18).

佾舞於庭 (LY 3.1); b tr: 皆衣文衣而舞康 樂 (SJ 47).

5

Classifier 137  舟

once’: 君苟有信,諸侯不貳,何患焉 (Zuo 10.13); 苟為後義而先利,不奪不 饜 (Meng 1A1); 苟富貴,勿相忘 (SJ 48). Cf. 15.4.1–㉒.

0

舟  zhōu,

*tu, nc, ‘boat’: 江海之魚吞 舟,大國之樹必巨 (SY 12).

5

船  chuán, (Mo 45).

*m-­ lon, nc, ‘ship’: 船木也

Classifier 138  艮 1

良  liáng, *raŋ, vst, 1 ‘be good, fine, excellent’, a itr: 夫賢君之治也,溫良而和

苟  gǒu, *kôʔ, conj, ‘if, if only, if really;

苦  kǔ, *khâʔ, vst, 1 ‘be bitter; unpleasant’, a itr: 疾則苦而不入 (Zhuang 13); b caus, ‘regard as bitter, resent’: 夫農,民 之所苦 (SJS 6); ‘embitter; exhaust’: 苦其 志意,勞其四肢,傷其五臟 (SJS 6); c comp: 民之內事,莫苦於農 (SJS 22); 2 nc, ‘bitter­ness’: 少嘗苦曰苦,多嘗苦曰 甘,則必以此人為不知甘苦之辯矣 (Mo 17). Ant. 甘 [99.0]; 樂 [75.11].

苗  miáo, *mau, nc, ‘shoots, sprouts’: 予 助苗長矣 (Meng 2A2).

499

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

若  ruò, *nak, pr

per, ‘you’, a S: 吾即沒, 若必師之 (SJ 47); b mod: 若國有妖乎 (Lü 15.1). Cf. Box 5.

若  ruò, *nak, vn, 1 ‘reach; be like, resemble, equal’, a tr: 病未若死 (GY 11.9); b tr with C, ‘when it comes to, as to’: 先王 將若之何 (Guan 15); c cop, with subject C: 十仞之木若箸 (Xun 21); 大巧若拙 (Zhuang 10); 吾勇不子若 (Yan 2.24); d with­in object C: 則鄰國之民仰之若父母 矣 (Meng 2A5); cf. 11.2–4; 2 conj, ‘if ’: 君 若以臣為有罪 (Zuo 10.31); cf. 15.4.1; syn. 如 [38.3]; 3 prdem, ‘such’, mod: 君子哉若人 (LY 5.3); 任公子得若魚 (Zhuang 26). 英  yīng, *ʔraŋ, n , ‘bloom, blossom; prosc

perity; heyday’: 大道之行也,與三代之 英,丘未之逮也 (Li 9); ‘person of great talent’: 賢不肖不雜則英傑至 (Xun 9).

苛  kē, *gâi, v , ‘be cruel, harsh, oppresi

sive, des­potic’, a itr: 令苛則民亂 (SY 7); 其政平,其吏不苛 (SY 7); b caus: 苛關 市之征,以難其事 (Xun 10); c mod: 苛 政猛於虎也 (Li 4). Syn. 暴 [72.11].

6

草  cǎo, *tshûʔ, n , ‘grass, herbs’: 草上之 c

風,必偃 (LY 12.19).

荊  Jīng, *kreŋ, np, alt. name of 楚 [75.9]. 茲/兹  zī, *tsə, var. of 滋 [85.10]. 7

莫  mò, *mâk, adv [< 無], ‘none, nobody, nothing’, a in vb clauses: 百姓莫敢不順上 之法 (Xun 12); with 之 as O: 莫之致而至 者,命也 (Meng 5A6); 身死莫之養也 (HFei 49); b in nom clauses: 尺地莫非其 有也,一民莫非其臣也 (Meng 2A1). Cf. 7.3.

莫  mò, *mâk, var. of 漠 [85.11]. 8

菅  jiān, *krân, var. of 姦 [38.6].

萌  méng, *mrâŋ, v , ‘sprout, bud, put i

out shoots’, a itr: 天地和同,草木萌動 (Li 6); 知者見於未萌 (SJS 1); b mod: 萌 芽未動,形兆未見 (SY 2).

著/箸  zhù, *trakh, v , ‘make manitr

fest, display’, a tr: 病鄭也,著鄭伯子罪 也 (GLiang 5.6); esp. ‘write down’: 於是老 子乃著書上下篇 (SJ 63); b pseudo-­tr2: 舉 之表旗,而著之制令 (Zuo 10.1); 請著 之竹帛,宣布天下 (SJ 10); c pass: 誠則 形,形則著,著則明 (Li 31); 如此,則 臣忠君明,治著而兵強矣 (SJS 9); esp. ‘be written’: 著於丹書 (Zuo 9.23); 著在周 之春秋 (Mo 31).

9

萬  wàn, *mâns, vi, ‘be ten thousand’, a itr: 民之所欲萬 (SJS 5); 禺筴之,商日 二百萬 (Guan 72); b mod: 天地與我並 生,而萬物與我為一 (Zhuang 2). Cf. Box 3.

葉  yè, *lap/lep, nc, ‘leaf (of a tree)’: 宋 人有為其君以象為楮葉者 (HFei 21). 葬  zàng,

*tsâŋh, vtr, ‘bury, inter, entomb’, a tr: 生事之以禮,死葬之以禮 (LY 2.5); b pass: 伯夷以將軍葬於 首陽山 之下 (HFei 33); 堯葬於穀林 . . . 舜葬於 紀市 . . . 禹葬於會稽 (Lü 10.3); c pseudo­tr2: 葬之中野,不封不樹 (Yi 8); but d itr (!): 孟子自齊葬於魯,反於齊 (Meng 2B7).

葆禱  bǎodǎo,

*pûʔ-­tûʔ, nc, ‘fort, stronghold’: 為高葆禱於王路 (Lü 22.3).

10

蓋  gài, *kâts, adv, ‘actually, presumably, proba­bly; ought to’, a in vb clause: 天下蓋 有不仁不祥者 (Mo 27); 技蓋至此乎 (Zhuang 3); 蓋取諸離 (Yi 8); b in nom clause: 蓋其母也 (Lü 9.5); 蓋非兼王之道 也 (Mo 1). Cf. 7.2.

蓏  luǒ, *rôiʔ, n , ‘fruit’: 民食果蓏蜯蛤 (HFei 49).

500

c

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

11

蔡  Cài, *tsrets n , a polity: 武王 . . . 封弟 p

叔鮮於管,弟叔度於蔡 (SJ 4). Note: 蔡, located in present-­ day Henan, was given as a fief to a brother of King Wu of Zhou in the eleventh century bc. Repeatedly haras­sed by Chu 楚 in the Chunqiu period, it was finally destroyed by Chu in 447 bc.

蔓  màn, *mans, v , ‘spread’, a itr: 無使 i

滋蔓,蔓,難圖也 (Zuo 1.1); 白莖青秀 以蔓 (Guan 58); b mod: 蔓草猶不可 除,況君之寵弟乎 (Zuo 1.1).

之愛人也,薄於聖人之愛人也 (Mo 44); 其虧彌甚者也,其尊彌薄 (Lü 2.2). Ant. 厚 [27.7].

蕭  Xiāo, *siû, n , a polity: 王怒,遂圍 p

蕭 (Zuo 7.12). Note: 蕭 was a small dependency of Song 宋, ruled by the Zi 子clan; it was annihilated by Chu 楚 in 597 bc.

薪  xīn,

*sin, nc, ‘brushwood, firewood’: 古之葬者,厚衣之以薪 (Yi 8); 明足以察秋毫之末,而不見輿薪 (Meng 1A7).

14

12

藏  cáng, *dzâŋ, v , ‘conceal; store, keep 蔽  bì, *pets, v , ‘cover, conceal, obscure, (in)’ , a tr: 輿人納之,隸人藏之 (Zuo n

tr

obfus­cate’, a tr: 女子出門,必擁蔽其面 (Li 12); 君子不蔽人之美, 不言人之惡 (HFei 30); b pass: 墨子蔽於用而不知 文,宋子蔽於欲而不知得 (Xun 21); 蔽 於委羽之山,不見日 (HNan 4).

10.4); b itr: 商賈皆欲藏於王之市 (Meng 1A7); 天子藏於四海之內,諸侯藏於境 內,大夫藏於其家,士庶人藏於篋櫝 (SY 20).

lán, *râm, n , ‘indigo; indigo plant’: 蕩/盪  dàng, *lâŋh, v , ‘move, shake’, a 藍  青,取之於藍而青於藍 (Xun 1). c

tr

tr: 齊侯與蔡姬乘舟于囿,蕩公 (Zuo 5.3); 剛柔相摩,八卦相盪 (Yi 7); b pass: 德蕩乎名,知出乎爭 (Zhuang 4); 海水 震蕩 (Zhuang 26).

蕃  fán, *ban, vi, ‘flourish, bloom’, a itr: 樹之而五穀蕃焉 (Xun 32); 男女同姓, 其生不蕃 (Zuo 5.23); 嘗為司職吏而畜 蕃息 (SJ 47); b caus: 山不槎蘗,澤不伐 夭 . . . 蕃庶物也 (GY 4.13). 蔬/疏  shū, *sra, n , 1 ‘grain; vegetac

ble’: 主人疏食水飲 (Li 22); 2 adv, ‘on grain’: 主人疏食水飲 (Li 22).

蕪  wú, *ma, vi, ‘be overgrown (with weeds)’, itr: 野蕪曠則民乃菅 (Guan 1); 楚四竟之田, 曠蕪而不可勝辟 (Mo 46). 13

薄  bó, *bâk, v , ‘be thin, shallow, insubst

stantial, sparse, meagre’, a itr: 知行淺薄 (Xun 3.5); b caus, ‘lessen, diminish’: 省刑 罰,薄稅斂 (Meng 1A5); ‘regard lightly’: 有知其厚厚而薄薄 (CQFL 1); c comp: 天

17

蘧蘧  qúqú, *ga-­ga, adv (?), ‘jumping (from sleep); awakening with a start’, hapax: 俄然覺,則蘧蘧然周也 (Zhuang 2). Classifier 141  虍 2

虎  hǔ, *hlâʔ, n , 1 ‘tiger’: 虎求百獸而食 c

之 (ZGC 14.3); 2 np, personal name: 陽虎 務取之 (HFei 33).

3

虐  nüè, *ŋauk, vi, 1 ‘be cruel, oppress’, a itr: 厲王虐,國人謗王 (GY 1.3); 將虐於 王身 (SY 1); b pseudo-­tr: 若此則上偪主 而下虐民 (HFei 13); c mod: 民之憔悴於 虐政,未有甚於此時者也 (Meng 2A1); 2 nc, ‘cruelty’: 兄敬愛弟謂之友,反友為 虐 (XS 8.3). Syn. 暴 [72.11].

501

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

5

9

處  chǔ, *k-­hlaʔ, vi, ‘reside, dwell’, a itr: 地辟舉則民留處 (Guan 1); 吾不忍 見,以是野處也 (XX 7.1); b caus, ‘place, put into place’: 早處之,使知其極 (GY 7.6); 此非吾所以居處子也 (LNZ 1.11): ‘treat’: 何以處我 (Li 4); c pseudo-­tr: 齊人 有一妻一妾而處室者 (Meng 4B33); d mod: 聖王不作,諸侯放恣,處士橫議 (Meng 3B9). Syn. 居 [44.5].

虢  Guó, *kwrâk, np, a polity: 諺所謂輔 車相依,脣亡齒寒者,其虞虢之謂也 (Zuo 5.5). Note: Guo was a small polity south of the He, whose rulers, belonging to the Ji 姬 clan, held the title of 公. It was destroyed by Jin in 655 bc, as was its neighbour, Yu 虞.

c [< preceding], ‘place, resi­dence’: 重徙則死處而無二慮 (Lü 26.3); 處雖闢, 行中正 (SJ 47). Syn. 所 [63.4].

*khwai, vi, ‘wane, decrease, dwindle’, a itr: 新穀熟而陳穀虧 (Lü 24.5); b caus, ‘lessen, deprive; harm’: 以虧 人自利也 (Mo 17); 虧姑以成婦,逆莫 大焉 (Zuo 9.2).

處  chù, *k-­hlah, n

6

虛  xū, *ha, vi, ‘be empty, void, futile’, a itr: 不信仁賢,則國空虛 (Meng 7B12); b caus: 虛其心,實其腹,弱其志,強 其骨 (Lao 3). Syn. 空 [116.3]; 盡 [108.9]; 竭 [117.9]; ant. 實 [40.11]; 滿 [85.11]; 盈 [108.4]. 7

號  hào, *ɦâuh, vtr, ‘call, name; instruct’, a tr with C: 號之曰有巢氏 (HFei 49); 號 其書曰新語 (SJ 97); c pass with C: 項羽 兵四十萬,號百萬 (SJ 8); 天子妃號曰 后,諸侯妃號曰后 (XS 1.8). Cf. note at 名 [30.3].

11

虧  kuī,

Classifier 142  虫 5

蛇  shé, *m-­lai, n , ‘snake’: 人民不勝禽 獸蟲蛇 (HFei 49).

c

6

蛤  gé, *kə̂p, n , ‘mussel’: 民食果蓏蜯蛤 (HFei 49).

c

8

蜯  bàng, *brôŋʔ, nc, ‘oyster’: 民食果蓏 蜯蛤 (HFei 49).

虞  yú, *ŋwa, v , 1 ‘worry, be anxious 蜡  Zhà, *dzrâkh, n , a sacrifice (at the (about); take pre­ cautions, prepared n

p

(against)’, a tr: 爾無我虞 (Zuo 7.15); b itr: 主虞而安,吏肅而嚴,民樸而親 (Guan 52); 2 nc, ‘anxiety; precaution’: 以 虞待不虞者勝 (Sun 3); 悔吝者,懮虞之 象也 (Yi 7).

虞  Yú, *ŋwa, n , a polity: 諺所謂輔車 p

相依,脣亡齒寒者,其虞虢之謂也 (Zuo 5.5). Note: Yu was a small polity located in the south-­ east of modern Shanxi. Its rulers, belonging to the Ji 姬 clan, held the title of 公. Yu was destroyed by Jin in 655 bc, after it had helped Jin attack Guo 虢.

end of the year): 昔者仲尼與於蜡賓 (Li 9); 子貢觀於蜡 (Li 21).

9

蝴蝶  /胡蝶húdié, *gâ-­lêp, nc, 1 ‘butterfly’: 昔者莊周夢為胡蝶 (Zhuang 2); cf. Focus 9; 2 vi, ‘be (like) a butterfly’, itr: 栩栩然胡蝶也 (Zhuang 2). 10

融融  róngróng, *luŋ-­luŋ, vi (?), ‘be united, harmonize’, hapax, itr: 大隧之 中,其樂也融融 (Zuo 1.1).

502

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

12

蟲  chóng, *druŋ, n , ‘insect, vermin’: 人 c

民不勝禽獸蟲蛇 (HFei 49).

之法,可乎 (HFei 43). Note: 術 is a technical term in political discourse, esp. of the 法家, denoting methods of governing.

9

18

Wèi, *wes, n , a polity: 鄭人以王師 蠹  dù, *tâkh, n , ‘worm, vermin, para- 衛  虢師伐衛南鄙 (Zuo 1.1). Note: Wei p

c

site’: 肉腐出蟲,魚枯生蠹 (Xun 1); 此 五者,邦之蠹也 (HFei 49).

Classifier 143  血 0

血  xuè,

*hwît, nc, ‘blood’: 凡有血氣 者,莫不尊親 (Li 31).

Classifier 144  行

Classifier 145  衣

0

行  xíng, *grâŋ, v , ‘perform, practise, tr

carry out’, a tr: 夫子加齊之卿相,得行 道焉 (Meng 2A2); 其所善者,吾將行之 (XX 4.16); b pass: 道行而志通 (Xun 22); 事行而不悖 (GY 1.3); 身不行道,不行 於妻子 (Meng 7B9); c mod: 世事變而行 道異也 (SJS 7).

行  xíng, *grâŋ, vi, 1 ‘go, walk, travel; act, conduct oneself ’, a itr: 終日在地上行止 (Lie 1); 身必安而後行 (YDan 1); b caus, ‘lead, guide’: 子行三軍,則誰與 (LY 7.11); c mod: 使其子狐庸為行人於吳 (GY 17.4). Ant. 止 [77.0]. 行  xìng/xíng, *grâŋh, n

c [< preceding], ‘conduct, behaviour’: 人有此三數行 者(Xun 3.5); esp. ‘proper conduct’: 廉而 不劌,行也 (Xun 30).

3

衍  yǎn, *janʔ, n , ‘lowlands’: 民之有口 c

也. . . 猶其有原隰衍沃也 (GY 1.3).

5

術  shù, *m-­lut, n , ‘method, tech­nique; c

(often written Wey to distinguish it from Wei 魏), located in the area of southern Hebei and northern Henan, was given as a fief to a brother of King Wu in the eleventh century bc in order to rule over several former Shang lineages. Originally an important polity, Wei lost significance in Chunqiu and Zhanguo times, until it was annihilated by Wei 魏 in 254 bc.

proce­dure’: 臣主失術而不亂者,未之有 也 (SJS 6); 主用申子之術,而官行商君

0

衣  yī, *ʔəi, nc,‘clothes, garment’: 至殺不 辜人也,扡其衣裘 (Mo 17). Cf. 服 [74.4].

衣  yì, *ʔəih, vn, ‘dress, wear’, a itr: 不耕而 食,不織而衣 (Zhuang 29); 衣冠而見之 (ZGC 11.1); b tr: 七十者衣帛食肉 (Meng 1A3); 田贊衣補衣而見荊王 (Lü 15.5). 衣  yì, *ʔəih, v , ‘clothe; cover’, a tr: 飢 n

者食之,寒者衣之 (Guan 80); 古之葬 者,厚衣之以薪 (Yi 8); b tr2: 殺囚,衣 之王服 (Zuo 10.13).

3

表  biǎo,

*pauʔ, nc, ‘outside; exterior, periphery; sign’: 虢,虞之表也 (Zuo 5.5); 明主之表易見 (HFei 27).

4

衰  shuāi,

*srui, vn, ‘decay, decline, diminish’, on­ly itr: 今周德既衰 (Zuo 5.24); 彌子色衰愛弛 (HFei 12); 氣衰則生 物不遂(Li 19). Ant. 興 [134.9], 盛 [108.6].

5

被  bèi, *baiʔ/h, v , ‘cover with, wear’, a n

tr: 且居學之士 . . . 有難不被甲 (HFei 32);

503

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

esp. ‘wear hair unbound’: 微管仲,吾其 被髮左衽矣 (LY 14.17); 東方曰夷,被髮 文身 (Li 5); b itr: 衛靈公襜被以與婦人 遊 (SY 13); c aux for pass: 國一日被 攻,雖欲事秦,不可得也 (ZGC 8.17); 信而見疑,忠而被謗 (SJ 84); cf. 4.4.2.

6

裂  liè, *rat/ret, vtr, ‘rip apart, tear’, a tr: 車裂商君 (XX 3.7); b pass: 國必裂矣 (ZGC 5.10); 商君車裂於秦 (HFei 13). 裁  cái, *dzə̂(h), v , ‘cut; curtail, restrict; n

rein in, control; decide’, tr: 故序四時,裁 萬物 (Xun 9); 故聖人裁財制物也 (HNan 11); 唯君裁之 (Zuo 5.15).

7

補  bǔ, *pâʔ, v , ‘to mend; to patch’, a tr: tr

有餘者損之,不足者補之 (Lao 77); b pass: 信衰誼缺,如羅不補 (SY 10); c mod: 田贊衣補衣而見荊王 (Lü 15.5).

裘  qiú, *gwə, nc, 1 ‘fur coat’: 至殺不辜 人也,扡其衣裘 (Mo 17); 2 vi, ‘wear a fur coat’, a itr: 是月也,天子始裘 (Li 6); b caus: 國有凍人,人主不裘 (XS 6.1). 裝  zhuāng,

Classifier 146  西 0

西  xī, *sî, n

loc, 1 ‘west’: 東西南北,莫敢 不服 (ZGC 11.5); 2 vi, ‘go/turn west’, a itr: 如此而民不西者,秦士戚而民苦也 (SJS 15); 日西則景朝多陰 (ZL 18); b caus: 主風吹萬物而西之 (SJ 25). Ant. 東 [75.4].

3

要  yào, *ʔiauh, nc, ‘essence’: 故省刑之 要,在禁文巧 (Guan 1); 故其治國 也,察要而已矣 (SJS 3). 12

覆  fù, *phuk, v

tr, ‘overturn, overthrow’, a tr: 太甲顛覆湯之典刑 (Meng 5A6); 毋覆巢,毋殺孩蟲 (Li 6); b pass: 三維 絶則覆 (Guan 1); 中流而遇風波也,船 必覆矣 (XS 3.1); c mod: 奔車 之上無仲尼,覆舟之下無伯夷 (HFei 25).

覆  fù, *phukh, v , ‘overspread, cover; tr

protect’, a tr: 仁覆天下矣 (Meng 4A1); 天 之所覆,地之所載 (Li 31); b pass: 萬物 覆焉 (Li 31).

*tsraŋ, nc, ‘baggage, luggage’: 於是約車治裝 (ZGC 11.1); 治裝衣 而下遊乎 (SY 19).

Classifier 147  見

8

0

裳  cháng,

*daŋ, nc, ‘lower garment, skirt’: 黃帝堯舜垂衣裳而天下治 (Yi 8); 童子不衣裘裳 (Li 1).

12

襁  qiǎng, *kaŋʔ, nc, ‘baby sling’: 夫如 是,則四方之民,襁負其子而至 矣,焉用稼 (LY 13.4). 16

襲  xí, *s-­ləp, v , ‘assail, raid, attack (by n

sur­prise)’, 凡師有鍾鼓曰伐,無曰 侵,輕曰襲 (Zuo 3.29); tr: 將襲鄭,夫 人將啟之 (Zuo 1.1); 鄭人襲胡,取之 (HFei 12); cf. note at 征 [60.5].

見  jiàn, *kêns, v , ‘see; meet’, a tr: 見其 n

生,不忍見其死 (Meng 1A7); esp. ‘have an audience with’: 田贊衣補衣而見荊 王 (Lü 15.5); b itr: 鄉也吾見於夫子而問 知 (LY 12.22); 吾固願見 (Meng 3A5); c aux for pass: 夫子何以知其將見殺 (Meng 7B29); 必見欺於張儀 (SJ 40); syn. 睹 [109.8]; cf. 4.4.2. Note: 見, as opposed to 視 [147.5], does not imply an inten­tional effort; one can watch without seeing: 視之不見 (Lao 14); 視之不見其 形 (Zhuang 14).

見  xiàn, *gêns, vi, a itr, ‘be visible, appear’: 龍忽不見 (LNZ 7.3); b caus, ‘make visible, present’: 不見可欲 (Lao 3).

504

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 149  言

4

規  guī, *kwe, n , ‘circle; pair of c

passes; rule’: 其曲中規 (Xun 1).

com-

0

視  shì, *giʔ, vn, ‘look at’, a tr: 視之弗 見,聽之弗聞 (Shi 1); b itr: 視乎不可 見,聽乎不可聞,為乎不可成 (Xun 12); c tr, with C, ‘regard as’: 文王視民如 傷 (Meng 4B20). Cf. note at 見 [147.0].

言  yán, *ŋan, vn, 1 ‘speak, talk (about)’, a itr: 子生五月而能言 (Zhuang 14); 國人 莫敢言 (GY 1.3); 吉凶者,言乎其失得 也,悔吝者,言乎其小疵也 (Yi 7); b tr: 夏禮,吾能言之 (LY 3.9); 2 nc,’word, speech, saying’: 子之所言,世俗之言也 (SJS 1). Syn. 语 [149.7].

9

2

c, 1 ‘kin, relatives; parents’: 莫不尊親 (Li 31); 2 vst, ‘be intimate, close to’, a itr: 父與夫孰親 (Zuo 2.2); b caus, ‘treat as close; be affection­ ate towards; care for’:鬼神非人實親,惟德 是依 (Zuo 5.5); 君子賢其賢而親其親 (Li 42); syn. 暱 [72.11].

量地而立國,計利而畜民 (Xun 10); 誰 習計會 (ZGC 11.1); b itr: 臣竊計 (ZGC 11.1); 君子計而後行,二三子其計乎 (SY 9); 2 nc, ‘plan, proposal’: 誰為此計者 乎 (SJ 79).

5

親  qīn, *tshin, n

13

覺  jué, *krûk, v , a itr, ‘be aware, conscious i

of’: 君人者亦可以覺矣 (Xun 10); b caus, ‘make aware of, alert to’: 予將以斯道覺斯 民也, 非予覺之,而誰也 (Meng 5A7).

覺  jiào, *krûkh, v , ‘wake up, awaken’, a n

itr: 俄然覺 (Zhuang 2); b tr: 覺寢而說 (HFei 7).

18

觀  guān, *kôn, vn, ‘look at, observe; contem­plate’, a tr: 父在,觀其志,父 沒,觀其行 (LY 1.11); b itr: 季桓子微服 往觀再三 (SJ 47); 孔子之周,觀於太廟 右陛之前 (SY 10). 觀  guàn, *kôns, v

[< preceding], 1 ‘show’, a tr2: 觀之伐齊以弊吳 (HFei 21); b tr with PrP: 嘉量既成,以觀四國 (ZL 78); 2 nc, ‘watchtower’: 事畢,出游於觀之上 (Li 9). n

Classifier 148  角 13

bump’, tr: 觸槐而死 (Zuo 7.2).

n

3

討  tǎo,

thûʔ, vn, ‘investigate; hold accountable, punish’, a tr: 慮世事之 變,討正法之本 (SJS 1); 子為正卿,亡 不越竟,反不討賊,非子而誰 (Zuo 7.2); b itr: 己則無禮,而討於有禮者 (Zuo 6.15).

託  tuō, *thâk, vtr, ‘entrust (sth to s.o.)’, a tr: 寡人託國於子 (HFei 34); 王之臣有託 其妻子於其友,而之楚遊者 (Meng 1B6); cf. 8.2.4; b pass: 士之託於諸侯,非 禮也 (Meng 5B6). Syn. 寄 [40.8]. 訓  xùn,

*hwəns, nc, ‘instructions, t­ eachings’: 予之法制,告之訓典 (Zuo 5.23).

記  jì, *kəh, vn, 1 ‘note, record, register’, a tr: 春秋記臣弒君者以百數 (ZGC 1.11); 上古之傳言,春秋所記 (HFei 17); b itr: 瞽史記曰 (GY 10.1); 2 nc, ‘note, notification, document’: 周記曰 (HFei 44); 後孟 嘗君出記 (ZGC 11.1). 4

觸  chù, *thok, v , ‘clash, collide, butt, n

計  jì, *kîh, v , 1 ‘calculate, reckon’, a tr:

設  shè, *nhet, v , ‘place, set up, estabtr

lish; put on; apply’, a tr: 設法度以齊 民,信賞罰以盡民能 (HFei 48); 無田祿

505

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

者不設祭器 (Li 2); 設禮容 (SJ 47); b pass: 天地設,而民生之 (SJS 7); 而公廟 之設於私家,非禮也 (Li 11).

許  xǔ, *hŋaʔ, v , ‘promise, agree, grant, tr

concede, allow’, a tr: 君其勿許 (Zuo 5.7); 君既許我矣 (Zuo 11.13); 不許惠王,而 遂殺之 (Lü 1.5); b tr with C: 君既許我殺 太子而立奚齊矣 (GY 8.1); c pass with C: 女子許嫁 (Li 1).

訟  sòng, *s-­loŋ, vn, ‘dispute, sue, litigate’, a itr: 聽訟,吾猶人也,必也使無訟乎 (LY 12.13); 閭里不訟於巷 (XY 8); b tr: 言 至而無所訟之也 (GY 8.1); 吾未見能見 其過而內自訟者也 (LY 5.27); c pseudo­tr: 辯訟公門之下 (YTL 27). 5

詐  zhà, *tsrâkh, v , 1 ‘cheat, de­ceive’, a tr: n

我無爾詐 (Zuo 7.15); b itr: 今之愚也詐而 已矣 (LY 17.16); 毋敢詐偽 (Li 6); 2 nc, ‘falsehood, deceit, fraud’: 久矣哉,由之 行詐也 (LY 9.12).

6

誠  chéng, *deŋ, v

i [perhaps < 成], 1 ‘be sincere, truth­ ful, true’: 志操精果謂之 誠,反誠為殆 (XS 8.3); a itr: 誠哉是言 也 (LY 13.11); b caus: 先誠其意 (Li 42); 2 adv, ‘truly, really; completely’, a in vb clause: 其所是焉誠美,其所得焉誠 大,其所利焉誠多 (Xun 10); b in nom clause: 子誠齊人也 (Meng 2A1); cf. 3.3.2. Syn. 信 [9.7].

詩  shī, *lhə, nc, ‘song, ode’: 詩者,中聲 之所止也 (Xun 1); 史為書,瞽為詩 (Zuo 9.14); especially the ‘(Canon of) Odes’: 詩書義之府也 (Zuo 5.27); 詩 者,志德之理,而明其指 (XS 8.5). 誅  zhū,

*tro, vtr, 1 ‘punish, execute’, a tr: 聞誅一夫紂矣,未聞弒君也 (Meng 1B8); b pass: 過匿則民勝法,罪誅則法勝 民 (SJS 5); 叔武賢而殺於兄,比干忠而 誅於君 (HSWZ 7); but c itr (!): 令尹誅而 楚姦不上聞 (HFei 49); 2 nc, ‘punishment’:

鬼神之誅,若此之憯遫也 (Mo 31); 寡人 將加大誅於子 (SY 7).

7

誡  jiè, *krə̂h, var. of 戒 [62.3]. 誨  huì, *hmə̂h, v , ‘teach, instruct’, a itr: n

瞽史教誨,耆艾修之,而後王斟酌焉 (GY 1.3); b tr: 母能食之,不能教誨之 (Xun 19).

語  yǔ, *ŋaʔ, vn, 1 ‘speak, converse, talk (about)’, a itr: 食不語, 寢不言 (LY 10.10); 有楚大夫於此,欲其子之齊語 也 (Meng 3B6); b tr: 坎井之蛙,不可與 語東海之樂 (Xun 18); 2 nc, ‘word, speech, saying’: 雍雖不敏,請事斯語矣 (LY 12.2). Syn. 言 [149.0].

語  yù, *ŋah, vn [< preceding], ‘tell (s.o. sth.)’, a tr2: 公語之故,且告之悔 (Zuo 1.1); 季桓子 . . . 語魯君為周道游 (SJ 47); b tr with PrP: 主亦有以語肥也 (GY 5.10).

說  shuō, *lhot, v , 1 ‘discuss, expound, n

explain’, a tr: 不務說其所以然 (Xun 12); 說其意,辯其辭 (SY 11); b itr: 繆公聞 之,素服廟臨,以說於眾曰 (Lü 16.4); 2 nc, ‘explanation, justification; doc­trine’: 有說則可,无說則死 (Zhuang 13); 寢兵 之說勝,則險阻不守 (Guan 4).

說  shuì, *lhots, v , ‘urge, exhort; conn

vince, per­suade’, a tr: 說大人,則藐之 (Meng 7B34); 所說出於為名高者 也,而說之以厚利 (HFei 12); b itr: 王 亦使賓滑執甘大夫襄以說於晉 (Zuo 10.9).

說  yuè, *lot, var. of 悅 [61.7]. 誦  sòng, *s-­loŋh, v , 1 ‘declaim, recite; n

recount, narrate’, a tr: 誦詩三百,授之 以政,不達 (LY 13.5); 子服堯之服,誦 堯之言 (Meng 6B2); b itr: 春誦夏弦 (Li 8); 孟子少時誦 (HSWZ 9).

誓  shì, *dats, v , 1 ‘swear, vow, make an n

oath (to s.o.)’, a tr: 舉賢而置之,聚眾而 誓之 (Li 10); b tr with C: 遂寘姜氏于城 潁,而誓之曰 (Zuo 1.1); c itr: 誓於牧

506

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

野,伐商紂 (SJ 32); d itr with C: 鮑叔乃 誓曰 (Guan 18).

8

論  lún,

*run, vn, ‘discuss, deliberate; judge, eval­uate, examine’, a tr: 論至德者 不和於俗 (SJS 1); b itr: 六合之外,聖人 存而不論,六合之內,聖人論而不議 (Zhuang 2).

請  qı̌ng, *tsheŋʔ, v , ‘beg, ask, request’, a n

tr: 請京,使居之 (Zuo 1.1); 臣請為王言 樂 (Meng 1B1); b itr: 亟請於武公,公弗 許 (Zuo 1.1); c pseudo-­tr2: 臣愚不識,請 占之五泰 (Xun 26).

誰  shéi/shuí, *dui, pr

9

諫  jiàn, *krâns, vn, ‘remonstrate, protest; warn’, a itr: 宣子驟諫 (Zuo 7.2); 事父 母幾諫 (LY 4.18); 諫於其君而不受 (Meng 2B12); b tr: 以是諫非而怒之 (Xun 13); 天下且孰肯諫女矣 (Xun 29); c mod [!]:夫人君無諫臣則失政 (SY 3).

謀  móu, *mə, vn, 1 ‘plan; plot/scheme (against)’, a itr: 為人謀而不忠乎 (LY 1.4); b tr: 彼則懼而協以謀我 (Zuo 2.6); 舍其 舊而新是謀 (Zuo 5.28); 2 nc, ‘plan; plot’, a head: 其舉果,其謀和 (SY 8); b mod: 謀 士盡其謀,百工盡其巧 (Guan 76). 謂  wèi, *wəs, v , ‘call (s.o. sth.), refer to tr

, ‘who?’, a P: 是誰 也 (Yan 1.18); b S in rhetorical questions: 誰非君臣 (Zuo 10.7); c O, in medial position: 吾誰欺 (LY 9.12); d mod: 是誰之過 與 (LY 16.1). Cf. 5.3.2.

(sth. as); say (sth. to s.o.)’, a tr: 我之謂也 (Zhuang 17); 何謂也 (LY 4.15); b tr with C: 吾必謂之學矣 (LY 1.7); 不知軍之不 可以進,而謂之進 (Sun 3); c pass with C: 是謂聖人 (HFei 8).

‘it in/on/at’: 舉 直 錯 諸 枉 (LY 12.22); cf. 8.2.5; 2 O and ptint, ‘it?’: 湯放桀,武王伐 紂,有諸 (Meng 1B15); cf. 5.3.1 and Box 16.

所謂輔車相依 (Zuo 5.5).

int

諸  zhū, *ta, fus [< 之乎], 1 O and prep, 諺  yàn, *ŋans/ŋrans, n , ‘proverb’: 諺

諸  zhū, *ta, [perhaps

< 多] prind, ‘all (the)’, only mod: 國之諸市,履賤踴貴 (Zuo 10.3); 諸先生以文學取 (HFei 49); 諸天下之士 (Lü 18.7). Cf. Box 28 and 諸 侯 [149.8].

c

謁  yè,

*ʔat, vn, ‘announce, proclaim; visit a supe­rior’, a tr: 微君言,臣故將謁 之 (HFei 36); 吾欲有謁於主君 (SJ 43); b itr: 未嘗得拜謁於前也 (SJ 105); c pseudo­tr2: 其父竊羊而謁之上 (HFei 49).

諾  nuò, *nâk, vn, ‘agree, approve, consent’, itr: 夫輕諾必寡信 (Lao 63); 公曰諾 (Meng 1B16).

諸侯  zhūhóu, *ta-­gô, n , ‘feudal lords’: 謝  xiè, *s-­lakh, v , ‘renounce, abstain c

n

諸侯各愛其國,不愛異國 (Mo 14).

談  tán, *lâm, v , ‘talk, chat’, a itr: 聞記 i

不言,無務多談 (SY 3); 從容談語,無 問不應 (XS 8.2); b mod: 公見夫談士辯 人乎 (SJ 127).

調  tiáo, *diû, v , ‘harmonize, accord, be i

in tune with’, a itr: 梨橘棗慄不同味,而 皆調於口 (HNan 17); 在內而合乎 道,出外而調於義 (HNan 8); b caus, ‘blend, tune, harmonize’: 盛五色,調五 聲,以誘其耳目 (CQFL 20); 其言足以 調陰陽,正四時,節風雨 (SY 2). Syn. 和 [30.5].

from; part, withdraw, bid farewell; apologize’, a itr: 齊人為殺彭生,以謝於魯 (Guan 18); 從而謝焉 (Li 4); 孟子謝,遂 留其婦 (LNZ 1.11); b tr: 湯乃謝罪請服 (SY 13); 乃謝客就車 (SJ 77).

10

謗  bàng, *pâŋh, vn, ‘complain about; slander, vilify’, a tr: 厲王虐,國人謗王 (GY 1.3); b itr: 管仲以公,而國人謗怨 (HFei 33). 謙  qiān, *khêm, v , 1 ‘be modest, unasi

suming, humble’, a itr: 不謙而失天

507

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

下,亡其身者,桀紂是也 (HSWZ 3); b caus, ‘humble’: 欲致賢者,必卑謙其身 (CQFL 22); 此之謂自謙 (Li 42); c mod: 能行謙德者,其惟周公乎 (HSWZ 8); 2 nc, ‘modesty’: 吾故曰謙也者,致恭以 存其位者也 (SY 10).

講  jiǎng, *krôŋʔ , v , ‘clarify, elucidate, n

explain; discuss’, a itr: 請講以所聞 (Zhuang 5); b tr: 天子乃命將帥講武 (Li 6); 選賢與能,講信修睦 (Li 9).

11

謾  màn, *mâns/mrâns, vn, ‘slander, insult’, tr: 鄉則不若,偝則謾之 (Xun 3.5). 謷/驁  áo,

*ŋâu, vn, ‘slander, vilify, mock’, a tr: 辟遠聖制,謷醜先王 (Lü 7.5); b pass, with aux: 有獨知之慮者,必 見驁於民 (SJS 1).

謹  jı̌n, *kənʔ, vn, ‘be careful, circumspect; consid­erate, polite’, a itr: 其在宗廟 朝廷,便便言,唯謹爾 (LY 10.1); 未有 不謹於禮者也 (Li 9); b tr: 謹權量,審 法度 (LY 20.1); 謹庠序之教,申之以孝 悌之義 (Meng 1A3). Syn. 慎 [61.10]. 12

譙  qiào, *dzau, vn, ‘criticize, rebuke, re­proach’, a tr: 鄉人譙之弗為動 (HFei 49); b itr: 里尉以譙于游宗,游宗以譙 于什伍,什伍以譙于長家 (Guan 4). 識  shì/shí, *lhək, v , ‘know, recognize’, a n

tr: 賜也,女以予為多學而識之者與 (LY 15.3); 行乞於市, 其妻不識也 (SJ 86); b itr: 弟子誡之,足以識矣 (HSWZ 9).

識/志  zhì, *təh, v , 1 ‘mark, rememn

ber, record’, tr: 默而識之 (LY 7.2) 由志 之,吾語女 (Xun 29); 2 nc, ‘record, document’: 安特將學雜識

志  (Xun

1); 大道之行也,與三代之 英 . . . 有志焉 (Li 9).

譏  jī,

*lhək, vn, ‘criticize, condemn, blame; ridi­cule’, a tr: 今日吾譏晏子 也,猶裸而訾高橛者 (SY 12); 稱鄭 伯,譏失教也 (Zuo 1.1); 春秋之於世事

也,善複古,譏易常 (CQFL 1); b itr: 於 讎者將壹譏而已,故擇其重者而譏焉 (GLiang 3.4).

13

議  yì, *ŋaih, v , ‘discuss, debate, argue; n

judge, criticize’, a tr: 恐天下之議我也 (SJS 1); 梁王贅其群臣而議其過 (SY 12); b itr: 寡人讀書,輪人安得議乎 (Zhuang 13).

警/䜘/儆  jı̌ng, *kreŋʔ [< 敬, 66.9], vi, ‘be alert, one one’s guard’, a itr: 退舍於 夫渠,不儆,鄭人覆之 (Zuo 8.16); 吾 邊鄙不儆,倉廩盈 (GY 7.8); b caus, ‘warn, alert, ad­ monish’: 故妖孽者,天 所以警天子諸侯也,惡夢者,所以警 士大夫也 (SY 10); 為父者以䜘其子 (Mo 31); 儆百官,誅不法 (HNan 3). Syn. 戒 [62.3]; cf. 驚 [187.13]. 14

譽  yù, *la, v , ‘praise, compliment’, a tr: n

好面譽人者,亦好背而毀之 (Zhuang 29); 吾之於人也, 誰毀誰譽 (LY 15.25); 從而譽之,謂之義 (Mo 17); b itr: 人主 孝,則名章榮,下服聽,天下譽 (Lü 14.1); ant. 毀 [79.9].

15

讀  dú, *lôk, v , ‘read (aloud)’, only tr: 桓 n

公讀書於堂上 (Zhuang 13); 昭王讀法十 餘簡而睡臥矣 (HFei 32).

16

變  biàn, *prons, vtr, ‘change, alter, modify’, a tr: 今吾欲變法以治 (SJS 1); b pass: 未聞 變於夷者也 (Meng 3A4); 雖體解吾猶未 變兮 (ChuC 1); 師長教之弗為變 (HFei 49); c mod: 遭變事而不知其權 (SJ 130). 讎/仇  chóu, *du, n , ‘enemy’: 解狐非 c

子之讎邪 (Lü 1.5); 相攻擊如仇讎 (SJ 6); syn. 敵 [66.11].

17

讒  chán, *dzrâm, nc, 1 ‘slander, defamation; slanderer’: 傷良曰讒 (Xun 2); 好言人

508

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

之惡,謂之讒 (Zhuang 31); a head: 是故 先施讒於申生 (GY 7.6); 管仲 歿矣,多 讒在側 (GY 10.1); b mod: 讒人困窮 (Lü 3.4); 故人君者,周則讒言至矣 (Xun 21); 2 vn, ‘slander, defame’, tr: 君子固讒人乎 (Xun 31); 人或讒之於孝王 (XX 3.7).

讓  ràng,

*naŋh, vn, 1 ‘decline, relinquish, abdi­cate (in favour of)’, a tr: 堯以 天下讓舜 (Lü 20.6); 其祖弗父何始有宋 而嗣讓厲公 (SJ 47); b itr: 三以天下 讓,民無得而稱焉 (LY 8.1); 2 nc, ‘deference, respect’: 故争奪生而辭讓亡焉 (Xun 23); 刑仁講讓,示民有常(Li 42).

10

谿  xī, *khê, n , ‘canyon, gorge’: 不臨深 c

谿,不知地之厚也 (Xun 1).

Classifier 151  豆 0

豆  dòu, *dôh, n , ‘sacrificial bowl’: 常陳 c

c

宋人有為其君以象為楮葉者 (HFei 21); 驅虎豹犀象而遠之 (Meng 3B9).

象  xiàng, *s-­jaŋʔ, nc, ‘figure, shape, appearance, image’: 象也者,像也 (Yi 8); 在天成象,在地成形 (Yi 7); 亦取象 於是矣 (Xun 7). Classifier 154  貝 2

*bəʔ, vn, ‘carry on the back/ shoulders’, tr: 四方之民襁負其子而至矣 (LY 13.4);負耒耜而自宋之滕 (Meng 3A5); fig: 而智日困,而自負其責也 (HNan 9).

負  fù, *bəʔ, vn, ‘turn one's back on, turn away from, ignore’, a tr: 負秦之日,太子 為糞矣 (ZGC 7.4); 且夫有高人之行 者,固見負於世 (SJS 1); b itr: 臣寧得罪 於王,無負於先王 (SY 9); 客何負於秦 哉 (SJ 87). Cf. 偝 [9.9]. 負  fù, *bəʔ, v , ‘lose, be defeated’, itr: 不 i

3

豈  qı̌, *khəiʔ, adv, ‘perhaps; how could?, surely not?’, in rhetorical questions: 我楚國 之為,豈為一人行也 (Zuo 9.28); 是豈 水之性哉 (Meng 6A2); 鬼神之有,豈可 疑哉 (Mo 31). Cf. 其 [12.6]; 庸 [53.8].

知彼而知己,一勝一負 (Sun 3); 中之者 勝,不中者負 (HFei 30). Ant. 勝 [19.10].

3

財  cái, *dzə̂, n , 1 ‘resources, material c

wealth’: 國多財則遠者來 (Guan 1); 2 vn, ‘utilize, exploit’, tr: 中國得而財之 (Xun 9); 故其形宜可得而財也(CQFL 18).

貢  gòng, *kôŋh, n , ‘tribute (item)’: 貢

Classifier 152  豕

c

之不入,寡君之罪也 (Zuo 5.4).

0

豕  shı̌, *lheʔ, n , ‘pig, swine’:

視生如 死,視富如貧,視人如豕,視吾如人 (Lie 4). c

4

象  xiàng, *s-­jaŋʔ, n , ‘elephant; ivory’:

負  fù,

Classifier 150  谷

俎豆 (SJ 47).

5

豚  tún, *lûn, nc, ‘piglet’: 至攘人犬豕雞 豚者,其不義又甚入人園圃竊桃李 (Mo 17).

4

貫  guàn, *kôns, vtr, ‘pierce, penetrate; string together’, a tr: 予一以貫之 (LY 15.3); b mod, ‘successive’: 若夫貫日而 治詳,一日而曲列之 (Xun 11); 端如貫 珠 (Li 19). 貨  huò, *hŋôih, n , ‘goods, products’: 不 c

貴難得之貨,使民不為盜 (Lao 3).

509

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

貧  pín, *brən, v , 1 ‘be poor’, a itr: 不知 st

節用裕民則民貧 (Xun 10); 孔子貧且賤 (SJ 47); b caus: 無以嗜欲貧其家 (Yan 3.26); 此貧國弱兵之教也 (SJS 3); c comp: 故墨術誠行,則天下尚儉而彌貧 (Xun 10); d mod: 貧民飢餓 (Wen 10); 2 nc, ‘pov­erty’: 視生如死,視富如貧,視人 如豕,視吾如人 (Lie 4). Dist. 貪 [154.4]; syn. 窮 [116.10], 乏 [4.4]; ant. 富 [40.9].

貪  tān, *rhə̂m, v , ‘be greedy, covet’, a n

itr: 君子…欲而不貪 (LY 20.2); 昔上世 之窮民,貪於飲食,惰於從事 (Mo 35); b tr: 上好貪利 (Xun 12); 虞公貪財 (CQFL 6). Dist. 貧 [154.4].

30); 賈人買百里奚以五羖羊之皮 (SY 2); b pass: 人不死則棺不買 (HFei 17); c pseudo-­tr2: 請買其方百金 (Zhuang 1); 馬 已死,買其骨五百金 (XX 3.4).

賀  hè, *gâih, vn, ‘congratulate’, a tr: 父母 之於子也,產男則相賀,產女則殺之 (HFei 46); 敢賀君 (HNan 12); b itr: 君臣 皆賀,而子獨不賀 (HNan 18). 6

賈  gǔ, *kâʔ, nc, ‘merchant, vendor’: 商 賈皆欲藏於王之市 (Meng 1A7); 因而賈 利之 (ZGC 11.1). Syn. 商 [30.8].

賈  jià, *krâh, var. of 價 [9.13]. 責  zhài, *tsrêkh, n , ‘charges, debt’: 誰 賂  lù, *râkh, v , ‘reward, bribe’, a tr: 賂 c

習計會,能為文收責於薛者乎 11.1); 使無券契之責(Guan 81).

(ZGC

5

貳  èr, *nis, vn, 1 ‘be double-­minded (to), dupli­citous; split loy­alty, make a ­second commitment’, a itr: 大叔命西鄙北鄙貳 於己 (Zuo 1.1); 晉信蠻夷而棄兄弟,其 執政貳也 (GY 5.8); b tr: 法不貳後王 (Xun 9); 為人臣者,無外交,不敢貳 君也 (Li 11). 貳èr, *nis, var. of二 [7.0]. 費  fèi, *phəts, n , 1 ‘expenditure, cost’: 古 c

之善賞者, 費少而勸眾 (HNan 13); 2 vtr, ‘expend, use up; squander, waste’, a tr: 老師 費財,亦無益也 (Zuo 5.33); 日費千 金,然後十萬之師舉矣 (Sun 2); b pass: 若是則戎甲俞眾,奉養必費 (Xun 15).

貴  guì, *kwəs, v , 1 ‘be expensive; valust

able, es­teemed; noble, honourable’, a itr: 國之諸市,履賤踴貴 (Zuo 10.3); 秦女 必貴而夫人斥矣 (SJ 70); b caus: 貴酒肉 之價,重其租 (SJS 2); 貴貴,尊賢,其 義一也 (Meng 5B3); c comp: 凡前貴於後 (Li 25); d mod: 貴人則為衣服 (Li 12); 2 nc, ‘valu­ables; (people of) honour, esteem, rank’: 富與貴是人之所欲也 (LY 4.5); 賤 而不肯事貴 (Xun 3.5). Ant. 卑 [24.6]; 賤 [152.8].

n

吾以天下,吾滋不從也 (Zuo 10.26); 驪 姬賂二五,使言於公 (GY 7.6); b itr: 以 璧馬賂於虞故也 (SJ 5).

賊  zéi, *dzə̂k, vn, 1 ‘murder, assassinate’, a tr: 公患之,使鉏麑賊之 (Zuo 7.2); b itr: 為主而賊,為父母而暴 (Guan 64); 2 nc, ‘murder(er), villain(y)’: 反不討賊 (Zuo 7.2); 故殘賊生而忠信亡焉 (Xun 23). Cf. note at 殺 [79.6]. 資  zī, *tsi v , ‘furnish, provide (with)’, a n

tr: 今逐客以資敵國 (SJ 87); b tr2: 陰行謀 臣,資之金玉 (SJ 87); 資之金玉寶器 (ZGC 15.5); c tr with PrP: 鎮之以辭 令,資之以幣帛 (HFei 44); 堯何以資汝 (Zhuang 18).

7

賓  bīn, *pin, n , ‘guest, visitor’: 出門如 c

見大賓,使民如承大祭 (LY 12.2); esp. ‘foreign advisor’: 則內外之費,賓客之 用 (Sun 2).

8

賜  cì, *sekh, v , ‘present, bestow; impart n

favour, pardon; ’, a tr: 王喜,多賜金錢 (SJ 118); b itr: 王雖為之賜,而令吏弗誅 (Lü 1.5); b tr2: 賜母金布 (LNZ 6.2); cf. 5.2.2.

買  mǎi, *mrêʔ, v , ‘buy’, 買鬻,易也 賤  jiàn, *dzans, v , 1 ‘be cheap; base, tr

(Mo 40); a tr: 貴即賣之,賤則買之 (SJ

st

mean, low’, a itr: 吾少也賤 (LY 9.6);

510

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

b  caus: 我將罪賤之 (Mo 13); 民不貴學 問,又不賤農 (SJS 2); c mod: 君命 召,雖賤人,大夫士必自御之 (Li 1); 2 nc, ‘base, low things/people’: 以道觀 之,物無貴賤 (Zhuang 17). Ant. 貴 [152.8].

賞  shǎng, *hjaŋʔ, n , 1 ‘reward’: 賞者所 c

以助禁也 (SJS 6); 2 vn, ‘reward’, a tr: 殺人 不為暴,賞人不為仁者,國法明也 (SJS 17); b itr: 今子賞罰而民且不仁 (Zhuang 12). Ant. 罰 [122.9].

賢  xián, *gîn, v , 1 ‘be worthy, capable, st

able’, a itr: 故君明而吏賢矣,吏賢而民 治矣 (XS 9.2); b caus: 君子賢其賢而親 其親 (Li 42); 楚王賢其言,辯其詞 (HSWZ 10); c mod:有商子者,賢人也 (SY 3.6); 2 nc, ‘worthy, man of ability’: 我 且賢之用,能之使 (HFei 33). Ant. 不 肖[1.3].

質  zhì, *tət, n , ‘substance, natural qualc

ity’: 質勝文則野,文勝質則史 (LY 6.18); 君子義以為質, 禮以行之 (LY 15.18); 性者,生之質也 (Zhuang 23). Syn. 樸 [75.12]; ant. 文 [67.0].

質  zhì, *təts, n , 1 ‘collateral; hostage’: 宋 c

及楚平,華元為質 (Zuo 7.15); 張丑為質 於燕 (ZGC 31.2); 2 vn, ‘give/take as hostage’, a tr: 歸之而質其大子, 必得大成 (Zuo 5.15); b itr: 於是為長安君約車百乘,質 於齊 (SJ 43); 故周鄭交質 (Zuo 1.3).

賦  fù, *pah, v , 1 ‘bestow, confer, contribn

ute; re­cite; collect (duties/taxes)’, a tr: 狙公 賦芧 (Zhuang 2); 賦詩斷章 (Zuo 9.28); b itr: 君子登高必賦 (HSWZ 7); 魯梁不賦 於民而財用足也 (Guan 84); 2 nc, ‘duties, contributions’: 輕田野之賦,平關市之 征 (Xun 10); 收水泉池澤之賦 (Li 6).

Classifier 155  赤 4

赦  shè, *lhakh, v , 1 ‘liberate, set free; n

pardon, grant amnesty’, a tr: 赦鄭囚,皆 禮而歸之 (Zuo 9.11); 赦其死罪,使臣

得言其說 (SY 2); b itr: 故王非能恤楚國 而赦 (SJ 41); 2 nc, ‘pardon, mercy, amnesty’: 才行反時者死無赦 (Xun 9).

Classifier 156  走 0

走  zǒu, *tsôʔ, vi, 1 ‘flee, run (toward)’, a itr: 獸見之皆走 (ZGC 14.3); 民走於山中 (Guan 80); b caus: 釁蚡黃生於楚,走之 晉 (SY 11); 走魏將暴鳶 ,得 魏 三縣 (SJ 72); c pseudo-­tr: 民之號呼而走之 (Lü 7.2); d mod: 麒麟之於走獸 (Meng 2A2); 北海則有走馬吠犬焉 (Xun 9); 2 nc, ‘run­ning’: 力不若牛,走不若馬 (Xun 9). Syn. 亡 [8.1], 奔 [37.5], 逃 [162.6]. 3

起  qı,̌ *khəʔ, v , ‘arise, rise, get up’, a itr: i

聖人復起,必從吾言矣 (Meng 2A2); 禮 起於何也 (Xun 19); b caus, ‘raise’: 於是 吳王起師 (GY 19.9); c pseudo-­tr: 當察亂 何自起,起不相愛 (Mo 14). Ant. 伏 [9.4], 居 [44.5].

5

越  yuè, *wat, vn, ‘traverse, cross, go beyond; transgress’, a tr: 子為正卿,亡 不越竟 (Zuo 7.2); 豈能踰五湖九江,越 十七厄 (Lü 14.5); b itr: 亡人不佞,失守 社稷,越在草莽 (Zuo 10.20); 外不得越 於皮膚 (HDNJ 61).

越  Yuè, *wat, np, a polity: 今王非越是 圖 (GY 19.3). Note: 越, located in modern Zhejiang (its capital Kuaiji 會稽 was located close to modern Shao­xing), was said to be the fief of descendants of the Xia dynas­ty, Si 姒 clan. Long considered a ‘barbari­an’ polity, it rose to power after defeating Wu 吳 in 473 bc. In 306 bc, it was in turn destroyed by Chu 楚. 8

趣  qù, *tshoh, v , 1 ‘hurry, hasten’, a itr: i

趣走進退,容貌不悖乎目 (SY 3); b

511

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

caus, ‘rush, spur on’ 乃命有司,趣民收 斂 (Li 6); 乃趣獄刑,毋留有罪 (Li 6); 2 adv, ‘quickly, hastily’: 公不悅,趣召晏子 (SY 2); 趣炊鑊湯煮之 (SY 9); 3 nc, ‘objective, pur­ pose, tendency’: 三子者不同 道,其趣一也 (SY 17); 由此觀之,則趣 行各異 (HNan 11).

趣  qǔ, *tshoʔ, var. of 取 [29.6]. 10

趨  qū, *tshro, vn, 1 ‘hurry, hasten, rush, run (to)’, a itr: 夫子亦放德而行,循道 而趨 (Zhuang 13); b tr: 德厚者士趨之 (SY 16); 2 adv, ‘swiftly’: 其右提彌明知 之,趨登 (Zuo 7.2). Classifier 157  足 0

足  zú, *tsok, vi, ‘be sufficient, adequate, suit­able’, a itr: 衣食足則知榮辱 (Guan 1); b caus: 今也父兄百官不我足也 (Meng 3A2); 百姓不安其處,不樂其鄉,不足 其上矣 (Xun 20); c aux for pass, ‘be enough, suitable, deserving (to be treated in a specified way)’: 丈夫不耕,草木之 實足食也 (HFei 49); 拘禮之人,不足與 言事 (SJS 1); cf. 4.4.2. 足  zú, *tsok, n , ‘foot’: 啟予足,啟予 手 (LY 8.3).

c

8

踐  jiàn, *dzanʔ, v , ‘step, tread on’, a tr: n

馬,蹄可以踐霜雪,毛可以禦風寒 (Zhuang 9); b itr: 擇土而踐,位平然後 處 (SY 18). Syn. 履[44.12].

踞  jù, *kah, vn, ‘sit; squat’, a itr: 孟子妻 獨居,踞(HSWZ 9); 莊子則方箕踞鼓 盆而歌 (Zhuang 18); b tr: 沛公方踞 床,使兩女子洗足 (SJ 8). Cf. note at 坐 [32.4]. 9

蹄  tí, *dê, nc, ‘hoof ’: 馬,蹄可以踐霜 雪,毛可以禦風寒 (Zhuang 9).

踰  yú, *lo, v , ‘pass over, transgress, go n

beyond, exceed’, a tr: 七十而從心所 欲,不踰矩 (LY 2.4); b itr: 公乃走踰於 北牆 (HFei 14); 信事,則不失天時,百 姓不踰 (HFei 32).

12

蹯  fán, *ban, n , ‘paw’: 宰夫胹熊蹯不 熟 (Zuo 7.2).

c

14

躍  yuè , *liauk, vi, ‘jump, leap’, a itr: 騏 驥一躍,不能十步 (Xun 1); b caus: 今夫 水,搏而躍之,可使過顙 (Meng 6A2). Classifier 158  身 0

4

身  shēn, *lhin, n , ‘torso, body’: 汝身非 跂  qı̌, *kheʔ/kheh, v , ‘stand on tiptoe’, 汝有也 (Zhuang 22); ‘him-/her-/oneself ’: c

n

itr: 吾嘗跂而望矣,不如登高之博見也 (Xun 1).

6

身為宋國笑 (HFei 49); ‘life’: 子路終身誦 之 (LY 9.27).

跪  gùi, *gwaiʔ, v , ‘kneel’, a itr: 登堂而

Classifier 159  車

路  lù, *râkh, nc, 1 ‘road, path’: 為高葆禱 於王路(Lü 22.3); 行於道路 (Li 21); syn. 道 [162.9]; 2 vn, ‘hit the road, run about’, itr: 是率天下而路也 (Meng 3A5).

車  chē/jū,

n

跪 (SY 3.6); b tr: 跪乘,執策分轡 (Li 1).

0

*k-­ lha nc, 1 ‘cart, waggon, carriage’: 顏路請子之車 (LY 11.8); 以木 為車 (SHJ 18); 2 adv, ‘with/by carts’: 商君 車裂 (HFei 13).

512

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

2

輿馬者,非利足也,而致千里 1).

(Xun

軍  jūn, *kwən, nc, ‘army’: 凡制軍, 萬 有二千五百人為軍 (ZL 54); 萬人為一 軍 (Guan 20); 三軍行止 (LT 4.5); 凡用兵 之法 . . . 全軍為上,破軍次之 (Sun 3).

轅  yuán, *wan, nc, ‘drawbar’: 與其父母 妻子同一轅 (Zuo 8.17).

6

Classifier 160  辛

載  zài, *tsə̂h, vn, ‘load, carry, transport’, a tr: 地能載人,不能治人也 (Xun 19); 天之所覆,地之所載 (Li 31); b itr: 頒白 者不負載於道路矣 (Meng 1A3); 使婦人 載以過朝 (Zuo 7.2).

辛  xīn, *sin, ntemp, eighth of the ‘Heavenly stems’: 八月辛未,治兵,建 而不旆 (Zuo 10.13). Cf. Focus 8.

7

5

輔  fǔ, *baʔ, n

‘prop (of a cart wheel)’: 諺所謂輔車相依 (Zuo 5.5); fig. ‘bulwark, support’: 夫將者,國之輔 也,輔周則國必強,輔隙則國必弱 (Sun 3). c,

輕  qīng, *kheŋ, vst, 1 ‘be light (weight), not impor­tant’: 無勢之謂輕 (HFei 21); a itr: 賞輕,刑威 (SJS 5); b caus, ‘take lightly, disrespect, slight’: 民游而輕其上 (SJS 6); 財者,君之所輕 (XX 2.8); c comp: 福輕乎羽 . . . 禍重乎地 (Zhuang 4); 是故得地而權彌輕 (Xun 15); 2 adv, ‘lightly, carelessly’: 吾唯不知務而輕用 吾身 (Zhuang 5). Syn. 傲 [9.11]; 慢 [61.11]; ant. 重 [166.2]. 8

輪  lún, *run, n

c, ‘wheel’: 斲輪,徐則甘 而不固 (Zhuang 13); ‘wheelwright’: 輪扁 斲輪於堂下 (Zhuang 13).

0

辜  gū, *kâ, nc, 1 ‘guilt; crime’: 天乎, 無 辜, 吾不弒君 (GYang 7.6); 弛罪廢 法,而伏其辜 (HSWZ 2); 2 vi, ‘be guilty’, a itr: 周宣王殺其臣杜伯而不辜 (Mo 8); b caus, ‘disgrace’: 朽者揚其灰,未朽者 辜其尸 (SY 11); c mod: 至殺不辜人也 (Mo 17). 6

辟  bì, *bekh, var. of 避 [162.13]. 辟  bì, *bek, n , ‘law, rule’: 士則學習法 令辟禁 (SJ 6).

c

辟/闢  pì, *bek, vtr, ‘open; develop (land)’, a tr: 今世主欲辟地治民而不審 數 (SJS 6); 將闢田野,實倉廩 (Xun 10); b pass: 上令行而荒草闢 (SJS 8); 晨 往,寢門闢矣 (Zuo 7.2). 9

辨/辯  biàn,

輮/揉  róu, *nuh, vn, ‘bend, shape’, a tr: 斲木為耜,揉木為耒 (Yi 8); 朝揉輪而 夕欲乘車 (Guan 6); 南山有竹,弗揉自 直 (SY 3); b itr: 木直中繩,輮以為輪 (Xun 1). Ant. 直 [109.3].

*brenʔ, vn, 1 ‘separate, distin­guish’, a tr: 明貴賤,辨等列 (Zuo 1.5); 辯義與不義之亂也 (Mo 17); b itr: 公 孫揮能知四國之為,而辨於其大夫之 族姓 (Zuo 9.31); 樂師辨乎聲詩 (Li 19); 2 nc, ‘dis­tinction’: 此可謂知義與不義之辯 乎 (Mo 17); 人之所以為人者 . . . 以其有 辨也 (Xun 5).

10

12

9

輿  yú, *la, n

c, ‘cart, waggon, coach’: 禮 國之幹也,敬禮之輿也 (Zuo 5.11); 假

辭  cí, *s-­lə, n , ‘word, statement, utterance’:

513

c

宰我子貢,善為說辭

(Meng

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

2A2); 說詩者,不以文害辭,不以辭害 志 (Meng 5A4); 所樂而玩者,爻之辭也 (Yi 7).

辭  cí, *s-­lə, vn, 1 ‘make excuses; refuse, reject, decline; bid farewell’, a itr: 萬物 作焉而不辭 (Lao 2); 晏子辭去,景公 拜送 (HSWZ 9); b tr: 為貧者, 辭尊居 卑,辭富居貧 (Meng 5B5); 田駢聽之 畢而辭之,客出 (Lü 26.1); 2 nc, ‘modesty, courte­sy’: 辭讓之心,禮之端也 (Meng 2A6); 故争奪生而辭讓亡焉 (Xun 23). 14

辯  biàn, *brenʔ, vst, 1 ‘be eloquent, articulate, glib’, a itr: 善者不辯,辯者不 善 (Lao 81); 子言非不辯也 (HFei 49); b caus, ‘make clear, convinc­ ing’ 辯其 言,說其義 (SY 1); 楚王賢其言,辯其 詞 (HSWZ 10); c mod: 雖有辯言,不得 以相干也 (SJS 13); 2 nc, ‘el­o­quence’: 論物 明辯謂之辯,反辯為訥 (XS 8.3); 予豈 好辯哉,予不得已也 (Meng 3B9); 夫以 孔 墨 之 辯 , 而 不 能 自 免 於 諂 諛 (X X 3.7); 3 var. of 辨 [160.9]. Classifier 161  辰

4

近  jìn, *gənʔ, vst, ‘be close, near’, a itr: 言 近而指遠者,善言也 (Meng 5A7); 齊滕 之路,不為近矣 (Meng 2B6); b caus: 遠 者近之,而舊者新之 (Mo 46); c comp: 莫近於仁,莫急於智 (CQFL 30); 撥亂 世反之正,莫近於春秋 (SJ 130); d mod, ‘near, close’: 近古之世,桀紂暴亂,而 湯武征伐 (HFei 49). Ant. 遠 [162.10].

近  jìn, *gəns, [< preceding], vn, ‘come close, ap­proach’, a tr: 知所先後,則近道 矣 (Li 42); b itr: 是故至孝近乎王 (Li 24). 迎  yìng, *ŋraŋh, v , ‘go to meet, weln

come’, a tr: 民扶老攜幼,迎君道中 (ZGC 11.1); b itr: 不親迎,則得妻,必 親迎乎 (Meng 6B1).

5

述  shù,

*m-­ lut, vn, ‘follow, proceed; pass on, hand down’, a tr: 子如不言,則 小子何述焉 (LY 17.16); 父作之,子述 之 (Li 31); b itr: 述而不作,信而好古 (LY 7.1).

6

0

辰  chén, *dən, n

temp, fifth of the ‘Earthly branches’: 戊辰,徙濟川王王梁 (SJ). Cf. Focus 8.

3

辱  rù, *nok, v , 1 ‘disgrace, humiliate’, a tr

tr: 使於四方,不辱君命 (LY 13.20); b pass: 南辱於楚 (Meng 1A5); 2 nc, ‘shame, disgrace’: 衣食足則知榮辱 (Guan 1).

6

Classifier 162  辵

農  nóng, *nûŋ, 1 nc, ‘peasant, farmer’: 所謂四民者,士農工商也 (SY 7); 2 vn, ‘farm, cultivate the land’, itr: 其民惰而不 農 (SJS 3).

逆  nì, *ŋrak, vn, ‘contravene, run counter to, violate’, a tr: 是逆天帝命也 (ZGC 14.3); b itr: 順於道不逆於倫,是之謂畜 (Li 25). Syn. 悖 [61.7].

逃  táo, *lâu, vi, ‘flee (from), evade’, a itr: 竊藏以逃 (Zuo 5.24); 王與太后奔于 莒,逃於城陽之山 (SY 12); b caus: 謹逃 其爪牙,不利其器 (Guan 78); 伍子胥 生於楚,逃之吳(SY 11); c pseudo-­tr: 逃墨必歸於楊,逃楊必歸於儒 (Meng 7B26); 少則能逃之 (Sun 3). Syn. 亡 [8.1]; 奔 [37.5]; 走 [156.0]. 退  tuì,

*thûs, vn, ‘retire, withdraw, retreat (from)’, a itr: 問其名居,不告而 退 (Zuo 7.2); 鯉退而學詩 (LY 16.13); b tr: 冉子退朝 (LY 13.14).

514

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

追  zhuī, *trui, v , ‘follow, chase, pursue’, 逮  dài, *lə̂s, v , ‘reach, arrive at; catch n

a tr: 追周宣王,射之車上 (Mo 31); 夏,追戎於濟西 (GYang 3.18); b itr: 子 羔入,追者罷 (SY 14).

7

造  zào/cào, *tshûh, v , ‘go to’, a itr: 凡四

n

up, meas­ure up to’, a tr: 吾勇不子若,功 不子逮 (Yan 2.24); b itr: 逮于袁婁而與 之盟 (GYang 8.2); 魏武侯謀事而當,群 臣莫能逮 (Xun 32).

逸  yì, *lit, var. of 佚 [9.5].

i

方之賓客造焉 (ZL 26); b caus: 君子深造 之以道 (Meng 4B14); c pseudo-­tr: 不幸而 有疾,不能造朝 (Meng 2B2).

速/遫  sù, *sôk, v , ‘be fast, quick’, a st

itr: 欲速則不達,見小利則大事不成 (LY 13.17); b caus, ‘speed up’: 所以速禍也 (Zuo 1.3); c comp: 德之流行,速於置郵 而傳命 (Meng 2A1); 危國憂主,莫速於 兵 (Guan 16). Syn. 疾 [104.5]; ant. 遲 [162.12].

通  tōng, *lhôŋ, vi, ‘be accessible; comprehensi­ble; be in contact; communicate; understand’: 我可以往,彼可以 來,曰通 (Sun 10); a itr: 道行而志通 (Xun 22); 及宋人盟于宿,始通也 (Zuo 1.1); 言語不通 (Li 5); 自以為不通乎命 (Zhuang 18); b caus, ‘make accessible, open up; penetrate’: 子不通功易事 (Meng 3B4); 開通道路,毋有障塞 (Li 6); c mod, fig. ‘comprehensive, universal’: 天地之通道 (XY 6). 通  tōng,

*lhôŋ, vn, ‘have an affair (with), commit adultery’, a tr: 齊襄公故 嘗私通魯夫人 (SJ 32); b itr: 公之優曰 施,通於驪姬 (GY 7.6); 先時慶父與哀 姜私通 (SJ 33).

逐  zhú, *driuk, vtr, ‘pursue, chase; expel, drive out’, a tr: 蒙恬為秦將,北逐 戎人 (SJ 7); b pass: 丘再逐於魯 (Zhuang 31); 斥乎齊,逐乎宋衛 (SJ 47). Dist. 遂 [162.9]. 8

進  jìn, *tsins, vi, ‘advance, proceed, move for­ ward; promote’, a itr: 禮不踰 節,義不自進 (Guan 1); b caus: 孔子 退,揖巫馬期而進之 (LY 7.31).

9

達  dá, *dât, vi, 1 ‘penetrate; arrive at, reach, suc­ceed’, a itr: 在邦必達,在家必 達 (LY 12.20); 自天子達於庶人 (Meng 2B7); b caus: 己欲達而達人 (LY 6.30); c pseu­do-­tr: 名聞不爭,未達人心 (Zhuang 4); d mod: 其後必有達人 (Zuo 10.7); 知仁勇三者,天下之達德也 (Li 31); ant. 窮 [116.10]; 2 nc, ‘success, achieve­ ment’: 何哉,爾所謂達者 (LY 12.20). 道  dào, *lûʔ, n , 1 ‘path, way, course’: 會 c

天大雨, 道不通 (SJ 48); syn. 塗 [32.10]; 路 [157.6]; fig, ‘way (of doing sth.), method, principle’: 父道聖,子道 仁,君道義,臣道忠 (SY 3); 治世不一 道, 便國不必法古 (SJS 1); 天下有道, 則 公侯能為民干城 (Zuo 8.12); in an absolute sense, ‘right way, universal princi­ple, the Dao’: 一陰一陽之謂道 (Yi 7); 故道 大,天大,地大,王亦大 (Lao 25).

道  dào, *lûh, v , < the preceding, ‘say, n

utter, ex­press’, tr: 仲尼之徒無道桓文之 事者 (Meng 1A7); 如切如磋者,道學也 (Li 42).

道  dǎo, *lûh, var. of 導 [41.13]. 遁  dùn, *lûnʔ, lûns, v , ‘turn away from, n

with­draw; flee’, itr: 諸侯救鄭,楚師夜遁 (Zuo 3.28); 遁逃反側之民 (Xun 9).

過  guō, *kôi, v , ‘pass by, walk past’, a tr: n

哭而過市 (Zuo 6.18); b itr: 有一人過曰 (ZGC 2.6).

過  guò, *kôih, nc [< preceding], 1 ‘mistake, fault’: 是誰之過與 (LY 16.1); 2 vn, ‘surpass, exceed; go too far, transgress’, a tr: 喪之法,不過三年 (CQFL 2); 由也好

515

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

勇過我 (LY 5.7); b itr: 師也過,商也不 及 (LY 11.16); 過而能改,善莫大焉 (Zuo 7.2).

遂  suì, *s-­jus, vi, 1 ‘go ahead, progress; succeed’, a itr: 功成名遂, 名譽不可虛 假 (Mo 2); b caus: 遂其賢良,順民所喜 (Lü 8.3); 遂賢良,舉長大 (Li 6); cf. 隨 [170.13]; 2 conj, ‘there­upon, then’: 遂為母 子如初 (Zuo 1.1); 孔子遂行 (SJ 47); cf. 16.3. Dist. 逐 [162.7]. 違  wéi,

*wəi, vn, 1 ‘be distant, apart from; de­part, leave; deviate from; disobey’, a itr: 無乃違乎

(GY 4.10); 當於法者賞之,違於法者誅 之 (Guan 67); b tr: 君子無終食之間違仁 (LY 4.5); 違穀七里,穀人不知 (Zuo 12.27); 不違農時,穀不可勝食也 (Meng 1A3); 2 nc, ‘devi­ance, wrongdoing’: 君人 者, 將昭德塞違 (Zuo 2.2); 今虢公動匱 百姓以逞其違 (GY 1.12).

遊  yóu, *ju, vn, ‘stroll, go for a walk; travel, wan­der’, itr: 父母在,不遠遊 (LY 4.19); 孔子西遊於衛 (Zhuang 14); 莊子 與惠子遊於濠梁之上 (Zhuang 17). 遇  yù, *ŋoh, vn, ‘meet, encounter’, a tr: 吾之不遇魯侯,天也 (Meng 1B16); ‘treat’: 秦僻在雍州,不與中國諸侯之 會盟,夷翟遇之 (SJ 5); b aux for pass: 靈 公遇賊 (SY 6); cf. 為 [87.8], 見 [147.0], 可 [30.2], 足 [157.0], 被 [145.5].

遍/徧  biàn, *pêns, vi, 1 ‘be all around, every­where; pervasive’, a itr: 其痛遍於體 (HNan 18); b caus: 徧其發揮 (Guan 33); c pseudo-­tr: 行賈遍郡國 (SJ 129); d mod: 遍國中無與立談者(Meng 4B33); 2 adv, ‘all’, a referring to O: 徧問於大夫 (GY 14.5); 出其書遍觀之 (SJ 69); cf. Box 30; b re­ferring to S: 券徧合 (ZGC 11.1). 運  yùn, *wəns, v , 1 ‘circulate, revolve; i

move’, a itr: 天道運而無所積,故萬物 成 (Zhuang 13); b caus: 桀為酒池,可以 運舟 (HSWZ 4); 言治天下如運諸掌然 (SY 7); c pseudo-­tr2: 猶運之掌也 (Meng 2A1); 2 adv, ‘circularly, in circles’: 日月運

行,一寒一暑 (Yi 7); 3 nc, ‘circulation, revolution’: 明於星辰日月之運 (HNan 15); 四時之運 (XS 6.2).

10

遠  yuǎn, *wanʔ, vst, 1 ‘be long, far, far away, dis­tant’, a itr: 聖人之行不同也,或 遠,或近,或去,或不去 (Meng 5A7); b comp: 今邯鄲去魏遠於市 (XX 2.2); c mod: 有朋自遠方來,不亦樂乎 (LY 1.1); 2 adv: 父母在,不遠遊 (LY 4.19). Ant. 近 [162.4].

遠  yuàn, *wans, [caus < preceding], v , tr

‘keep at a distance; remove; stay far from; consider far’, a tr: 敬鬼神而遠之 (LY 6.22); 是以君子遠庖廚也 (Meng 1A7); 驪姬既遠大子,乃生之言 (GY 7.6); 叟 不遠千里而來 (Meng 1A1); b pass: 則見 下節而遇卑賤,必棄遠矣 (HFei 12).

11

適  shì, *lhek, vn, 1 ‘go to, reach; be convenient, suit; satisfy’, a tr: 孔子適周,將 問禮於老子 (SJ 63); 自喻適志與 (Zhuang 2); b itr: 於是伊尹接履而 趨,遂適於湯 (HSWZ 2); 然其名尊 者,以適於時也 (SJS 18). 遫  sù, *sôk, var. of 速 [162.7]. 12

遲  chí, *dri, v , 1 ‘be slow, tardy, late’, a st

itr: 庸遲乎 (SJ 47); b comp: 痛甚者其愈 遲 (Li 38); c mod: 湍水人輕,遲水人重 (HNan 4). Ant. 速 [162.7]; 疾 [104.5].

選  xuǎn, *sonʔ, v , ‘choose, select’, a tr: n

於是選齊國中女子好者八十人 (SJ 47); b itr: 舜有天下,選於眾,舉皋陶 (LY 12.22).

遺  yí,

*wi, vtr, 1 ‘neglect, lose, leave behind’, a tr: 齊桓公飲酒醉,遺其冠 (HFei 37); b mod, ‘lost, left behind; bequeathed’: 衛之遺民男女七百有三十 人 (Zuo 4.2); 不聞先王之遺言,不知學 問之大也(Xun 1); 2 nc, ‘lost goods’: 塗不 拾遺 (SJ 47).

516

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

遺  wèi, *wih, v

[< preceding] ‘present, bequeath; send’, a tr2: 巫臣自晉遺二子書 (Zuo 8.7); b only Oi: 故書其言以遺後世 (Mo 17); 於是選齊國中女子好者八十 人 . . . 遺魯君 (SJ 47); 小人有母,. . . 未 嘗 君 之 羹 ,請以遺之 (Zuo 1.1). n

遷  qiān,

*tshan, vi, ‘move, migrate, relocate’, a itr: 吾聞出於幽谷遷于喬木 者 (Meng 3A4); esp. ‘relo­cate the capital’: 周乃東遷 (SY 18); b caus, ‘trans­fer, transport’: 毀其宗廟,遷其重器 (Meng 1B11); 遷其民於臨洮 (SJ 6); esp. ‘banish’: 盡遷之於邊城 (SJ 68); c mod: 秦之遷人 皆居蜀 (SJ 7). Syn. 徙 [60.8].

13

避/辟  bì,

*bekh, vn, ‘avoid, evade, dodge, es­cape’, a tr: 姜氏欲之,焉辟害 (Zuo 1.1); 從臺上彈人,而觀其辟丸也 (Zuo 7.2); b itr: 君去,臣亦且亡避 (SY 6); 秦子梁子以公旗辟于下道 (Zuo 3.9).

15

義,人之性邪 (Zhuang 13); b caus: 生時 不邪,死而邪之 (LNZ 2.11); c mod: 邪 辭,知其所離 (Meng 2A2); 2 nc, ‘evil; deprav­ity, cor­ruption’: 政以治民,刑以 正邪 (Zuo 1.11). Syn. 姦 [38.6].

邪  yé, *la/ja, var. of 耶 [128.3]. 6

郊  jiāo, *krâu, n , 1 ‘outskirts’: 小學在 c

公宮南之左,大學在郊 (Li 5); cf. note at 郭 [163.8]; 2 np, name of a sacrifice (in the outskirts): 於郊,故謂之郊 . . . 郊所以明 天道也 (Li 11); 3 vn, ‘sacrifice (in the outskirts)’, a itr: 魯今且郊 (SJ 47); b tr: 鬼神 視之無形,聽之無聲,然而郊天,望 山川 (HNan 20).

8

都  dū, *tâ, n , ‘city, residence’: 凡邑,有 c

宗廟先君之主曰都,無曰邑 (Zuo 3.28); 都城過百雉,國之害也 (Zuo 1.1); 秦必 興兵而圍王一都 (HFei 2).

邊  biān, *pên, n , 1 ‘border, periphery’: 郭  guō, *kwâk, n , ‘outer city wall, ramc

故上失其道則邊侵於敵 (Lü 3.3); 國多 盜賊,邊境不寧 (Li 6); 2 vn, ‘border on, be next to’, a itr: 何危爾,邊乎齊也 (GLiang 11.12); b tr: 齊邊楚,權輕 (SJ 8). Cf. note at 境 [32.11].

Classifier 163  邑 0

邑  yì, *ʔəp, nc, 1 ‘city, town’: 孔子生魯 昌平鄉 陬邑 (SJ 47); 2 vi, ‘found a city, take residence’, a itr: 邑于岐山之下居焉 (Meng 1B15); b caus: 邑之秦,使復續嬴 氏祀 (SJ 6). 4

邦  bāng,

*prôŋ, nc, ‘country, polity, state’: 在邦無怨,在家無怨 (LY 12.2); syn. 國 [31.8].

邪  xié,

*s-­la/s-­ja, vi, 1 ‘be crooked, askew; per­ verse, corrupt, evil’, a itr: 仁

c

part; sub­urbs’: 三里之城, 七里之郭, 環而攻之而不勝 (Meng 2B1); 卒之東郭 墦間之祭者 (Meng 4B33); 城郭溝池以 為固 (Li 9). Note: In the Classical pe­riod, cities, were often surrounded by two walls: an inner wall, 城 [32.7], which surrounded the city proper, and an outer wall, 郭, which sur­rounded the suburbs. Just like 城, by extension, came to designate the city itself, 郭 also came to mean ‘suburb’. The outskirts beyond the 郭 were called 郊 [163.6].

9

鄉  xiāng, *haŋ, nc, ‘community, district’: 何不嘗入一鄉一里而問之 (Mo 31). 鄉  xiàng, *haŋh, vn, ‘turn toward, face’, a tr: 君不鄉道,不志於仁,而求富 之,是富桀也 (Meng 6B9); b itr: 鄉則不 若,偝則謾之,是人之二必窮也 (Xun 3.5).

517

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

11

鄙  bı̌, *prəʔ , vst, ‘be shabby, mean, lowly’, a itr: 肉食者鄙,未能遠謀 (Zuo 3.10); b caus: 過我而不假道,鄙我也 (Zuo 7.14); 易初本迪兮, 君子所鄙 (ChuC 4); c mod: 吾少也賤,故多能鄙事 (LY 9.6). 12

鄰  lín, *rin, nc, ‘neighbour’, a head: 吾誰 與為鄰 (Zhuang 20); 德不孤,必有鄰 (Meng 1B17); b mod: 鄰國相望 (Lao 80); 攻其鄰家,殺其人民 (Mo 49). 鄭  Zhèng, *dreŋh, n , a polity: 鄭武公 p

娶于申,曰武姜 (Zuo 1.1); 晉侯秦伯圍 鄭,以其無禮於晉,且貳於楚也 (Zuo 5.30). Note: Zheng was original­ ly founded in present-­day Shanxi in 806 bc by a brother of the Zhou king Xuan. After the fall of the Western Zhou capital, it was relocated to the area of Henan. Scions of the royal Ji 姬 clan, the dukes of Zheng were powerful rulers in early Chunqiu times. But their fortunes faded in later centuries, until Zheng was finally annihilated by Han in 365 bc.

spirits were only introduced much later during the imperial period.

8

醉  zuì, *tsuts, vi, ‘get drunk, be drunk’, a itr: 醉而卧 (HFei 19); b caus: 醉之以酒而 觀其側 (Zhuang 32); 齊襄公與魯君 飲,醉之 (SJ 32). Ant. 醒 [164.9]. 9

醒  xı̌ng, *sêŋ(ʔ/h), vn, ‘be(come) sober, awaken’, itr: 眾人皆醉,我獨醒 (ChuC 7). Ant. 醉 [164.8]. 18

釁  xìn, *həns, v , ‘smear with blood (as n

a sacri­fice)’, a tr: 縛之,殺以釁鼓 (HFei 23); 將以釁鐘(Meng 1A7; b itr: 湯得伊 尹 . . . 釁以犧猳 (Lü 14.2); 且使死者而無 知也,又何釁於鐘 (SY 12).

Classifier 165  釆 13

釋  shì, *lhak, v , ‘loosen; let go of, leave, tr

鄹/鄒  Zōu, *tsro, n , a place in the

put aside, give up, neglect’, a tr: 釋椎鑿而 上 (Zhuang 13); 臣亦釋宋之圍 (Zuo 5.28); b pass: 時雪不降,冰凍消釋 (Li 6).

Classifier 164  酉

Classifier 166  里

0

0

14

p

polity of Lu, home of Confucius: 孰謂鄹人 之子知禮乎 (LY 3.15).

酉  yǒu, *juʔ, n

, tenth of the ‘Earthly branches’: 己酉,師于牛首 (Zuo 9.10). Cf. Focus 8. temp

3

酒  jiǔ, *tsiu/tsiuʔ, n , ‘alcohol, ale’: 晉侯 c

飲趙盾酒 (Zuo 7.2); 終日飲酒 (Li 19). Note: In ancient China, 酒 was a kind of ale brewed from millet, wheat, rice, or other cereals, to which diverse ingredients were added. Grape wine and distilled

里  lı,̌ *rəʔ, nc, ‘village, settlement, neighbour­hood’: 百家為里 (Guan 57); 五 家為鄰, 五鄰為里 (ZL 29); 里仁為美 (LY 4.1). 里  lı̌ , *rəʔ, n , ‘(Chinese) mile’, about one c

quar­ter of a mile: 叟不遠千里而來 (Meng 1A1); 不積跬步,無以至千里 (Xun 1).

2

重  zhòng, *droŋʔ, vst, 1 ‘be heavy; important’, a itr: 禮與食孰重 (Meng 6B1); fig.

518

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

‘lethargic’: 民農則重,重則少私義 (Lü 26.3); b caus, ‘consider important, take seriously’: 使民重死而不遠徙 (Lao 80); ‘make heavy’: 重罰輕賞,則上愛民 (SJS 4); c comp: 禍重乎地,莫之知避 (Zhuang 4); 是故得地而權彌重 (Xun 15); d mod: 重罰 不用而民自治 (HFei 49); 2 nc, ‘heavy load, burden’: 故不大其棟,不能任重,任重 莫若棟,任國莫若德 (Wen 11). Ant. 輕 [159.7].

made of bronze, as well as swords, axes, and other weap­ons. With the spread of iron in late Chunqiu times, however, the latter became the dominant metal for tools and weapons.

4

鈍/頓  dùn, *tûns, v , ‘be blunt, dull; i

used up’, a itr: 其衣煖而無綵,其兵鈍 而無刃 (Wen 1); b caus: 其用戰也勝,久 則鈍兵挫銳 (Sun 2); c mod: 鈍金必將待 礱厲然後利 (Xun 23). Ant. 利 [18.5].

重  chóng, *droŋ [< preceding], vtr, 1 ‘double, mul­tiply; repeat’, a tr: 殺十人十 重不義,必有十死罪矣 (Mo 17); 因而 重之,爻在其中矣 (Yi 8); b pass: 聲禁 重,色禁重 (Lü 1.5); 2 adv, ‘twice, re­peat­ edly’: 其以喪服見, 是重受弔也 (Zuo 10.10).

鉅  jù, *gaʔ, var. of 巨 [48.2]. -鉅子  jùzĭ, *gaʔ-­tsəʔ, n ,

4

6

c a title used by the Mo­hists, ‘grandmaster’: 墨者 有鉅子腹䵍 (Lü 1.5).

野  yě, *laʔ, v , 1 ‘barbaric; uncultivated, 銘  míng, *mêŋ, n , ‘inscription’: 故鼎銘 i

rustic, wild’, a itr: 質勝文則野,文勝質 則史 (LY 6.18); 野哉由也 (LY 13.3); 2 nc, ‘uncultivated land, wilderness’: 禹往見 之,則耕在野 (Zhuang 12); 3 adv, ‘in the wilderness’: 上古穴居而野處 (Yi 8). Ant. 田 [102.0].

5

量  liàng, *raŋh, n , ‘measure’: 度量,所 c

以立公審也 (Shen 1); 夫商君為孝公平 權衡,正度量,調輕重 (ZGC 5.17); fig. ‘moderation’: 唯酒無量,不及亂 (LY 10.8); 上無量則民乃妄 (Guan 1).

Classifier 167  金 0

云 (SJ 47).

c

7

銳  duì,

*lôts, nc, ‘lance’: 其用戰也 勝,久則鈍兵挫銳 (Sun).

鋏  jiá, *kêp, nc, ‘sword hilt; sword’: 彈 其鋏,歌曰,長鋏歸來乎 (ZGC 11.1). 8

錯/措  cù/cùo, *tshâkh, vtr, ‘put, place, settle; lay to rest; implement’, a tr: 舉直錯 諸枉 (LY 12.22); 錯法務民主張,臣之行 也 (SJS 1); 苟錯諸地而可矣 (Yi 7); b pass: 明將也,而措於屯伯 (HFei 42); c pseudo-­tr2: 措之廟,立之主 (Li 2). 錦  jı̌n,

金  jīn, *kəm, nc, ‘metal’, esp. ‘bronze’: 以 金賜其母 (HSWZ 9); a monetary unit, 20 兩, i.e. approx. 317 g of coins: 日費千 金,然後十萬之師舉矣 (Sun 2). Note: Whereas gold and silver were very rare, bronze was the metal par ex­cellence in ancient China. It was used especially for sacrifi­cial vessels, bells, and other ritual im­plements; mirrors and coins, too, were

*kəmʔ, nc, ‘brocade’: 錦衣狐 裘,諸侯之服也 (Li 13).

錢  qián, *dzan, n , ‘coin’: 景王二十一 c

年,將鑄大錢 (GY 3.5); 厚賞賜金錢財 幣 (SJ 60).

9

鍾  zhōng, *toŋ, nc, ‘bowl’, a capacity measure, approx. 128 l: 齊舊四量,豆區 釜鍾, 四升為豆,各自其四,以登於

519

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

釜,釜十則鍾 (Zuo 10.3); 食敵一 鍾,當吾二十鍾 (Sun 2). Cf. Box 26.

12

鐘  zhōng, *toŋ, n , ‘bell’: 百姓聞王鐘鼓 c

之聲 (Meng 1B1); 魯莊公鑄大鐘 (Shen 8). Cf. note at 鐸 [167.13].

13

鐸  duó, *lâk, nc, ‘bell’: 天將以夫子為 木鐸 (LY 3.24). Note: 鐸 are bronze bells with a clapper, probably used in the military; other types of bells include the 鐘, sets of which were suspended from wooden racks and struck with mallets to play ritual music, and the larger 鎛, which was individually suspended. 19

長  zhǎng, *traŋʔ, v , 1 ‘be old, senior’, a st

itr: 齊侯長矣 (GY 10.1); 太后春秋長 (SJ 51); b caus: 彼長而我長之 (Meng 6A4); c comp: 以吾一日長乎爾, 毋吾以也 (LY 11.24); 鄉人長於伯兄一歲 (Meng 6A5); d mod: 長子死焉 (Meng 1A5); e pseu­ do-­tr, ‘be superior to, pre­side over’: 天帝 令我長百獸 (XX 2.4); 2 nc, ‘elder one, senior; leader’: 長幼有序,朋友有信 (Meng 3A4); 君行仁政, 斯民親其上, 死其長矣 (Meng 1B12). Ant. 幼 [52.2].

Classifier 169  門 0

門  mén, *mə̂n, nc, 1 ‘door, gate’: 季孟自 南門入,出自東門 (Zuo 11.6); 2 vn, ‘guard/keep the gate’, itr: 勇士入其大 門,則無人門焉者 (GYang 7.5).

鑽  zuān, *tsôn, v , ‘drill, pierce, bore’, a -門人  ménrén, *mə̂n-­nin, n , ‘distr: 有聖人作,鑽燧取火 (HFei 49); 鑽 n

c

龜陳卦 (Xun 9); b itr: 乃刳龜,七十二 鑽而無遺筴 (Zhuang 26).

ciple, follow­er’: 又嘗為孔子門人 (Yan 7.11); cf. 門下 and note at 弟子 [57.4].

ménxià, *mə̂n- grâʔ, n , 鑿  zuò, *dzâuk, n , ‘chisel’: 釋椎鑿而上 -門下  ‘retinue; follow­ers’: 使人屬孟嘗君,願 (Zhuang 13).

c

c

Classifier 168  長

寄食門下 (ZGC. 11.1); 吾門下食客三千 餘人 (XX 1); 廚中有臭肉,則門下無死 士矣 (SY 8); cf. 門人.

0

3

長  cháng, *draŋ, [< 張, 57.8], v , 1 ‘be 閉  bì, *pît(s), v , ‘lock, close’, a tr: 明君 tr

st

long, tall; long-­last­ing, old’, a itr: 蓋帝堯 長,帝舜短,文王長,周公短,仲尼 長,子弓短 (Xun 5); 陰陽長短 (Li 24); 日有短長,月有死生 (Sun 6); b comp: 福莫長於無禍 (Xun 1); 今亂而無 責,則亂愈長矣 (Lü 17.3); c mod: 人皆 謂之長人而異之 (SJ 47); 吳為封豕長蛇 (Zuo 11.4); 2 adv: 君長有齊,奚以薛為 (HFei 23). Ant. 短 [111.7].

者閉其門,塞其塗 (Guan 13); b pass: 門 已閉矣 (Zuo 12.15).

4

開  kāi, *khə̂i, vtr, ‘open, lay open, clear; begin’, a tr: 開通道路 (Li 6); 秦人開關延 敵 (XS 1.1); b pass: 田野開闢 (GY 21.1); 戶開亦開 (Li 1); 引其綱而萬目開矣 (HNZ 10).

長  zhǎng, *traŋʔ, [< 張, 57.8], v , 1 ‘grow 間/閒  jiān, i

(long/old)’, a itr: 禽獸成群,草木遂長 (Zhuang 9); 予助苗長矣 (Meng 2A2); b caus: 勝如卵,余翼而長之 (Zuo 12.16); 2 nc, ‘growth, development’: 六畜皆得其 長 (Xun 9).

*krên, nloc, ‘interval, intervening space/time; (in the) midst; (in)between, among’, 居有間 (Lü 1.5); 河 水間耳,君不能自渡 (SY 17); 困於陳蔡 之閒,於是反魯 (SJ 47); 有數存焉於其 間 (Zhuang 13).

520

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

閒  jiān, *krên, var. of 間 [169.4]. 閔  mı̌n, *mrən, var. of 憫 [61.12].

降  xiáng, *grûŋ, v , ‘submit, surrender’, i

a itr: 郕降於齊師 (Zuo 3.8); 仲尼賞而魯 民易降北 (HFei 49); b caus: 齊人降鄣 (GYang 3.30).

9

闇  àn, *ʔəms, vi, ‘be dark; benighted, ignorant’, a itr: 愚者闇於成事 (SJS 1); b caus: 不下比以闇上 (Xun 3); 進則揜蔽 賢良以陰闇其主 (HFei 44); c mod: 明君 之所賞,闇君之所罰也;闇君之所 賞,明君之所殺也 (Xun 13). Ant. 明 [72.4]. 闊  kuò, *khôt, v , ‘be vast, extensive’, itr: i

不知其禍福之深大而道闊遠若是也 (HFei 20).

10

7

陛  bì, *bîʔ, nc, ‘stairs, staircase’: 孔子之 周,觀於太廟右陛之前,有金人焉 (SY 10).

除  chú, *d-­la, vtr, ‘remove, eradicate, eliminate’, a tr: 除其害者以持養之 (Xun 1); 救患除禍,則莫若明分使群 矣 (Xun 10); b pass: 如是則天下之害除 (Xun 6).

闕  jué, *got, var. of 掘 [64.8]. 13

闢  pì, *bek, var. of 辟 [160.6].

8

Classifier 170  阜 0

阜  fù, *buʔ, nc, 1 ‘hill’: 有物於此,生 於山阜 (Xun 26); syn. 丘 [1.4], 山 [46.0]; 2 vi, ‘be massive, emi­nent’, a itr: 不義則利 不阜,不祥則福不降 (GY 2.1); 韓氏其 昌阜於晉乎 (Zuo 9.26); b caus, ‘accumu­ late’: 行善而備敗,其所以阜財用衣食 者也 (GY 1.3).

陳  Chén, *drin, vn, 1 ‘present, set forth, display; deploy’: a tr: 雖有甲兵,無所陳 之 (Lao 80); b tr2: 上陳之教,而先服 之,則百姓從風矣 (HSWZ 3); c itr with PrP: 我非堯舜之道,不敢以陳於王前 (Meng 2B2); 陳之以禮樂,風之以歌謠 (HNan 9); 2 nc, ‘stairs, passage (leading to a hall)’: 堂途謂之陳(Erya 5); 美人充下 陳 (ZGC 11.1). 陳  chén, *drin, n , 1 a polity: 於是陳亂 p

4

防  fáng, *baŋ, v , ‘dike, dam up, block; tr

guard against, restrain’, a tr: 防民之 口,甚於防川 (GY 1.3); 政以一其 行,刑以防其姦 (SY 19); b pass: 障防而 清,其似知命者 (SY 17).

6

陋  lòu, *rôh, vi, ‘be shabby, narrow; narrow-­minded, ig­norant; poor’: 多見曰 閑,少見曰陋 (Xun 2); a itr: 彼眾人 者,愚而無說,陋而無度者也 (Xun 5); b mod: 一簞食,一瓢飲,在陋巷 (LY 6.11); 不明鬼神則陋民不悟 (Guan 1).

降  jiàng, *krûŋ, vi, ‘come down, descend’, a itr: 若時雨降,民大悅 (Meng 3B5); b caus, ‘send down’: 天降禍 于周 (Zuo 10.32).

(Zuo 2.5); 2 family name: 陳良之徒陳相 與其弟辛,負耒耜而自宋之滕 (Meng 3A4). Note: Chen was located in the east of mod­ern Henan and parts of Anhui. Its rulers, en­feoffed by King Wu, belonged to the clan of Gui 媯, which claimed descent from the legendary ruler Shun 舜. Chen was annihi­lated by Chu in 479 bc.

陷  xiàn, *grə̂ms, vi, ‘fall into’, a itr: 臣聞 聖王公之先封者,遺後之人法,使無 陷於惡 (GY 4.3); 若以邪臨民,陷而不 振 (GY 4.15); b caus, ‘bring to fall’: 三長史

521

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

皆害湯, 欲陷之 (SJ 122); 貪利以陷其 身(YTL 57).

陰  yīn, *ʔjəm, n , 1 ‘shadow, darkness; c

Yin’: a head: 一陰一陽之謂道 (Yi 7); 夫 水旱俱天下陰陽所為也 (SY 19); b mod, ‘dark, sinister’ 此湯之陰謀也 (Guan 80); 2 adv, ‘covertly, se­cretly’: 則陰用其言顯 棄其身矣 (HFei 12). Note: 陰 and 陽, origi­ nally denoting a shadowy and a sunny place, respectively, came to mean diverse oppositions—cold vs. warm, weak vs. strong, fe­male vs. male, etc.—the interplay of which was thought to characterize all beings: it was adduced to explain floods, earthquakes, and other natural phenome­na. The concept is particularly promi­nent in Yijing commentaries. Zou Yan is credited with having founded an entire school of thought on the principles of Yin and Yang, and in imperial times, it became a foundation of medi­cine, astro­ logy, architecture, science, and many other fields of Chinese culture.

9

陽  yáng, *laŋ, n , 1 ‘light, brightness; c

Yang’, a head: 不與陰陽俱往來 (CQFL 16); b mod: 行冬令,則陽氣不勝 (Li 6); 2 adv: 陽收其身而實疏之 (HFei 12). Cf. note at 陰 [170.8].

10

隙  xì, *khrak, nc, 1 ‘gap, crack, rift, divide’: 蠹衆而木析,隙大而牆壞 (SJS 14); 寬於死而遠於憂, 則可以上下無 隙矣 (GY 2.9); 2 vn, ‘be divided, in discord’, itr: 輔周則國必強,輔隙則國必 弱 (Sun 3). 11

13

隨  suí, *s-­wai, vn, ‘follow, go along with’, a itr: 主倡而臣和,主先而臣隨 (SJ 130); b tr: 吾為子先行,子隨我後 (ZGC 14.3); 國人爭隨之 (SJ 99). Cf. 遂 [162.9]. 隨  Suí, *s-­wai, n , a

polity: 漢東之 p 國,隨為大 (Zuo 2.6). Note: Sui was a small polity in the region of modern Hubei. Allotted to members of the royal Ji 姬 clan in Western Zhou times, it be­came a dependent of Chu 楚 in late Chunqiu times.

險  xiǎn, *hŋramʔ, v , ‘be

steep, danst gerous’, a itr: 彼其道遠而險,又有江山 (Zhuang 20); b caus: 脩其土田,險其走 集 (Zuo 10.23); c comp: 是姦人將以盜名 於晻世者也,險莫大焉 (Xun 3). Ant. 平 [51.2].

隧  suì, *s-­jus, nc, 1 ‘(underground) passage, tun­nel’: 鑿隧而入井 (Zhuang 12); 2 vn, ‘dig a tunnel’, a itr: 若闕地及泉,隧 而相見 (Zuo 1.1); b tr: 矢隧地而塵不揚 (Lie 4). 14

隰  xí, *s-­ləp, nc, ‘marshland’: 猶其有原 隰衍沃也 (GY 1.3); 善相丘陵阪險原隰 土地所宜 (Li 6). 隱  yı̌n, *ʔənʔ, v , 1 ‘hide, conceal, be n

secretive’, a itr: 父為子隱,子為父 隱,直在其中矣 (LY 13.18); 古之良史 也,書法不隱 (Zuo 7.2); b tr: 吾無隱乎 爾 (LY 7.24); 昔帝鴻氏有不才子,掩義 隱賊,好行兇德 (Zuo 6.18); 2 adv: 隱居 深宮,若心之藏於胸 (CQFL 78). Ant. 彰 [59.11].

Classifier 172  隹

障  zhàng, *taŋh, vn, ‘dike; block, hinder, ob­struct’, a tr: 欲聞枉而惡直 言,是障其源而欲其水也 (Lü 23.1); b itr: 哀公有臣外障距內比周以愚其君 (HFei 38).

5

雉  zhì, *driʔ, nc, a unit of measurement for walls, c. 2.3 m × 6.9 m; 都城過百 雉,國之害也 (Zuo 1.1); 古制國不過千 乘, 都城不過百雉 (Li 30).

522

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

9

雖  suī, *swi, conj, 1 ‘although, even if, even though’: 國一日被攻,雖欲事 秦,不可得也 (ZGC 8.17); cf. 16.2; 2 ptemp, ‘even’: 雖小道,必有可觀者焉 (LY 19.4); 雖我亦將非之 (Mo 16). 雖然  suīrán, *swi–nan, conj, ‘never-

theless’: 雖然,未聞道也 (Meng 3A2).

(Yi 7); b mod: 孟春行夏令,則雨水不 時 (Li 6).

雨  yǔ/yù,

*wah, vn ‘rain, precipitate; shower’, a itr: 自十月不雨至于五月 (Zuo 5.3); b tr: 燒荅覆之,沙石雨之 (Mo 63).

3

雪  xuě, *sot, n , ‘snow’: 馬,蹄可以踐 c

霜雪,毛可以禦風寒 (Zhuang 9).

10

雞  jī, *kê, n , ‘chicken’: 至攘人犬豕雞 c

豚者,其不義又甚入人園圃竊桃李 (Mo 17). Cf. note at 畜 [102.5].

4

雲  yún, *wən, n , ‘cloud’: 不義而富且 c

貴,於我如浮雲 (LY 7.16); 今天油然作 雲,沛然下雨 (SY 6).

雜  zá, *dzə̂p, vi, 1 ‘be mixed, diverse, hetero­genous’, a itr: 黃白雜則堅且牣 (Lü 25.2); 苦痛雜於腸胃之間 (HFei 20); b caus: 九雜天下之川 (Zhuang 33); 畫繢 之事,雜五色 (ZL 79); 2 adv: 子女不雜 處 (XS 8.1).

雷  léi, *rûi, nc, ‘thunder’: 雷乃發聲,始 電 (Li 6); 天下應之如雷霆 (Xun 8).

11

霆  tíng, *lêŋ, n , ‘lightning’: 鼓之以雷

5

6

c

離  lí, *rai, v , ‘leave, separate’, a itr: 兄弟 n

妻子離散 (Meng 1A5); 亂則離,離則弱 (Xun 9); b tr: 其離仁義亦遠矣 (Lü 21.3); 夫離法者罪 (HFei 49).

難  nán, *nân, v , ‘be difficult, problemst

atic’, a itr: 夫勉於仁者不亦難乎 (Li 32); b caus, ‘find difficult, doubtful, have a problem (with)’: 吾欲作大事,而難三公子 之徒 (GY 7.6); 先主欲立太子而卿難之 (SJ 88); c comp: 民之外事,莫難於戰 (SJS 22); 乃難乎而也 (GYang 7.8); d aux for pass.: 難得而易祿也,易祿而難畜 也 (Li 41); cf. 4.4.2. Ant. 易 [72.4].

難  nàn, *nâns, n , ‘trouble, difficulty’: c

驪姬果作難 (GY 7.3); 苟遇其時,何難 之有 (Xun 28).

Classifier 173  雨 0

雨  yǔ, *waʔ, n

c ‘rain’, a head: 若時雨 降,民大悅 (Meng 1B11); 潤之以風雨

霆,潤之以風雨 (Yi 7); 天下應之如雷 霆 (Xun 8).

9

霜  shuāng, *sraŋ, n , ‘frost’: 霜露所隊 c

(Li 31); 馬,蹄可以踐霜雪,毛可以禦 風寒 (Zhuang 9).

13

霸  bà, *prâkh, nc, 1 ‘overlord, hegemon’: 五霸桓公為盛 (Meng 6B7); 成霸安 彊,自宋始矣 (Zuo 8.18); 2 vn, ‘rule (as hegemon/over­ lord)’, a itr: 五霸不同法而霸 (SJS 1); 孔子 為政必霸,霸則吾地近焉 (SJ 47); b tr: 管仲相桓公,霸諸侯 (LY 14.17). Note: The ‘Five Hege­mons’, powerful rulers who allegedly directed the concert of rulers in lieu of the weak Zhou kings in Chunqiu times, were Huan of Qi (r. 685–643), Wen of Jin (636–628), Zhuang of Chu (613–591), Helü of Wu (514–496), and Goujian of Yue (496–470); other sources name Huan of Qi, Xiang of Song (650–637), Wen of Jin,

523

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Mu of Qin (659–621), and Zhuang of Chu (613–591); cf. note at 主 [3.4].

露  lù, *râkh, n , ‘dew’: 霜露所隊 (Li 31); c

故天為之下甘露,朱草生 (CQFL 6).

16

靈  líng,

*rêŋ, nc, 1 ‘magical power; grace’: 以君靈不死 (Zuo 10.14); 2 np, ‘Numinous’, a posth. name: 晉靈公不君 (Zuo 7.2).

Classifier 174  青 0

青  qīng, *tshêŋ, v , ‘be blue/green’, a itr: st

‘wrong’: 以是為非 (Lü 18.4); ant. 是 [72.5]; 3 conj, ‘if not, unless’: 我非堯舜之 道,不敢以陳於王前 (Meng 2B2); 故君 子非禮而不言,非禮而不動 (CQFL 82); 非吳喪越,越必喪吳 (Lü 14.5); cf. 15.4.1 and 微 [60.10].

Classifier 177  革 0

革  gé, *krə̂k, nc, ‘skin; leather; armour’: 馳車千駟,革車千乘 (Sun). Classifier 180  音

又似練絲,染之藍則青 (XS 5.3); b mod: 載青旗,衣青衣 (Li 6); 野無青草,而 民無飢色 (XS 3.9); c comp: 青,取之於 藍而青於藍 (Xun 1). Note: 青 does not equal what we call ‘blue’ or ‘green’ in English; rather, it spans the entire spectrum of the two. Hence, the sky can be called 青, and so can grass.

0

8

Classifier 181  頁

靜  jìng, *dzeŋʔ, v , ‘be quiet, tranquil, i

immobile’, a itr: 定而後能靜,靜而後能 安 (Li 42); b caus: 決裂阡陌, 以靜生民 之業而一其俗 (SJ 79).

音  yīn, *ʔəm, n , ‘sound, (musical) note, c

tone’: 音聲相和 (Lao 2); 清水音小,濁 水音大 (HNZ 4).

2

竟  jìng, *kraŋʔ, var. of 境 [32.11].

2

頂  dı̌ng, *têŋʔ, nc, ‘crown (of the head)’: 生而首上圩頂 (SJ 47). 頃  qı̌ng, *khweŋʔ, n , ‘a while, moment’: c

Classifier 175  非

有頃,聞齊將伐魯 (Mo 39); 不可少頃 舍禮義之謂也 (Xun 9).

0

非  fēi, *pəi, pt [< 不唯], negates P, a in

nom. clauses, ‘is not’: 是非君子之道 (Meng 3B8); 非酒也 (HFei 15); cf. 3.3.1; b in vb clauses, ‘by no means’: 聲非加疾 也,而聞者彰 (Xun 1).

非  fēi, *pəi, vst, 1 ‘be wrong’, a itr: 非也 (LY 15.3); 合於我,未必不非於俗也 (HNan 11); b caus, ‘re­gard as wrong, criticize, condemn, reject’: 為下則好非其 上(Xun 3.5); 法之所非,君之所取 (HFei 49); 眾聞則非之 (Mo 17); 2 nc,

3

順  shùn, *m-­luns, vn, ‘follow; comply, obey; cor­respond, be appropriate’, 行歸 而過謂之順,反順為逆 (XS 8.3); 能以 事上謂之順 (Xun 9); a tr: 人不我順 (Zuo 9.30); 順風而呼,聲非加疾也 (Xun 1); b itr: 名不正,則言不順 (LY 13.3). 須  xū, *sno, vn, ‘wait (for), await; hesitate; be slow, tardy’, a itr: 子不少須,眾 懼盡 (Zuo 8.2); 職而教之,須而待之

524

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

(Xun 9); 梁之亡可立而須也 (SJ 70); b tr: 故君子疾學修身端行,以須其時也 (SY 17).

須臾  xūyú, *sno-­jo, nc, 1 ‘moment; short while’, mod: 不如須臾之所學也 (Xun 1); 2 adv: 不可以須臾忘於法 (SJS 20). 4

頓  dùn, *tûns, var. of 鈍 [167.4]. 頌  sòng, *s-­loŋh, n , ‘hymn’: 右國頌 c

(Guan 1).

7

頸  jı̌ng, *keŋʔ, nc, ‘neck’: 折頸而死 (HFei 49). 9

顏  Yán, *ŋrân, n , family name: 紇與顏 p

氏女野合而生孔子 (SJ 47).

顯  xiǎn, *hênʔ, vst, ‘be clear, evident, conspic­ uous; prominent, notable, eminent’, a itr: 商鞅顯於西秦 (XY 3); b caus: 相秦而顯其君於天下 (Meng 5A9); c comp: 先王之教,莫榮於孝,莫顯於 忠 (Lü 4.1). Classifier 182  風 0

風  fēng, *pəm, nc, 1 ‘wind’: 草上之 風,必偃 (LY 12.19); 2 vn, ‘expose one­self to the wind’, itr: 不可以風 (Meng 2B2). Classifier 183  飛 0

飛  fēi, *pəi, v , ‘fly’, a itr: 魚游于水,鳥 i

10

類  lèi, *rus, nc, 1 ‘similarity; kind, type; class, category’: 麒麟之於走獸,鳳凰之 於飛鳥,. . . 類也 (Meng 2A2); 皆守株之 類也 (HFei 49); 2 vn, ‘be similar (to), resemble’, a tr: 於是始作八卦,以通神 明之德,以類萬物之情 (Yi 8); b itr: 人 與天地相類 (Wen 3); 此與彼同類 (Mo 45). 顙  sǎng, *sâŋʔ, nc, ‘forehead’: 今夫 水,搏而躍之,可使過顙 (Meng 6A2).

願  yuàn, *ŋons, v , ‘wish to (do sth.), n

hope for’, a tr: 願車馬,衣輕裘 (LY 5.27); 為人臣常譽先王之德厚而願之 (HFei 51); b aux: 願無伐善,無施勞 . . . 願聞子 之志 (LY 5.27). Cf. 4.4.1.

12

顧  gù, *kâh, v , ‘turn around, look back; n

14

regard, consider; have regard for’, a itr: 車中,不內顧,不疾言,不親指 (LY 10.17); 莊公奮乎勇力,不顧于行義 (Yan 1.1); b tr: 君亟定變法之慮,殆無 顧天下之議之也 (SJS 1).

飛於雲(HNan 4); b caus, ‘let fly’: 公輸子 削竹木以為鵲,成而飛之 (Mo 49); c mod: 鳳凰之於飛鳥 . . . 類也 (Meng 2A2).

Classifier 184  食 0

食  shí, *m-­lək, v , 1 ‘eat, feed on’, a tr: 夫 n

馬,陸居則食草飲水 (Zhuang 9); b itr: 子食於有喪者之側 (LY 7.9); 主人疏食 水飲 (Li 22); c pseudo-tr: 食敵一鍾,當 吾二十鍾 (Sun 2); 2 nc, ‘food, meal’: 禮與 食孰重 (Meng 6B1).

食  sì, *s-­ləkh, v

tr [caus < preceding], ‘feed, nour­ish’, a tr: 母能食之,不能教 誨之 (Xun 19); b pass: 治人者食於人 (Meng 3A4). Syn. 啗 [30.8].

2

飢  jī, *kri, v , ‘starve, be hungry’, a itr: 黎 i

民不飢不寒 (Meng 1A3); b caus: 飢之渴 之,馳之驟之,整之齊之 (Zhuang 9); c mod: 國無飢民 (Guan 83).

525

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

4

飲  yı̌n, *ʔəmʔ, vn, ‘drink’, a tr: 終日飲酒 (Li 19); b itr: 智伯與襄子飲 (SY 3); 主人 疏食水飲 (Li 22). 飲  yìn, *ʔəms, v

[caus < preceding], ‘give to drink’, a tr2: 與犬肉 . . . 飲小臣酒 (GY 8.1); b tr, only Oi: 齊王方大飲左右 (Lü 20.6). n

5); b mod: 於是發而用之,有餘金矣 (HFei 10); 武留十餘歲 (XX 7.29).

9

餬  hú, *gâ, n , 1 ‘gruel’: 餬, 饘也 (Erya c

2); 2 vn, ‘feed, fill (with gruel)’, tr: 以餬余 口 (SJ 47).

12

飭/飾  chì, *rhək, v , ‘bring in order, 饑  jī, *kəi, n , ‘starvation, hunger, famtr

arrange; instruct’, a tr: 守國之度,在飾 四維 (Guan 1); b pass: 聖人在前,貞廉 在側,競稱於義,上下皆飾 (Guan 33).

飯  fàn, *bans, nc, ‘food, meal’, esp. cooked cereals: 裹飯而往食之 (Zhuang 6); 嬰兒 . . . 以塵為飯,以塗為羹 (HFei 32). 5

飾  shì, *lhək, v , ‘decorate, embellish, tr

adorn; display’, a tr: 越人飾美女八 人,納之太宰嚭 (GY 20.1); 飾邪說,文 姦言 (Xun 4); fig., ‘raise (prices)’: 粥羔豚 者弗飾賈 (SJ 47); b pass: 婦人不飾,不 敢見舅姑 (Li 4); 小恥不飾於國 (Guan 3). Syn. 文 [67.0]; ant. 素 [120.4].

飾  chì, *rhək, var. of 飭 [184.4]. 6

養  yǎng, *jaŋʔ, v , 1 ‘feed, raise, look after’, tr

a tr: 父能生之,不能養之 (Xun 19); 養桑 麻, 育六畜也 (Guan 1.4); b pass: 五十養 於鄉,六十養於國,七十養於學 (Li 5); syn. 育 [130.4]; 2 nc, ‘nourish­ment, food’: 不 事力而養足 (HFei 49).

7

餓  è, *ŋâih, v , ‘starve’, a itr: 父母凍餓 i

(Meng 1A5); b caus: 餓其體膚,空乏其 身 (Meng 6B15); 餓主父百日而死 (HFei 14); c mod: 翳桑之餓人也 (Zuo 7.2).

餘  yú,

*la, nc, ‘remnant, surplus’, a head: 其餘不足觀也已 (LY 8.11); 無為 也,則用天下而有餘 (Zhuang 13); 人民 少而財有餘 (HFei 49); with numerals, ‘more than, over’: 淮泗之閒小國十餘 (SJ

c

ine’: 數奪民時,大饑乃來 (Lü 26.3).

饋  kuì, *gus, vn, ‘present; provision, supply (with food)’, a tr: 湯使亳眾往為 之耕,老弱饋食 (Meng 3B5); 昔者有饋 生魚於鄭子產(Meng 5A2); b tr2: 齊人饋 之餼 (Zuo 2.6); c itr: 當此之時,一饋而 十起 (HNan 13). 13

饘  zhān, *tan(ʔ)/tian(ʔ), nc, ‘gruel’: 饘 於是 (SJ 47). 14

饜  yàn, *ʔems, vn, ‘eat one’s fill; be satisfied (with), con­ tent’, a itr: 不奪不饜 (Meng 1A1); b tr: 良人出,則必饜酒肉 而後反 (Meng 4B33). Classifier 185  首 0

首  shǒu, *lhuʔ, n , ‘head’: 周公拂其首 c

(SY 3.6); fig. ‘leader’: 成者為首,不成者 為尾 (Zhuang 29); fig. ‘first’: 賦大明之首 章 (Zuo 10.1).

Classifier 186  香 0

香  xiāng, *haŋ, nc, 1 ‘fragrance, smell’: 香禁重,味禁重 (Lü 1.5); 目好色,耳 好聲,鼻好香,口好味 (Wen 3); 2 vi, ‘be fragrant, sweet-­scented’, a itr: 小米鉗 而不香 (Lü 26.6); b mod: 十步之澤,必 有香草 (SY 16). Ant. 臭 [132.4].

526

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 187  馬 0

馬  mǎ, *mrâʔ, n , ‘horse’: 其馬力竭矣 c

(Zhuang 19).

3

馳  chí, *drai, vi, ‘gallop, hurry, speed; flee’, a itr: 嗇夫馳,庶人走 (Zuo 10.17); 馳矣而兵不止 (ZGC 26.10); b caus: 飢之 渴之,馳之驟之 (Zhuang 9); c mod: 凡 用兵之法,馳車千駟,革車千乘 (Sun 2). Syn. 驅 [187.11]. 5

駕  jià, *krâih, vn, ‘yoke, harness; ride’, a tr: 衛國之法,竊駕君車罪刖 (SY 17.4); b itr: 三駕而楚不能與爭 (Zuo 9.9); 為之 駕,比門下之車客 (ZGC 11.1). 駟  sì,

*sis/slis, nc, ‘quadriga, team of four horses’: 文馬三十駟 (SJ 47).

6

駭  hài,

*grə̂ʔ, vtr, ‘startle, frighten; panic’, a tr: 以惡駭天下 (Zhuang 5); 景公 射鳥,野人駭之 (Yan 1.24); b pass: 人見 蛇則驚駭 (HFei 23); 公聞之大駭 (SY 6); c pseudo-­tr: 馬駭輿,則莫若靜之;庶 人駭政,則莫若惠之 (Xun 9); d mod: 彼有駭形而無損心 (Zhuang 6); 康叔有 駭色 (SY 3.6). Syn. 驚 [187.13].

11

驅  qū, *kho(h), vi, ‘gallop, race ahead’, a itr: 驅而之薛 (ZGC 11.1); 車驅而騶 (Li 1); b caus, ‘spur on, drive on/out’: 驅車至茆 門 (HFei 34); 周公兼夷狄,驅猛獸 (Meng 3B9); c mod: 其驅馬舒,其民依 依 (HSWZ 2). Syn. 馳 [187.3]. 驁  ào, *ŋâuh, var. of 傲 [9.11]. 驁  áo, *ŋâu, var. of 謷 [149.11]. 12

驕  jiāo, *kau, v , ‘be haughty, arrogant’, st

a itr: 君子泰而不驕,小人驕而不泰

(LY 13.26); 驕而不亡者,未之有也 (Zuo 11.13); 民固驕於愛,聽於威矣 (HFei 49); b comp: 齊桓公益驕,不務 德而務遠略 (SJ 39); c pseu­do-­tr: 驕其妻 妾 (Meng 4B33); 志意修則驕富貴 (Xun 2); 予虛爵而驕之 (Guan 35); d mod: 驕 君使疲民,則國危 (Guan 8). Syn. 傲 [9.11].

13

驚  jīng, *kreŋ, v ,‘be surprised, shocked, i

star­tled, afraid’, a itr: 襄子至橋而馬驚 (ZGC 18.4); 犬吠不驚,命曰金城 (SY 16); b caus: 莊公寤生,驚姜氏 (Zuo 1.1); 馬方駭而重驚之 (SY 9); syn. 駭 [187.6]; c mod: 如造父之遇驚馬,牽馬推車則不 能進 (HFei 35).

14

驟  zòu, *dzroh, adv, ‘repeatedly’: 宣子 驟諫 (Zuo 7.2).

19

驪  Lí, *rê/re, np, a clan: 驪姬果作難 (GY 7.3). Classifier 188  骨 0

骨  gǔ, *kût, n , ‘bone’: 弱其志,強其 骨 (Lao 3).

c

Classifier 189  高 0

高  gāo, *kâu, vst, ‘be high, tall’, a itr: 嘗 與汝登高山,履危石 (Zhuang 21); b caus: 舍農游食,而以言相高也 (SJS 3); 聖人不高山,不廣河 (HNan 19).

高  gāo/gào *kâuh, nc, ‘high place, height’: 登高而招,臂非加長也,而見 者遠 (Xun 1); 高下相傾 (Lao 2). Ant. 卑 [24.6], 下 [1.2].

527

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 191  鬥 10

鬪  dòu, *dôh, var. of 鬬 [191.14]. 14

鬬/鬪  dòu, *dôh, vi, ‘fight, struggle’, a itr: 鬬且出 (Zuo 7.2); 遇諸市朝,不反 兵而鬭 (Li 3); 若夫越趙魏而鬭於 燕,則豈楚之任也哉 (ZGC 26.22); b caus: 強者必剛鬬其意 (SJS 11); 鷄可鬭 已乎 (Lie 2); c mod: 紀渻子為周宣王養 鬭鷄 (Lie 2); 其人不讓,皆有鬭心 (Wu 2); d pseudo-­tr: 王公必將乘人而鬭其捷 (Zhuang 4). Classifier 193  鬲

Note: Located in modern Shanxi, Wei was originally given to a scion of the Zhou family as a fief in the eleventh century bc. It was annexed by Jin in 661, but reemerged as an independent state upon the partition of Jin in the fifth century bc. During the Zhanguo period, it was one of the seven great powers, until it was finally destroyed by Qin in 225 bc.

Classifier 195  魚 0

魚  yú, *ŋa, nc, 1 ‘fish’: 魚,我所欲 也,熊掌,亦我所欲也 (Meng 6A10); 2 vi, ‘be a fish’, itr: 微禹,吾其魚乎 (Zuo 10.1). 4

12

鬻/粥  yù,

*luk, vtr, ‘sell, vend’: 買 鬻,易也 (Mo 10); a tr: 鬻羔豚者弗飾 賈 (SJ 47); fig.: 是必與衍鬻吾國矣 (ZGC 9.3); b pass: 傅說轉鬻 (HFei 3); 田里不 粥,墓地不請 (Li 5).

Classifier 194  鬼 0

鬼  guı̌,

*kwəiʔ, nc, 1 ‘ghost, spirit, demon’: 致孝乎鬼神 (LY 8.21); 2 vi, ‘behave like a ghost’, itr: 故明主之行制也 天,其用人也鬼 (HFei 48). Note: Whereas 鬼 (perhaps related to 歸, hence ‘revenant’) is the most general word for ghosts (of deceased people), 神 [113.5] often refers to righteous spirits, de­ scending from heaven, that bestow rewards or mete out punishments.

5

魄  pò, *phrâk, var. of 粕 [119.5].

6

鮮  xiǎn, *senʔ, v , 1 ‘be rare, scarce, few’, i

a itr: 富而不驕者,鮮 (Zuo 11.13); 巧言 令色,鮮矣仁 (LY 1.3); b caus, ‘reduce’: 作重幣以絕民資,又鑄大鍾以鮮其繼 (GY 3.6); ‘have few’: 君之所得山戎之寶 器者,中國之所鮮也 (SY 13); 2 adv, ‘rarely’: 不度之人,鮮不為患 (Zuo 9.31).

10

鰥/矜  guān,

8

魏  Wèi, *ŋuih, n , a polity: 韓魏趙三子 p

魯  Lǔ, *râʔ, np, a polity: 武王 . . . 封弟周 公旦於曲阜,曰魯 (SJ 4); 魯,小國也 (Guan 18); 孔子為魯司寇 (Meng 6B6). Note: Lu, a small polity located in modern Shandong, was given as a fief to the duke of Zhou in the eleventh century bc. Despite being the home of Con­fucius and his school, Lu’s impor­tance declined in the Chun­qiu period, when political power was taken over by three cadet families of the ducal house. Lu was finally extin­ guished by Chu in 256 bc.

分晉 (HFei 44); 魏趙韓列為諸侯 (SJ 44).

*kwrə̂n, nc, ‘widower’: 老而無妻曰鰥 (Meng 1B5); 恤鰥寡,問 疾病 (Guan 10); 矜寡孤獨廢疾者,皆 有所養 (Li 9). Syn. 寡 [40.11].

528

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Classifier 196  鳥 0

鳥  niǎo,

*tiûʔ, nc, ‘bird’: 鳥飛於空 (Shen 9.15); 山林者,鳥獸之居也 (Xun 14). Cf. 禽 [114.8], 獸 [94.15].

3

鳴  míng, *mreŋ, vi, ‘crow, caw, moo, cry’, a itr: 牛鳴於門外 (HFei 20); b caus, ‘sound’: 小子鳴鼓而攻之,可也 (LY 11.17); c mod: 前有塵埃,則載鳴鳶 (Li 1). Note: 鳴 is a generic word used to de­scribe the sound of diverse ani­mals: birds, chicken, cattle, frogs, cicadas, etc., as well as that of drums and bells. Only for dogs, 吠 [30.4] is used instead.

黃帝  Huángdì, *gwâŋ-­têh, np, ‘Yellow Thearch’, a mythical ruler: 黃帝者,少典 之子,姓公孫,名曰軒轅 (SJ 1); 黃帝 遊乎赤水之北,登乎崑崙之丘 (Zhuang 12). Cf. Focus 7. Classifier 202  黍 0

黍  shǔ, *nhaʔ, nc, ‘(glutinous) millet’: 殺雞為黍而食之 (LY 18.7). 3

黎  lí, *rî, v ,‘be black’, a itr: 手足胼胝,面 i

目黎黑(SJ 87); b mod: 黎民不飢不寒 (Meng 1A3). Syn. 黑 [203.0]; cf. 民 [83.1].

Classifier 203  黑

7

鵠  hú,

*gûk, nc, ‘swan’: 鵠不日浴而 白,烏不日黔而黑 (Zhuang 14).

0

8

鵬  Péng, *bə̂ŋ, n , a mythical giant bird:

不日浴而白,烏不日黔而黑 (Zhuang 14); 顏色黎黑而不失其所 (Xun 32); b caus: 有知其…白白而黑黑也 (CQFL 1). Syn. 黎 [202.3]; ant. 白 [106.0].

Classifier 198  鹿

4

p

化而為鳥,其名為鵬 (Zhuang 1).

st

黔  qián,

8

麒麟  qílín, *gə-­rin, n , a mythical anic

mal, ‘uni­corn’: 麒麟之於走獸,鳳凰之 於飛鳥,. . . 類也 (Meng 2A2).

*grəm, vtr, ‘darken, blacken; tattoo’: 今瞽曰 . . . 黔者黑也 (Mo 12); a tr: 豫讓乃自黔劓,敗其形容 (HFei 14); b pass: 鵠不日浴而白,烏不日黔而黑 (Zhuang 14); c mod: 黔首安寧 (Lü 5.25).

4

Classifier 201  黃 0

黑  hēi/hè, *hmə̂k, v , ‘be black’, a itr: 鵠

黨  dǎng, *tâŋʔ, n , 1 ‘group, followers, c

黃  huáng, *gwâŋ, vst, ‘be yellow, brown’, a itr: 染於黃則黃 (Mo 3); b comp: 地有 黃,而絲假之,黃於地 (HSWZ 5); c mod: 黃帝 . . . 乃備黃冕,帶黃紳 (SY 18); 野夫 黃冠;黃冠,草服也 (Li 11). Cf. also 黃帝 [50.6],黃泉 [85.5]. Note: 黃 does not equal what we call ‘yellow’ or ‘brown’ in English; rather, it is a blend of the two. The muddy waters of the ‘Yellow River’ are just that, 黃.

party; class’: 人之過也,各於其黨 (LY 4.7); 2 vn, ‘be partisan, prejudiced’, int: 吾 聞君子不黨,君子亦黨乎 (LY 7.31).

Classifier 206  鼎 0

鼎  dı̌ng, *têŋʔ, n , ‘cauldron, tripod’: 武 c

王克商,遷九鼎于雒邑 (Zuo 2.2); 果以 鼎生烹文摯 (Lü 11.2); 故鼎銘云 (SJ 47).

529

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

g lo s s a ry

Note: 鼎 were bronze food vessels used for ancestral sacrifices (and cruel punishments, as the Lü quote indicates). Especially in Western Zhou times, they often bore long inscriptions on the inside, which apparently were no longer understood in Classical times. Thousands of these ‘tripods’ (some of which actually have four legs) have been excavated, the smallest measuring just a few inches, the biggest standing 1.37 m tall and weighing 875 kg.

Classifier 207  鼓 0

鼓  gǔ, *kâʔ, nc, 1 ‘drum’: 鼓所以任 也,所以起也,所以進也 (Guan 17); 樂於鐘鼓琴瑟 (Xun 5); 2 vn, ‘beat (a drum)’, tr: 填然鼓之 (Meng 1A3). Note: Old Chinese drums had pottery or wooden shells, covered with a leather drumhead; or they were made entirely of bronze (which may explain the association with the sound 填 *dîn). In the mili­ tary, drums were struck as an attack signal. Classifier 210  齊

齊  Qí, *dzî, n , 1 a polity: 武王 . . . 封尚 p

父於營丘,曰齊 (SJ 4); 齊大於魯 (HFei 39); 2 vi, ‘behave/speak in the way of Qi’, itr: 雖日撻而求其齊也,不可得矣 (Meng 3B6). Note: Qi, a large polity located in present-­ day Shandong, was given as a fief to the Jiang 姜 clan in the eleventh century bc. Under the rule of Duke Huan (桓公, r. 685–643 bc) it became the domi­nant power of its age. In the fifth century bc, the Tian 田 family usurped power, ending the reign of the Jiang. One of the last great Zhanguo states, Qi was annexed by Qin in 221 bc.

Classifier 211  齒 0

齒  chı̌, *k-­hjəʔ, nc, ‘teeth; old age, seniority’: 脣亡齒寒 (Zuo 5.5); 天下有達尊 三,爵一,齒一, 德一 . . . 鄉黨莫如齒 (Meng 2B2). Classifier 212  龍 0

0

齊  qí, *dzî, v , ‘regulate, order , arrange, tr

下 (Xun 10); c mod: 上以忠於世主,下 以化於齊民 (Zhuang 31).

make even’, a tr: 道之以政,齊之以刑 (LY 2.3); 必將脩禮以齊朝,正法以齊 官,平政以齊民 (Xun 10); b pass: 然後 節奏齊於朝,百事齊於官,眾庶齊於

龍  lóng, *roŋ, nc, 1 ‘dragon’: 積水成 淵,蛟龍生焉 (Xun 1); 夫龍之為蟲 也,柔可狎而騎也,然其喉下有逆鱗 徑尺,若人有嬰之者則必殺人 (HFei 12); 2 adv, ‘like a dragon’: 君子 . . . 尸居而 龍見 (Zhuang 11).

530

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Bibliography For the editions of Classical texts cited in this book, cf. the list on pp. xiv–xvi. Further editions are listed in the respective chapters in Part II. Abel-Rémusat, Jean-Pierre. Élémens de la grammaire chinoise, ou principes généraux du Kou-Wen ou style antique, et du Kouan-Hoa, c’est-à-dire, de la langue commune généralement usitée dans l’empire chinois. Paris: Imprimerie Royale, 1822. Allan, Sarah and Crispin Williams (eds.). The Guodian Laozi: Proceedings of the International Conference, Dartmouth College, May 1998. Berkeley: Society for the Study of Early China, 2000. Allan, Sarah and Xing Wen (eds.). Studies on Recently Discovered Chinese Manuscripts, Beijing: Wenwu chubanshe, 2000. Ames, Roger (tr.). Sun-tzu: The Art of Warfare. New York: Ballantine Books, 1993. Ames, Roger and Henry Rosemont, Jr. The Analects of Confucius: A Philosophical Translation, New York: Ballantine Books, 1998. Bagley, Robert. ‘Anyang Writing and the Origin of the Chinese Writing System’. In: The First Writing: Script Invention as History and Process, ed. by Stephen D. Houston. Cambridge: Cambridge University Press, 2004, 190–249. Bao Tingyi 包廷毅. Siya 死雅. Beijing: Zhongguo da baike quanshu chubanshe, 2006. Barnard, Noel. ‘The Nature of the Ch’in “Reform of the Script” as Reflected in Archeological Documents Excavated under Conditions of Control’. In: Ancient China: Studies in Early Chinese Civilization, ed. by David T. Roy and Tsien Tsuen-hsuin. Hong Kong: Chinese University Press, 1978, 181–213. Bauer, Wolfgang. Der chinesische Personenname: Die Bildungsgesetze und hauptsächlichsten Bedeutungsinhalte von Ming, Tzu und Hsiao-Ming. Wiesbaden: Harrassowitz, 1959. Bauer, Wolfgang. Die Besonderheiten der chinesischen Schrift: Ihr Einfluß auf Gesellschaft und Kultur Chinas in Vergangenheit und Gegenwart. Munich: Typographische Gesellschaft, 1991. Baxter, William H. A Handbook of Old Chinese Phonology. Berlin/New York: Mouton De Gruyter, 1992. Baxter, William  H.  and Laurent Sagart. Old Chinese: A New Reconstruction. Oxford University Press, 2014.

531

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Behr, Wolfgang. Reimende Bronzeinschriften und die Entstehung der chinesischen Endreimdichtung. PhD diss., Frankfurt/M., 1996. Behr, Wolfgang, Martin Kern, and Dirk Meyer (eds.). Reading Early Chinese Manuscripts: Texts, Contexts, Methods. Leiden: Brill, forthcoming. Bellos, David. Is that a Fish in Your Ear? Translation and the Meaning of Everything. New York: Faber and Faber, 2011. Biot, Édouard. Le Tcheou-Li ou Rites des Tcheou. 3 vols. Paris: Imprimerie Nationale, 1851. Birrell, Anne. Chinese Mythology: An Introduction. Baltimore and London: Johns Hopkins University Press, 1993. Bloom, Irene. Mencius. New York: Columbia University Press, 2009. Bodde, Derk. ‘Punctuation: Its Use in China and Elsewhere’. Rocznik Orientalistyczny 47.2 (1991), 15–23. Boltz, William G. ‘Word and Word History in the Analects: The Exegesis of Lun yü IX.1’. T’oung Pao 69.4–5 (1983), 261–71. Boltz, William  G.  ‘Textual Criticism and the Ma wang tui Lao tzu’. Harvard Journal of Asiatic Studies 44.1 (1984), 185–224. Boltz, William  G.  ‘The Etymology of the Old Chinese Numeral “Two”: Grammatical and Semantic Considerations’. Journal of the American Oriental Society 107.3 (1987), 395–9. Boltz, William G. ‘Notes on the Textual Relation between the Kuo yü and the Tso chuan’. Bulletin of the School of Oriental and African Studies 53.3 (1990), 491–502. Boltz, William  G.  The Origin and Early Development of the Chinese Writing System. New Haven: American Oriental Society, 1994. Boltz, William  G.  ‘Textual Criticism more sinico’. Early China 20 (1995), 393–405. Boltz, William  G.  ‘Manuscripts with Transmitted Counterparts’. In: New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts, ed. by Edward  L.  Shaughnessy. Berkeley: Institute of East Asian Studies, 1997, 253–83. Boltz, William  G.  ‘The Invention of Writing in China’. Oriens Extremus 42 (2000), 1–17. Boltz, William G. ‘The Composite Nature of Early Chinese Texts’. In: Text and Ritual in Early China, ed. by Martin Kern. Seattle: University of Washington Press, 2005, 50–78. Boltz, William  G.  Elementary Classical Chinese. Unpublished MS. Seattle, 2011. Boltz, William G. ‘Ergativity in Classical Chinese’. In: Encyclopedia of Chinese Language and Linguistics, ed. by Rint Sybesma, et al. Brill Online, 2015[a]. .

532

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Boltz, William  G.  ‘Binomes’. In: Encyclopedia of Chinese Language and Linguistics, ed. by Rint Sybesma, et al. Brill Online, 2015[b]. . Bonsall, G.  W.  (tr.). Records of The Warring States. Typescript, University of Hong Kong Libraries, 2005. . Brandt, J. J. Wenli Particles. Peiping: The North China Union Language School, 1929. Broman, Sven. ‘Studies on the Chou li’. Bulletin of the Museum of Far Eastern Antiquities 33 (1961), 1–89. Brooks, Bruce and Taeko Brooks. The Original Analects. Sayings of Confucius and his Successors. New York: Columbia University Press, 1998. Broschat, Michael Robert. Guiguzi: A Textual Study and Translation. Diss., University of Washington, 1985. Carson, Michael F. The Language of the Lü shih ch’un ch’iu: Some Characteristic Features of Grammar and Style in a Third Century B.C. Text. Diss., University of Washington, 1980. Chan, Alan  K.  L.  (ed.). Mencius: Contexts and Interpretations. Honolulu: University of Hawai’i Press, 2002. Chan Wing-tsit (ed.). A Source Book in Chinese Philosophy. Princeton: Princeton University Press, 1969. Chang Tsung-Tung. Metaphysik, Erkenntnis und Praktische Philosophie im Chuang-Tzu. Frankfurt/M.: Klostermann, 1982. Chavannes, Édouard. Les mémoires historiques de Sema Ts’ien. 5 vols. Paris: Leroux, 1895–1905. Chavannes, Édouard. ‘Les livres chinois avant l’invention du papier’. Journal Asiatique (1905), 1–75. Chen Cuizhu 陳翠珠. Hanyu rencheng daici kaolun 汉语人称代词考论. Beijing: Guangming ribao chubanshe, 2013. Chen Fuhua 陳復華 et al. (eds.). Gudai Hanyu cidian 古代汉语词典. Beijing: Shangwu yinshuguan, 2002. Chen Mengjia 陳夢家. Shangshu tonglun 尚書通論. Beijing: Zhonghua shuju, 1985. Chen Zhi. The Shaping of the Book of Songs: From Ritualization to Secularization, Sankt Augustin: Monumenta Serica, 2007. Cheng, Anne. Entretiens de Confucius. Paris: Éditions du Seuil, 1981. Cheng, Anne. ‘ “Un Yin, un Yang, telle est la Voie”: les origines cosmologiques du parallélisme dans la pensée chinoise’. Extrême-Orient/Extrême-Occident 11 (1988), 35–43. Cikoski, John S. Classical Chinese Word-Classes. Diss., Yale University, 1970. Cikoski, John S. ‘Three Essays on Classical Chinese Grammar, I: An Outline Sketch of Word-Classes and Sentence Structure in Classical Chinese’.

533

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Computational Analysis of Asian and African Languages 8 (1978[a]), 17–152. Cikoski, John S. ‘Three Essays on Classical Chinese Grammar II: The Existence and Function of the So-Called Modal chyi in Classical Chinese’. Computational Analysis of Asian and African Languages 9 (1978[b]), 77–131. Cikoski, John S. ‘Three Essays on Classical Chinese Grammar III: An Analysis of Some Idioms Commonly Called “Passive” in Classical Chinese’. Computational Analysis of Asian and African Languages 9 (1978[c]), 133–208. Cikoski, John S. Lexicon of Classical Chinese. Typescript. 3 vols. 1994–2011. Cook, Scott. ‘The Lüshi chunqiu and the Resolution of Philosophical Dissonance’. Harvard Journal of Asiatic Studies 62.2 (2002), 307–45. Cook, Scott. Hiding the World Within the World: Ten Uneven Discourses on Zhuangzi. Albany: State University of New York Press, 2003. Cook, Scott. ‘Chujian yunwen fenlei tanxi’ 楚簡韻文分類探析. Chutu wen­ xian yu guwenzi yanjiu 出土文獻與古文字研究 4 (2011), 215–58. Cook, Scott. The Bamboo Texts of Guodian: A Study and Complete Translation. 2 vols. Ithaca, NY: Cornell University, 2012. Coutinho, Steve. Zhuangzi and Early Chinese Philosophy: Vagueness, Transformation, and Paradox. London: Ashgate, 2004. Couvreur, Séraphin. Cheu King: Texte chinois avec une double traduction en français et en latin. Ho kien Fou: Imprimerie de la Mission Catholique, 1892. Couvreur, Séraphin. Chou King, Les annales de la Chine. Ho kien Fou: Imprimerie de la Mission Catholique, 1897. Couvreur, Séraphin. Dictionnaire classique de la langue chinoise. Ho kien Fou: Imprimerie de la Mission Catholique, 1904. Couvreur, Séraphin. Tch’ouen Ts’iou et Tso Chuan: Textes chinois avec traduction française. 3 vols. Ho kien Fou: Imprimerie de la Mission Catholique, 1914. Couvreur, Séraphin. Cérémonial. Hsien hsien: Imprimerie de la Mission Catholique, 1916. Crone, Thomas. Between Disaster, Punishment, and Blame: The Semantic Field of Guilt in Early Chinese Texts. Wiesbaden: Harrassowitz, 2020. Crump, James I. ‘The Chan-kuo Ts’e and Its Fiction.’ T’oung Pao 47.4–5 (1960), 305–75. Crump, James  I.  (tr.). Chan-kuo Ts’e. Rev. ed. Ann Arbor: The University of Michigan, 1996. Csikszentmihalyi, Mark and Tae Hyun Kim. ‘The Formation of the Analects’. In: The Analects, tr. by Simon Leys, ed. by Michael Nylan. New York: Norton, 2014, 152–65. Csikszentmihalyi, Mark and Michael Nylan. ‘Constructing Lineages and Inventing Traditions through Exemplary Figures in Early China’. T’oung Pao 89.1–3 (2003), 59–99.

534

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Dawson, Raymond. A New Introduction to Classical Chinese. Oxford: Oxford University Press, 1993. De Bary, William Theodore, and Irene Bloom (eds.). Sources of Chinese Tradition. 2nd ed., vol. 1. New York: Columbia University Press, 1999. Defoort, Carine and Nicolas Standaert (eds.). The Mozi as an Evolving Text: Different Voices in Early Chinese Thought. Leiden: Brill, 2013. DeFrancis, John. The Chinese Language: Fact and Fantasy. Honolulu: University of Hawai’i Press, 1984. Demiéville, Paul. ‘Archaïsmes de prononciation en chinois vulgaire’. T’oung Pao 40.1–3 (1950), 1–59. Dobson, W.  A.  C.  H.  Late Archaic Chinese. Toronto: University of Toronto Press, 1959. Downer, G. B. ‘Derivation by Tone-Change in Classical Chinese’. Bulletin of the School of Oriental and African Studies 22.1–3 (1959), 258–90. Downing, Angela and Philip Locke. English Grammar: A University Course. London and New York: Routledge, 2006. Durrant, Stephen, Li Wai-yee, and David Schaberg (trs.). Zuo Tradition/ Zuozhuan: Commentary on the ‘Spring and Autumn Annals’. 3 vols. Seattle: University of Washington Press, 2016. Duyvendak, J. J. L. (tr.). The Book of Lord Shang. London: Probsthain, 1928. Egan, Ronald  C.  ‘Narratives in the Tso Chuan’. Harvard Journal of Asiatic Studies 37.2 (1977), 323–52. Egerod, Søren and Else Glahn (eds.). Studia Serica Bernhard Karlgren Dedicata: Sinological Studies Dedicated to Bernhard Karlgren on His Seventieth Birthday. Copenhagen: Ejnar Munksgaard, 1959. Elman, Benjamin A. and Martin Kern (eds.). Statecraft and Classical Learning: The Rituals of Zhou in East Asian History. Leiden: Brill, 2009. van Els, Paul. Van Orakelbot tot Weblog: Lesboek Klassiek Chinees. Leiden: Leiden University Press, 2011. Emmerich, Reinhard and Hans Stumpfeldt (eds.). Und folge nun dem, was mein Herz begehrt: Festschrift für Ulrich Unger zum 70. Geburtstag. 2 vols. Hamburg: Hamburger Sinologische Gesellschaft, 2002. van Ess, Hans. Politik und Geschichtsschreibung im alten China: Pan-Ma i-tung 班馬異同. Wiesbaden: Harrassowitz, 2014. van Ess, Hans, Olga Lomová, and Dorothee Schaab-Hanke (eds.). Views from Within, Views from Beyond: Approaches to the Shiji as an Early Work of Historiography. Wiesbaden: Harrassowitz, 2015. Fischer, Paul. ‘Intertextuality in Early Chinese Masters-Texts: Shared Narratives in Shi Zi’. Asia Major (third series) 22.2 (2009), 1–34. Fölster, Max. The Imperial Collection of the Former Han and the Origins of Philology in China: A Study of the Bielu, Qilüe and Hanshu Yiwenzhi. Diss., Universität Hamburg, 2016.

535

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Forke, Alfred. Mê Ti des Sozialethikers und seiner Schüler philosophische Werke. Berlin: Vereinigung Wissenschaftlicher Verlage, 1922. Foucault, Michel. The Birth of the Clinic: An Archaeology of Medical Perception, tr. by A. M. Sheridan. London and New York: Routledge, 2003. Friedrich, Michael. ‘Chinesische Mythen und chinesische Mythologie’. In: Mythos und Mythologie, ed. by Reinhard Brandt and Steffen Schmidt. Berlin: Akademie-Verlag, 2004, 237–53. Friedrich, Michael. ‘Der editorische Bericht des Liu Hsiang 劉向 zum Chankuo ts’e 戰國策’. In: Den Jadestein erlangen. Festschrift für Harro von Senger, ed. by Monika Gänssbauer and Harro von Senger. Frankfurt/M.: Lembeck, 2009, 239–57. von der Gabelentz, Georg. Chinesische Grammatik: Mit Ausschluss des niederen Stiles und der heutigen Umgangssprache. Leipzig: T. O. Weigel, 1881. Galambos, Imre. ‘The Myth of the Qin Unification of Writing in Han Sources’. Acta Orientalia Academiae Scientiarum Hung 57.2 (2004), 181–203. Galambos, Imre. Orthography of Early Chinese Writing: Evidence From Newly Excavated Manuscripts. Budapest: Department of East Asian Studies, Eötvös Loránd University, 2006. Galambos, Imre. ‘Punctuation Marks in Medieval Chinese Manuscripts’. In: Manuscript Cultures: Mapping the Field, ed. by Jan-Ulrich Sobisch, Dmitry Bondarev, and Jörg Quenzer. Berlin/New York: De Gruyter, 2014, 341–58. Gassmann, Robert  H.  Cheng Ming. Richtigstellung der Bezeichnungen. Bern: Peter Lang, 1988. Gassmann, Robert  H.  Antikchinesisches Kalenderwesen: Die Rekonstruktion der chunqiu-zeitlichen Kalender des Fürstentums Lu und der Zhou-Könige. Bern: Peter Lang, 2002. Gassmann, Robert H. Verwandtschaft und Gesellschaft im alten China: Begriffe, Strukturen und Prozesse. Bern: Peter Lang, 2006. Gassmann, Robert  H.  and Wolfgang Behr. Antikchinesisch—Ein Lehrbuch in drei Teilen. 3 vols. Bern: Peter Lang, 2005. Gentz, Joachim. Das Gongyang zhuan: Auslegung und Kanonisierung der Frühlings- und Herbstannalen (Chunqiu). Wiesbaden: Harrassowitz, 2001. Gentz, Joachim. ‘Zwischen den Argumenten lesen: Doppelt gerichtete Parallelismen zwischen Argumenten als zentrale Thesen in frühen chines­ ischen Texten’. In: Komposition und Konnotation—Figuren der Kunstprosa im Alten China, ed. by Wolfgang Behr and Joachim Gentz. Munich: Iudicium, 2006, 35–56. Gentz, Joachim. ‘Zum Parallelismus in der chinesischen Literatur’. In: Parallelismus Membrorum, ed. by Andreas Wagner. Göttingen: Vandenhoeck and Ruprecht, 2007, 241–69. Gentz, Joachim. ‘Mohist Traces in the Early Chunqiu fanlu Chapters’. Oriens Extremus 48 (2009), 55–70.

536

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Gentz, Joachim and Dirk Meyer (eds.). Literary Forms of Argument in Early China. Leiden/Boston: Brill, 2015. Giele, Enno. ‘Excavated Manuscripts: Context and Methodology’. In: China’s Early Empires. A Re-appraisal, ed. by Michael Nylan and Michael Loewe. Cambridge: Cambridge University Press, 2010, 114–34. Giles, Herbert  A.  A Chinese-English Dictionary. Shanghai: Kelly and Walsh, 1892. Goldin, Paul R. Rituals of the Way: The Philosophy of Xunzi. Chicago/La Salle: Open Court, 1999. Goldin, Paul R. ‘Xunzi in the Light of the Guodian Manuscripts’. Early China 25 (2000), 113–46. Goldin, Paul R. ‘The Old Chinese Particles yan 焉 and an 安’. Journal of the American Oriental Society 123.1 (2003), 169–73. Goldin, Paul R. ‘Persistent Misconceptions about Chinese “Legalism” ’. Journal of Chinese Philosophy 38.1 (2011), 88–104. Goldin, Paul  R.  ‘Ancient Chinese Manuscripts: Bibliography of Materials in  Western Languages’. Last updated 24 January 2021. . Goldin, Paul R. (ed.). Dao Companion to the Philosophy of Han Fei. Dordrecht: Springer, 2013. Graham, Angus C. ‘A Probable Fusion-Word: 勿 wuh = 毋 wu + 之 jy’. Bulletin of the School of Oriental and African Studies 14.1 (1952), 139–48. Graham, Angus  C.  ‘The Final Particle “Fwu” 夫’. Bulletin of the School of Oriental and African Studies 17.1 (1955), 120–32. Graham, Angus C. ‘The Relation between the Final Particles yu 與 and yee 也’. Bulletin of the School of Oriental and African Studies 19.1 (1957), 105–23. Graham, Angus  C.  ‘The Archaic Chinese Pronouns’. Asia Major 15 (1969), 17–61. Graham, Angus C. ‘The Classical Chinese Topic-Marker fu 夫’. Bulletin of the School of Oriental and African Studies 35.1 (1972), 85–110. Graham, Angus C. Later Mohist Logic, Ethics and Science. Hong Kong: Chinese University Press, 1978. Graham, Angus C. Chuang Tzu: The Inner Chapters. London: George Allen and Unwin, 1981. Graham, Angus  C.  Disputers of the Tao: Philosophical Argument in Ancient China. Chicago/La Salle: Open Court, 1989. Granet, Marcel. Fêtes et chansons anciennes de la Chine. Paris: Leroux, 1919. Gu Jiegang 顧頡剛 et al. (eds.). Gushi bian 古史辨. 7 vols. Shanghai: Shanghai guji chubanshe, 1982. Gu Jiegang 顧頡剛 et al. Gushi lunwenji 古史論文集. Beijing: Zhonghua shuju, 1996.

537

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Guan Xiechu 管燮初. Zuozhuan jufa yanjiu 左传句法研究. Hefei: Anhui jiaoyu chubanshe, 1994. Guan Xihua 管錫華. Zhongguo gudai biaodian fuhao fazhan shi 中國古代標 點符號發展史. Chengdu: Ba Shu shushe, 2002. Guo Xiliang 郭錫良. ‘Jieci “yu” de qiyuan he fazhan 介词“于”的起源和发展’. Zhongguo yuwen 2 (1997), 131–8. Handa wenku 漢達文庫. . Hanshu 漢書, by Ban Gu 班固, Beijing: Zhonghua shuju 1997. Harbsmeier, Christoph. Aspects of Classical Chinese Syntax. London/Malmö: Curzon Press, 1981. Harbsmeier, Christoph. ‘An Alphabetized Cornucopia of Chinese Death Terminology’. Unpublished MS. Oslo, 1995. Harbsmeier, Christoph. ‘Xunzi and the Problem of Impersonal First Person Pronouns’. Early China 22 (1997), 181–220. Harbsmeier, Christoph. Language and Logic [= Science and Civilisation in China, vol. VII:1]. Cambridge: Cambridge University Press, 1998. Harbsmeier, Christoph. ‘Notes on the Syntax and Semantics of Pre-Buddhist Chinese Numerals’. In: Und folge nun dem, was mein Herz begehrt: Festschrift für Ulrich Unger zum 70. Geburtstag, ed. by Reinhard Emmerich and Hans Stumpfeldt. 2 vols. Hamburg: Hamburger Sinologische Gesellschaft, 2002, 253–72. Harbsmeier, Christoph (ed.). Thesaurus Linguae Sericae: An Historical and Comparative Encyclopaedia of Chinese Conceptual Schemes. . Harbsmeier, Christoph. ‘Irrefutable Conjectures. A Review of William  H. Baxter and Laurent Sagart, Old Chinese. A New Reconstruction’. Monumenta Serica 64.2 (2016), 445–504. Hardy, Grant. Worlds of Bronze and Bamboo: Sima Qian’s Conquest of History. New York: Columbia University Press, 1999. de Harlez, M.  C.  ‘Koue-yü (discours des royaumes)’. Journal Asiatique 9.2 (1893), 373–419, and 9.3 (1894), 5–91. Hart, James. The Philosophy of the Chou Yü. Diss., University of Washington, 1973. He Jin 何晉. ‘Zhanguo ce’ yanjiu 《战国策》研究. Beijing: Beijing daxue ­chubanshe, 2001. He Leshi 何乐士, Ao Jinghao 敖镜浩, Wang Kezhong 王克仲, Mai Meiqiao 麦 梅翘, and Wang Haifen 王海棻. Gudai Hanyu xuci tongshi 古代汉语虚词 通释. Beijing: Beijing chubanshe, 1985. Henderson, John  B.  Scripture, Canon and Commentary: A Comparison of Confucian and Western Exegesis. Princeton: Princeton University Press, 1991. Henry, Eric (tr.). Shuo Yuan. Seattle: University of Washington Press, forthcoming.

538

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Herforth, Derek Dane. ‘Philological Marginalia on a Grammatical Study of Old Chinese. A Review of Aspects of Classical Chinese Syntax, by Christoph Harbsmeier’. Early China 11–12 (1985–87), 209–42. Herforth, Derek Dane. Conditional sentences in Old Chinese. Diss., University of California, Berkeley, 1994. Herforth, Derek Dane. ‘A Sketch of Late Zhou Chinese Grammar’. In: The SinoTibetan Languages, ed. by Graham Thurgood and Randy J. LaPolla. London and New York: Routledge, 2003. Herforth, Derek Dane. ‘Working Out Basic Patterns in Classical Chinese Syntax: Further Data on the Benefactive Ditransitive in Late Zhou’. In: Proceedings of the 20th North American Conference on Chinese Linguistics, ed. by Marjorie  K.  M.  Chan and Hana Kang. Columbus: Ohio State University, 2008, 455–71. Hertzer, Dominique. Das Mawangdui-Yijing: Text und Deutung. Munich: Diederichs, 1996. Hoang, Pierre. Concordance des chronologies néoméniques chinoise et europé­ enne. Shanghai: Imprimerie de la Mission Catholique, 1910. Hong Chengyu 洪成玉. Gu Hanyu changyong tongyici cidian 古漢語常用同 義詞詞典. Beijing: Shangwu yinshuguan, 2009. Hong Zhanhou 洪湛侯. Shijing xueshi 詩經學史. Beijing: Zhonghua shuju, 2002. d’Hormon, André and Rémi Mathieu. Guoyu, Propos sur les principautés I: Zhouyu. Paris: Institut des Hautes Études Chinoises, 1985. Hsiao, Kung-chuan. A History of Chinese Political Thought. vol. 1: From the Beginnings to the Sixth Century A.D., tr. by F. W. Mote. Princeton: Princeton University Press, 1979. Hu Pu’an 胡樸安. Zhongguo wenzixue shi 中國文字學史. Shanghai: Shanghai shudian, 1984. Hua Xuecheng 華學誠. Zhou Qin Han Jin fangyan yanjiu shi 周秦漢晉方言硏究 史. Shanghai: Fudan daxue chubanshe, 2007. Huang Junjie 黃俊傑. Mengxue sixiangshi lun 孟學思想史論. 2 vols. Taipei: Sanmin shudian, 1991–97. Hucker, Charles O. A Dictionary of Official Titles in Imperial China. Stanford: Stanford University Press, 1985. Huddleston, Rodney and Geoffrey K. Pullum. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press, 2002. Hulsewé, A. F. P. Remnants of Ch’in Law: An Annotated Translation of the Ch’in Legal and Administrative Rules of the 3rd Century BC. Leiden: Brill, 1985. Hunter, Michael. Confucius Beyond the Analects. Leiden: Brill, 2017. Hunter, Michael and Martin Kern (eds.). Confucius and the Analects Revisited: New Perspectives on Composition, Dating, and Authorship. Leiden: Brill, 2018.

539

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Hutton, Eric  L.  Xunzi: The Complete Text. Princeton: Princeton University Press, 2014. Hutton, Eric L. (ed.). Dao Companion to the Philosophy of Xunzi. Dordrecht: Springer, 2016. Imber, Alan. Kuo Yü. An early Chinese Text and its Relationship with the Tso Chuan. Diss., University of Stockholm, 1975. Institut Ricci (ed.). Grand dictionnaire Ricci de la langue chinoise. Paris: Desclée de Brouwer, 2001. Ivanhoe, Philip J. Ethics in the Confucian Tradition: The Thought of Mencius and Wang Yangming. Atlanta: American Academy of Religion, 1990. Jiang Shanguo 蔣善國. Shangshu zongshu 尚書綜述. Shanghai: Shanghai guji chubanshe, 1988. Johnston, Ian. The Mozi: A Complete Translation. Hong Kong: Chinese University Press, 2010. Jonker, D.  R.  Wenyan Partikels: Woordenboek van de Klassiek Chinese Functiewoorden. Leiden: Stichting Shilin, 2011. Kalinowski, Marc. ‘Les justifications historiques du gouvernement ideal dans le Lüshi chunqiu’. Bulletin de l’École française d’Extrême-Orient 68 (1980), 155–208. Kamata Tadashi 鐮田正. Saden no seiritsu to sono tenkai 左傳の成立とそ の展開. Tokyo: Tashukan, 1963. Kamenarovic, Ivan P. Printemps et automnes de Lü Buwei. Paris: Le Cerf, 1998. Karlgren, Bernhard. Études sur la phonologie chinoise. Leiden and Stockholm: Norstedt and Söner, 1915. Karlgren, Bernhard. ‘On the Authenticity and Nature of the Tso Chuan’. ́ Göteborgs Högskolas Årsskrift 32 (1926), 1–65. Karlgren, Bernhard. ‘The Early History of the Chou-li and Tso-chuan Texts’, Bulletin of the Museum of Far Eastern Antiquities 3 (1931), 1–59. Karlgren, Bernhard. ‘Glosses on the Book of Odes’, Bulletin of the Museum of Far Eastern Antiquities 14 (1942), 71–247, 16 (1944), 25–169, and 18 (1946), 1–198. Karlgren, Bernhard. ‘Legends and Cults in Ancient China’. Bulletin of the Museum of Far Eastern Antiquities 18 (1946), 199–366. Karlgren, Bernhard. ‘Glosses on the Book of Documents’. Bulletin of the Museum of Far Eastern Antiquities 20 (1948), 39–315, and 21 (1949), 63–206. Karlgren, Bernhard. ‘The Book of Documents’. Bulletin of the Museum of Far Eastern Antiquities 22 (1950), 1–81. Karlgren, Bernhard.‘Compendium of Phonetics in Ancient and Archaic Chinese’. Bulletin of the Museum of Far Eastern Antiquities 26 (1954), 211–367. Karlgren, Bernhard. Grammata Serica Recensa. Stockholm: Museum of Far Eastern Antiquities, 1957.

540

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Kennedy, George  A.  ‘A Study of the Particle Yen’. Journal of the American Oriental Society 60.1 (1940), 1–22, and 60.2 (1940), 193–207. Kennedy, George A. ‘The Monosyllabic Myth’. Journal of the American Oriental Society 71.3 (1951), 161–6. Kennedy, George A. ‘Another Note on Yen’. Harvard Journal of Asiatic Studies 16.1–2 (1953), 226–36. Kennedy, George A. ‘The Butterfly Case (Part I)’. Wennti 8 (1955), 1–47. Kennedy, George A. ‘Zai lun wu wo (The Classical Pronoun Forms ngo and nga)’. Bulletin of the Institute of History and Philology, Academia Sinica 28 (1956), 273–81. Kennedy, George A. ‘Interpretation of the Ch’un-ch’iu’. In: Selected Works of George  A.  Kennedy, ed. by Li Tien-yi. New Haven: Yale University Press, 1964, 79–103. Kern, Martin. ‘Shi jing Songs as Performance Texts: A Case study of “Chu Ci” (Thorny Caltrop)’. Early China 25 (2000), 49–111. Kern, Martin (ed.). Text and Ritual in Early China. Seattle: University of Washington Press, 2005. Kern, Martin and Dirk Meyer (eds.). Origins of Chinese Political Thought: Studies in the Classic of Documents. Leiden: Brill, 2017. King, Brandon Russell. Adapting with the Times: Fajia Law and State Development. Diss., Hong Kong, 2015. Klein, Esther. ‘Were there Inner Chapters in the Warring States? A New Examination of Evidence about the Zhuangzi’. T’oung Pao 96 (2010), 299–369. Kline, T. C. and Philip J. Ivanhoe (eds.). Virtue, Nature, and Moral Agency in the Xunzi. Indianapolis/Cambridge, MA: Hackett, 2000. Klöpsch, Volker. Sunzi: Die Kunst des Krieges. Frankfurt/M.: Insel Verlag, 2009. Knechtges, David R and Taiping Chang. Ancient and Early Medieval Chinese Literature: A Reference Guide. 4 vols. Leiden: Brill, 2010–14. Knoblock, John. Xunzi: ‘Chronology of Xunzi’s Works’. Early China 8 (1982–83), 29–52. Knoblock, John. Xunzi: A Translation and Study of the Complete Works. 3 vols. Stanford: Stanford University Press, 1988–94. Knoblock, John and Jeffrey Riegel. The Annals of Lü Buwei: A Complete Translation and Study. Stanford: Stanford University Press, 2000. Knoblock, John and Jeffrey Riegel. Mozi: A Study and Translation of the Ethical and Political Writings. Berkeley: Institute of East Asian Studies, 2013. Kohn, Livia and Michael LaFargue. Lao-tzu and the Tao-te-ching. Albany: State University of New York Press, 1998. Köster, Hermann. Hsün-Tzu. Kaldenkirchen: Steyler, 1967. Kroker, Eduard Josef. Der Gedanke der Macht im Shang-Kün-shu. Diss., Universität Zürich, 1950.

541

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Kroll, Paul W. A Student’s Dictionary of Classical and Medieval Chinese. Leiden: Brill, 2015. Lau, D. C. Mencius. Harmondsworth: Penguin Books, 1970. Lau, D.  C.  ‘A Study of Some Textual Problems in the Lü-shih ch’un-ch’iu’. Bulletin of the Institute of Chinese Literature and Philosophy 1 (1991), 45–87. Lau, D.  C.  Confucius: The Analects. Hong Kong: Chinese University Press, 1992[a]. Lau, D. C. ‘The Doctrine of Kuei Sheng 貴生 in the Lü-shih ch’un-ch’iu’. Bulletin of the Institute of Chinese Literature and Philosophy 2 (1992[b]), 51–90. Lau, D. C. and Roger Ames (trs.). Sun Pin: The Art of Warfare. Albany: State University of New York Press, 2003. Le Blanc, Charles and Susan Blader (eds.). Chinese Ideas about Nature and Society: Studies in Honour of Derk Bodde. Hong Kong: Hong Kong University Press, 1987. Legge, James. The Shoo King or: The Book of Historical Documents [= The Chinese Classics, vol. 3]. Oxford: Oxford University Press, 1865. Legge, James. The She King or: The Book of Poetry [= The Chinese Classics, vol. 4]. Oxford: Oxford University Press, 1871. Legge, James. The Ch’un Ts’ew with the Tso Chuen [= The Chinese Classics, vol. 5]. Oxford: Oxford University Press, 1872. Legge, James. The Li Ki. 2 vols. Oxford: Clarendon, 1885. Legge, James. Confucian Analects, the Great Learning, and the Doctrine of the Mean [= The Chinese Classics, vol. 1]. Oxford: Oxford University Press, 1893. Levi, Jean (tr.). Le livre du prince Shang. Paris: Flammarion, 1981. Levi, Jean (tr.). Han-Fei-tse, ou Le tao du Prince. Paris: Seuil, 1999. Levi, Jean (tr.). Sun Tzu: L’art de la guerre. Paris: Hachette, 2000. Levi, Jean (tr.). Les Œuvres de Maître Tchouang. Paris: L’Encyclopédie des Nuisances, 2006. Levi, Jean (tr.). Les deux arbres de la voie, vol. 2: Les Entretiens de Confucius. Paris: Les Belles Lettres, 2018. Lévy, André (tr.). Mencius. Paris: You-Feng, 2003. Leys, Simon (tr.). The Analects, ed. by Michael Nylan. New York: Norton, 2014. Li Chor-Shing. Beiträge zur kontrastiven Aspektologie: Das Aspektsystem im modernen Chinesisch. Frankfurt/M.: Lang, 1991. Li Fanggui 李方桂. ‘Shanggu yin yanjiu 上古音研究’. The Tsing Hua Journal of Chinese Studies IX.1–2 (1971), 1–61 [tr. by Gilbert  L.  Mattos, Monumenta Serica 31 (1974–75), 219–87]. Li Junming 李均明 et al. (eds.). Dangdai Zhongguo jianboxue yanjiu 当代中 国简帛学研究. Beijing: Zhongguo shehui kexueyuan chubanshe, 2011. Li Wai-yee. The Readability of the Past in Early Chinese Historiography. Cambridge, MA: Harvard University Asia Center, 2007.

542

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Li Xinkui 李新魁. Zhonggu yin 中古音. Beijing: Shangwu yinshuguan, 1991. Li Yin. The Diachronic Development of Passive Constructions from Archaic Chinese to Modern Mandarin. Diss., University of Washington, 2015. Li Yu-ning (ed.). Shang Yang’s Reforms and State Control in China. White Plains: M. E. Sharpe, 1977. Lianmian zidian 聯緜字典. By Fu Dingyi 符定一. 4 vols. Beijing: Zhonghua shuju, 1954. Liao Wenkui. The Complete Works of Han Fei Tzǔ. 2 vols. London: Probsthain, 1959. Liu Qiyu 劉起釪. Shangshu xueshi 尚書學史. Beijing: Zhonghua shuju, 1989. Liu Xiaogan (ed.). Dao Companion to Daoist Philosophy. Dordrecht: Springer, 2015. Liu Yucai and Luke Habberstad. ‘The Life of a Text: A Brief History of the Liji 禮記 (Rites Records) and Its Transmission’. Journal of Chinese Literature and Culture 1.1–2 (2014), 289–308. Loewe, Michael. Early Chinese Texts: A Bibliographical Guide. Berkeley: Institute of East Asian Studies, 1993. Loewe, Michael. and Edward L. Shaughnessy (eds). The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. Cambridge: Cambridge University Press, 1999. van der Loon, Piet. ‘On the Transmission of the Kuan tzu’. T’oung Pao 41.4–5 (1952), 357–93. Long Yuchun 龍宇純. ‘Xian Qin sanwen zhong de yunwen’ 先秦散文中的韻 文. Chongji xuebao 崇基學報 2.2 (1962), 137–69, and 3.1 (1963), 55–87. Lu Deming 陸德明. Jingdian shiwen 經典釋文. Shanghai: Shanghai guji chubanshe, 1985. Lundahl, Bertil. Han Fei Zi: The Man and the Work. Stockholm: Institute of Oriental Languages, Stockholm University, 1992. Lunheng jiaojian 論衡. 2 vols., ed. by Yang Baozhong 楊寶忠. Shijiazhuang: Hebei jiaoyu chubanshe, 1999. Luo Zhufeng 羅竹風 et al. (eds.). Hanyu da cidian 漢語大詞典. 13 vols. Shanghai: Hanyu da cidian chubanshe, 1986–94. Ma Jianzhong 馬建忠. Mashi wentong kanwu 馬氏文通刊誤, ed. by Yang Shuda 楊樹達. Shanghai: Shanghai guji chubanshe, 2007. Ma Shinian 馬世年. Han Feizi de chengshu ji qi wenxue yanjiu 韓非子的成書 及其文學研究. Shanghai: Shanghai guji chubanshe, 2011. Maas, Paul. Textual Criticism, tr. by Barbara Flower. Oxford: Clarendon Press, 1958. McCraw, David. Stratifying Zhuangzi: Rhyme and Other Quantitative Evidence. Taipei: Institute of Linguistics, Academia Sinica, 2010. Mair, Victor H. Wandering on the Way: Early Taoist Tales and Parables of Chuang Tzu. New York: Bantam Books, 1994.

543

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Major, John S. and Constance A. Cook. Ancient China: A History. New York and London: Routledge, 2017. Makeham, John. ‘The Formation of Lunyu as a Book’. Monumenta Serica 44 (1996), 1–24. Malmqvist, Göran. ‘On the Functions and Meanings of the Graph 其 Chyi in the Tzuoojuann’. Proceedings of the International Conference on Sinology, August 15–17, 1980. Section on Linguistics and Palaeography. Taipei: Academia Sinica, 1981, 365–89. Malmqvist, Göran. Bernhard Karlgren: Portrait of a Scholar. Bethlehem and Lanham: Lehigh University Press, 2011. Martzloff, Jean-Claude. A History of Chinese Mathematics, tr. by Stephen S. Wilson. Berlin and Heidelberg: Springer, 2006. Maspéro, Henri. ‘Légendes mythologiques dans le Chou-king’. Journal Asiatique 204 (1924), 1–100. Maspéro, Henri. ‘La Composition et la date du Tso Tchouan’. Mélanges chinois et bouddhiques 1 (1931–32), 137–215. Mathews, Robert Henry. Chinese-English Dictionary. Shanghai: China Inland Mission and Presbyterian Mission Press, 1931. McNeal, Robin. ‘The Development of Naturalist Thought in Ancient China: A Review of W. Allyn Rickett’s Guanzi’. Early China 28 (2003), 161–200. Meisterernst, Barbara. ‘Eine Übersetzung der “Yüeh yü”-Sektion des Kuo yü’. In: Und folge nun dem, was mein Herz begehrt: Festschrift für Ulrich Unger zum 70. Geburtstag, ed. by Reinhard Emmerich and Hans Stumpfeldt. 2 vols. Hamburg: Hamburger Sinologische Gesellschaft, 2002, 509–42. Meisterernst, Barbara. ‘Some Remarks on the Syntax and the Semantics of the So-Called Aspectual Markers ji and yi’. Journal of Chinese Linguistics 33.1 (2005), 68–113. Meisterernst, Barbara. ‘The Negative Wei in Hàn Period Chinese’. Journal of Chinese Linguistics 36.1 (2008), 121–54. Meisterernst, Barbara. ‘Aspect, Premodern’. In: Encyclopedia of Chinese Language and Linguistics, ed. by Rint Sybesma, et al. Brill Online, 2015. . Meisterernst, Barbara (ed.). New Aspects of Classical Chinese Grammar. Wiesbaden: Harrassowitz, 2016. Meyer, Dirk. Philosophy on Bamboo: Text and the Production of Meaning in Early China. Leiden: Brill, 2012. Mögling, Wilmar. Die Kunst der Staatsführung: Die Schriften des Meisters Han Fei. Leipzig: Kiepenheuer, 1994. Moritz, Ralf. Konfuzius. Gespräche. Stuttgart: Reclam, 1998. Morohashi Tetsuji 諸橋轍次. Dai Kan-Wa jiten 大漢和辭典. 13 vols. Tōkyō: Taishūkan shoin, 1955–60.

544

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Nienhauser, William H. Jr. (ed.). The Grand Scribe’s Records. 6 vols. Bloomington: Indiana University Press, 1994–. Nienhauser, William H. Jr. ‘Sima Qian and the Shiji’. In: The Oxford History of Historical Writing, vol. 1, ed. by Andrew Feldherr and Grant Hardy. Oxford: Oxford University Press, 2011, 463–84. Nijōan shujin, 二畳庵主人 [Kaji Nobuyuki 加地伸行]. Kanbun bunpō kiso: Hontōni wakaru Kanbun nyūmon 漢文文法基礎:本当にわかる漢文入門. Tōkyō: Kōdansha, 2012. Van Norden, Bryan W. Mengzi: With Selections from Traditional Commentaries. Indianapolis/Cambridge, MA: Hackett, 2008. Norman, Jerry. Chinese. Cambridge: Cambridge University Press, 1988. Nylan, Michael. The Five “Confucian” Classics. New Haven and London: Yale University Press, 2001. Nylan, Michael. ‘Kongzi and Mozi, the Classicists (Ru 儒) and the Mohists (Mo 墨) in Classical-Era Thinking’. Oriens Extremus 48 (2009), 1–20. Nylan, Michael and Michael Loewe (eds.). China’s Early Empires. A Re-appraisal. Cambridge: Cambridge University Press, 2010. Ōnishi Katsuya 大西克也. ‘Shishou tongci chuyi—Shiji zhong de “zhongxing dongci” he “zuoge dongci” 施受同辭芻議—《史記》中的 “中性動詞” 和 “作格動詞” ’. In: Meaning and Form: Essays in Pre-Modern Chinese Grammar, ed by Ken-ichi Takashima and Jiang Shaoyu. Munich: Lincom, 2004, 375–94. Pan Wuyun 潘悟雲. Hanyu lishi yinyunxue 汉语历史音韵学. Shanghai: Shanghai jiaoyu chubanshe, 2000. Pelliot, Paul. ‘Le Chou King en caractères anciens et le Chang chou che wen’. Mémoires concernant l’Asie orientale 2 (1916), 123–77. Petersen, Jens Østergaard. The Distribution of ‘于’ and ‘於’ in Zuozhuan 左傳: A Stylistic Approach. Unpublished paper, 2001. Petersen, Jens Østergaard. ‘The Zuozhuan Story about Qi Xi’s Recommendations and Its Sources’. Sino-Platonic Papers No. 255 (2015), . Pi Xirui 皮錫瑞. Jingxue lishi 經學歴史, ed. by Zhou Yutong 周予同. Shanghai: Shanghai shudian, 1996. Pian Yuqian 駢宇騫 and Duan Shu’an 段書安 (eds.). Ershi shiji chutu jianbo zongshu 二十世纪出土简帛综述. Beijing: Wenwu chubanshe, 2006. Pines, Yuri. Foundations of Confucian Thought: Intellectual Life in the Chunqiu Period, 722–453 B.C.E. Honolulu: University of Hawai’i Press, 2002. Pines, Yuri. Envisioning Eternal Empire: Chinese Political Thought of the Warring States Era. Honolulu: University of Hawai’i Press, 2009. Pines, Yuri. ‘Social Engineering in Early China: The Ideology of the Shangjun shu (Book of Lord Shang) Revisited’. Oriens Extremus 55 (2016), 1–37. Pines, Yuri. The Book of Lord Shang: Apologetics of State Power in Early China. New York: Columbia University Press, 2017.

545

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Plaks, Andrew H. ‘Where the Lines Meet: Parallelism in Chinese and Western Literatures’. Poetics Today 11.3 (1990), 523–46. Pulleyblank, Edwin G. ‘Fei 非, Wei 唯 and Certain Related Words’. In Egerod and Glahn 1959, 178–89. Pulleyblank, Edwin G. ‘Emphatic Negatives in Classical Chinese’. In: Ancient China: Studies in Early Chinese Civilization, ed. by David T. Roy and Tsien Tsuen-hsuin. Hong Kong: Chinese University Press, 1978, 115–35. Pulleyblank, Edwin  G.  Middle Chinese: A Study in Historical Phonology. Vancouver: University of British Columbia Press, 1984. Pulleyblank, Edwin G. ‘The Locative Particles Yü 于, Yü 於, and Hu 乎’. Journal of the American Oriental Society 106 (1986), 1–12. Pulleyblank, Edwin G. Lexicon of Reconstructed Pronunciation in Early Middle Chinese, Late Middle Chinese, and Early Mandarin. Vancouver: University of British Columbia Press, 1991. Pulleyblank, Edwin G. ‘How Do We Reconstruct Old Chinese?’ Journal of the American Oriental Society 112.3 (1992), 365–82. Pulleyblank, Edwin  G.  Outline of Classical Chinese Grammar. Vancouver: University of British Columbia Press, 1996. Qiu Xigui. Chinese Writing. Berkeley: Society for the Study of Early China, 2000. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. A Comprehensive Grammar of the English Language. London and New York: Longman, 1985. Ramsey, S. Robert. The Languages of China. Princeton: Princeton University Press, 1989. Richter, Matthias. The Embodied Text: Establishing Textual Identity in Early Chinese Manuscripts. Leiden: Brill, 2013. Richter, Matthias. ‘Punctuation, Premodern’. In: Encyclopedia of Chinese Language and Linguistics, ed. by Rint Sybesma, et al. Brill Online, 2015. . Rickett, Walter Allyn. Guanzi: Political, Economic, and Philosophical Essays From Early China. 2 vols. Princeton: Princeton University Press, 1985–98. Rickett, Walter Allyn. ‘Kuan-tzu and the Newly Discovered Texts on Bamboo and Silk’. In: Chinese Ideas about Nature and Society: Studies in Honour of Derk Bodde, ed. by Charles Le Blanc and Susan Blader. Hong Kong: Hong Kong University Press, 1987, 237–48. Riegel, Jeffrey. ‘Do Not Serve the Dead as You Serve the Living: The Lüshi chunqiu Treatises on Moderation in Burial’. Early China 20 (1995), 301–30. Robinet, Isabelle. Histoire du taoïsme des origines au XIVe siècle. Paris: Le Cerf, 1991. Roth, Harold. The Textual History of the Huai-nan Tzu. Ann Arbor: Association of Asian Studies, 1992.

546

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Roth, Harold. A Companion to A. C. Graham’s Chuang Tzu: The Inner Chapters. Honolulu: University of Hawai’i Press, 2003. Roy, David  T.  and Tsien Tsuen-hsuin (eds.). Ancient China: Studies in Early Chinese Civilization. Hong Kong: Chinese University Press, 1978. Rusk, Bruce. Critics and Commentators: The Book of Poems as Classic and Literature. Cambridge, MA: Harvard University Asia Center, 2012. Rutt, Richard. The Book of Changes (Zhouyi): A Bronze Age Document. London and New York: Routledge Curzon, 2002. Sagart, Laurent. The Roots of Old Chinese. Amsterdam and Philadelphia: John Benjamins, 1999. Sargent, Howard  W.  A Preliminary Study of the Kuo yü. Diss., University of Chicago, 1975. Satō Masayuki. The Confucian Quest for Order: The Origin and Formation of the Political Thought of Xun Zi. Leiden: Brill, 2003. Satō Masayuki. ‘The Idea to Rule the World: The Mohist Impact of Jian 兼 on the Xunzi’. Oriens Extremus 48 (2009), 21–54. Saussy, Haun. ‘Repetition, Rhyme, and Exchange in the Book of Odes’. Harvard Journal of Asiatic Studies 57.2 (1997), 519–42. Sawyer, Ralph  D.  The Seven Military Classics of Ancient China. Boulder: Westview, 1993. Schaberg, David. ‘Song and the Historical Imagination in Early China’. Harvard Journal of Asiatic Studies 59.2 (1999), 305–61. Schaberg, David. A Patterned Past: Form and Thought in Early Chinese Historiography. Cambridge, MA: Harvard University Asia Center, 2001. Schafer, Edward H. ‘What and How is Sinology’. T’ang Studies 8–9 (1990–91), 23–44. Schilling, Dennis. Yijing: Das Buch der Wandlungen. Frankfurt: Verlag der Weltreligionen, 2009. Schmidt-Glintzer, Helwig. Gegen den Krieg. Düsseldorf: Diederichs, 1975[a]. Schmidt-Glintzer, Helwig. Mo Ti: Solidarität und allgemeine Menschlichkeit. Düsseldorf: Diederichs, 1975[b]. Schmidt-Glintzer, Helwig. Mo Ti: Von der Liebe des Himmels zu den Menschen. Munich: Diederichs, 1992. Schuessler, Axel. ABC Etymological Dictionary of Old Chinese. Honolulu: University of Hawai’i Press, 2007. Schuessler, Axel. Minimal Old Chinese and Later Han Chinese: A Companion to Grammata Serica Recensa. Honolulu: University of Hawai’i Press, 2009. Schwartz, Benjamin. The World of Thought in Ancient China. Cambridge, MA: Harvard University Press, 1985. Schwarz, Ernst. Konfuzius: Gespräche des Meisters Kung. Munich: dtv, 1985. Schwermann, Christian. ‘Dummheit’ in altchinesischen Texten: Eine Begriffsgeschichte. Wiesbaden: Harrassowitz, 2011.

547

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Sellmann, James D. Timing and Rulership in Master Lü’s Spring and Autumn Annals (Lüshi chunqiu). Albany: State University of New York Press, 2002. Serruys, Paul L.-M. The Chinese of Han Time According to Fang Yen. Berkeley and Los Angeles: University of California Press, 1959. Shaughnessy, Edward  L.  The Composition of the ‘Zhouyi’. Diss., Stanford University, 1983. Shaughnessy, Edward L. I Ching: The Classic of Changes. New York: Ballantine, 1996. Shaughnessy, Edward  L.  (ed.). New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts. Berkeley: Institute of East Asian Studies, 1997. Shaughnessy, Edward  L.  ‘The Guodian Manuscripts and Their Place in Twentieth-Century Historiography on the Laozi’. Harvard Journal of Asiatic Studies 65.2 (2005), 417–57. Shaughnessy, Edward L. Rewriting Early Chinese Texts. Albany: State University of New York Press, 2006. Shaughnessy, Edward  L.  (ed.). Unearthing the Changes: Recently Discovered Manuscripts of the Yi Jing (I Ching) and Related Texts. New York: Columbia University Press, 2014. Shen Yucheng 沈玉成 and Liu Ning 劉寧. Chunqiu Zuozhuan xueshi gao 春秋 左傳學史稿. Nanjing: Jiangsu guji chubanshe, 2000. Shun Kwong-loi. Mencius and Early Chinese Thought. Stanford: Stanford University Press, 1997. Simon, Walter. ‘Die Bedeutung der Finalpartikel 矣’. Mitteilungen des Seminars für Orientalische Sprachen 37 (1934), 143–68. Simson, Wojciech Jan.  ‘Applying Stemmatology to Chinese Textual Traditions’. Asiatische Studien/Études Asiatiques 56.3 (2002), 587–608. Simson, Wojciech Jan. Die Geschichte der Aussprüche des Konfuzius (Lunyu). Bern: Peter Lang, 2006. Smith, Kidder. ‘Sima Tan and the Invention of Daoism, “Legalism”, et cetera’. Journal of Asian Studies 62.1 (2003), 129–56. Smith, Richard  J.  The I Ching: A Biography. Princeton: Princeton University Press, 2012. Steele, John. The I li, or Book of Etiquette and Ceremonial. 2 vols. London: Probsthain, 1917. von Strauß, Victor. Schî-kîng: Das kanonische Liederbuch der Chinesen. Heidelberg: Carl Winter’s Universitätsbuchhandlung, 1880. Stumpfeldt, Hans. ‘Ein Diener vieler Herren. Einige Bemerkungen zum Yentzu ch’un-ch’iu’. In: Und folge nun dem, was mein Herz begehrt: Festschrift für Ulrich Unger zum 70. Geburtstag, ed. by Reinhard Emmerich and Hans Stumpfeldt. 2 vols. Hamburg: Hamburger Sinologische Gesellschaft, 2002, 183–208.

548

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Stumpfeldt, Hans. ‘Ein Lied der “Lieder”?—Vorläufige Bemerkungen zu einem Passus in Erh-ya 3’. Oriens Extremus 46 (2007), 28–47. Stumpfeldt, Hans. ‘Thinking Beyond the “Sayings”: Comments About Sources Concerning the Life and Teachings of Confucius (551–479)’. Oriens Extremus 49 (2010), 3–27. Sun Liangming 孫良明. Gudai Hanyu yufa bianhua yanjiu 古代漢語語法變 化研究. Beijing: Yuwen chubanshe, 1994 Sun Yuwen 孫玉文. Hanyu biandiao cigou kaobian 漢語變調詞構考辨. 2 vols. Beijing: Shangwu yishuguan, 2015. Sybesma, Rint et al. (eds.). Encyclopedia of Chinese Language and Linguistics. Brill Online. 2015. . Tai, James H. Y. and Marjorie K. M. Chan. ‘Some Reflections on the Periodization of the Chinese Language’. In: Studies in Chinese Historical Syntax and Morphology: Linguistic Essays in Honor of Mei Tsu-lin, ed. by Alain Peyraube and Chaofen Sun. Paris: École des Hautes Études en Sciences Sociales, 1999, 223–39. Takashima, Ken-ichi and Jiang Shaoyu (eds.). Meaning and Form: Essays in Pre-Modern Chinese Grammar. Munich: Lincom, 2004. Tang Jialong (tr.). Sunzi: L’ art de la guerre. Sun Bin: L’ art de la guerre. Paris: Payot and Rivages, 2004. Tchang, Mathias. Synchronismes chinois: Chronologie complète et concordance, Shanghai: Imprimerie de la Mission Catholique, 1905. Tchang Tcheng-ming. Le parallélisme dans les vers du Cheu King. Shanghai: T’ou Sè-wè and Paris: P. Geuthner, 1937. Tharsen, Jeffrey R. The Digital Etymological Dictionary of Old Chinese (Version 1.0). 2012. . Tharsen, Jeffrey  R.  Chinese Euphonics: Phonetic Patterns, Phonorhetoric and Literary Artistry in Early Chinese Narrative Texts. Diss., University of Chicago, 2015. Thompson, P. M. The Shen Tzu Fragments. Oxford: Oxford University Press, 1979. Timpanaro, Sebastiano. The Genesis of Lachmann’s Method, tr. by Glenn W. Most. Chicago and London: University of Chicago Press, 2005. Tong Shuye 童書業. Chunqiu Zuozhuan yanjiu 春秋左传研究. Shanghai: Renmin chubanshe, 1980. Tsien Tsuen-hsuin. Written on Bamboo and Silk: The Beginnings of Chinese Books and Inscriptions. Chicago: University of Chicago Press, 2004. Tu Ching-I (ed.). Classics and Interpretations: The Hermeneutic Traditions in Chinese Culture. New Brunswick: Transaction, 2000. Unger, Ulrich. ‘Zur Morphologie des altchinesischen Verbs’. Unpublished MS (= Hao-ku 20), 1983.

549

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Unger, Ulrich. Einführung in das Klassische Chinesisch. Wiesbaden: Harrassowitz, 1985. Unger, Ulrich. Glossar des Klassischen Chinesisch. Wiesbaden: Harrassowitz, 1989. Unger, Ulrich. ‘Der shang-sheng-Faktor—Nachgedanken zur Serie von 己’. Unpublished MS (= Hao-ku 41), 1993. Unger, Ulrich. Rhetorik des Klassischen Chinesisch. Wiesbaden: Harrassowitz, 1994. Unger, Ulrich. Abriß der Literatur des chinesischen Altertums: Prodesse aut delectare? Unpublished MS. Münster, 1998. Unger, Ulrich. Grammatik des Klassischen Chinesisch, ed. by Reinhard Emmerich. Heidelberg: CrossAsia-eBooks, 2019. Van Auken, Newell Ann. ‘The Modal Negative wu in Classical Chinese’. In: Meaning and Form: Essays in Pre-Modern Chinese Grammar, ed by Ken-ichi Takashima and Jiang Shaoyu. Munich: Lincom, 2004, 191–210. Van Auken, Newell Ann. ‘Could “Subtle Words” Have Conveyed “Praise and Blame”? The Implications of Formal Regularity and Variation in Spring and Autumn (Chūn qiū) Records’. Early China 31 (2007), 47–111. Van Auken, Newell Ann. The Commentarial Transformation of the Spring and Autumn. Albany: State University of New York Press, 2016. Vandermeersch, Léon. La formation du légisme. Recherche sur la constitution d’une philosophie politique caractéristique de la Chine ancienne. Paris: École française d’Extrême-Orient, 1965. Vochala, Jaromír and Ruzhen Vochalová. Einführung in die Grammatik des Klassischen Chinesisch. Leipzig: VEB Verlag Enzyklopädie, 1990. Vogelsang, Kai. ‘Inscriptions and Proclamations: On the Authenticity of the “gao” Chapters in the Book of Documents’. Bulletin of the Museum of Far Eastern Antiquities 74 (2002[a]), 138–209. Vogelsang, Kai. ‘Introduction’. In: Textual Scholarship in Chinese Studies: Papers from the Munich Conference 2000, ed. by Kai Vogelsang. Asiatische Studien/ Études Asiatiques 56.3 (2002[b]), 529–32. Vogelsang, Kai. ‘Prolegomena to Critical Zuozhuan Studies: the Manuscript Tradition’. Asiatische Studien/Études Asiatiques 61.3 (2007), 941–88. Vogelsang, Kai. ‘Beyond Confucius: A Socio-Historical Reading of the Lunyu’. Oriens Extremus 49 (2010), 29–62. Vogelsang, Kai. ‘From Anecdote to History: Observations on the Composition of the Zuozhuan’. Oriens Extremus 50 (2011). 99–124. Vogelsang, Kai. ‘Conceptual History: A Short Introduction’. Oriens Extremus 51 (2012), 9–24. Vogelsang, Kai. ‘Getting the Terms Right: Political Realism, Politics, and the State in Ancient China’. Oriens Extremus 55 (2016), 39–72. Vogelsang, Kai. Shangjun shu: Schriften des Fürsten von Shang. Stuttgart: Kröner, 2017.

550

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Wagner, Rudolf G. The Craft of a Chinese Commentator: Wang Bi on the Laozi. Albany: State University of New York Press, 2000. Wagner, Rudolf G. ‘A Building Block of Chinese Argumentation: Initial Fu 夫 as a Phrase Status Marker’. In: Literary Forms of Argument in Early China, ed. by Joachim Gentz and Dirk Meyer. Leiden/Boston: Brill, 2015, 37–66. Waley, Arthur. ‘ The Book of Changes’. Bulletin of the Museum of Far Eastern Antiquities 5 (1933), 121–42. Waley, Arthur. The Book of Songs. London: Allen and Unwin, 1937. Wang Fengyang 王鳳陽. Guci bian 古辭辨. Jilin: Wenshi chubanshe, 1993. Wang Guowei 王國維. Jiandu jianshu kaojiao zhu 简牍检署考校注, ed. by Hu Pingsheng 胡平生 and Ma Yuehua 馬月華. Shanghai: Shanghai guji chubanshe, 2014. Wang Hong 王宏 et al. (tr.). Discourses of the States. Changsha: Hunan renmin chubanshe, 2012. Wang Li 王力. Hanyu Shigao 漢語史稿. Beijing: Zhonghua shuju, 1980. Wang Li 王力 (ed.). Gudai Hanyu 古代漢語. 4 vols. Beijing: Zhonghua shuju, 1981. Wang Li (ed.). Gu Hanyu zidian 古汉语字典. Beijing: Zhonghua shuju, 2000. Wang Qiming 汪啟明. Xian Qin liang Han Qiyu yanjiu 先秦兩漢齊語研究. Chengdu: Ba Shu shushe, 1998. Wang Yankun 王彦坤 (ed.). Lidai bihuizi huidian 历代避讳子汇典. Zhengzhou: Zhongzhou guji chubanshe, 1997. Wang Yinzhi 王引之. Jingyi shuwen 經義述聞. Nanjing: Jiangsu guji chubanshe, 1985[a]. Wang Yinzhi 王引之. Jingzhuan shici 經傳釋詞. Nanjing: Jiangsu guji chubanshe, 1985[b]. Watson, Burton. Hsün Tzu: Basic Writings. New York: Columbia University Press, 1963[a]. Watson, Burton. Mo Tzu: Basic Writings. New York: Columbia University Press, 1963[b]. Watson, Burton. Han Fei Tzu: Basic Writings. New York: Columbia University Press, 1964. Watson, Burton. The Complete Works of Chuang Tzu. New York: Columbia University Press, 1968. Watson, Burton. Records of the Grand Historian. Rev. ed. 3 vols. Hong Kong: Chinese University Press, 1993. Wei Desheng 魏德胜. ‘Shuihu di Qinmu zhujian’ yufa yanjiu 《睡虎地秦墓竹简》 语法研究. Beijing: Shoudu Shifan daxue chubanshe, 2000. Weigand, Jörg. Staat und Militär im altchinesischen Militärtraktat Wei Liao Tzu. Diss., Universität Würzburg, 1969. Weingarten, Oliver. ‘The Singing Sage: Rhymes in Confucius Dialogues’. Bulletin of the School of Oriental and African Studies 79.3 (2016), 581–607.

551

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Wilhelm, Hellmut. Sinn des I Ging. Düsseldorf: Diederichs, 1972. Wilhelm, Richard. Kungfutse: Gespräche (Lun Yü). Jena: Diederichs, 1921[a]. Wilhelm, Richard. Mong Dsi (Mong Ko). Jena: Diederichs, 1921[b]. Wilhelm, Richard. Dschuang Dsi: Das wahre Buch vom südlichen Blütenland. Jena: Diederichs, 1923. Wilhelm, Richard. I Ging: Das Buch der Wandlungen. 2 vols. Jena: Diederichs, 1924. Wilhelm, Richard. Frühling und Herbst des Lü Bu We. Jena: Diederichs, 1928. Wilhelm, Richard. Li Gi: Das Buch der Sitte. Jena: Diederichs, 1930. Wilkinson, Endymion. Chinese History: A New Manual. 5th ed. N.p.: Endymion Wilkinson, 2018. Wu Geng, Die Staatslehre des Han Fei: Ein Beitrag zur chinesischen Idee der Staatsräson. Vienna: Springer, 1978. Wu Sue-mei. The Coverbs in Classical Chinese. Diss., Ohio State University, 1997. Wu Yuemei. The Early History of the Chinese Passive. Diss., Stanford University, 2007. Xiao, Richard and Tony McEnery. Aspect in Mandarin Chinese: A CorpusBased Study. Amsterdam and Philadelphia: John Benjamins, 2004. Xu Renfu 徐仁甫. Zuozhuan shuzheng 左傳疏證. Chengdu: Sichuan renmin chubanshe, 1981. Yang Bojun 楊伯峻. Gu Hanyu xuci 古汉语虚词. Beijing: Zhonghua shuju, 1981. Yang Bojun 楊伯峻 and He Leshi 何樂士. Gu Hanyu yufa ji qi fazhan 古汉语 语法及其发展. 2 vols. Beijing: Yuwen chubanshe, 2003. Yang Shuda 楊樹達. ‘Inversion in Classical Chinese’, tr. by Achilles Fang. Monumenta Serica 7.1–2 (1942), 267–84. Yang Shuda 楊樹達. Ciquan 詞詮. Shanghai: Shanghai guji chubanshe, 1986. Yates, Robin. ‘The Mohists on Warfare: Technology, Technique, and Justification’. Journal of the American Academy of Religion 47 (1979), 549–603. Yates, Robin. ‘Early Modes on Interpretation of the Military Canons: The Case of the Sunzi bingfa’. In: Classics and Interpretations: The Hermeneutic Traditions in Chinese Culture, ed. by Tu Ching-I. New Brunswick: Transaction, 2000[a], 65–79. Yates, Robin. ‘Texts on the Military and Government from Yinqueshan: Introductions and Preliminary Translations’. In: Studies on Recently Discovered Chinese Manuscripts, ed. by Sarah Allan and Xing Wen. Beijing: Wenwu chubanshe, 2000[b], 334–87. Yu, Pauline R. ‘Allegory, Allegoresis, and the Classic of Poetry’. Harvard Journal of Asiatic Studies 43.2 (1983), 377–412. Zádrapa, Lukáš. Word-Class Flexibility in Classical Chinese: Verbal and Adverbial Uses of Nouns. Leiden: Brill, 2011.

552

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

b i b l i o g r a p hy

Zhang Qiyun 張其昀 et al. (eds.). Zhongwen da cidian 中文大辭典. 10 vols. Taipei: Zhongguo wenhua daxue, 1990. Zhang Shilu 張世祿 and Yan Xiu 嚴修 (eds.). Gudai Hanyu jiaocheng 古代漢 語教程. Shanghai: Fudan daxue chubanshe, 1990. Zhang Xincheng 張心澂. Weishu tongkao 偽書通考. Shanghai: Shangwu yinshu guan, 1939. Zhang Yiren 張以仁. ‘Lun Guoyu yu Zuozhuan de guanxi’ 論國語與左傳的 關係. Lishi yuyan yanjiusuo jikan 33 (1962), 233–86. Zhao Tong 趙彤. Zhanguo Chu fangyan yinxi 戰國楚方言音系. Beijing: Zhongguo xiju chubanshe, 2006. Zheng Liangshu 鄭良樹. Han Feizi zhushu ji sixiang 韓非之著書及思想. Taipei: Xuesheng shuju, 1993. Zhou Fagao 周法高. Zhou Qin mingzi jiegu huishi bubian 周秦名字解詁匯釋 補編. Beijing: Zhonghua congshu 中華叢書, 1968. Zhou Fagao 周法高. Zhongguo gudai yufa 中國古代語法. 3 vols. Taipei: Institute of History and Philology, Academia Sinica, 1972.

553

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

Index of linguistic terms References are to lessons (L), section numbers, boxes (B), foci (F), reviews (R), or notes (section No. followed by n). Classical Chinese words are not indexed here; pertinent references are to be found in the glossary. active voice  4.2.1; 4.4 adjuncts  1.3.5; 3.3; 4.3; L7–10 adverbs  1.2.5; 4.3; 10.4–5; 24.3–B24 pronominal 7.5 appositions 20.3–B21 aspect see time bracket constructions  23.5–B23 Chinese language  F1 Classical  pp. xvii, xxi; F1 Literary  p. xvii Middle Chinese  F1; F2 Old Chinese  F1; F3; F6; F11 pre-Classical  p. xvii; F1; 3.2n; F6; F11 clauses  1.4; R2 causal 16.2 concessive 16.2 conditional 15.4 coordinated 15.2 demoted 6.2.2–⑬ final/consecutive 16.5 interrogative see questions nominal L3 nominalized  12.3; R3 non-canonical  L13–14; R4 subject/object  12.3; 14.3 subordinated 15.3–4 temporal 16.3 verbal L4 see also questions: rhetorical comparative  4.3; 24.3–B24 comparisons  7.4; 8.2.2; 11.4 complements  1.3.4; L11 conjunctions  1.2.6; 7.7; L15–16 constituents  1.1; 1.3

see also adjuncts; complements; objects; predicator; subject coordination  2.2; 2.3; 4.2.2 copula 3.1 deletion  1.3.1; 3.2; 5.2.3; 15.3–⑩; 22.5–B22 dialects 25.2–⑦; F26 disposal construction  9.2.2; 16.3–⑰ euphonic rules  2.2–B4 fanqie F2 faux amis B2 form vs. function  1.1 genitives  2.4.1; 2.4.2; 6.2; 6.2–B9 head 2.4 hypallage 16.4–㉑; 24.2–⑱ interjections 1.2.8 measure words  10.3–B13 modal auxiliaries  4.4 modification  2.4; 4.3; 5.3–B8; 10.3–B13; 24.2–⑤ morphology  p. xxvi; 4.2.1n; F3; 5.2.1n names  F5; 12.3–B15 negation  3.3.1; 5.2.3; 24.2–⑮ nominalization  6.3; L12 nouns  1.2.1; L2; L6 compound 2.4.1 plural 28.3–B28 temporal 10.2 see also phrases: noun numerals  1.2–B3; 10.3–B13; 28.3–B28 objects  1.3.3; L5; 12.2; 30.1–B30 anteposed  14.2; 14.3–B17; R4 direct  L5; 8.2.4; 9.2.2; 11.3 dummy  13.3; 14.2

555

OUP CORRECTED PROOF – FINAL, 19/07/21, SPi

index of linguistic terms

exposed  13.3; 14.2; R4 indirect  5.2.2; 9.2.2 infinitive 12.3 prepositional  L8–9; 13.3.1 topicalization 13.3 parallelism  R2; 15.2; F 21 participles 2.4.3 particles  1.2.7; 14.3; 18.4–B19 passive voice  4.2.1; 4.4.2; 8.2.3 phonetic series  F3 phonology  p. xxi; F1; F2; 2.2–B4; 3.4–B6; F3; F4 phrases four-syllable 6.2.2 noun  L2; L6; 7.4; L12; R3; 20.3–B21; 24.2–⑮ prepositional L8–9 verb  4.1; 6.2.2; 7.6 see also coordination; euphonic rules; subordination predicate 1.3 inversion 14.3 predicator 1.3.2 nominal L3 verbal L4 and passim prepositions  1.2.4; L8–9; 12.2.2; 23.5–B23 pronouns  1.2.2; 2.3; 2.4.2 indefinite 5.2.1 interrogative  3.4.3; 5.3.2; 8.2.5 personal  2.4–B5; 5.2.3 questions  3.4; 5.3 rhetorical  3.4.3; 5.3.2; 7.2; 16.2–⑦; 16.2–⑩ rhymes  F2; F3; 25.2–⑦; F22 sentences 1.1 and passim complex L15–16 subject  1.3; 1.3.1 dummy 16.2–⑥

infinitive 12.3 inversion 14.3 locative 6.2.2–⑱ topicalization 13.2 subordination  2.4; 6.2; 7.6 superlative 9.3–㉛; 13.4–㉓; 24.3–B24 synonyms F25 syntax  p. xxvi time  3.1; L10; F8 aspect  10.4; 10.5 tense 10.2 topics  L13; R4 verbs  1.2.3; 2.4.3; 4.2.1; 9.4–B12 causative usage  5.2.1 ditransitive usage  5.2.2 ergative 1.2.3n intransitive 5.2.1 negation  5.2.3; 24.2–⑮ neutral  5.2.2; 9.2.2 nominal usage  6.2.2 pseudo-transitivity 8.2–B11 stative  5.2.1; 8.2.2 see also comparisons; copula; numerals; superlative; time words content  1.2; R1 function  1.2; R1 fusion  5.2.3; 5.3.1; 13.4–B16; 7.5n; 8.2.5 lexical properties  p. xxiv; 1.1 two-syllable F9 vs. characters  p. xxiv; F4 word classes  1.1; 1.2; 4.2.1; 6.2.2; 7.6–B10; 9.4–B12 see also adverbs; conjunctions; interjections; measure words; nouns; particles; prepositions; pronouns; synonyms; verbs writing  p. xxiii; F3; F4; F12

556